Are

Re

~ eo e

-

al

TAS. ng vs

— arene
: a ie Se ane Sal agente

= mst abel 4

V4

resesl ee

Accession No,

Given by (Liotz

Place, +/Z

*,* No book or pamphlet is to be removed from the Lab-
oratory without the permission of the Trustees.

UZICKA,
3INDING
TAW ST.
Mo.

y i i
ny
mh)

Hi

iM ; (a
y i fi nit! i NA
Wi Rte
i ‘e Ny
Ny Davee atan atta
PRU TC MiNi
yi ral { i i

Nis
f

CA
4 Aven)

UB exh
NOAM

NG,
CN SRR

igh ¥)

‘i ij i Ny i
i i AR
WA

ki ya 4

une, ft My )
Reais ital

K a
i 4

eae
(art
i

i |

Nu

ii,

wy

AM
ARAN i

Ht
te Nt

RAY AO
Met

THE AMERICAN JOURNAL

OF

ANATOMY

EDITORIAL BOARD

LEWELLYS F. BARKER,

Johns Hopkins University.

THOMAS DWIGHT,

Harvard University.

JOSEPH MARSHALL FLINT,

University of California.

SIMON H. GAGE,

Cornell University,

G. CARL HUBER,

University of Michigan.

GEORGE 8. HUNTINGTON,
Columbia University.
FRANKLIN P. MALL,
Johns Hopkins University.
J. PLAYFAIR McMURRICH,
University of Michigan.
CHARLES 8. MINOT,
Harvard University.
GEORGE A. PIERSOL,
University of Pennsylvania.

HENRY McE. KNOWER, SEcRETARY,
Johns Hopkins University.

VOLUME IV
1905

THE AMERICAN JOURNAL OF ANATOMY
BALTIMORE, MD., U. S. A.

The Fricdenwmaflo Company

12 ae

BALTIMORE, MD.,

Tit

LE

VE

vers VEE

Vit
VIII.

CONTENTS ©F VOLahy

No. 1. DeEcEMBER 20, 1904.

FRANKLIN P. Matz. On the Development of the

Blood-Vessels of the Brain in the Human Embryo. 1
With 3 double plates and 4 text figures.

THomas Dwicut. The Size of the Articular Surfaces
of the Long Bones as Characteristic of Sex; an
Anthropological Study . 19

With 6 plates.

J. PuayrarR McMourricu. The Roe of the

Crural Flexors . 33
With 14 fect eatin es.

JOSEPH MarsHALL Fiint. The Framework of the

Glandula Parathyroidea . . v7
With 3 text figures.

Grorce L. Streeter. The Development of the Cranial
and Spinal Nerves in the Occipital Region of the
Human Embryo . 83

With 4 plates and 14 feat fee.

Grorce C. Price. A Further Study of the Develop-

ment of the Excretory Organs in Bdellostoma Stouti . 117
With 31 text figures.
No. 2. FEBRUARY 28, 1905.

FRANKLIN P. Matyi. Wilhelm His . . 139

Cuartes R. Bardeen. The Development of dhe
Thoracic Vertebre in Man. . 163

With 7 plates.

Dean D. Lewis. The Elastic Tissue of the Human

Larynx ae ye?

With 5 plates.

lv

XAT.

OE

XVII.

>. VABNT

Contents

R. H. WuireHeap. Studies of the Interstitial Cells of
Meydig. 9. Qe a al ee
With 5 5 ey seaeen

KATHARINE Foor and E. C. Stropett. Prophases and
Metaphase of the First Maturation Spindle of Allolo-
bophora Fetida .. . Mere ee eee 4k

With 9 Hiner

CHARLES-SEDGWIcK Minor. Genetic Interpretations
in the Domain of Anatomy. (|... . . <)> 2) ipeeoeae

No. 8.. May 25, 1905.

CHARLES R. BARDEEN. Studies of the Development of
the: Efuman ,Skeleton’ . 0.0 6)... <
With 13 plates.

Ropert Bennett Bean. A Composite Study of the
Subclavian Artery in Man. . . oon ket
With 7 figures and 18 nies

Irving Harpesty. On the Occurrence of Sheath Cells
and the Nature of the Axone Sheaths in the Central
Nervous System . . 9. + « 2 2 =. 2) goes

FLORENCE R. Sastn. The Development of the Lym-
phatic Nodes in the Pig and Their Relation to the
eymiph, Hearts ice. Sk wo ee

With 17 fest dere

No. 4. SEPTEMBER —, 1905.

FRANKLIN P. Matt. On the Angle of the Elbow . . 391
With 1 figure and 8 tables.

EK. Linpon Meuuus. A Study of the Location and
Arrangement of the Giant Cells in the Cortex of the
Right Hemisphere of the Bonnet Monkey (Macacus
SiniCus) cess. * ao a a ee. ae

With 3 figures.

NA DX.

XX.

wo XXT.

XXII.

XX.

Contents

SUSANNA PueEtps Gace. A Three Weeks’ Human
Embryo, with Especial Reference to the Brain and the
Nephric-Systemesaass osm

With 5 plates.

Writt1am Snow Miiier. The Blood and Lymph Ves-
sels of the Lung of Necturus Maculatus .
With 3 text figures and 2 plates.

FRANK A. Strromsten. A Contribution to the Anatomy
and Development of the Venous System of Chelo-

. 409

. 445

1M, 5 2 fol: Dé a oe 5 ESS

With 12 text figures.

PROCEEDINGS OF THE ASSOCIATION OF
AMERICAN ANATOMISTS. #IGHTEENTH

SHSSION, Dec Gipees, 29.1902: «4, 2 EX VOIT

SUPPLEMENT TO VOL. IV, JUNE 1, 1905. G. Carn
Huser. On the Development and Shape of Urini-
ferous Tubules of Certain of the Higher Mammals

With 24 figures.

. 1-98

NOTE.

Table 15, p. 322, summarizes 60 cases. The figures under the head-
ings “* Common trunk” and “ Absent ” do not enter into the totals in the
last column, which were made by adding the figures in the other columns.
In the fifth line the number of cases (3) in which the artery occurred
double should be added, making the total 62 instead of 59.

Table 16, p. 323, summarizes 69 cases. The figures under the headings
“Common trunk,” “ Absent” and ‘“ A. cervicalis superficialis” do not
enter into the totals given in the last column. In the line opposite “ A.
cervicalis ascendens ” transpose 9 and 5.

The total number of cases in which any given artery has been
worked out is often less than the total number of cases dissected, because
for various reasons it was impossible to work out each individual artery
in every case.

ON THE DEVELOPMENT OF THE BLOOD-VESSELS OF THE
BRAIN IN THE HUMAN EMBRYO.

BY
FRANKLIN P. MALL.
From the Anatomical Laboratory of the Johns Hopkins University.

WITH 3 DOUBLE PLATES AND 4 TEXT FIGURES.

During the past year, while studying sections through the heads of
the embryos in the collection at this laboratory, it was noticed that
in some of the specimens the blood-vessels were unusually well marked,
for they were well distended with blood. This natural injection made
it possible to reconstruct the blood-vessels in a satisfactory manner
down to the capillaries. At the same time I obtained from Mr. Brédel.
a number of embryos’ brains in which the arteries had been injected
with Prussian blue, which, together with numerous embryo pigs injected
alive or immediately after death, form the basis of this study.

TABLE OF EMBRYOS STUDIED.

| | | |
Length Thickness of | Direction ot

Number. | jn mm. Sectionin». | Section. From Whom Obtained.
= = ae |e = a ase ES

2 | 7 15 Transverse | Dr. C. O. Miller.
163 i) 20 | fs Dr D. SS damib.
109 | 11 20 ce | Dr. Harvey Cushing.
144 14 40 Sagittal Dr. Watson.

7 19 50 ee | Dr. Irving Miller.
145 33 50 | os | Dr. W. T. Watson.
225 46 ye Injected Dr. Wegefarth.
237 48 | Ae ce Dr. Todd. (Brédel Collection).
235 59 sf | 24 Dr. Linthicum. ze .*
234b | 65 AD | ee Brédel Collection.
ates 80 36 a Brédel Collection. _
234a | 80 50 Transverse Dr. Ashby (Brédel Collection)
238 90 ae Injected | Dr. Smart re ce.
<36 92 ae te Dr. Wilson “ i

|

‘The blood-vessels of five human embryos were reconstructed from
serial sections, and eight older embryos which had been injected were
dissected. The brains of pigs which had been injected with India ink
proved to be of great value to control the studies of the human. It is
quite easy to make single or even double injections of young embryos
by injecting them either before or after death, or both. In case India
ink is injected into the liver of a live pig with a hypodermic syringe, the

AMERICAN JOURNAL OF ANATOMY.—VOL. IY.
1

2 Development of Blood-Vessels of Human Brain

fluid is taken up by the heart and is pumped through the arterial system.
When all the arteries are full the beat of the heart may be arrested by
cooling the embryo. A second injection into the liver with a different
fluid (and for this purpose I usually employed aqueous Prussian blue)
fills the entire venous system. More frequently single injections were
made of the arteries or of the veins by injecting India ink into the liver
either before or after the heart had ceased to beat. India ink, being
resistant, is preferred, for embryos injected with it can be hardened in
alcohol and cleared in a one per cent solution of caustic potash and
preserved in glycerine. Such specimens are perfectly transparent, show-
ing the arrangement of the vessels beautifully and their relation to the
structures within the head. Sagittal sections of whole embryos are also
very valuable for study, for the half brain is easily peeled out, leaving
the injected membranes intact within the head.

It is difficult to make complete injections of the veins of the head in
dead embryos without extravasations into the arachnoid spaces. So
frequent is this extravasation that one is inclined to think that the vessels
of the brain, especially the veins, have open communication with these
spaces. But since the arachnoid spaces are always free from blood, and
since complete injections with India ink made by the contraction of the
heart in live embryos do not form extravasations, it must be concluded
that the vessels are closed in life. A similar communication has been
demonstrated by Key and Retzius* in the adult brain by injecting Prus-
sian blue into the arachnoid spaces. Frequently the fluid passes over
into the sinuses through the Pacchionian bodies, showing that here again
the communication is easily established. This question will be taken up
again in the description of the specimens.

I shall first describe the blood-vessels of the brains of eight embryos
of the third month, which had been injected, then take them up in
regular order, beginning with a reconstructed specimen of the fourth
week. At this point I wish to express my great obligations to Mr. Brédel
for much of this valuable material. He has an exceptional opportunity
to obtain many fresh specimens which can be injected, and I sincerely
hope that physicians will continue to send him all the embryos they
obtain, for Mr. Brédel makes the greatest possible use of them.

INJECTIONS.
Unfortunatley there is a tendency for the Prussian blue which has
been injected to extravasate over the surface of the brain, interfering

1Key and Retzius: Studien in der Anatomie des Nervensystems und des
Bindegewebes. Stockholm, 1875, p. 218.

Franklin P. Mall 3

very much with the sharpness of the arterioles and making it impossible
to define the veins, or embryonic sinuses. So constant is this extravasa-
tion in position and degree that it often seems as if the arachnoid spaces
communicate freely with the veins, but, as will be shown presently, this
is not the case.

In the smallest specimen (No. 225, 46 mm. long) the middle cerebral
artery and the arteries to the mid-brain are well injected, but in no case
does the injection extend into the brain substance. The arachnoid spaces
are filled evenly with the blue injecting fluid, but there is none within
the ventricles. Since the fluid does not reach the capillaries, it is evident
that the extravasation took place from the arterioles, and this seems to be
the case, for the arterioles are easily torn at the point they enter the brain
substance. In the early stages the brain is attached only slightly to the
embryonic pia mater, and it is practically impossible to remove the
brain with its pia mater intact, as can be done in older embryos or in
the adult. At the point the vessels leave the pia mater to enter the brain
substance the blood-vessels have but a single endothelial wall, and it is
here that the rupture and extravasation take place when these arteries
are injected.

In an embryo a little older, No. 237, Fig. 1, the injection of
the artery is practically perfect, and I have therefore given a drawing
of it. The brain was peeled out with its pia mater only with difficulty
and over the region of the lateral cerebral fissure (Sylvius) some of the
vessels separated and remained attached to the dura. This portion was
drawn inverted and redrawn upon the brain, and the point at which the
main trunks are torn off is indicated in the drawing in the region of the
island. The injection is practically a complete arterial injection with
but little extravasation into the arachnoid and none into the ventricles.
An extravasation is over the region of the island, on both sides, and to
a slight extent over the mid-brain on one side. The arteries divide and
subdivide in regular fashion until the terminal branches are reached,
when they turn at right angles to enter the brain substance. There are
from five to ten of these cortical arteries to each square millimeter of
brain surface. Around some of them there is some extravasation of
Prussian blue, indicating the way the blue enters the arachnoid spaces.

Over the surface of the brain of an embryo 65 mm. long (No. 234>)
there are numerous blue spots, about one to each square millimeter.
Where the spots are larger there is a tendency for them to run together,
but in general the brain is only spotted rather than being covered evenly
with an extravasation. There is no extravasation in the ventricle. In
another brain of about the same age (No. 235, 59 mm. long) the ex-

4 Development of Blood-Vessels of Human Brain

travasation is complete, filling all the arachnoid spaces and the whole
ventricle. After the extravasation was brushed off, the brain substance
was still found to be spotted, showing that the extravasation penetrated
the brain substance.

In an embryo of the beginning of the fourth month (80 mm. long)
the whole brain was evenly spotted, about one spot to each square milli-
meter. Another specimen of the same age and of the same general
appearance (No. 234%, 80 mm. long) (Fig. 2) was cut into serial sections
in order to study the relation of the spots to the surrounding tissues and
to the cortical arteries. Around the large cortical arteries (possibly the
medullary arteries) there is an extravasation which encircles the vessel
as a small spherical body. There is no rupture of the vessel. It indi-
cates that at this point the vessel is at least very pervious. There is no
extravasation into the ventricle.

In specimen No. 238 (90 mm. long) both the arteries and the veins
were injected without injecting the capillaries. There was no injection
of the brain substance, and there is no extravasation of the cortex nor
into the ventricles. At the base of the brain and in the falx there is
considerable extravasation, apparently coming from the veins. In an
embryo of the same age (No. 236, 92 mm. long) the arterial injection
is complete again, with the usual spots of extravasation in the cortex of
the cerebral vesicle. The extravasation fills all of the arachnoid spaces as
well as the cavities of the ventricle. The injection passes through the
medial opening into the fourth ventricle (Majendie), and apparently
the ventricles are injected through this opening from the arachnoid.

It is apparent from the description of the injected embryos that as a
rule the extravasation into the arachnoid spaces takes place from the
arteries as they penetrate the cortex of the brain, and that in case the
veins are injected the extravasation is directly from them. This con-
clusion was reached in part by making corresponding injections of em-
bryo pigs, many being constantly at my disposal. In general the
extravasation is the same in the pig as it is in the human embryos. It
frequently appeared, however, as if the India ink injected leaked with
even greater ease from the veins and sinuses of the pig’s brain. In
embryos in which the heart had just stopped beating the injected fluid
would first fill the jugular veins, then the sinuses, from which the
arachnoid spaces filled as readily as did the capillaries.

When the arachnoid spaces were filled by injecting directly into the
lateral ventricles of perfectly fresh embryos, the injected fluid would not
pass over into the veins. I made this test repeatedly with live embryos
from 3 to 8 em. long, always with the same result. It is best to inject

Franklin P. Mall ~ 5

ordinary India ink into the ventricle of a live embryo with a hypodermic
syringe. The ink spreads at once throughout the central canal of the
brain and cord and escapes through the medial opening of the fourth
ventricle and fills the spaces of the arachnoid of the whole brain and cord.
From the cord the ink extended for a short distance along the main
trunks of the spinal nerves. In the larger embryos the ink invariably
flowed freely from the mouth of the pig as soon as all of the arachnoid
spaces had been filled. After hardening the specimens in formalin, razor
sections showed that it had reached the mouth through the Eustachian
tube. It had entered the middle ear along the trunks of the seventh and
eighth nerves. In younger embryos (5 em. long) the fluid came out of the
mouth in only half of the tests, while in the smallest ones injected
(3 em. long) it did not come out of the mouth at all.

In all of these tests the India ink or the Prussian blue should have
passed over into the veins were the communications with them free. In
all instances the pigs were still alive or just dead when the tests were
made, for it is known that extravasations take place with the greatest of
ease after the embryo has been dead for some time. While in these tests
injections could be made with ease from the veins into the arachnoid
spaces, but not in the opposite direction, in embryos still alive it was
found that in no instance would an injection into the artery pass into the
arachnoid spaces. The live embryo may be injected using its own heart
to inject the India ink. If the uterus is kept warm the embryo will
remain alive for an hour or longer, giving ample time. The ink is to
be injected directly into the liver with a hypodermic syringe and then
by means of gentle massage or by gravity it is forced into the heart,
which gradually pumps it all over the body. The arteries to the brain
fill slowly and the granules pass over into the veins. If at this time the
embryo is cooled the heart will stop, thus giving a single injection of the
arteries. If it is continued, the veins will fill through the capillaries,
vhich confuses more or less. Yet this double injection is desired in this
test, for the result is always the same: in no instance is there an extrava-
sation into the arachnoid space. In case too great a quantity of ink is
injected into the liver, it is forced directly into all of the veins of the
body and then the ink granules will leave the veins and enter the arach-
noid spaces. If the injection of the arteries and veins of the brain is
made through the arteries, using the embryo’s heart to do the pumping,
all of the granules remain within the blood-vessels, showing conclusively
that there are no free communications between the vessels and the arach-
noid spaces. When the granules do leave the spaces by injecting them
directly into the veins, we must conclude that artificial openings are

6 Development of Blood-Vessels of Human Brain

made in their walls by the excessive pressure, no matter how careful
we are in making the injection.

The delicate veins and capillaries can be injected without extravasation
of the fluid in case it is done by injecting a small quantity of India ink
into the liver and allowing it to run from there to the head by gravity.
In this way I have often obtained beautiful specimens which are clear
and sharp. This pressure, which is often not over one centimeter of water,
is so small that it cannot possibly be imitated with a syringe. In fact it
is similar to that produced by the embryo’s heart, and with these normal
pressures no extravasation takes place.

I may add that in all cases the embryos were placed at once in the
strongest alcohol in order to prepare them so that they may subsequently
be cleared in a one per cent solution of potassium hydrate. Specimens
of this sort are beautiful and instructive; a black vascular system shows
through a translucent embryo. These specimens proved to be a most
valuable control in the study of the sections of the human embryos, for
I had them in abundance, and I also got some ideas of the variations of
the blood-vessels and their general relation to the surrounding structures.

ARTERIES.

It has been shown during recent years that in the embryo a series of
segmental arteries arise from the aorta, which in the head-end of higher
vertebrates unite on their distal ends to produce the two vertebral arteries.
These in turn unite at the middle line to produce the basilar artery, as
has been shown by His in his monograph on the human embryo. By
this process of loop-throwing we have produced in the human embryo of
four weeks two vertebral arteries which unite to form the basilar and on
the anterior end join with the internal carotids. So as soon as the
vertebral arteries unite to form the basilar we have marked off the circle
of Willis, and considering its relation to the neural tube we can identify
its branches to the brain as they arise. In the specimen four weeks old
(Fig. 3) the arteries are not well marked, and it is difficult to outline the
primary circle of Willis, let alone the branches arising from it. A
specimen a little older has in it all of the circle of Willis with the
primary arteries to the brain beautifully outlined (Fig. 4), and it is
possible to follow them through the capillaries over to the veins. Were
it not for the great number of variations found in the arrangement
of blood-vessels it would be easy to identify most of the arteries in this
specimen by considering them in relation to the cranial nerves and other
structures.

Franklin P. Mall i

The circle of Willis is fully formed in this specimen and extends from
the bifurcation of the basilar artery to the anterior communicating. At
the point the carotid enters the circle there is a short ophthalmic which
is also present in the embryo of the fourth week (No. 2), and is shown
in No. 74, Fig. 5. Throughout the region of the brain branches rise at
quite regular intervals from the anterior communicating to the vertebral
arteries. So regular are these branches that they might be spoken of
as the segmental arteries to the brain. These are then gradually shifted,
some becoming enlarged and others disappear.

The anterior and middle cerebral arteries (Fig. 4) arise as a common
stem and form a main branch encircling the optic stalk from which
small branches pass on the lateral side of the cerebral vesicle while the
main stem continues to the front of the brain and communicates with
its fellow on the opposite side immediately behind the olfactory pit. It
is easy to imagine the anterior cerebral pushed into place when the
cerebral vesicle protrudes over it in every direction. In embryo No. 74
(Fig. 5) the middle cerebral is much better marked, while the anterior
cerebral cannot be followed to its end. In embryo No. 145 (Figs. 6-8)
the adult form of these two vessels is well given. Numerous radiating
branches mark the middle cerebral over the embryonic island (Fig. 6)
and the anterior cerebral extends to the mesial side of the hemisphere
as it does in the adult (Fig. 7). The anterior cerebral artery is pic-
tured by His in his last great monograph on the brain? in an embryo °
of about the same age as my embryo No. 145. This illustration is from -
a sagittal series like the one from which my Figs. 6-8 were reconstructed.
In His’ paper this vessel is called “die vorder Bogenvene die das Blut
aus dem vordern Abschnitte der Hemisphiirenwand sammelt2?* When
the direction of this vessel is considered, and especially when it is re-
constructed, it is easily shown that His was in error in calling it a vein.

The anterior choroidal artery is next in order, for it also arises from
the carotid artery and its destination, the choroid plexus, is well marked
in young embryos. In Fig. 4, the artery which takes this position is
intimately associated with the middle cerebral and lies between the
cerebral hemisphere and the optic thalamus. It may be that what I have
termed the anterior choroidal is in reality the middle cerebral, and that
the artery more dorsalwards is in reality the choroidal, for it is well
known that arteries often shift a great deal in young embryos. Not

*His: Die Entwicklung des menschl. Gehirns. Leipzig, 1904.
® Page 79, Fig. 51.
12) PAR

8 Development of Blood-Vessels of Human Brain

until their walls are fairly well developed are the arteries well fixed.
The arrangement shown in Fig. 4 is again present in embryo No. 74,
Fig. 5, and here the choroid plexus is well developed. In an older em-
bryo, Fig. 8, the anterior choroidal artery is well marked, and it arises
farther back,—from the posterior communicating artery. For this rea-
son I often thought that the more dorsal artery in Fig. 4 represented
the choroidal, and only in a much later stage than that shown in Fig. 8
is the anterior choroidal shifted from the posterior communicating to
the carotid. To test this question further I examined numerous trans-
parent pigs of corresponding stages in which the arteries only had been
injected, and in all cases the anterior choroidal arose in common with
the middle cerebral, and after this I was strongly inclined to consider
the origin of the anterior choroidal from the posterior communicating
in Fig. 8 as a variation.

The posterior cerebral artery is relatively late to develop, and in early
embryos its place of origin is taken by a number of large branches to
the mid-brain. This is very marked in Figs. 1, and 4 to 8. In the
large embryo (Fig. 1) the vessels have been injected and in drawing it
the cerebrum was pulled forward to show the large artery to the mid-
brain. There is also a small posterior cerebral artery present showing
that for a long time the artery to the mid-brain is much more prominent
than the posterior cerebral. In the adult the posterior cerebral arteries
mark the terminations of the basilar and lie immediately in front of
the third and fourth cranial nerves. The arteries which fulfill these
requirements supply the dorsal portion of the mid-brain, corresponding
to the posterior quadrigeminal body in the adult brain. But in the
adult the posterior cerebral artery, in addition to its main branches to
the cerebrum, supphes much more than this, for it also sends branches
to the crus, posterior part of the thalamus walls of the third ventricle,
as well as elsewhere. In fact, all of the branches together arising from
the circle of Willis between the third and fourth nerves behind and the
origin of the middle cerebral in front (compare Figs. 4 and 9) must
become united to produce the posterior cerebral artery. The region
supplied by these numerous branches in the embryo is supplied by the
posterior cerebral in the adult, and in its development these branches
must be gradually drawn together into one stem to produce the final
condition. And this is to be expected. It is only after the terebrum
makes its appearance and reaches the great prominence it does in man
that the condition found in the lower vertebrates is overshadowed. At
a relatively late stage, later than the one shown in Fig. 1, all of these
arteries arise from a single trunk. In this there are still two main
trunks which must unite subsequently to form the posterior cerebral.

Franklin P. Mall 9

The posterior communicating branch must be formed by shifting
nearly all of its branches, as shown in Fig. 4, back to the third and fourth
nerves to produce the posterior cerebral artery, while the cerebral hemi-
sphere is growing over the thalamus and mid-brain. Or by a series of
arches, as indicated in Fig. 4, the arteries to the more dorsal portions of
the thalamus and mid-brain, as well as to the structures which wander
into this region, are gradually transferred towards the basilar, leaving
the small branches of the posterior communicating to supply the imme-
diate neighborhood, as it does in the adult. At any rate the large
vessel arising from the posterior communicating in Fig. 1 arises from
the posterior cerebral, probably as the posterd-lateral set, in the adult.
The question is further complicated by variations, which are quite
numerous, the most common variation of the posterior cerebral being
in its origin, which is transferred from the basilar to the internal
carotid. I have also foynd all kinds of combinations of this artery
with neighboring arteries in the embryo pig, which I interpret in part as
transformation stages. Furthermore, it may be possible that shifting
of arteries takes place until the individual is fully grown, for Bean*
has shown recently that the branches of the subclavian artery of the
infant differs from those of the adult.

The branches from the basilar and vertebral arteries are more easily
followed, for in this region there is less shifting, and the landmarks
prove to be of more value. In the upper part of the mid-brain there is a
cluster of branches which are destined to become the superior cerebellar
arteries (Fig. 4). This group is reduced to a single artery in Figs. 5
and 6, where it is just behind the isthmus. The next group in Fig. 4
is the transverse pontine between the superior cerebellar and the otic
vesicle. Then the anterior cerebellar between the seventh and eighth
nerves near the otic vesicle. Finally, the group which perforates the
root of the twelfth nerve is destined to form the posterior inferior cere-
bellar. This branch is also shown in Fig. 1. Between this and the otic
vesicle there are a couple of branches, shown in Fig. 4, the fate of which
is uncertain.

VEINS.

While the arteries of the brain undergo many changes in their devel-
opment, their history is relatively simple when compared with the gyra-

5 According to Windle (Journal of Anatomy and Physiology, XXII, 1888),
this variation occurred 28 times in 400, or in 7 per cent of the cases.
‘Bean: American Journal of Anatomy, Vol. 4.

10 Development of Blood-Vessels of Human Brain

tions the veins undergo. The subject is, however, simplified to a great
extent by the excellent studies of Hochstetter and his pupils, Salzer, and
Groszer and Brezina, who have unraveled many of the tangles of the
anterior cardinal vein while it is being transformed into the brain
sinuses. The study of the development of the veins of the head of
the guinea-pig by Salzer‘ is especially of value to me, for it takes up a
number of points which would be difficult to interpret properly from
my material.

It is generally believed since the time of Luschka * that the blood from
the veins of the brain leaves the embryonic skull through a foramen in
front of the temporal bone—the foramen jugulare spurium—and emp-
ties into the external jugular vein. A secondary communication is
formed with the internal jugular vein, which in man and in monkeys
is the only outlet of the brain sinuses, both communications remaining
open to a greater or less degree in many vertebrates. Luschka also
found a human skull with a foramen jugulare spurium present be-
tween the temporal bone and the glenoid fossa. This opening is referred
to frequently in the various text-books on anatomy,’ and it is explained
by stating that it is the remains of a channel through which the blood
poured in the foetus. This explanation may be correct as far as it goes,
but when it is asserted that the brain sinuses at first communicate with
the external jugular vein through the foramen spurium and later with
the internal jugular vein, a conclusion is drawn which the facts do not
warrant. Although Salzer showed conclusively that the internal jugu-
lar vein receives all the blood from the brain from the very earliest
stage, and that the connection with the external jugular is of much later
formation, Luschka’s statement is still retained in the text-books.”
While Salzer corrected the erroneous interpretation of Luschka, he also
discovered that in the embryo the veins first leave the embryonic skull
through a canal near the seventh nerve, and then emptied into the internal
jugular vein. All this takes place long before there is a trace of an
external jugular vein present, so the idea of Luschka that the external
jugular vein is the primary vein for the blood from the brain is un-
tenable, and should be removed from the text-books as soon as possible.

Salzer’s work shows that the anterior cardinal veins of mammals are
placed on either side of the chorda ventral to the brain between the

*Salzer: Morph. Jahrbuch, XXIII, 1895.

®Luschka: Denkschriften der Wiener Akademie, XX, 1862.
®See for instance Cunningham’s Anatomy, p. 116.
Cunningham’s Anatomy, Figs. 603 to 606.

Franklin P. Mall 11

roots of the cranial veins. They soon begin to shift lateralwards, and
by a process of sprouts encircle the cranial nerves successively and soon
come to lie to the lateral sides of the nerves. In other words, the nerve
trunks have wandered through the veins and changed positions with
them. This all takes place without interrupting the circulation through
the vein. In Fig. 3 the vein is shown partly wandered out, the otic
vesicle, seventh and ninth nerves now being on the medial side of the vein.
A number of sprouts encircles the tenth nerve, which in Fig. 9 is also on
the medial side of the vein. Finally the fifth nerve wanders through the
vein. In Fig. 13 this has taken place in part, and in Fig. 11 the
change in position is complete. What I have here given in rapid order
comprises the results found by Salzer in the guinea-pig, and I repeat
it only to illustrate his point with my figures. Now Salzer calls the
vein which lies to the medial side of the nerves the anterior cardinal
vein, and that portion which moves to the lateral side of the nerves the
vena capatis lateralis.” After the vein crosses the twelfth nerve passing to
the heart he calls it the internal jugular. These terms I shall retain,
for they are most useful in the description of the fate of the anterior
cardinal vein. So finder wir, says Salzer (J. c., p. 248), bei den unter-
suchten Saiigern sehr schén tibereinstimmende Verhaltnisse in Bezug
auf die erste Entwicklung der Venen des Kopfes. Ueberall wird die
urspriinglich medial von den Kopfnerven gelegene Vene durch ein
Gefisz ersetzt, das eine laterale Lage den Nerven gegeniiber einnimmt.
Diese Lageverinderung geht durch Inselbildung vor sich, und zwar
bilden sich derartige Inseln zuerst um Acustico-facialis, fast zu gleicher
Zeit auch um den Vagus herum, dann erst erfolgt die Verlagerung der |
Venenbahn dem Hypoglossus gegeniiber; dem Trigeminus gegentiber
behilt die Vena verhiltnismaszig lange ihre urspriingliche Lage bei.
Ist das knorpelige Skelet angelegt so verliisst das Blut der vorderen
Hirnabschnitte gemeinsam mit dem Facialis die Schidelhohle, wahrend
das Blut des Hinter- und Nachhirns von einer Vene gesammelt wird,
die durch das Foramen jugulare an der lateralen Seite des Vagus nach
auszen zieht; hier verbinden sich beide Gefiisze zur Vena jugularis
interna. Bald jedoch obliterirt nach Ausbildung einer Anastomose
dorsalwiirts vom Gehérorgan die neben dem Facialis austretende Vene,
so dass die neben dem Vagus austretende Vene die einzige abfiihrende
Blutbahn des Schiidels darstellt. An dieses Verhalten schlieszen sich
die nun sekundiir auftretenden Verbindungen der Gefisze des Schade-
linneren theils mit den Gesichtsvenen, theils mit den Venen des Riick-

1 His (Entwickl. des menschl. Gehirns) calls this vein Basalvene.

12 Development of Blood-Vessels of Human Brain

enmarkes an. Dabei geht die Bahn durch das Foramen jugulare
entweder vollstindig oder zum Theil zu Grunde. Die bei den meisten
Saiigern auftretende sekundiire Verbindung ist die, welche das Foramen
jugulare spurium zum Austritte benutzt, doch giebt es auch 'Thier-
formen, z. B. die Katze, bei denen ein solches gar nicht zur Ausbildung
kommt, obwohl die Vena jugularis interna fast vollstandig zu Grunde
gegangen ist. Hier treten eben die sekundiren Verbindungen, welche
die Orbital- und Nachhirnvenen eingehen, fiir diese Gefasze ein. Mithin
kann man wohl behaupten, dass die Vena jugularis interna als Fort-
setzung des Sinus transversus, wie sie beim Menschen und beim Affen
am schénsten ausgebildet ist, ein primitiveres Verhalten darstellt, bei
welchen die Vena jugularis externa die hauptsiichliche, wenn nicht die
einzige abfiihrende Bahn des Schiidelinneren darstellt.

It is apparent from the above that the beginning of the internal
jugular vein is marked by the twelfth nerve crossing the anterior cardinal
vein and that it extends to the lateral sinus through a number of
sprouts behind the otic vesicle (Fig. 13). The vena capatis lateralis
is that portion of the anterior cardinal vein which wanders to the lateral
side of the cranial nerves and extends from the twelfth to the fifth nerve.
The portion of the anterior cardinal vein which lies medial to the fifth
nerve retains that position throughout its development and marks the
cavernous sinus (Fig. 10). Into the embryonic cavernous sinus there
empties the ophthalmic and anterior cerebral veins. The latter soon
extend to the embryonic superior sagittal sinus. Between the fifth
and 7th nerves a vein extends to the region of the cerebellum, the vena
cerebralis media (Figs. 9, 10 and 13), and behind the otic vesicle there
extends through the embryonic jugular foramen the vena cerebralis
posterior (also shown in Fig. 11).

We have, therefore, in the embryos of the second month an arrange-
ment of the veins in the head similar to that found in the reptiles as
shown by Groszer and Brezina,” and I shall employ the nomenclature
of these authors. The anterior cardinal vein shifts lateralwards and
by the end of the first month is partly lateral to the otic vesicle as shown
in Fig. 3. The process is shown more advanced in Fig. 9, and is com-
plete in Fig. 13. The condition shown in Fig. 13 is that which is per-
manent in selachians. Now the reptilian stage is entered. From the
lateral vein—the vena capatis lateralis—three veins extend into the
head and encircle the brain. The first—the vena cerebralis anterior
(shown well in Fig. 10)—passes up and over the cerebral vesicle, 2. @.,

12 Groszer and Brezina: Morph. Jahrb., XXIII, 1895.

Franklin P. Mall a:

the region of the island. The second—the vena cerebralis media (Figs.
9, 10 and 13)—arises at the anterior juncture of the vena capatis lateralis
with the anterior cardinal vein between the fifth and seventh nerves. The
third—the vena cerebralis posterior—arises from the ‘posterior part of
the vena capatis lateralis with the internal jugular behind the otic
vesicle and enters the embryonic skull through the jugular foramen
(Fig. 12), and ultimately becomes the transverse sinus. From this simple
reptilian stage the mammalian is formed, and in man but little must be
added to, and but little subtracted from, the general plan.

The anterior end of the anterior cardinal vein remains in large part
en the medial side of the fifth nerve in the human embryo and is ulti-
mately transformed into the cavernous sinus. From the earliest stages
the ophthalmic vein enters this sinus as is shown in all of the embryos
studied. Although I have no evidence regarding the development of
intercavernous sinus, it is easy to understand its development by branches
from the cavernous sinus growing to encircle the hypophysis, and then
to unite, thus forming a plexus around it. So also by an extension of
the cavernous sinus forwards the spheno-parietal sinus must be formed.
In the early embryos the anterior cardinal vein or the portion which
forms the cavernous sinus is extended forward to form the vena
cerebralis anterior, which ends in the bilateral superior sagittal sinus as
shown in Fig. 3. With this the veins from the region of the island
communicate as shown in Fig. 10, the basal portions of which are evi-
dently retained to form the middle cerebral (superficial Sylvian) vein.
So also the superior sagittal sinus, the superior and inferior petrosal
sinuses and the vena capatis lateralis are directly continuous with the
cavernous sinus from their beginning.

I have spoken enough about the vena capatis lateralis above, and wish
enly to add its relation to the permanent brain sinuses in man at this
point. It may be defined as that portion of the anterior cardinal vein
which is transferred to the lateral sides of the cranial nerves extending
in the human embryo from the fifth to the twelfth cranial nerves, being
directly continuous in front with the anterior cardinal vein, or better
the cavernous sinus, and behind with the internal jugular vein. This
vein is clearly outside of the skull, leaving it between the fifth and seventh
nerves (Fig. 12), and then communicating with the internal jugular.
It is this vein which Kolliker believed to be the external jugular, and
apparently confirmed Luschka’s notion regarding the relation of the
external jugular vein to the brain sinuses. It certainly does leave the
skull along the root of the seventh nerve, a line in common with the so-
ealled foramen jugulare spurium, but it disappears long before

14 Development of Blood-Vessels of Human Brain

external jugular vein is formed, as shown by Salzer. The vena capatis
lateralis is fully developed during the fifth week, as is best shown in
Fig. 13. In this embryo it is irregular in shape, ending in a lakelet
behind, a condition which may also be due to the way the blood accu-
mulated in this vein just before the death of the embryo. In Fig. 10,
which is from an embryo in which these veins were gorged with blood,
the lakelet is not present. Soon the vena capatis disappears, and veins
more dorsalwards carry blood from the brain, as shown in Fig. 11.

Fia. 14.
Fria. 14. Diagram of the veins of the head of an embryo four weeks old. ACV, anterior
cardinal vein; VCL, vena capatis lateralis; SUS, superior sagittal sinus; AV, auditory
vesicle; V, fifth nerve; L, eye.

Fia. 15,

Fia. 15. Diagram of the veins of the head during the fifth week. VCP, vena cerebralis
posterior; VOM, vena cerebralis media; WVCA, vena cerebralis anterior; TH, torcular
Herophili; OV, ophthalmie veins; VJ, jugular vein.

When the vena capatis is well developed it sends from its two extremi-
ties two main veins to encircle the brain and to collect its blood. The
first of them, the vena cerebralis media, arises at the point of juncture

between the vena capatis lateralis and the cavernous sinus and extends
between the fifth and seventh nerves towards the region of the cerebellum.

Franklin P. Mall 15

The vena cerebralis posterior arises from the vena capatis lateralis more
dorsalwards, at its juncture with the vena jugularis interna, and encircles
the hind-brain in the region of the twelfth nerve. These two veins are
well shown in Figs. 9 and 13.

Is

wok

Fi4q. 16.

Fig. 16. Diagram of the veins of the head at the beginning of the third month. Letter-
ing as before.

Fi4q. 17.

Fid. 17. Diagram of the veins of the brain of an older foetus. Lettering as before. LS,
transverse sinus; SR, sinus rectus; ILS, inferior sagittal sinus; VG, great cerebral vein
(Galen); CS, cavernous sinus; SS, middle cerebral vein (Sylvian); SPS, spheno-parietal
sinus; Sup. Pet., superior petrosal; Inf. Pet., inferior petrosal sinus.

16 Development of Blood-Vessels of Human Brain

The supertor sagittal sinus is formed by an accumulation of small
veins over the dorsal side of the cerebral vesicle (Fig. 3), which some-
times appear as a tuft (Fig. 13) and at other times as a lakelet (Fig.
9). Soon the sinuses of the two sides communicate (Fig. 10), and from
now on the paired sinuses are single. At first the sinuses com-
municate with the anterior cardinal vein (the cavernous sinus) through
the vena cerebralis anterior, and these two veins take up all of the small
veins from the cerebral vesicle (Fig. 10). With the growth of the
cerebrum the superior sagittal sinus is shifted downwards and its
communication with the cavernous is broken. It now communicates
with the vena capatis lateralis through the vena cerebralis media, a tran-
sitional form being shown in Fig. 10. With great rapidity the commu-
nication is transferred to the jugular through the vena cerebralis pos-
terior, which leaves the skull through the jugular foramen. This stage
is shown in part in Fig. 11, which is a reconstruction from a partial
natural injection. However, even in this case the superior sagittal
sinus must be shifted more dorsalwards, for in this embryo it still
passes lateral to the otic vesicle, and therefore in this region it is outside
of the skull. The same criticism can be made of Salzer’s figure of the
corresponding stage in the guinea-pig.” Here also the vena capatis
lateralis, as well as the first dorsal anastomosis, is lateral to the otic
vesicle, and therefore cannot possibly be the permanent vein in this
animal. In order to reach the permanent form, as shown in Salzer’s
Fig. 5, a second dorsal anastomosis must be established, and this is
well begun, as Salzer’s Fig. 4 shows. So in order to complete the super-
ior sagittal and transverse sinuses a more dorsal anastomosis must be
established than that shown in Fig. 11, and the indications for this are
present in this figure, as well as in Fig. 10. In this latter figure the
superior sagittal sinus must be transferred completely from the vena
cerebralis anterior to the vena cerebralis posterior, and in so. doing
the vena capatis lateralis is obliterated. In case they all remained open,
we would have the condition found in Tropidonotus,” but this is not the
case, as is indicated in Fig. 11. The complete condition of the superior
sagittal sinus is shown in Fig. 8. Here the internal jugular com-
municates through the vena cerebralis posterior with the dorsal end of
the superior sagittal sinus along the line of the hind-brain and mid-brain.
The steps towards this are all indicated in Fig. 10. Therefore the main
portion of the transverse sinus is formed directly from the vena cerebralis
posterior.

#8 Salzer: Morph. Jahr., XIII, Taf. XVIII, Fig. 4.
“Groszer and Brezina: Morph. Jahr., XXIII, Taf. XXI.

Franklin P. Mall 17

If now the vena cerebralis media, as shown in the human embryo, is
compared with that in Tropidonotus, and in turn with that of the adult
sinuses, it is seen that the vena cerebralis media is the superior petrosal
snus. They all communicate with the cavernous sinus between the fifth
and seventh nerves, they lie lateral to the cranial nerves behind the fifth,
and they are also medial to the otic vesicle, 7. ¢., they are within the skull.
This latter condition is not yet the case in Fig. 11, but is indicated in
Fig. 10, and it is marked by the stub vein near the pons in Fig. 8.

The dilatation at the posterior end of the superior longitudinal sinus
marks the beginning of the torcular Herophili (Fig. 8), and from it the
sinus rectus extends towards the choroid plexus, where it ends in the
great cerebral vein. Between the straight sinus and the superior sagittal
sinus a small vein enters the falx and ends at once in a capillary plexus.
This vein no doubt marks the beginning of the inferior sagittal sinus.
Behind the transverse sinus (Fig. 8) there is a second venous anastomosis
extending from the region of the mid-brain to the internal jugular vein,
and no doubt marks the occipital sinus, which in the adult is as an anas-
tomosing channel between the upper and lower ends of the transverse
sinus. Extending forwards from the juncture of the transverse sinus
with the internal jugular vein, a venous sprout is shown in Fig. 8, which
passes on the outside of the skull towards the seventh nerve, and marks
the remnant of the vena capatis lateralis.

In the youngest human embryos the anterior cardinal veins run on
the medial side of all of the cranial nerves before the vena capatis
lateralis is formed. In case the whole of the anterior cardinal vein
remained permanently, that portion between the cavernous sinus and the
internal jugular vein would become the inferior petrosal sinus, for
they both hold the same position. But it appears that the inferior pet-
rosal sinus is of new formation, for in none of the intermediate stages
can a trace of it be found.

A résumé of the development of the sinuses of the brain from the
anterior cardinal vein is given in Figs. 14 to 17. They explain them-
selves.

EXPLANATION OF PLATES.
PLATE I.

Fie. 1. Surface of the brain with the arteries injected in an embryo 48
mm. long (No. 237). Enlarged 5 times; injected by Mr. Brédel. The dorsal
end of the cerebral vesicle has been drawn forward to show better the vessels
of the mid-brain.

Fie. 2. Surface view of the brain of an embryo 80 mm. long (No. 234a).
Slightly enlarged from a photograph by Dr. Mellus. Over the region of the

2

18 Development of Blood-Vessels of Human Brain

island the extravasation is extensive, while over the rest of the brain it is
in spots along the arterioles, as they penetrate the brain.

Fic. 3. Embryo of the fourth week (No. 2). Enlarged 16 times. The
external form is from nature. The structures within the head have been
reconstructed; the tenth and twelfth cranial nerves and the first cervical
nerve by Dr. Streeter.

Fic. 4. ~Brain and its arteries of embryo No. 163. Enlarged 15 times. The
picture of the brain is from a wax-plate model by Dr. Lewis. The arteries
are from a graphic reconstruction. The basilar artery extends throughout
the length of the hind-brain and the circle of Willis throughout that of the
fore-brain. The position of the cranial nerves and otic vesicle is given in
Fig. 9.

PuaTE II.

Fic. 5. Graphic reconstruction of the head of embryo No. 75, showing the
arteries and the brain. Enlarged 7 times.

Fic. 6. Graphic reconstruction of the brain and arteries of embryo No.
145. Enlarged 7 times.

Fic. 7. Same as Fig. 6. The right cerebral hemisphere has been removed,
showing the anterior cerebral artery throughout its extent.

Fic. 8. Same as Fig. 7 with the choroid plexus and large veins added.

Fic. 9. Embryo No. 163. Enlarged 13 times. The surface view is from
an excellent photograph, and the structures in the head are from a graphic
reconstruction.

PLATE III.

Fic. 10. Graphic reconstruction of the veins and brain of embryo No. 74.
Enlarged about 10 times.

Fie. 11. Graphic reconstruction of veins of the head and brain of embryo
No. 144. Enlarged about 7 times.

Fic. 12. Section through the head of embryo No. 109. Enlarged 12%
times. H, hypophysis; S. Ob, superior oblique muscle; Ez. R, lateral oblique
muscle; Jr, trapezius; VJ, jugular vein. The cranial nerves are numbered
with Roman numerals. On the side lettered the section is nearer the mouth
than on the other side, showing that the vena capatis lateralis which con-
nects with the jugular vein is on the outside of the skull.

Fic. 13. Head of embryo No. 109. Enlarged 1214 times. The form of the
arm and body are from photographs. The face, brain and nerves are from a
wax-plate reconstruction by Dr. Lewis. The veins are from a graphic recon-
struction. The brain and face are somewhat distorted, but are given in this
way to complete Fig. 5, Plate IV, in the publication of Bardeen and Lewis in
vol. I of this journal.

DEVELOPMENT OF THE BLOOD-VESSELS OF THE HUMAN BRAIN
FRANKLIN P. MALL

KEK Ue

FIG. 2.

AMERICAN JOURNAL OF ANATOMY=VOL, IV

PLATE |

FIG. 4, Kline, del.

DEVELOPMENT OF THE BLOOD-VESSELS OF THE HUMAN BRAIN
FRANKLIN P. MALL

gg

FIG. 5.

FIGS if

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE Il

FIG. 9.

Kline, del.

DEVELOPMENT OF THE BLOOD-VESSELS OF THE HUMAN BRAIN
FRANKLIN P. MALL

FIG. 10.

FIG. 11.

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE Ill

FlGaaai 2s

FIG. 13. Kline, del.

THE SIZE OF THE ARTICULAR SURFACES OF THE LONG
BONES AS CHARACTERISTIC OF SEX; AN
ANTHROPOLOGICAL STUDY.

BY
THOMAS DWIGHT, M.D., LL. D.,

Parkman Professor of Anatomy at the Harvard Medical School.

WITH 6 PLATES.

The pelvis has long been recognized as a reliable guide to the sex
of the skeleton and still longer as the greatest peculiarity of the female
figure. From twenty to thirty years ago several papers appeared on
the means of determining the sex of the skull. It is, I think, now
generally admitted that the skull is of value in the hands of an expert;
but the late Professor Brinton very near the end of his life declared
that apart from the pelvis there is no guide to the sex among the bones.

Hyrtl (1) long ago wrote: “TI find the difference between the male and
female sternum so clearly expressed by the proportion of the manu- ,
brium to the body that it is hardly possible to err in determining the sex.
The manubrium of the female sternum exceeds in length that of half
the body; while in the male sternum it is at least twice as long as the
manubrium. I (2) was able to show on sufficiently large series that
while this was true of the average male and female sterna, it was not
true of about 40 per cent of the individual instances, so that it was
very possible indeed to err in determining the sex by that means. Prob-
ably the rule applies to well-formed bodies, but not to a large proportion
of those that we meet with. The femur again is a bone that is to the
expert of much value. A typical male and a typical female femur can
hardly be mistaken; but practically there are a great many thigh bones,
perhaps 75 per cent, on which an expert would be unwilling to give an
opinion by methods hitherto in use. Without going so far as Professor
Brinton, we may say that with our present methods, excepting the
pelvis, and even this is not always conclusive, in the great majority of
cases the expert must form his opinion of the sex of bones from their
general appearance, and that comparatively rarely can he speak (still
excluding the pelvis) with any great certainty.

AMERICAN JOURNAL OF ANATOMY.—VOL. TY.

20 Size of Articular Surfaces of Long Bones

The purpose of this paper is to present a new method, which indeed
T have suggested before, but which I had not established by a sufficient
series of observations; namely, the relatively small size of the articular
surfaces of the long bones in the female. If this be true it certainly
deserves a place among the laws of anthropology. While I believe that
this applies to the long bones in general, I have limited the demon-
stration of the principle to the heads of the humerus and femur.

In the Shattuck lecture on the Range and Significance of Variation
in the Human Skeleton, which I had the honor of giving before the
Massachusetts Medical Society in 1894 (3), I advanced the opinion that
the size of the articular surfaces of the limbs has an important sexual sig-
nificance. J mentioned that I had studied the dimensions and propor-
tions of the glenoid cavity of the scapula on 63 male and 27 female
bones. Its average length in the male bones was 3.92 em. and in those
of the female 3.36 cm. Very few male ones were less than 3.6 cm. and
very few female as long. Though I had made observations on the bones
of the arm and forearm I had no series large enough to quote; but I
spoke more in detail of the observations on 64 femora on which many
measurements had been taken, and which came from bodies that had
been measured before dissection. After discussing some of the more
common features as guides to determine the sex, I said: “Some other
measurements seem to throw more light on this matter. They tend
to establish the theory that the small size of joints is characteristic of
woman. They are the greatest diameter of the head of the femur and
the greatest transverse breadth through the condyles. The average diam-
eter of the male head is 4.8 cm., that of the female 4.15. My tables
show one marked difference between the sexes; namely, that in the
women there is a fairly regular increase in the size of the head cor-
responding with the increase in length of the femur. Among men this
is not so. While it is true that most of the largest heads are found in
the longer half of the bones and most of the smallest in the shorter half,
the correpondence is far less evident. I find, moreover, that but two male
heads have a diameter of less than 4.5 cm., and but two of the female a
greater. Both these female bones were among the longest, but the two
male were but little below the average. Thus it would seem that the
actual measurement of the head of the femur is a pretty good criterion
of the sex. The measurements of the knee are less conclusive. The
average difference is just under one centimeter (8.5 and 7.3), but there
are more that overlap.”

Dr. Hepburn (4) published in 1896 measurements of femora of many
races, one of the measurements being that of the head. He did not,

Thomas Dwight — . 21

however, make any attempt to consider the sexual significance of the head.
In fact, there were but few females among the bodies of Europeans from
which the bones were taken. He mentions, however, incidentally, that
the diameter of the head of the male femur was never below 40 mm.,
except among the Andamans; and also that a diameter below 40 was
found in several female bones of various races.

Dr. Dorsey (5) published in 1897 a paper recording observations on the
long bones of American aborigines in which he tested the accuracy of
my views, taking the greatest diameter of the head of the humerus and
of the femur and also the breadth of the upper end of the tibia. The
sex was first decided from the pelvis. His results from the heads. of
the bones of 135 skeletons of various races of both North and South
America were very strikingly in confirmation of the value of the size of
the joints as a sexual characteristic. ‘“ Thus, if the maximum diameter
of the head of the humerus of any American skeleton measure 44 mm.,
the chances are extremely great that it is a male; if it measure 45 mm.,
it is a male to a practical certainty. The inference to be drawn from the
measurements of the femur seem almost, if not quite, equally valuable;
and it would almost seem that we could determine the sex from the femur
alone with a great deal more certainty than we could from the skull.
After Professor Dwight’s disparaging remarks about his results from
measurements through the condyles of the femur, I was quite unpre-
pared for the results which have been derived from the tibia. The range
of variation is, to be sure, greater than it is for either the humerus
or the femur; and this, it may be repeated, is largely due to the discrep-
ancy in stature between the North and South American skeletons, but
the dividing line for the two sexes, between 71 mm. and 72 mm., is almost
as sharp as it is for the femur, and makes the tibia a valuable aid for the
determination of sex.”

Although I was satisfied that the principle that the small size of joints
is characteristic of woman is correct, I felt that it should be established
by a series of measurements large enough to be beyond question. Accord-
ingly, I undertook to make the measurements of the head of one humerus
and one femur of 100 male and as many female bodies. Those of white
adults only were used. The head of the humerus was measured in both
the vertical and the transverse diameter, the object being to get the great-
est diameter for each, even if it deviated somewhat from the strictly verti-
eal or transverse plane. In the femur the greatest possible diameter was
carefully sought for. The measurements were made with blunt calipers.
The bodies were those used for anatomy and surgery in the Harvard
Medical School. I took the measurements when the cartilage was still

22 Size of Articular Surfaces of Long Bones

fresh. This is certainly proper for the purpose of an anthropological
study as it represents the size of the joints as they are in life. Moreover,
on many dried bones, the cartilage remains as a very thin layer, which,
though amounting to little, causes a discrepancy between those bones
and others in which it has been quite removed. The question of what
deduction from diameters thus obtained should be made in comparing
them with those from dry bones shall be considered later.

When I had obtained these measurements on 100 male and 100 female
bones I tabulated the results and drew the curves. While the results
seemed to establish the law, the curves were so irregular that it seemed
certain that they could hardly show a true mean. I then made 50 more
examinations in each sex, and again was dissatisfied with the curves,
and undertook 50 more. Thus I have now the measurements of 400
bones equally divided between the sexes. Owing to the relative scarcity
of female subjects, I think the additional hundred measurements of
female bones has retarded me by nearly three years. The bones were
those of white adults, and in every case the humerus and femur were
both measured, so that comparisons can be made between the upper and
lower extremity. It was not possible to restrict the measurements to
bones of one side, as post-mortem injury or some pathological blemish
often rendered at least one of the joints unavailable. They seem to
establish the point at issue.

The averages are as follows:

Head of Humerus. Fead of Femur.
Vertical. Transverse.
Miaile's ccc vegeeorchens 48.76 mm. 44.66 mm. 49.68 mm.
Hemale: ss... 42.67 88.98 43.84
Difference..... 6.09 mm. 5.68 mm. 5.84 mm.

The above average measurements of the female are to the respective
male ones as 87.51, 87.28; and 88.24 are to 100. (Plates I, II and III.)

It is easy to see by the curves (Plates I, II and III) that there is only 1
male with a vertical diameter of the head of the humerus below that of
the average female, and only 2 females with the same diameter above
that of the average male. Taking the transverse diameter of the head
of the humerus we find 2 males below the female average, and 3 females
above the male average. With the head of the femur we have but 1 male
below the female average and but 1 female above the male average.

Taking separately the three series, each of 400 measurements, the fol-
lowing deductions may be made:

Thomas Dwight 23

Head of Humerus, vertical diameter.

In the 36 smallest 0 male. In the 51 largest 0 female
ee Q4 ee ie: “ 85 “e 1 a
“e 188 4 4 “c iG lhl “ 9 “a
“ atte ath cc QQ «& a3 135 « 8 73
ag 165 “c 10 a3
Head of Humerus, transverse diameter.
In the 55 smallest 0 male. In the 42 largest 0 female.
“c Q4 “c DIE (a3 69 ve 1 ce
ce = P26 “cc 3 «c cc 107 73 8 «c
«c 155 “ rd a4 (73 139 x3 5 3
“ a la hg a3 9 cc
Head of Femur, greatest diameter.
In the 386 smallest 0 male. In the 51 largest 0 female.
ce $38 “c 1 «“ 73 WOR ““c 1 (7
ce 119 “e 4 “c cc 1338 (a3 38 “ce
“ 154 “ce 6 cc ce 168 ce iltr¢ “cc

Continuing this line of comparison I was anxious to divide the bones
into a smaller and a larger half, and to see how many male bones were
among the 200 smaller and how many female among the 200 larger,
taking each of these diameters successively. Unfortunately the groups
did not allow this division to be made without putting some of a group
of equals into the smaller and some in to the larger half. It seems to me
that this can be done very properly; but that all may judge of this for
themselves I give the process in detail. Thus in the vertical measure-
ments of the head of the humerus there were 204 bones measuring 45 mm.
and less, and 196 measuring more than 45 mm. The number measuring
45 mm. was nearly equal in both sexes, there being 16 male and 17 female.
Thus if 4 of this group, 2 of each sex, were transferred from the smaller
bones to the larger, there would be two divisions of 200 each, obtained by
the transfer of only 1 per cent of the whole. In the smaller 200 there
would be 23 (11.5 per cent) of the male and 177 (88.5 per cent) of the
female. In the second half these figures would be reversed. In the
series of the transverse diameter of the head of the humerus there were
191 bones measuring 41 mm. or less and 209 measuring more than 41
mm. Among the male bones were 14 of 42 mm., and among the female
18; by transferring 9 of these, 4 male and 5 female (2.25 per cent of
the whole), to the smaller bones, we had 200 in each division arranged
as follows: In the smaller 200 there were 22 male (11 per cent), and
178 female (89 per cent). In the Jarger 200 these figures were reversed.
In the case of the head of the femur there were 195 of 46 mm. or
less, and 205 of more than 46 mm. There were 37 measuring 47 mm.,
of which 22 were male and 15 female. By transferring 3 male and 2
female from the larger to the smaller bones, we once more have 200

24 Size of Articular Surfaces of Long Bones

in each division, the transfer being 1.25 per cent. In the smaller 200
there were 30 male (15 per cent), and 170 female (85 per cent). In the
larger 200 these figures were reversed.

Surely these results are very convincing, and the manipulation by which
two even halves are obtained quite justifiable.

Let us now inspect the curves. It stands to reason that there must
be some overlapping. It is self-evident that the joints of all males
cannot be larger than those of all females, even of the same race. I have
already given the figures which show how surprisingly few of either sex
pass beyond the average of the other sex; that is, how few male bones
are below the female average and how few female bones above the male
average. Now these curves show that if we suppress a small percentage
composed of erratic individuals, the overlapping is remarkably small and
restricted to a very narrow debatable ground. The curve of the vertical
diameter of the humerus (Plate I) shows that the smallest male measure-
ment is 41 mm. and the largest female 50 mm. Thus there is an over-
lapping extending through half the breadth of the two curves. There
are 313 individual measurements overlapping (78.25 per cent). But
the chart shows clearly that this wide spread of overlapping is due to a
few aberrant specimens. If we take away only 9 male and 10 female
(4.75 per cent), the number of overlapping bones is reduced to 64, or
16.80 per cent of the remaining 381. What is most remarkable is that
after this elimination of extreme formations, the overlapping is limited
to diameters of 45 and 46 mm.”

The curve of the transverse diameter of the head of the humerus
(Plate II) is very similar; 303 overlap (75.75 per cent), but if 7 male
and 9 female (4 per cent) are thrown out only 68 bones (17.71 per cent)
of the remainder overlap; and the overlapping is limited to bones meas-
uring 41 and 42 mm.

The curve of the head of the femur (Plate III) is interesting inas-
much as there are fewer aberrant bones to remove and yet greater ulti-
mate overlapping. Originally 313 bones (78.25 per cent), (precisely the
same as in the vertical diameter of the humerus) overlap, but of these
only 6 male and 3 female (2.25 per cent) are sufficiently isolated to
justify their removal, after which 115 (28.90 per cent) of the remainder
still overlap. Moreover, the overlapping includes three millimeters,
namely, 46, 47 and 48 mm., instead of only two, as in both diameters
of the humerus.

1That part of the curves represented by a continuous line shows them as
they would be after this elimination.

Thomas Dwight 25

Even the last is far from a bad result and shows that the size of
the head of the femur has a great sexual significance, but distinctly less
than that of the head of the humerus. The averages show the same
thing, though less strikingly.

The main thesis seems thus to be established.

Let us now consider whether any particular shape of the articular
head is more characteristic of either sex than another. As we have
begun by assuming that the head of the femur is spherical, there can be
no question about that; but as the head of the humerus has a long and
a short diameter the question is possible. From some old observations on
the glenoid cavity I had come to the conclusion that the head of the
female bone is narrower than that of the male, and this is supported by
the averages; but to such a minute degree as to be unworthy of con-
sideration. As already stated, the female head of the humerus meas-
ures 87.51 per cent of the male in the vertical direction, and 87.28 in
the transverse.. The average difference in the former direction is 6.09
mm., and in the latter 5.65 mm. The transverse diameter is 91.59 per
cent of the vertical in the male, and 91.35 per cent in the female. I then
went to work on the individual differences between the vertical and the
transverse diameters. The range of differences extends from —1 to 8
mm. ‘The former means that in one single female bone the transverse
diameter was 1 mm. greater than the vertical. No difference whatever
was found in 3 males and 2 females. The difference was 1 mm. in 10
males and 13 females; the greatest difference, 8 mm., was found in 2
of each sex. Differences of 6 and 7 mm. were found much more fre-
quently among the males than the females, as is to be expected from
the greater size of the former. Running through a number of cases in
both sexes in which the difference was slight and of others in which
it was large, I was quite unable to see anything that pointed to a sexual
difference in this respect sufficiently marked to be worth recording.
I was unequally unsuccessful in trying to ascertain whether, regardless
of sex, a long or a round head of the humerus was to be expected more
frequently in large or small bones.

I endeavored to ascertain whether the difference between the head of
the humerus and that of the femur was greater in one sex (probably
the female) than the other, but I failed again. The average differ-
ence of the vertical diameters showed a difference to the advantage of
the femur of .92 mm. for the male, and 1.17 mm. for the female.
On the other hand, the transverse diameter gave the femur an advan-
tage of 5.02 mm. in the male, and 4.86 in the female. Study of indi-

26 Size of Articular Surfaces of Long Bones

vidual cases of large and small bones did not give any encouragement to
undertake the labor of elaborate tabulation.

We come now to the very important feature of this series of observa-
tions that the measurements were made on bones with the articular
cartilage not only in place, but not dried. As above mentioned, this is
the proper method, as it shows the parts as they are in life, giving the
true size of the joints; but there is the serious consideration that most
observations are made on dried bones, so that one needs to know what
allowance is to be made for the absence of the cartilage. What compli-
cates the matter is that the conditions for the humerus and for the femur
are not the same. In measuring the practically globular head of the
femur, the greatest diameter passes through the centre of the sphere and
traverses the whole thickness of the cartilage on both sides of the head.
With the humerus the conditions are very different. Both the long and
the short diameter, especially the former, run through the bone just at
the insertion of the shaft into the head, that is, at the border of the
articular cartilage which narrows around the margin of the head so as
to be extremely thin. I have been at great trouble to find a method
of determining how much to allow for the cartilage and can find none
that is satisfactory. I think that from 2 mm. to 3 mm. should be allowed
for the femur, and that .5 or 1 mm. is enough for the humerus.

When the work was far advanced I regretted that I had contented
myself with measuring only the heads instead of taking the length and
perhaps the thickness of the bones. Although as a practical anatomist
I know that no one would think of determining the sex of either of these
bones by its length, I felt that it would be difficult to answer anyone
who might ask how I could be certain that there is a greater discrepancy
between the articular heads of the bones than between their lengths. I
had recourse to Dr. Hrdlicka of the National Museum of Washington,
who came to my rescue with measurements of bones of 200 white adults,
100 of each sex, made by him at the Medical Department of Colum-
bia University of New York. The sources from which these bones came
are perhaps a little more diverse than those that lead to Boston, but not,
I believe, very much so. I have already shown that the dissecting room
material in Boston does not in the least represent any single race.” That
from New York is only somewhat more heterogenous. In short this
collection of measurements of the length is, failing that of the bones
on which the joints were measured, as good as could be expected. It is
quite good enough for the very general conclusions I shall draw. I wish
to express my deep obligations to Dr. Hrdlicka for his kind generosity
in this matter, by which I am enabled to compare with this series of meas-

Thomas Dwight ras

urements of 400 articular heads of humerus and as many of that of the
femur a series of the length of 200 humeri and 200 femora, equally
divided between the sexes. The length of the thigh bone was taken by
the bicondylar method. The greatest possible length of the arm bones
was recorded. All the measurements were of bones of the right side.
It is very evident that the difference between the bones of the arm and
thigh in the matter of length are much less important sexually than those
of the diameters of the heads. In comparing the curves (Plates IV and
V) it is to be remembered that the observation of the lengths are but
half as numerous as those of the joints.

The average length of the male humerus is 32.46 cm., that of the
female 29.98 cm., which is 92.36 per cent of that of the male. The
average male femur measures 44.95 cm., the female 41.55 cm., or 92.44
per cent of the former. The diameter of the head of the female femur
approaches that of the male rather more closely than either of the
diameters of the head of the humerus, yet its percentage is only 88.24.
Before plotting the curves, the millimeters were suppressed, each bone
being recorded as at the nearest centimeter. Cases which came at pre-
cisely half a centimeter were put at the lower mark.

I pointed out in the case of the joints how very few male ones were
below the female average, and how very few female ones above the male
average. It is easy to see by consulting the curves of the lengths, in
which the averages have been marked, that though there are only half
as many observations there are, especially in the case of the female
femora, decidedly more beyond the line. The contrasts between the two
sets of series when divided into a smaller and a larger half are very
instructive. Let us take first the length of the humerus. If the dividing
line be put between 31 and 32 cm., we find 119 in the first division
and 81 in the second. In the former there are 22 male and 23 female
bones measuring 31 cm. If we transfer 9 males and 10 females (9.5
per cent of the whole) we have 100 in each division. There are 24 males
in the shorter 100 and 24 females in the longer 100, 1. e., 24 per cent of
each sex in the wrong half. If we draw the dividing Jine for the femora
between 43 and 44 cm., we have 118 in the shorter division and 82 in the
longer. There are 16 male and 16 female bones measuring 43 cm. If
we transfer 9 of the former and 9 of the latter (9 per cent) to the longer
division we again have two divisions of 100. Among the shorter 100 are
2” (27 per cent) male, and among the longer 27 female (27 per cent).
A glance at the former statement of this manipulation with the diameters
of the joints will show that the percentage of specimens transferred was
insignificant, never over 2.25 per cent, and the result much better.

28 Size of Articular Surfaces of Long Bones

Let us now compare the curves. At first sight they give no hint of
the difference which analysis reveals. The curves of the joints show an
original overlapping in the order in which they are given of 78.25, 75.75
and 78.25 per cent repectively. The curve of the length of the humerus
shows 89 per cent overlapping, and that of the femur 82 per cent, but
in the case of the joints the amount of overlapping was wonderfully
reduced by the elimination of a few stragglers, respectively 4.75, 4. and
2.25 per cent, the percentages of overlapping dropping to 16.80, 17.71
and 28.90 of the remainder. In the tables of the length a much larger
elimination brings about much less satisfactory results. Thus with the
humerus the elimination of 15 bones (7.5 per cent), reduced the over-
lapping of the remainder only to 41.62 per cent; and with the femur
the rejection of 15 bones (7.5 per cent) still leaves the overlapping of |
the remainder at 46.49 per cent. This is a very significant difference.

Thus it is demonstrated that the difference in the size of the articular
surfaces in the sexes is very much more marked than that of the length
of the respective bones. Although I have not established it by figures,
J have no hesitation in saying that it is also much more marked than the
difference in the thickness of bones. A striking illustration of this is
furnished by the photograph (Plate VI) of a male and a female humerus,
side by side, so placed as to show the articular heads as well as possible.
There is very little difference in the length and in the thickness, but the
much greater size of the joint shows at a glance which is the male.
What is most interesting is that the male bone came from the body of a
very puny young man of nineteen, who being blind, had passed his life
in an almshouse doing very little work. I remember him particularly
from the fact that he had but one kidney. My personal recollection of
the body of the woman from which the female bone was taken is less
sharp; but she is said to have been of uncommonly powerful make. The
muscular ridges on the bone confirm this, yet a glance at the head is so
conclusive that it is needless to mark which is which.

I have devoted a good deal of time to the glenoid cavity of the scapula,
and more or less to other joints in the extremities. Although I cannot
speak by the book, I feel very sure that the law which is deduced from
the humerus and femur will be found to apply, though perhaps with
more exceptions, to all the joints of the extremities. It should not need
to be said that it is the province of such a law to be a guide to the expert
who will apply with discretion. Absolute certainty as to the sex of bones
does not exist in all cases. The judgment matured by long observation
is certainly better than any rigid adherence to a mathematical law. None
the less, in many cases, the law of the relative smallness of the female

Thomas Dwight 29

joints will go far to show the probability of what one may hesitate to
affirm as absolutely certain.

The following conclusions seem justified by this study of 400 humeri
and femora of white adults:

The heads of the humerus and femur are relatively small in woman.
Probably the same may be said of other joints.

Dorsey’s investigations show that this anthropological law applies also
to savage races (or at least to some of them).

The number of measurements of male joints smaller than the average
female joint and of female ones larger than the average male is insigni-
ficant. In the transverse diameter of the head of the humerus the com-
bined number is only 1.25 per cent, and in the head of the femur only
.05 per cent.

By rejecting a few aberrant specimens the overlapping in the curves
of both diameters of the humerus is reduced to about 17 per cent, and
is limited to joints measuring 45 and 46 mm, vertically and to those
measuring 41 and 42 mm. transversely.

The head of the femur is somewhat less characteristic, but still very
valuable as a guide to the sex.

(These measurements were made with the articular cartilage in place
and still fresh.)

BIBLIOGRAPHY.
. Hyrrt.—Topographische Anatomie.
. DwieHt.—Journal of Anat. and Phys., Vol. XV and XXIV.
. DwieHtT.—Publications of the Mass. Med. Society.
HeEpsBurRN.—Journ. of Anat. and Phys., Vol. XXXI, p. 116.

. Dorsrty.—Boston Medical and Surgical Journal, July 22, 1897.
. DwicHt.—Anatomischer Anzeiger, Bd. X, s. 209.

anrwnr

DESCRIPTION OF PLATES I TO VI.

PLate I.—The binomial curves of the vertical diameter of the heads of 200
male and 200 female humeri expressed in millimeters. The female curve is
on the left.

PLATE II.—Ditto of the transverse diameter of the same.

PLate III.—Ditto of the greatest diameter of the head of the femur.

Pirate I1V.—The binomial curves of the length of 100 male and 100 female
humeri (Dr. Hrdlicka’s) expressed in centimeters. The female curve is on
the left.

PLATE V.—Ditto of the length of the femur (Bicondylar) (Dr. Hrdlicka’s).

The part of the curves drawn with a continuous line represents what they
would be after the elimination of certain aberrant overlapping bones.

PLATE VI.—The humerus of a strong woman and of a puny man. The sex
is evident from the size of the heads of the bones.

30 Size of Articular Surfaces of Long Bones

APPENDIX.

It is hoped that these figures may be of use. The humerus and femur in the same line always
came from the same body, but no regard has been paid to the side. The measurements include
the articular cartilage which was in good condition.

DIAMETERS OF HEAD OF HUMERUS AND FEMUR.

MALE.
HUMERUS. FEMUR. HUMERUS. FEMUR. HUMERUS. FEMUR.
Vert. Trans. Vert. Trans. Vert. Trans.

1h sR? A7 49 68 46 45 50 | 135 46 AL aul

2 jay 18 48 | 69 46 40 46 136 49 46 sal

os AS 39 46 70 50 44. 50 137 49 44 16

4 45 110. AT 71 48 43 48 138 49 49 56

a 45 44 AZT 72 54 52 52 139 52 50 51

6 46 43 416 73 52 A7 52 140 AZT 44 AG

ri tee 47 54 74 51 45 54 141 52 46 50

8 45 AO AG 7) 5d 48 53 142 50 46 spl

9 50 46 49 76 AZT 44 AT 143 49 43 49
10 46 43 48 77 46 43 49 | 144 48 AZT 5L
11 «648 46 48 78 45 43 46 145 49 43 49
12 Zi 45 48 79 53 48 ay 146 0) 43 50
13 50 45 A9 sO 45 45 49 147 51 7 51
14 49 44. 50 S81 46 44. 54 148 ail! 46 52
15 A7 44 Sil) 82 A9 45 5A 149 ie 45 48
16 45 42 48 83 AZT 45 46 150 a2 49 52
thes ast) 43 46 S4 46 42 46 151 48 A2 48
18 44 43 49 85 50 44 52 152 53 46 53
19 Al 38 42 S6 45 41 46 1538 90 45 SY)
20 48 44 AT S87 51 46 52 154 51 45 52
21 As 44 52 sS 46 A3 46 155 49 44 do
22 AS 42 45 s9 51 7 48 156 S51 7 52
23 A48 46 49 90 51 46 53 157 7 A3 50
24 46 42 A7 91 52 46 52 158 7 44 50
25 43 4% 44 | 92 46 43 50 159 53 48 spl
26 46 43 48 93 53 44 52 160 49 44 50
27 45 43 56 94 45 43 AT 161 45 41 49
28 48 43 A7 95 50 43 46 162 46 43 46
29 3 Al AT 96 48 44 50 163 55 ti 54
30 52 46 54. 97 46 42 46 164 46 AL 49
31 50 46 51 98 46 40 51 165 44 42 16
32 44 42 46 99 54 49 53 166 48 46 5il!
33 51 44. 50 100 ai 44 AT 167 53 51 55
34 46 41 48 101 45 44 AT 168 50 AT 52
35 7 3 50 102 53 46 51 169 sis 43 48
36 52 7 51. 103 46 41 46 170 51 48 50
37 44 41 45 104 52 7 52 171 of 44 49
38 52 44 49 105 50 43 Sill 172 50 45 50
39 49 42 46 106 45 44. 4S 173 50 AT 51
40 7 43 AZT 107 52 50 pill 174 49 43 49
41 54 A9 54 108 48 15 AZT 175 a 45 48
42 49 45 49 109 48 45 48 176 7 43 48
43 52 48 50 110 51 48 54. 177 50 45 53
44 50 45 sul 111 52 49 54. 178 A7 45 50
45 50 50 51 112 48 43 AT 179 ff 43 48
46 51 AG 51 113 50 3 5) 180 03 AZT oo
47 45 41 AT 114 50 45 49 181 50 44 Sill
48 51 48 52 115 A9 AD iW 182 52 4G 51
49 -48 415 52 116 45 38 49 1S: 19 43 sal
50 «48 As? tai 117 48 45 5O 184 50 46 93
o1 49 48 52 118 50 46 51 185 AT AL 48
52 49 16 47 119 49 45 19 186 50 13 51
a3 ik 45 AT 120 48 45 50 187 49 45 50
04 50 44. 49 121 AQ 44 50 188 A4 42 46
dD 650 16 49 122 50 49 5 189 A6 13 46
56 48 13 51 123 al 49 50 190 AZ 44 48
a7 A9 44. 53 124 7 15 AZT 191 53 46 go 2
dS 50 45 A¢ 125 49 45 A9 192 49 46 56
599 48 42 AZT 126 53 49 53 193 ail 45 50
60 51 18 51 127 5 14 52 194 49 46 51
61 53 47 53 128 56 50 36 195 53 re 51
62 49 16 53 129 51 AZT 7 196 49 45 19
63 48 44 48 1390 50 7 52 197 48 AZ 52
64 45 40 44 131 46 11 14 198 50 45 49
65 48 15 51 132 50 46 52 199 50 45 53
66 48 43 53 1338 50 45 50 200 50 45 50
67 if 44 48 1384 47 3 AZT

SDS im Oo 0

HUMERUS.
Vert. Trans.
42 39
43 39
42 37
43 38
42 39
46 42
43 38
44 42
44 40
41 38
45 40
46 45
40 37
7 41
41 36
Al 37
42 39
38 37
43 41
43 AL
42 37
41 39
44 39
4O 37
43 39
38 37
43 38
42 38
39 37
44 39
42 37
50 4G
40 37
46 AL
43 39
44 41
46 42
44 41
42 38
42 4O
42 38
39 36
42 37
39 37
45 Al
49 42
38 36
43 39
42 38
42 38
45 44
43 AL
43 40
38 36
42 39
AS 40
46 Al
42 38
AO 36
41 38
41 38
43 42
39 36
7 Al
42 38
44 40
39 3

DIAMETERS OF HEAD OF HUMERUS AND FEMUR.

Thomas Dwight

FEMALE.

~)

mhek

fon)

led deeded

~~

ab

jesk jemk jm kjk kph fh ff fk fk fk fk

fo bk
we
—)

«
v

RWW SODA Oe we

WHWWWWNW NW We

WHww Ht

HUMERUS.
Vert. Trans.

45 40
43 40
38 36
40 36
40 38
44 40
39 36
43 38
45 42
38 35
42 7
45 39
42 38
44 39
39 36
42 38
43 38
43 42
44. 36
Al 37
44. A3
40 3
41 35
42 41
43 40
40 35
AL 39
44 40
40 41
41 38
38 339/
40 38
et! 37
44. 37
44 38
42 39
AL 37

7 45
43 39
43 41
43 AO
44 AO
AO 37
43 AL
42 39
AL 36

3 39
44. 41
Al 36
4O 38

3 38
39 36
43 37
40 36
42 36
42 36
41 38
44 42
42 39
48 42
41 39
44 39
42 39
44. 41
AL 37
40 38
46 40

FEMUR. |

HUMERUS.
Vert. Trans.
43 39
44. 39
AT 43
44 39
AT A3
40 38
44 39
42 40
44. 41
43 40
45 37
41 39
44 38
40 36
AT 42
41 37
A3 38
42 36
AL 37
44. 39
39 35
44 4]
43 By
46 42
44 40
45 42
43 38

3 38
44 39
41 37
40 86
46 42
43 37
45 40
45 42
45 42
46 41
45 43
43 38
39 38
42 38
45 40
44. 41
40 36
43 Al
AL 38
43 38
45 40
42 38
45 Al
47 42
46 42
A2 AO
42 40
44. 39
44. 38
42 Al
46 44
A4. 42
44. AO
45 40
42 37
Al 40
43 39
45 41
42 Al

ol

FEMUR.

oe fe eho |

To
cs

_ a
7 7 =——
cand
~*

=
=

PLATE |

SIZE OF ARTICULAR SURFACES OF LONG BONES

T. DWIGHT

aTVW

SNYSWNH AHL JO QV3SH AHL JO YSLAWVIO TVOILYSA

Hel

a1IVAGs

IV

AMERICAN JOURNAL OF ANATOMY--VOL.

SIZE OF ARTICULAR SURFACES OF LONG BONES PLATE Il
Wn DWIGHT —

|
MALE

TRANSVERSE DIAMETER OF THE HEAD OF THE HUMERUS

pee
FEMALE

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE IIP

T. DWIGHT

DIVIN

YNWad AHL SO GV3SH SHL SO Y3LAWVIG

TIVNGA

11

11

IV

SIZE OF ARTICULAR SURFACES OF LONG BONES

AMERICAN JOURNAL OF ANATOMY--VOL.

paAy unts os

we te!
ia, ; ——.
: iN " " ay j a
*) : e paul i

ee

'

PLATE IV

T. DWIGHT

SIZE OF ARTICULAR SURFACES OF LONG BONES

SNYSAWNH SHL SO HLSN3T
cuaeyt QTIVWOA

| TT Tol
HODES

K2nm

FH

a
=
a
a

He
Hi

Haren
H

co
Ee

scstt

IV

AMERICAN JOURNAL OF ANATOMY--VOL.

PLATE V

SIZE OF ARTICULAR SURFACES OF LONG BONES

T. DWIGHT

4

aHIVN

YNW34 SHL AO HLON]AI

aATVNAA

v

AMERICAN JOURNAL OF ANATOMY--VOL.

SIZE OF ARTICULAR SURFACES OF LONG BONES PEATE Vi
T. DWIGHT

FEMALE MALE

AMERICAN JOURNAL OF ANATOMY-=-VOL. IV

THE PHYLOGENY OF THE CRURAL FLEXORS.

BY
J. PLAYFAIR McMURRICH.
From the Anatomical Laboratory of the University of Michigan.

WitTH 14 Text FIGURES.

In an earlier paper (1903) I presented the results of a comparative
study of the flexor muscles of the antibrachial region and showed that it
was possible to trace step by step the changes by which the arrangement
occurring in the Urodelous amphibia was converted into that charac-
teristic of the mammalia. In the amphibia the muscles in question
possess a definite arrangement in layers and it was shown that these
layers have a fundamental significance, since, notwithstanding the almost
innumerable modifications and differentiations which they present in
higher forms and the apparently enormous differences which exist be-
tween the amphibian and mammalian forearm musculatures, yet the
layers could be recognized throughout and consequently afforded a basis
for the reconstruction of the phylogeny of the mammalian muscles.

It became of interest, consequently, to ascertain whether a compara-
tive study of the crural flexors would reveal a similar fundamental
arrangement in layers and so afford a basis for their phylogenetic re-
construction, and, if so, an opportunity for a satisfactory consideration
of the much-discussed question of the serial homology of the arm and
leg musculature. In the following pages the results of such a study
are recorded in so far as they bear upon the first of the two problems men-
tioned, namely, the phylogeny of the crural flexors. The question of
the serial homology of the arm and leg musculature I hope to discuss
later in connection with some other general questions relating to the
morphology of the vertebrate limb.

The methods and forms employed in the present study were essen-
tially the same as those made use of in the investigation of the arm
muscles. The arrangement and relations of the muscles were studied
in serial transverse sections and the forms employed were Amblystoma
tigrinum as a representative of the Urodele amphibia, Scincus sp? as

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

34 "The Phylogeny of the Crural Flexors

a representative of the reptilia and the opossum (Didelphys virginiana) ,’
the mouse, the cat and man as representatives of the mammalia.

I. Tor CruraL FLEXxorS OF THE UropELOUS AMPHIBIA.

A transverse section through the upper part of the crus of Ambly-
stoma shows the arrangement of parts represented in (Fig. 1). Super-
ficially upon the posterior surface of the section is seen a strong, some-
what crescentic, muscular mass, which, employing a terminology con-
sistent with that used in the description of the antibrachial muscles,
may be termed the plantaris swuper-
jicialis medialis (Psm). It may be
remarked in passing that while the
terms used by KEisler, 95, in his
careful and suggestive paper on the
homology of the extremities are also
employed here, their application is
very different, since Hisler has failed
to recognize the muscle now referred
to as a distinct constituent of the
crural flexor mass, that muscle which
he terms the plantaris superficialis
lying beneath the muscle now under
consideration and forming what I
shall describe as the plantaris pro-

: fundus III.
Fig. 1.—Transverse section through the

upper part of the crus of Amblystoma i+] ] He]
tigrinum. F, fibula; f, ramus superficialis Tn addition to this most super ficial
fibularis; FY, fibulo-tarsalis; J, interos-
seus; m, yams SU ee eae YP,
ramus profundus; Pp /-///, plantares pro- :

fundi I-III; PpL/m, plantaris profundus of the leg another muscle (Ps!)

III minor; Psl, plantaris superficialis la- :
teralis; Psm, plantaris superficialis medi- which must also be referred to the
alis ; T, tibia. . Z :

ie Bias superficial layer and which may be
termed the plantaris superficialis lateralis.

Beneath the superficial layer formed by these two muscles lies a larger

muscle, there is upon the outer side

1As on a former occasion, I am indebted to my friend, Dr. C. F. W. McClure,
for the opossum material which was used, and I desire to express my apprecia-
tion of his courtesy in placing at my disposal material without which my
studies would necessarily have been very incomplete.

To my assistant, Mr. F. S. Bachelder, I am greatly indebted for assistance
in the work, since he kindly undertook the preparation of all the serial sec-
tions that were required. He also studied with me the arrangement of the
muscles and nerves of the amphibian crus and so much of this paper as
treats of these structures is to be regarded as our joint work.

J. Playfair MceMurrich 35

muscle mass whose fibres have a markedly oblique direction and which
will be spoken of as the plantaris profundus III (Pp). Beneath the
fibular border of this there is a small oval muscle whose fibers are cut
transversely and which is the fibulo-tarsalis (FT), while opposite its
tibial border there is a very slender muscle which may be termed the
plantaris profundus III minor (Pp™!m). Still more deeply seated
are two layers of muscle whose fibers are directed obliquely downward
and tibia-wards, the plantares profundi II and I (Pp" and Pp"), and
finally, extending almost directly across between the fibula and tibia is
a muscular sheet which may be termed the m. interosseus (1).

A comparison of such a section with one taken through the forearm
of Amblystoma will show that a remarkable similarity exists between
the two. There is exactly the same number of layers and the same
direction of the fibers in the different layers. Indeed, the resemblance
is so close that the two sections might easily be confused in a casual
examination. A discussion of their resemblances and the significance
of these will, however, be postponed until a later occasion, and I shall
pass on now to a consideration of the various muscles mentioned above.

And first as to the plantaris superficialis medialis (Psm). As already
noted, this muscle is not regarded by Hisler as a part of the crural mus-
culature and in this view he is in agreement with his predecessors. The
muscle is continuous above with the lower end of the muscle named
ischio-flexorius by Hoffmann, 73, the caudo-pedal by Humphry, 72, and
the external flexor of the crus by Perrin, 93, and, indeed, is the terminal
part of that muscle. From dissections there is little reason to regard
it otherwise than as these authors have done, but its relations as seen in
sections, especially when these are compared with sections through the
forearm, speak so strongly for the view here set forth that I have no
hesitation in advancing it. And especially so since there are two other
facts bearing favorably upon it, namely (1) the insertion of the muscle
into the plantar aponeurosis, which occurs a short distance below the
knee joint, and (2) the fact that in Amblystoma the ischio-flexorius
is crossed at the knee joint by a well-marked tendinous inscription, which
marks, I believe, the line of junction of the ischio-flexorius proper with
the plantaris superficialis medialis. The ischio-flexorius of Hoffmann
is, according to this view, a compound muscle formed by the end to end
union of a true ischio-flexorius with a plantaris superficialis medialis.’

21If I understand aright Humphry’s, 72, description of the caudo-pedal
muscle of Cryptobranchus, there is in this form also a tendinous inscription
in the muscle in the neighborhood of the knee joint. If this be so, it is

36 The Phylogeny of the Crural Flexors

The plantaris superficialis lateralis (Psl) is a muscle which has in-
variably been described as the femoral head of the large superficial
muscle which I term the plantaris profundus III. From the remark-
able similarity of the muscles of the forearm and crus in the urodele
amphibia it might be expected that the same constancy in the relations
of the superficial and deep flexors of the leg to the femur and crural bones
as obtained in the corresponding muscles of the crus in their relations
to the humerus and antibrachial bones would be found, and it was on
this ground that I was first led to refer this muscle to the superficial
layer and to regard it as distinct from the plantaris profundus III.
Further study only served to confirm the correctness of this view by
revealing a consistent phylogeny for the crural flexors based upon it.

The muscle takes its origin from the flexor surface of the external
condyle of the femur and is separated by a distinct interval from the
upper part of the origin of the plantaris profundus III. It passes down
the fibular side of the leg, quite distinct from the plantaris profundus
and passes over into a rather feeble tendon, which is inserted into the
outer border of the fibula near its lower extremity. Throughout its
whole extent, therefore, it is distinct from the plantaris profundus III
in Amblystoma.

The plantaris profundus III (Pp) is the largest of all the muscles
of the crus and is described by Humphry, 72, as the flexor sublimis
digitorum, by Hoffmann, 73, as the femoro-fibule-digiti I-V, by Perrin,
93, as the external flexor of the digits and by Hisler, 95, as the plantaris
superficialis major, all these authors, as has already been noted, including
in the muscle the plantaris superficialis lateralis. It arises in Ambly-
stoma from the posterior surface of the upper part of the fibula and its
fibers are directed downwards and somewhat tibially to be inserted into
the under (dorsal) surface of the plantar aponeurosis, through which it
acts upon the digits.

From the upper part of the tibial border of this muscle a slender slip
(Pp''m) separates and passes almost vertically downwards to fade out
in connective tissue in the neighborhood of the ankle joint, in close
proximity to the tibial border of the plantaris profundus I. This is
evidently the muscle described by Hisler as the plantaris superficialis
minor and has apparently been overlooked by Perrin. It seems to be,

interesting to note that in the caudo-pedal of Cryptobranchus there is not
only an end to end union of the ischio-flexorius proper and the plantaris
superficialis medialis, but also of the former and what may be termed an
ischio-caudalis.

J. Playfair MeMurrich 37

in Amblystoma, a derivative of the plantaris profundus III and may be
termed the plantaris profundus III minor.

The fibulo-tarsalis (/'7') arises from the posterior surface of the upper
part of the fibula and extends vertically down the crus, lying immedi-
ately behind the fibula, to be inserted into a strong tendinous band which
extends transversely across the sole of the foot at about the level of the
distal row of tarsal bones. This muscle is the fibulo-plantaris of Hisler,
the deep common flexor of the phalanges of Perrin, the fibule-metatarsi
et digiti I-V of Hoffmann and the flexor profundus digitorum of
Humphry.

The plantaris profundus II (Pp!) also arises from the posterior
surface of the fibula and takes an oblique direction downwards and
tibially to be inserted into the deeper (dorsal) surface of the plantar
aponeurosis. It is the plantaris profundus I of Hisler, the internal
flexor of the digits of Perrin and the femoro-fibule metatarsi I-III of
Hoffmann.

The plantaris profundus I (Pp") arises from almost the whole length
of the fibula and from the tarsus and is directed downward and tibially
to be inserted into the lower end of the tibia, into the tibiale and the
tarsale I; I did not find any insertion into the plantar aponeurosis in
Amblystoma. This is the muscle described by Perrin as the direct
rotator of the foot, and is apparently represented in Menopoma, accord-
ing to Hisler, by four more or less distinct muscles which have been
named the plantares profundi II and III, the fibulo-tibialis and the
fibulo-tarsalis. Humphry and Hoffmann have not recognized it as dis-
tinct from the plantaris profundus II.

Finally, the interosseus (I) is a strong band of muscle fibers which
extend almost directly across between the tibia and fibula, occupying
the position of the interosseous membrane of the higher mammalia. It
is the pronator tibizee of Humphry and the fibule-tibialis of Hoffmann.

In the study of the arm flexors much light was thrown upon their
phylogenetic modifications by their nerve supply and the same holds good
for the crural flexors. It must be remembered, however, that with the
modifications which the muscles undergo in the various groups, a modi-
fication of the nerve trunks also occurs, and in making use of the nerve
supply for the identification of muscle equivalents in the different groups,
these changes in the paths followed by the nerve fibers must be taken into
consideration. The final test in the identification of a motor nerve is its
end organ, the muscle; that is a definite quantity in the problem. But
the path by which a given nerve reaches its end organ is not necessarily
the same in all cases; the nerve, as a rule, will seek the most direct route

38 The Phylogeny of the Crural Flexors

to its destination, but that route need not be exactly the same in all forms.

The tendency, however, is largely towards conservatism, and even when

the bulk of the fibers composing a given nerve trunk, adopt, in the higher

vertebrates, a new path, some will be apt to retain the original course
and so serve as guides for the determination of topographic relationships.

I have elsewhere (1903, pp. 466-7) expressed in general terms the con-

clusions in regard to the value of nerve supply in

determining muscle homologies to which my studies
ae of the muscle and nerves of the forearm have led me.

Pst In Amblystoma, immediately above the knee joint,

two distinct nerve trunks occur upon the posterior

surface of the leg (Fig. 2). They are formed by the
pp division of the sciatic nerve after it has given off the
peroneal nerve and are what Humphry, 72, has

tH 2 :

Ye’. termed the internal and external popliteal nerves.
They do not, however, correspond in composition to
the nerves so named in the maminalia, and for this
reason they will be spoken of here as the ramus plan-

PE taris profundus (rp) and the ramus plantaris super-

Fie. 2.— Diagram 5 Ode . 5 :

of the crural nerves ficialis (rs). They lie, at first, one on either side of

in Amblystoma tigri- ne

num. C, cutaneous the sciatic artery, but as they are traced downwards

branch; # P, external A

plantar; Ex, branches the ramus profundus passes slightly laterally so as to

to extensor surface ; ei 2. 5¢

I, branch to inter- come to lie in front of the ramus superficialis, amd a

osseus; IP, inter-

nal plantar; Ppl, little later the two stems fuse, only to separate again,
branches to plantares

protundi I-III; Psm, some interchange of fibers apparently taking place,
branch to plantaris R é :
superticialis medial- however, during the fusion, and a further interchange
is; rp, ramus pro- , A 5
fundus; rs, ramus js carried out by means of a cross connection between
superficialis. A

the two stems a little lower down.

From the ramus superficialis above the fusion branches are given off
to the plantaris superficialis medialis (Psm) and to the plantaris super-
ficialis lateralis (Psl), and below the fusion to the plantaris profundus
III (Pp) and the plantaris profundus II (Pp). Just as these
nerves are given off the stem is passing over the upper border of the
plantaris profundus III and its course is then downwards and outwards
between the fibulo-tarsalis and the plantaris profundus II (Fig. 1, f),
or to a certain extent through the substance of the fibulo-tarsalis. It
thus reaches the fibular side of the crus and descends towards the foot.
lying between the lateral border of the fibulo-tarsalis and the origin of
the plantaris profundus II. Beyond this it will be unnecessary to follow
it at present.

The ramus profundus (rp) gives off above the fusion a branch to

ts

J. Playfair MeMurrich 39

the plantaris profundus III (Pp!’) and below the fusion a branch
(E£x.) which passes downward and forward through a notch on the crest
of the tibia and is supplied to the muscle which has been termed the
tibialis anticus. The main stem then gives off a branch (J) to the
interosseus, and having in its downward course passed successively over
the upper border of the plantares profundi III-I, it passes over the upper
border of the interosseus and is continued downward on the extensor sur-
face of that muscle (Fig. 1, p). Before reaching the foot it gives off
a branch and then divides into two stems, one of which, together with
the branch, passes to the muscles upon the dorsum of the foot, while the
other passes backwards beneath the lower border of the interosseous
muscle, gives off a branch to the plantaris profundus I and continues
onward to be distributed to the plantar surface of the foot.

In order to understand the significance of this arrangement of the
rerves it will be necessary to compare it with what occurs in the arm.
In this but a single nerve trunk, the brachialis longus inferior, enters the
flexor surface of the antibrachium and it divides into a ramus profundus
and a ramus superficialis. The former has a course almost identical
with that of the ramus profundus of the crus and supplies the pronator
quadratus and the palmaris profundus I, which have the same topo-
graphical relations as the interosseus and plantaris profundus I supplied
by the ramus plantaris profundus. The latter nerve, however, also con-
tains some extensor fibers which are lacking in the deep nerve of the
antibrachium, the separation of the preaxial and postaxial fibers having
taken place higher in the arm than in the leg.

The ramus superficialis of the antibrachium divides into two portions,
a ramus superficialis medialis and a ramus superficialis ulnaris, the
latter of which possesses relations similar to those of the ramus plantaris
superficialis after it has given off its branches to the plantares profundi
III and II. It would seem, therefore, that these branches may well be
regarded as equivalent, in part at all events, with the ramus superficialis
medialis of the arm, while the main stem below their origin may be
considered the equivalent of the ramus superficialis ulnaris and be termed
the ramus superficialis fibularis.

But the ramus superficialis medialis of the arm supplies not only the
palmares profundi III and II, but also the palmaris superficialis. In
the leg the branches which are distributed to the muscles which I have
identified as forming the plantaris superficialis, are given off from the
ramus superficialis above the point of its fusion with the ramus pro-
fundus, so that a difference from the arrangement in the arm exists in

that there is no concrete ramus superficialis medialis, its branches aris-
4

40 The Phylogeny of the Crural Flexors

ing independently at different levels. And, furthermore, a more import-
ant difference exists in the origin of a branch to the plantaris profundus
III from the ramus profundus above its point of fusion with the ramus
superficialis. This might seem to vitiate any direct homology between
the ramus plantaris profundus and the ramus palmaris profundus, but,
on the other hand, it may be a part of the same lack of differentiation of
the plantar nerves which is evidenced in the retention of extensor fibers
in the ramus plantaris profundus. In Cryptobranchus, according to
Humphry, 72, both the ramus profundus and the ramus superficialis
send branches to the plantaris profundus III, and in Menopoma, to judge
from Hisler’s figures, 95, the two stems separate only at the upper border
of the plantaris profundus II, from which it may be presumed that the
branches to the plantaris profundus III are given off from the common —
stem above the bifurcation. Whether the high or the low bifurcation be
the more primitive condition, it is difficult to say, but it is at least plausi-
ble to suppose that the fusion of the two trunks in Amblystoma presents
opportunities for the transference of fibers destined for the plantaris
profundus III (and possibly II) from the ramus profundus to the ramus
superficialis, since, apparently, the fibers which form the lower cross
connection between the two stems are destined for the supply of the
plantaris profundus IT.

However that may be, it seems clear that in the plantar nerves there
is less definiteness in the differentiation of the nerve fibers into special
trunks than occurs in the palmar nerves, a fact which is shown by the
inclusion of preaxial fibers in the same trunk with postaxial ones
throughout the entire length of the crus and by the inclusion of fibers
destined for superficial muscles in the same trunk with others for the
deep muscles.

Tabulating the nerve supply of the plantar muscles according to the
origin of the fibers from the two main stems the following arrangement
is obtained :—

Plantaris superficialis medialis,
Plantaris superficialis lateralis, |
Plantaris profundus III (in part), }

Fibulo-tarsalis, |
Plantaris profundus II, J

Ramus superticialis.

Plantaris profundus III (in part),
Plantaris profundus I, Ramus profundus.
Interosseus,

but, if the interpretation of the plantar nerves given above on the basis
of a comparison with the arm nerve be accepted, the tabulation will be as
follows :—

J. Playfair MeMurrich 41

Plantaris superficialis medialis,
Plantaris superficialis lateralis,
Plantaris profundus III,
Plantaris profundus II,

Ramus superticialis medialis.

Fibulo-tarsalis, Ramus superficialis fibularis.

Plantaris profundus J,

Ramus profundus.
Interosseus,

Il. THe CrurAL FLEXORS IN THE LACERTILIA.

A goodly number of papers dealing with the myology of the hind limb
of members of the group Lacertila have appeared, that of Gadow, 82,
being one of the most comprehensive. It has been the custom, however,
to employ for the various muscles a terminology based upon that used
for the mammalia, a procedure which carries with it implications of
homologies which in some cases do not exist and in the majority of cases
are at best merely partial ones. Since in the present study the lacer-
tilian myology is being approached from below, rather than from above,
and since in the amphibia the characteristic feature of the crural muscles
is their arrangement in layers, I propose to employ for the reptilian
muscles a terminology which will indicate their relations to the amphi-
bian condition, using the terms employed by Gadow, for instance, only
for purposes of identification.

A transverse section through about the middle of the crus of Scincus
presents the appearance shown in Fig. 3. At first sight the differences
from the arrangement in Amblystoma are very apparent, but a closer
_ inspection will reveal marked similarities, which a study of the nerve
supply will but serve to emphasize. The topographical relations of the
muscles may, however, first be considered, with a view to determining
how far a layered condition can be recognized.

It is a characteristic of the amphibian superficial plantar layer that it
arises from the femur and is inserted below into the plantar aponeurosis.
In Scincus one finds superficially upon the posterior surface of the crus
three muscles, a plantaris superficialis medialis (Psm), a plantaris
superficialis lateralis (Psl), and between the two a long slender muscle
which may be termed the plantaris superficialis tenuis (Pst). Of these
the plantaris superficialis medialis differs from the other three in that it
arises from the head of the tibia, instead of from the femur as might be
expected if it be really a portion of the superficial plantar layer. Ex-
amining its origin more closely it will be seen to arise not only from the
head of the tibia but also from the posterior surface of a strong tendon
which passes from the head of the tibia to the internal condyle of the
femur. The existence of this tendon and the relation of the muscle

42 The Phylogeny of the Crural Flexors

to it is of great importance in determining the true significance of the
muscle, for they indicate, apparently, a primary attachment of the
muscle to the femur. The tendon, indeed, represents the proximal
portion of the muscle which has undergone degeneration in association
with a new attachment made by the muscle below the knee joint, the
resulting condition being strictly comparable with what has. occurred
in connection with the peroneus longus of man, this muscle having
similarly shifted its upper attachment from the femur to the fibula, its
upper part being represented by the external lateral ligament of the knee
joint.

Fie. 3.—Transverse section through the upper part of the crus of Scincus sp.
F, fibula; f, ramus superficialis fibularis; fc, fibular cutaneous nerve; J interosseus;
m, ramus superficialis medialis; p, ramus profundus; pc, cutaneous branch from ramus
profundus; P//-J///, plantaris profundus II-III; Psa, plantaris superficialis accessorius ;
Psl, plantaris superficialis lateralis; Psm, plantaris superficialis medialis; Pst, plantaris _
superficialis tenuis; 7’, tibia.

An examination of the origin of the plantaris superficialis medialis
as described for other lacertilia seems to give support to this view. It
is true that throughout the reptilia in general the muscle takes its origin
from the tibia. In Euprepes, however, Firbringer, 70, describes it,
under the name of the gemellus internus (epitrochleo-tibio-metatarsalis
ventralis) as arising both from the head of the tibia and from the inter-
nal condyle of the femur, and, according to Gadow, 82, it (gastroc-
nemius, caput internum) arises in Ophryoessa principally from the pos-
terior surface of the internal condyle, only a few fibres taking origin
from the tibia. It is clear then that one is dealing here with a muscle
which was either primarily attached to the femur and in the majority
of the reptilia has made a secondary connection with the tibia, or else was
primarily attached to the tibia and has secondarily migrated, so far as
its origin is concerned, upward to the femur.

There seems to be little question but that the former of these two

J. Playfair McMurrich ; 43

possibilities is the easier of accomplishment; it is not a migration, but
the formation of a new attachment in the course of the muscle and the
degeneration of the part above. And, as already noted, there is evidence
of the occurrence of such a process in at least one of the muscles found .
in man. On this view the tibial superficial muscle of the lacertilian
crus is to be regarded as having in reality a femoral origin and agrees
in its primary relations with the other superficial muscles.

Traced downwards the plantaris superficialis medialis becomes a
broad band which inserts into the tibial border of an aponeurotic sheet
(Fig. 4, a) which represents a portion of the superficial layer of the
plantar aponeurosis and receives also the insertion of the plantaris super-
ficialis lateralis.

Fic. 4.—Transverse section through the lower part of the crus of Scincus sp.
a, superficial portion of plantar aponeurosis; a’, deep portion of plantar aponeurosis ;
f, ramus superficialis fibularis; F, fibula; fe, fibular cutaneous nerve; J, interosseus ;
I’, vertical fibers of interosseus: p, ramus profundus; pce, cutaneous branch of ramus
profundus; Pp/,plantaris profundus I; Pp//-/Z/, plantaris profundus II-III; Psl,plantaris
superficialis lateralis; Psm, plantaris superficialis medialis; 7, tibia.

The plantaris superficialis tenuis (Fig. 3, Pst) takes its origin above
the knee joint from a sesamoid cartilage developed in a tendon arising
from the fibular border of the flexor tibialis externus (Gadow) and
passes downwards to unite with the tibial border of a portion of the
plantar aponeurosis which covers the posterior surface of the plantaris
profundus II-III (Fig. 4, a’). The muscle is slender throughout its
entire course. At first it lies superficially between the medial and lateral
superficial plantars, but lower down it is covered by the fibular edge of
the medialis and at about the middle of the crus fuses with the posterior
surface of the deep plantar mass, or, as it is better expressed above,
inserts into the tibial border of a portion of the plantar aponeurosis which
covers the deep plantar mass.

This muscle does not seem to be present in all lacertilia. Thus Perrin,
93, fails to find it in Uromastix and I have not succeeded in observing

t4 The Phylogeny of the Crural Flexors

it in dissections of Phrynosoma. On the other hand Perrin finds it in
Varanus and apparently in Lacerta and Gongylus, and Gadow, 82, re-
gards it as a typical portion of his flexor longus digitorum of which it
forms the caput accessorium. It seems to be a muscle separated from the
fibular border of the plantaris superficialis lateralis, a view which re-
ceives confirmation from the statement of Gadow that it is sometimes
fused above with that muscle. Its apparent absence in certain forms
may upon this view be regarded as due to its failure to separate from the
parent muscle.

The plantaris superficialis laterals (Fig. 3, Psl) is a rather large
inuscle which takes its origin by a tendon from the posterior surface of
the lateral condyle of the femur. A sesamoid cartilage is imbedded in
the tendon just above the line where the muscle fibres begin to make
their appearance and, as the tendon is traced downwards, it is found
to broaden out into a thin sheet covering the anterior (deep) surface
of the muscle and gradually fading out below, with the exception of a
narrow band which continues on to the region of the ankle joint, be-
coming enclosed by the muscle substance (Fig. 4). Just when the
tendon begins to fade out an aponeurotic layer (Fig. 4, @) appears on
the posterior (superficial) surface and increases in strength as it passes
downwards, becoming a part of the plantar aponeurosis. It is with the
inner border of this that the plantaris superficialis medialis and the ten-
don which descends from the border of the flexor tibialis externus and
gives rise to the plantaris superficialis tenuis, unites.

As the muscle substance is traced downwards it is seen to be continued
past the ankle joint into the plantar region of the foot. In the upper
part of the muscle the fibers are all parallel, arising from the tendon of
origin, but lower down fibers arise from the slender tendon which con-
tinues the tendon of origin downwards and have a somewhat radiating
arrangement (Fig. 4). Tracing out the two sets of fibers, it is found
that the upper ones insert into the upper part of the plantar aponeurosis,
while it is those which arise from the prolongation of the tendon of
origin that form exclusively the lower part of the muscle and are con-
tinued over the tarsus to be inserted into the plantar aponeurosis, and
the sesamoid cartilage developed in it over the fifth metatarsal.

In addition to these three muscles there is yet a fourth (Psa) which
is apparently to be reckoned as a portion of the plantaris superficialis.
It arises from the posterior surface of the external condyle of the
femur below the plantaris superficialis lateralis and passes downward
under cover of that muscle to about the middle of the crus, where it
unites with the plantaris profundus ITI-IT, or rather, inserts into the por-
tion of the plantar aponeurosis covering that muscle (Fig. 4, a’).

J. Playfair MeMurrich 45

Gadow, 82, and Perrin, 93, both describe this muscle as a portion of
the flexor longus digitorum (fléchisseur des quatre premiers doigts),
that is to say of the plantaris profundus III-II, making it a femoral head
of that muscle. It may be that it is really a portion of the plantaris
profundus group of muscles which has secondarily extended its origin
to the femur and that its absence in Ophryoessa and Cnemidophorus,
as noted by Gadow, is due to this upward migration not having taken
place. On the other hand it seems more probable that the origin from
the femur is a primary condition, the muscle being a separation of
the deeper portions of the plantaris superficialis lateralis. Its union
with the plantaris profundus presents no more obstacle to this view than
the similar union of the superficialis tenuis; both the plantaris profun-
dus and the plantaris superficialis insert primarily into the plantar
aponeurosis, so that a union of the two sets of muscles is not at all im-
possible. On account of its associations with the deep plantar muscles
it will be spoken of as the plantaris superficialis accessorius.

The plantaris profundus group of muscles is represented in the lacer-
tilia by three distinct muscles, one of which is to be regarded as repre-
senting the plantares profundi IIT and II of the amphibian crus, while
the other two represent the plantaris profundus I. The plantaris pro-
fundus ITI-II (Fig. 4, Pp") is the muscle termed by Gadow the
flexor longus digitorum, caput internum, and by Perrin the téte interne
du fléchisseur des quatre premiers doigts. It takes its origin from the
upper half of the fibula and to a slight extent from the outer surface of
the head of the tibia, and increases rapidly in size as it descends the
crus, forming the most voluminous muscle of the calf of the leg. At
about one-third of the length of the crus an aponeurotic layer appears
upon its posterior surface (Fig. 4, a’) and into this the plantaris super-
ficialis accessorius and the plantaris superficialis tenuis insert. As it
approaches the ankle joint the aponeurosis increases in strength and be-
comes tendon-like, the fibres of the muscle terminating upon it, and at
the ankle joint a sesamoid bone (Fig. 5, s) is developed upon its tibial
border, the last remaining muscle fibers and also the plantaris profundus
I accessorius inserting into this. With the development of the sesamoid
bone the whole aponeurosis or tendon becomes thick and almost carti-
laginous, but as it is traced onward into the foot it again becomes ten-
dinous and gives off a slip from its fibular border. This passes to the
fifth digit, sending off a slip to the fourth, and the main portion of the
tendon passes on beneath the superficial muscles of the planta to divide
eventually into tendons which pass to the three inner digits. All the
five tendons pass to the terminal phalanges of their respective digits and
give origin in their course to the lumbrical muscles.

46 The Phylogeny of the Crural Flexors

The amphibian plantaris profundus I is represented in Scincus by
two distinct muscles. The first of these, which may be termed simply
the plantaris profundus I (Figs. 4 and 5, Pp’), arises from the posterior
surface of the lower part of the fibula and is directed obliquely downward
and inward. It passes over into a flat tendon which lies beneath (ante-
rior to) the tibial border of the tendon of the plantaris profundus ITI-II,
and, indeed, is to a certain extent connected to the deep surface of the
sesamoid bone developed in that tendon. It separates from it again,
however, and is continued on over the large tarsal bone of the first row
(astragalo-caleaneus) and is inserted into the two inner bones of the
second row of the tarsus. This is the muscle which has very generally
been recognized as the tibialis posticus.

The second muscle, which may be
termed the plantaris profundus I acces-
sorius (Fig. 5, Pp'a), arises from the
plantar surface of the fibular portion of
the large astragalo-calcaneous (fac)
and is directed obliquely inward and
downward, passing posteriorly to the
lower part of the plantaris profundus I
(Pp'), to be inserted into the sesamoid
bone (s) developed in the tendon of
the plantaris profundus III-II.

Finally, there is a well developed
interosseus rouscle (Figs. 3 and 4, J)

Fic. 5.—Transverse section through
the ankle of Scincus sp. a, superficial
portion of plantar aponeurosis; aJ,
deep portion of plantar aponeurosis;
f, ramus superticialis fibularis; FAC,
fibulare-astragalo-calcaneus; fe, fib-
ular cutaneous nerve; p, ramus pro-
fundus; Ppl, plantaris profundus I;
Ppla, plantaris profundus I accesso-
rius; s, Sesamoid bone in deep portion
of plantar aponeurosis; 7, tibia.

which passes across from the fibula to
the tibia, filling up the interval between
the two bones through almost its entire
length. In the upper part of the mus-
cle (Fig. 3) the fibers have an almost
vertical direction, but, as it is traced

downwards, the lower fibers, which pass
over to the tibia anterior to the higher ones, become more and more
oblique, until finally in the lower part of the crus (Fig. 4) all the fibers
are exceedingly oblique, some almost transverse, and the vertical upper
fibers are seen as a small bundle (I) lying upon the posterior surface
of the tibia, completely isolated from the oblique ones. The higher
vertical fibers are inserted into the outer (fibular) and posterior sur-
faces of the lower half of the tibia, while the lower oblique fibers pass
te its anterior and inner surfaces, wrapping around the lower end of
the bone.

J. Playfair MceMurrich 47

Before passing to a consideration of the nerve-supply of these muscles
a few remarks may be made in the way of a comparison of the plantar
aponeurosis of the lacertilia with that of the amphibia. In the latter,
just as was the case with the palmar aponeurosis, it forms a continuous
sheet which receives the insertion of the crural flexors and gives origin
to the plantar muscles of the pes, the only indications of a layered con-
dition to be seen in it being at its upper and lower borders, where it
becomes partly divided into subjacent layers corresponding to the layers
of muscles inserting into or arising from it. In the lacertilia the con-
ditions are slightly different. Covering the posterior surfaces of the
plantaris superficialis lateralis there occurs a distinct aponeurotic layer
(Figs. 4 and 5, a) which receives the insertion of the fibers of that
muscle and is also joined by the tendon of the plantaris superficialis
medialis. As it is traced downwards this aponeurosis separates in the
neighborhood of the ankle joint into a thinner and narrower superficial
layer and a thicker and deeper layer. The former gradually verges
towards the fibular side as it passes into the foot and is. finally lost over
the outer side of the fifth metatarsal bone. The deeper layer gives rise
from the deep surface of its medial half to the superficial layer of the
plantar muscles, while its lateral portion, developing a sesamoid cartilage
which receives the insertion of the fibers of the plantaris superficialis
lateralis, inserts into the fifth metatarsal.

In addition to this superficial layer a deeper layer of the aponeurosis
also occurs (Figs. 4 and 5, a’), this being the aponeurosis with which the
plantaris profundus III-II becomes connected and which is continued
enward as the tendons of that muscle to be inserted into the terminal
phalanges of the digits.

There are then in the lacertilia two principal portions of the plantar
aponeurosis as compared with the continuous aponeurosis of the amphi-
bia. A deeper portion has separated from a more superficial one to
form the tendons of the plantaris III-II and having also inserted into
it portions of the plantaris superficialis. Probably too the tendon of
insertion of the plantaris profundus I, on account of its attachment to
the sesamoid bone developed in the tendon of the plantaris profundus
III-IT, is to be regarded as a separated portion of the original aponeu-
rosis and, if this be the case, all the crural flexors primarily insert into
the plantar aponeurosis as in the amphibia.

Turning now to the nerves of the crus. In sections just above the
knee the sciatic nerve is represented by three trunks. One of these
(Fig. 6, A), when traced onwards, curves around the outer border of
the fibula to the dorsal surface of the crus and need not concern us
further. The other two are supplied to the flexor surface.

48 The Phylogeny of the Crural Flexors

One of them (/’) passes down into the crus upon the fibular side of
the plantaris superficialis lateralis and gives off a couple of large cuta-
neous branches (C) which are distributed to the fibular side of the leg,
one of them (Figs. 4 and 5, fc) passing far down the crus in the groove

Mp FA

Fia.6.—Diagram of the
erural nerves of Scincus
sp. -A,nerve to extensor
surface; ¢, cutaneous
nerves; EHP, external
plantar; F’, ramus super-
ficialis fibularis; J,branch
to interosseous; IP, in-
ternal plantar; m, ramus
superficialis medialis;
Mp, common trunk of
ramus superficialis med-
jalis and ramus pro-
fundus; Pp/, branch to
plantaris profundus I;
Ppa, branch to plantaris
profundus I accessorius ;
PpJI-I1T, branch to plan-
taris profundus II-IIT;
Psa, branch to plantaris
superficialis accessorius ;
Psl, branch to plantaris
superficialis lateralis ;
Psm, branch to plantaris
superficialis medialis ;
Pst, branch to plantaris
superficialis tenuis; rp,
ramus profundus.

between the fibular border of the plantaris super-
ficialis lateralis and the peroneus. The main stem
in its downward course gradually verges medially,
so that it comes to lie beneath the plantaris super-
ficialis lateralis (Figs. 4 and 5, /) and, indeed, be-
comes partly enclosed in the substance of that
muscle. It appears to give off no branches in the
crus, nor could any twig to the plantaris super-
ficialis lateralis be found arising from it. At the
ankle it passes into the foot towards its fibular
border, and is then recognizable as the external
plantar nerve.

This nerve is evidently that referred to by Ga-
dow, 82, as stem III, and, notwithstanding its
somewhat different course in its upper part, is ap-
parently equivalent in part to the ramus super-
fictalis fibularis of Amblystoma, but, unlike it, is
quite distinct from the branches which represent
the ramus superficialis medialis, a difference which
may well be correlated with the absence of an
anastomosis between the flexor stems in the lacer-
tilia.

The other main stem, before leaving the thigh,
divides into two trunks (Fig. 6, m. and rp.), both
of which pass downwards to the tibial side of the
plantaris superficialis lateralis. The more posterior
trunk (m), the stem I of Gadow, may from its dis-
tribution be termed the ramus superficialis medialis.
Shortly after entering the crus it divides into

several branches which are entirely confined to the crus and supply all
the portions of the plantaris superficialis, as well as the plantaris pro-
fundus III-II.

The deeper branch (rp), the stem II of Gadow, may be termed the
ramus profundus. It passes towards the tibia and divides into two
branches, the posterior of which (Fig. 6, c; Figs. 3 and 4, pe) is cuta-
neous and is distributed to the skin over the tibial surface of the crus.
The deeper branch enters the substance of the the m. interosseus (Fig.

J. Playfair McMurrich 49

3, p) on its posterior surface and is continued downward through that
muscle, to which it gives branches, and, emerging from it at the ankle,
(Fig. 6, p) it sends twigs to the plantaris profundus I and to the plan-
taris profundus I accessorius and is then continued into the foot as
the internal plantar nerve.

Accepting the interpretation of the nerves of the amphibian crus
given above and comparing on the basis of their nerve supply the muscles
of the amphibian and lacertilian crus the following result is obtained.

Nerve. Amphibia. Lacertilia.

Plantaris superf. medialis, Plantaris superf. medialis.
Plantaris superf. tenuis.
Plantaris superf. lateralis.
Plantaris superf. accessorius.

Plantaris superf. lateralis,
R. superf. medialis,

Plantaris profundus III, minor,

Plantaris profundus III, |

Plantaris profundus II, J

Plantaris profundus III-II.

R. superf. fibularis, Fibulo-tarsalis, 4 £4 424 4 = ———— «ssseees

Plantari atath Plantaris profundus I.
R. profundus, i prdiaad aa 8 uy yam { Plantaris profundus I, accessorius.
Interosseus, Interosseus.

It is clear from this that a close comparison based both upon the
topographical relations and the nerve supply can be made between the
crural flexors of the amphibia and those of the lacertilia, there being,
liowever, in the latter a greater amount of differentiation of the original
layers. It is interesting to note that just as in the lacertilian arm no
representative of the ulno-carpalis could be distinguished, so too in the
crus there appears to be no representative of the amphibian fibulo-tarsalis.

Ill. Ture CruraAL FLExors IN THE MAMMALIA.

In considering the crural flexors of the mammalia it will be con-
venient to depart from the method of description and nomenclature
followed in the preceding pages, and to consider the various muscles as
independent structures, employing the terms usually assigned to them
in mammalian myology. In other words, the primary layers will be
temporarily neglected, the reference of the individual muscles to them
being considered later on.

A considerable amount of confusion seems to have existed with
reference to the soleus and gastrocnemius. Thus, the latter muscle has
been described as possessing but a single head in certain forms, the
lateral head being described as the soleus; in others the soleus is sup-
posed to be included in the lateral head of the gastrocnemius; and in
one case even, the medial head of the latter muscle has been termed

50 The Phylogeny of the Crural Flexors

the soleus. It becomes a question then what shall be termed a soleus
and what a gastrocnemius, and since the human arrangement is the
type with which all other mammalia are directly or indirectly com-
pared, it will be advisable to base the definitions of the two muscles on
that arrangement, and this is, essentially, that the gastrocnemius takes
its origin from the femur and is a two-joint muscle, while the soleus
has its origin from the bones of the crus and is a one-joint muscle. A
more satisfactory distinction could be made by referring the two muscles
to their respective primary layers, but for the present that given above
may suffice.

Fig. 7. Fi@a. 8.

Fic. 7.—Transverse section through the lower part of the crus of Didelphys virgini-
ana. ep, external plantar nerve; F, fibula; FF, flexor fibularis; ff, nerve to flexor fibu-
laris; FT, flexor tibialis; Ge, gastrocnemius lateralis; Gi, gastrocnemius medialis; ip,
internal plantar nerve; Pl, plantaris; PZ, pronator tibie; rp, ramus profundus;
T, tibia; T'P, tibialis posticus.

Fic. 8.—Transverse section through the upper part of the crus of Didelphys virgini-
ana. F, fibula; FF, flexor fibularis; Ge and GeJZ, outer and inner portions of gastroc-
nemius lateralis; Gi, gastrocnemius medialis; Pl, plantaris; PY, pronator tibie; s,
soleus; 7’, tibia; 7'P, tibialis posticus.

The medial head of the gastrocnemius is a practically constant ele-
ment of the mammalian crus and presents little variation except in
relative size. In the opossum (Figs. 7 and 8, G1) it arises from the
internal condyle and quickly passes over into a flattened tendon which
descends the leg, gradually verging toward its outer border, until near
the ankle joint it comes to lie to the outer side of the tendon of the
gastrocnemius lateralis, in close proximity to which it is inserted into
the os calcis. No union occurs between the two tendons except immedi-
ately at their insertion. In both the mouse (Fig. 10, Gz) and the cat
(Fig. 9, Gi) the muscle unites with the gastrocnemius lateralis high up
in the crus and the conjoined tendon inserts into the os calcis.

J. Playfair MceMurrich 51

In contrast to the extreme simplicity of structure presented by the
gastrocnemius medialis is the complexity of the gastrocnemius lateralis
in all three forms here under consideration. In the opossum the
muscle near its origin was found to consist of four bundles. Two of
these (Fig. 8, Ge) arose close together from the outer surface of the
lateral sesamoid cartilage of the knee joint and from the ligament
extending from this to the external condyle, and were distinguishable
not only by being separated by a band of connective tissue, but also by
a difference in the direction of their fibers. A third portion (Ge‘)
took its origin from the inner (tibial) surface of the lateral sesamoid
cartilage, while the fourth portion (s) arose from the posterior sur-
face of the head of the fibula, or, to be more precise, from the posterior
surface of a tendon which arises from the posterior surface of the head
of the fibula and is continued downwards to beyond the middle of
the crus upon the deep surface of the compound muscle.

Fic. 9.—Transverse section through the upper part of the crus of the Cat. F, fibula;
Gel and Gem, lateral and medial portions of the gastrocnemius lateralis; Gi, gastroc-
nemius medialis; P! and P?, oblique and vertical portions of popliteus; Pl, plantaris ;
pt, posterior tibial nerve; rm, ramus superficialis medialis; 7p, ramus profundus;
s, soleus; 7, tibia.

Below these four bundles became more or less confused, the connec-
tive tissue partition between the portions from the outer and inner
surfaces of the sesamoid cartilage persisting for a greater distance than
the others, and before its disappearance a tendon appears in the center
of the outer sesamoid portions (Fig. 7) and gradually increases in size
to become the tendon of the muscle. This is continued down the leg
quite independent of the tendon of the gastrocnemius medialis, with
which it is inserted into the tuberosity of the os calcis.

In the cat (Fig. 9) the lateral gastrocnemius arises together with the
plantaris (P/) from the patella by a strong aponeurotic sheet which is
continued backward from the lateral border of that bone, and also
from the downward continuation of this sheet which forms an invest-

t
©

The Phylogeny of the Crural Flexors

ment for the outer surface of the plantaris for more than its proximal
half. The muscle also takes origin from the lateral surface of the lateral
sesamoid bone. The large muscle mass which results is intimately re-
jated to the plantaris, and in its upper part a thin sagittal aponeurotic
plate appears dividing the muscle into almost equal portions (Ge! and
Ge”). This sheet is continued all the way down the leg and below
receives the insertion of the muscle fibers, becoming the tendon. In its
upper part the muscle comes into contact with the gastrocnemius
medialis (Gi), an aponeurotic plate intervening between them, however, .
and below, the tendons of the two muscles fuse completely to be inserted
into the tuberosity of the os calcis, the tendon of the soleus also joining
them shortly before their insertion, to form a typical tendo Achillis.’

In the mouse I cannot state the exact origin of the muscle (Figs. 10
and 12), but it has essentially the same structure as in the cat and has
similar relations with the gastrocnemius medialis and plantaris.

Two other muscles are intimately associated with the gastrocnemius,
the one more especially at its origin and the other at its insertion.
These are the plantaris and the soleus.

The plantaris, notwithstanding its variability in man, is of very con-
stant occurrence throughout the entire mammalian series, and has as
a rule a much greater development and a more important role than in
man. It is always closely associated at its origin with the gastrocnemius
jateralis and is inserted below into the plantar fascia (occasionally into
the os caleis) by which its action is transmitted to the digits. In the
three forms here under consideration it forms what may be regarded
as the medial anterior portion of the muscular mass formed by it and
the gastrocnemius lateralis. It arises in the opossum (Fig. 8, Pl) from
the medial half of the posterior surface of a tendon which extends down-
wards from the lateral fabella, and in the cat, from the fabella and from

’In sections through the tendo Achillis one sees to the medial side a dis-
tinct tendon, connected to the true tendo by thin fascia beneath which lies
the tendon of the plantaris as in a groove. This tendon might readily be
mistaken for that of the gastrocnemius medialis, but it is in reality a
thickening of the crural fascia and is quite independent of the muscle. It is
attached below to the os calcis medially to the insertion of the tendo Achillis.
Upon the lateral border of the tendo a similar thickening of the crural fascia
occurs, but this fuses with the tendon of the gastrocnemius lateralis shortly
before it is joined by the soleus. These fascial thickenings have been described
for the dog by Ellenberger and Baum, 91, who trace them to the semi-mem-
branous and biceps muscles, but they do not seem to have been noted for
the cat, at least they are not mentioned in any of the works on that form
to which I have access at present.

J. Playfair McMurrich 53

the strong aponeurosis which passes backwards from the outer border
of the patella. Throughout the greater portion of its extent it is barely
separable from the gastrocnemius lateralis (Figs. 9 and 10), but below
it becomes tendinous and lies below (1. e., anterior to) and to the inner
side of the tendon of the gastrocnemius lateralis, from which, however,
it is quite distinct. At the ankle joint it hes to the medial side but
posterior to the gastrocnemius and soleus tendon and spreads out into
the thin but dense plantar aponeurosis. This covers the insertion of
the tendo Achillis or its representatives and passes downward over the
tuberosity of the os calcis, being attached to the outer surface of that
hone by its outer border, but its inner border and the greater part of its
central portion is free. Passing on into the foot it gives rise upon its
deeper surface to the flexor brevis minimi digiti and may be continued
onward as a series of fascial slips to the bases of the digits.

Fie. 10.—Transverse section through the upper part of the crus of the Mouse.
F, fibula; FF, flexor fibularis; ff, nerve to flexor fibularis; Ge, gastrocnemius lateralis ;
Gi, gastrocnemius medialis; P, popliteus; Pl, plantaris; pt, posterior tibial nerve;
rp, ramus profundus; S, soleus; 7’, tibia; 7'P, tibialis posticus.

The important point about the muscle, so far as its insertion is con-
cerned, is its connection with the plantar fascia. That this is its true
termination becomes evident in those forms such as Cuscus (Cunning-
ham, 81) in which a plate of cartilage is developed in the fascia, the
plantaris being inserted into the proximal border of this cartilage. To
describe the plantaris as being continuous with the flexor brevis digi-
torum, as is sometimes done, merely leads to confusion; this muscle
really arises from the plantar aponeurosis and the slips which extend
to the bases of the digits are also portions of the plantar aponeurosis
and have no primary relation to the plantaris.

The soleus, unlike the plantaris is not always distinguishable in the
lower mammals. In both the cat and the mouse (Fig. 11, s) it is a well
developed muscle which arises from the posterior surface of the upper
part of the fibula (Figs. 9 and 10, s) and descends the leg beneath the

54 The Phylogeny of the Crural Flexors

plantaris and lateral gastrocnemius, eventually becoming tendinous and
uniting with the tendon of the latter muscle. In the opossum, however,
it does not appear to exist as a distinct muscle and the conditions in
this form probably serve to explain its apparent absence in others.

In the description of the gastrocnemius lateralis of the opossum it
was noted that it possessed an origin from the head of the fibula. This
head seems to be unrepresented in the gastrocnemius of the cat and
mouse, and its relations to the rest of the muscle in the opossum present
some interesting peculiarities. When it is first seen in tracing a series
of sections downward it consists of a thin band of fibers (Fig. 8, s) which
arise from a tendon extending downward from the head of the fibula,
a portion of the flexor longus digitorum lying between the tendon and
the bone. This muscle band is separated at this level from the deeper
surfaces of the gastrocnemius lateralis and the plantaris by the tendon
which extends downward from the lateral fabella and gives origin to the
plantaris. As this tendon gradually fades out below a distinction be-
tween the muscle band under consideration and the gastrocnemius
lateralis becomes less and less, until, finally, there is complete union of
the two.

This fibular head of the opossum seems to represent the soleus of the
higher mammalia, and the supposition of Cunningham, 81, and others
that the so-called gastrocnemius lateralis of the marsupials includes also
the soleus is correct, and the same is probably true of the dog and the
other higher mammalia in which the soleus is stated to be lacking. As
regards the monotremes it is to be noted that the lateral superficial
crural flexor has been termed the soleus, and the gastrocnemius is re-
garded as lacking (Westling in Leche, 98), this nomenclature being
adopted no doubt in view of the fact that the muscle arises from the
peculiar process developed upon the upper end of the fibula in these
forms and has no connection with the femur. If the fibular process
represents a true outgrowth of that bone such a nomenclature would be
justified, but it seems really to be an epiphysial structure and in all prob-
ability represents the lateral fabella of other forms. On this view the
distinction which is made between the marsupial and monotreme muscle
practically vanishes and it seems necessary to regard it as representing
in both groups the gastrocnemius and soleus of higher forms.

The long flexors of the digits in the mammalia have been thoroughly
discussed by F. E. Schulze, 66, and by Dobson, 83, and the former has
pointed out that the arrangement occurring in man is quite different
from that characteristic of the majority of mammals and as a conse-
quence the nomenclature employed in human anatomy cannot be con-

J. Playfair MeMurrich 55

sistently applied in the lower forms. He proposed, accordingly, and his
proposition was accepted by Dobson, to speak of the two muscles usually
recognized as the flexor longus hallucis and the flexor longus digitorum as
the flexor digitorum fibularis and the flexor digitorum tibialis. respect-
ively. The proposition is certainly worthy of general acceptance and
is almost necessary from the comparative standpoint, since in the majority
of mammals the flexor digitorum fibularis (fl. longus hallucis) is the
principal muscle and the flexor tibialis the subordinate one.

Dobson has pointed out that the relations of the two flexors is accord-
ing to one of two types and that all the members of any family, if not
order, of mammalia will present the same type. In one type the tendons
of the two muscles fuse, while in the other they remain distinct, and not-
withstanding that he found the aplacental mammalia presenting the
second type of relation, Dobson concludes that the first is the more
primitive, since, as he states it, “it is difficult to conceive that in any
animals in which a definite separation of the tibial from the fibular
flexors had once taken place—symmetrical reunion of these tendons
could subsequently occur.” With such a view the phylogenetic plan
here being traced agrees, for an important part of this plan is the recog-
nition of the plantar aponeurosis of the lower forms in the tendons of
the long flexors, all the post-axial muscles of the crus, except the inter-
csseus, having their insertion primarily into that aponeurosis, through
which their action is extended to the digits.

The descriptions of the long flexors which Dobson has given for so
many species of mammals are sufficiently thorough to warrant the omis-
sion of a detailed description of the arrangement observed in the forms I
have studied, but for the sake of completeness and to bring out especially
their relations to the plantar aponeurosis, or rather its mammalian repre-
sentatives, a description of the arrangement observed in the opossum may
be given.

The flexor fibularis digitorum (Figs. 7 and 8, FF) arises from the
inner and posterior surfaces of the greater portion of the fibula. In its
upper part it is separated by a strong aponeurosis from the adjacent
tibialis posticus, and at about the middle of the leg a strong aponeurosis
appears upon its posterior surface, separating the muscle from the more
superficial plantaris. Traced downwards this aponeurosis gives rise upon
its posterior surface to a muscle which increases rapidly in breadth,
while the aponeurosis diminishes in that dimension, although thicken-
ing to form a structure to which the term tendon is applicable. The
muscle is the flexor brevis digitorum, or rather a considerable portion of
it, and need not concern us any further except in so far as its origin from

5

56 The Phylogeny of the Crural Flexors

what is usually described as the tendon of the flexor fibularis serves to
confirm the homology of that tendon with part of the plantar aponeurosis
of lower forms. Eventually all the fibers of the flexor fibularis insert
into the tendon, the last of them disappearing some distance above the
ankle joint.

The tendon is continued onward into the foot, lying in the median
line between the os calcis and the inner malleolus, and at about the
level of the distal row of tarsal bones the tendon of the flexor accessorius
passes across it to be attached to its tibial border. This portion of the
tendon then separates to pass on to an insertion into the base of the termi-
nal phalanx of the first digit, and later the remainder of the tendon
divides into four nearly equal tendons, which pass to the remaining
digits. The relations of the lumbricales to the tendons will be con-
sidered on another occasion.

The flexor digitorum tibialis is, in contrast to the flexor fibularis,
a rather slender muscle. It has usually been described as arising from
the upper part of the tibia, but in my preparations I have not been
able to trace it to that bone. I find it (Fig. 7, FT) taking its origin
trom the strong aponeurosis which covers the posterior surface of the
strong pronator tibie (P7'), and although it thus comes very close to
the upper part of the tibia, no definite connection with that bone could
be made out. The difference may be due to the fact that the individuals
I studied were advanced fetuses, and that with advancing age the
insertion reaches the bone, a process which, if it really occurs, is interest-
ing as denoting a migration of the muscle tibia-wards. Its belly forms
an irregularly quadrilateral mass lying between the pronator tibie
internally and the flexor fibularis externally, and resting upon the
tibialis posticus. At about the middle of the crus its tendon begins to
appear upon its outer surface and into it the muscle fibers gradually
insert, until in the lower part of the crus only the tendon remains, rest-
ing directly upon that of the tibialis posticus, by which it is separated
from the posterior surface of the tibia. At the ankle joint it rests upon
the internal malleolus and as it passes onward into the foot it separates
from the tibialis posticus tendon and approaches the tendon of the flexor
fibularis. At the level of the junction of the proximal and distal rows of
the tarsal bones it gives origin to muscle fibers which represent a por-
tion of the flexor brevis digitorum and pass downward and inward to
join the rest of that muscle which arises from the tendon of the flexor
fibularis.

A little farther on the flexor tibialis tendon becomes connected by
fibrous tissue of varying density with the inner border of the flexor

J. Playfair McMurrich 57

fibularis tendon, but the actual tendon can be traced uninterruptedly
onward and, as Dobson states, does not really unite with the flexor
fibularis tendon, although the connection between the two may be suffi-
ciently strong as to make a practical ynion. The tendon: then begins
to flatten out into a broad band which fades out gradually at the sides
into the layer of the plantar aponeurosis with which the plantaris is
associated, and eventually associates itself with the flexor brevis hallucis,
inserting, in part at least, into the under surface of the cartilaginous
spur.

In the other two mammals which I studied the general arrangement
of the two flexors was similar to the above, except that the flexor brevis
digitorum did not arise from their tendons. The flexor fibularis (Fig.
11, FF) is much the larger of the two muscles and sends tendons to all
five digits, while the flexor tibialis, in
the cat, unites with the tendon of the
fiexor fibularis before it divides into
the terminal tendons. In the mouse
the flexor tibialis (Fig. 11, #7’) arises
in common with the tibialis posticus
and its tendon remains completely
separate from that of the flexor fibu-

Fia. 11.—Transverse section through

ia Q ] about the middle of the crus of the
laris and fades out into the plantar Mouse. F, fibula; FF, flexor fibularis;

7 FT, flexor tibialis; Ge, gastrocnemius
fascia. In the cat the muscle has an lateralis; Gi, gastrocnemius medialis;

independent origin from the back of the Pl, plantaris ; pt, posterior tibial nerve ;

S,soleus; 7, tibia, TP, tibialis posticus.

upper part of the tibia.

The tibialis posticus has also been described for a large number of
forms by Dobson, 83, and I shall indicate only briefly its arrangement in
the forms I have studied. In the opossum (Fig. 8, 7’P) it arises from
the upper part of the fibula and from a strong aponeurosis which separ-
ates it from the adjacent flexor fibularis, and quickly passes over into
a tendon which is continued down the leg, under cover of the tendon of
the flexor tibialis, and passing behind the inner malleolus is inserted into
the scaphoid bone. In the cat it arises from the upper part of the pos-
terior surface of the tibia, becomes tendinous at about the middle of the
crus and, passing into the foot in a groove on the inner surface of the
tibia, is inserted into the scaphoid. In the mouse (Fig. 11, TP) it also
arises from the upper posterior part of the tibia as a muscular mass from
which later the flexor tibialis separates. It is a slender muscle, soon be-
coming a tendon and inserting into the internal cuneiform bone.

The flexor accessorius pedis, although apparently a muscle of the pes,
is considered here with the crural flexors, since its affinities are altogether

58 The Phylogeny of the Crural Flexors

with these muscles; the evidence for this statement will be presented
later. In the opossum it is a well-developed muscle forming what has
been termed by Coues, 72, the flexor brevis pollicis obliquus. Leche, 98,
however, records it as wanting in the marsupials and Cunningham, 81,
remarks that the muscle is wanting in Thylacinus and Dasyurus.
Dobson, 83, on the other hand, finds in Dasyurus a band passing from the
os calcis to the under surface of the flexor fibularis tendon and identi-
fies it, probably correctly, with the flexor accessorius and Young, 82,
in his account of the musculature of Phascolarctos, while stating that
“there is no flexor accessorius in the foot,” goes on to say that a muscu-
lar bundle which arises from the os calcis and passes to a fibro-cartilagi-
nous backward prolongation of the plantar fascia is regarded by
Macalister as similar to it in its nature. McCormick, 87, suggests the
identity of one of the heads of his flexor brevis digitorum with the flexor
accessorius in Dasyurus viverrinus, but the brevity of his description of
this head and the absence of explanations of his figures prevent an
opinion as to the correctness of the suggestion. There cannot be the
slightest question as to the existence of the muscle in Didelphys virgin-
vana, and on account of its importance in the fundamertal plan of the
crural muscles, to be discussed later, it seems quite probable that it may
be found in a rudimentary condition in the majority of the marsupials.

In the opossum it arises from the outer surface of the os calcis as a
distinct bundle of fibers which are directed inwards and distally. They
early pass over into a tendon which crosses the plantar surface of the
tendon of the flexor fibularis and unites with its outer border, that
portion with which it unites immediately separating to form the tendon
for the hallux. This description differs somewhat from that of Coues,
72, who regards the tendon of the hallux as representing the direct
continuation of the accessorius. Sections show very clearly, however,
that this is not the true state of affairs and that the arrangement is as
described above. In the cat the accessorius is a strong muscle of con-
siderable size, arising from the outer surface of the os calcis. Its thin
tendon passes obliquely across the tendon of the flexor fibularis and unites
with the greater part of its plantar surface, including the united flexor
tibialis tendon. In the mouse it is also well-developed, arising from the
outer surface of the os calcis and passing obliquely to the tendon of the
flexor fibularis, especially to that portion of it which becomes the long
flexor tendon of the hallux.

There still remains for consideration the muscle which has been
termed the pronator tibie in the monotremes and marsupials and in
the higher mammals the popliteus, assuming for the present that the
two muscles are identical.

J. Playfair MceMurrich 59

In the opossum the pronator tibie is a muscular sheet which extends
obliquely from the fibula to the tibia throughout the greater part of the
Jength of those bones. It takes its origin partly from the inner border
of the fibula, but mainly from the strong aponeurosis which separates
ii from the tibialis posticus and the flexor digitorum tibialis above and
the flexor digitorum fibularis below. In its upper part the fibers are
directed very obliquely, indeed, almost directly tibia-wards, to the upper
part of the tibia, and in this upper portion the muscle is composed of
two fairly distinct sheets of fibers, one lying anterior to the other and
separated from it by a distinct layer of areolar tissue. Below (Fig.
%, PT), however, there is no such separation of two layers, and the fibers
have a more vertical course. The partial separation above, already
noted by Young, 81, is apparently of “ prophetic” interest in fore-

Fic. 12.—Transverse section through the crus of the Mouse just below the knee
joint. F, fibula; Ge, gastrocnemius lateralis; Gi, gastrocnemius medialis; P! and P?,
oblique and vertical portions of popliteus; Pl, plantaris; 7', tibia.

shadowing the differentiation of the muscle into an upper or popliteal
portion and a lower or pronator tibial portion.

In the mouse the popliteus arises from a strong fibro-cartilaginous
band attached above to the outer condyle of the femur. Those fibers
which arise from the tibial side of the band (Fig. 12, P’) have a much
more oblique direction than the rest (P*) and are inserted into the tibia
above them. No distinct indications could be discovered of a repre-
sentative of the pronator tibiw, t.¢., a lower portion of the muscle,
although it is possible that some scattered fibers which lie anterior to
the main mass of the fiexor digitorum fibularis and have an oblique
direction, may represent it. A separation between these fibers and the
flexor was, however, at best indistinct.

In the cat the popliteus takes its origin from a sesamoid bone which
is attached by a strong tendon to the outer condyle of the femur. The

60 The Phylogeny of the Crural Flexors

muscle passes obliquely downward and inward over the knee joint
(Fig. 9) and shows quite distinctly a composition from two masses of
fibers, one of which (P'), as in the mouse, has an oblique direction,
while the other (P?) is more vertical. No indications of a lower portion
of the muscle could be found in the individual I studied, although it
may be noted that both in the cat and in the mouse the interosseous
membrane is more strongly developed than in the oppossum.

IV. THe NERVES OF THE MAMMALIAN CRUS.

In the opossum, at the level where my sections began, there was a
main nerve stem, the internal popliteal (Fig. 13), and on one side of
it a stem for the internal gastrocnemius (GJ) and on the other side

GE

IP
Fria. 183. Fig. 14.

Fic. 13.—Diagram of the crural nerves of Didelphys virginiana. EP, external
plantar; FF, branch to flexor fibularis; F7, branch to flexor tibialis; GH, branch to
gastrocnemius lateralis; GZ, branch to gastrocnemius medialis; JP, internal plantar ;
pl, branch to plantaris; P7, branch to pronator tibie; 7P, branch to tibialis posticus.

Fic. 14.—Digram of the crural nerves of the Mouse. HP, external plantar; FF,
branch to flexor fibularis; FZ’, branch to flexor tibialis; Ge, branch to gastrocnemius
lateralis; Gi, branch to gastrocnemius medialis; JP, internal plantar; P, branches to
popliteus ; pl, branch to plantaris; s, branch to soleus; 7'P, branch to tibialis posticus ;
tp, posterior tibial. ~

a branch for the external gastrocnemius (and soleus) (GH), and more
externally the external popliteal which wound around the head of the
fibula to the front of the leg. The internal popliteal descended into
the crus between the two gastrocnemil, and soon after divided into five
branches, of which two (#P and JP) were quite large, two others were
much smaller, one of them (pl) passing exclusively to the plantaris,
while the other one (//’) was distributed to the flexor digitorum fibu-
laris.

The fifth branch (Fig. 7%, rp) was of moderate size and passed

J. Playfair MeMurrich 61

obliquely inwards, giving off branches to the tibialis posticus (7’P), the
pronator tibie (P7') and the flexor digitorum tibialis (#7’), and then
continued its course downward between the tibialis posticus and the
flexor digitorum fibularis, without supplying either, and terminated near
the ankle joint, apparently in the periosteum of the lower part of the
tibia. The lowest of the branches which passed to the pronator tibie
could be traced downwards in the muscle almost to the ankle joint and
seemed to end there in periosteal branches to the lower end of the fibula.

The branch to the plantaris (pl) might be described as arising from
the internal plantar nerve (JP) but with this exception neither the inter-
nal nor the external (/P) plantar nerves takes any part in the innerva-
tion of the muscles of the crus. After they have passed into the foot
the external plantar gives off a branch, which passes mainly to the
abductor minimi digiti, but also gives two twigs to the flexor digitorum
accessori1us.

In the mouse (Fig. 14) the internal gastrocnemius is supplied by a
branch (GI) given off above the level of my highest section. The inter-
nal popliteal descends between the two gastrocnemii, gives off branches
to the gastrocnemius externus (GH) and to the popliteus (P), and
Givides, opposite the knee joint, into three branches, two smaller ones
and one large one. The latter (tp) is comparable to the posterior tibial
nerve of human anatomy in many respects, although it takes no part in
the innervation of the crural muscles but descends unbranching to behind
the inner malleolus, where it divides into the external and internal
plantar nerves (HP and IP), the former sending a branch to the
flexor accessorius.

Of the two smaller branches, one (Fig. 14, P/’; Fig. 10, ff) shortly after
its formation gives branches to the plantaris (pl) and the soleus (s), but
passes mainly to the flexor fibularis. The other smaller branch (Fig. 10,
rp) gives off early in its course a branch to the popliteus (Fig. 14, P) and
probably supplies the fiexor tibialis also, although neither one nor the
other of my series of sections permitted of perfect certainty of this point
in this form. The branch then descends between the flexor fibularis and
the tibialis posticus, giving off a branch to the latter, and, passing more
deeply between the two muscles as it descends, finally rests upon the
interosseous membrane and seems to terminate in the periosteum of the
lower part of the fibula.

In the cat the arrangement of the nerves is in general the same as in
the mouse. A branch is given off from the sciatic, before its division,
to the internal gastrocnemius and another from the internal popliteal
soon after its formation passes to the external gastrocnemius. A little

62 The Phylogeny of the Crural Flexors

jater the internal popliteal gives branches to the popliteus and to the
plantaris, and shortly thereafter divides into two main trunks each of
which is composed of subordinate bundles. These two main trunks le
one behind the other (Fig. 9), and the posterior larger one (pt) descends
the leg without taking any part in the innervation of its muscles and
below the ankle divides into the external and internal plantar nerves.

The other trunk is clearly composed of two portions. From one of
these (rm) branches are distributed to the soleus and to the flexor
fibularis, while the other (7p) early divides into four branches, one of
which is distributed to the popliteus, another to the flexor tibialis,
a third to the tibialis posticus, while the fourth, which is very small,
passes downward in the aponeurosis between the tibialis posticus and
the flexor fibularis, gradually becoming smaller. I was not able to trace
this last nerve to its termination, but in all its relations it corresponds
to the branch to the periosteum in the mouse.

It may be recalled that in the lower vertebrates the nerves of the
fiexor surface of the crus were divisible into superficial and deep branches,
and that of the former there were two main trunks, one of which, the
ramus superficialis medialis, was entirely devoted, so far as its muscular
branches were concerned, to the supply of the plantaris superficialis and
the plantares profundi III and II. The other superficial trunk, the
ramus superficialis fibularis, on the contrary, passed downward, supply-
ing only the fibulo-tarsalis in the amphibia, and became the external
plantar nerve. The deep branch, the ramus profundus, was distributed
to the plantaris profundi I and the interosseus, and then was continued
into the foot to form the internal plantar nerve.

Comparing with this the arrangement described above for the opos-
sum, considerable similarity will be noticed. Thus descending the
entire length of the crus there are two nerves, the external and internal
plantar, the former of which has practically identical relations with the
ramus superficialis fibularis of the lacertilia. In addition there is given
off from the internal popliteal at or shghtly above its division into the
two plantar nerves a smaller stem which supplies the deep muscles of the
crus and is continued down to the ankle joint as an exceedingly fine
nerve, which is not, however, continued into the foot. Im its topo-
graphical relations and in its crural muscular distribution this nerve
seems to be the homologue of the reptilian ramus profundus, from
which, however, it differs in being limited in its distribution to the crus.

In my study of the nerves of the antibrachium (MeMurrich, 03) it was
shown that the ramus profundus of the amphibia and reptilia extended
into the manus, supplying in general the radial part of its palmar

J. Playfair McMurrich 63

surface, but that in the mammalia its palmar fibers became associated
with the median nerve, its antibrachial portion persisting as the anterior
interosseous nerve. Apparently a somewhat similar process has taken
place in the crus. The tibial plantar fibers have separated themselves
from the ramus profundus and have taken a more superficial course to
form, in the opossum, the internal plantar nerve, though it can hardly
be said that they have united with ramus superficialis medialis, which is
represented by the branches to the plantaris soleus and flexor fibularis,
together with the branches given off higher up to the two gastrocnemil.

The condition in the opossum does not, however, complete the re-
arrangement which is characteristic of the mammalia as a group, a
further modification consisting in the union of the internal plantar fibers
of the marsupial with the ramus superficialis fibularis (external plantar)
to form the posterior tibial nerve. It is noteworthy, however, that even
although this fused stem appears to be the prolongation of the internal
popliteal, yet, in the mouse and cat, the ramus profundus arises from
_it at the knee joint and that in these forms it is proper to describe the
internal popliteal as dividing into the ramus profundus and the posterior
tibial, notwithstanding the discrepancy in the sizes of the two nerves.
Furthermore the branches for the superficial muscles arise high up,
some of them from the internal popliteal before it branches, while others
may arise either at the point of bifurcation or even from the upper part
of the ramus profundus.

Finally, it may be added, that in man a further modification occurs
in the inclusion in the posterior tibial of certain of the fibers of both the
ramus superficialis medialis and the ramus profundus, namely, of the
former branches to the soleus and to the flexor fibularis, and of the latter
a branch to the tibialis posticus and that to the flexor tibialis. Indica-
tions of the original conditions are, however, still to be seen in the origin
in the popliteal space of a nerve which sends a branch to the popliteus
and another to the tibialis posticus and is then continued down the crus,
partly in the substance of the interosseus membrane, to end in the neigh-
borhood of the ankle joint. This nerve, whose terminal prolongation
down the crus was first thoroughly described by Halbertsma, 47, as the
n. interosseus cruris, is very evidently equivalent in its topographical
relations to the ramus profundus of the lower forms, although some of its
fibers destined for the tibialis posticus have separated from it and have
joined the posterior tibial nerve. It therefore represents one of the
primary branches of the internal popliteal and is deserving of more special
mention than is accorded to it in the text-books of human anatomy.

There occurs then in the vertebrate series a progressive modification

64 The Phylogeny of the Crural Flexors

of the paths followed by the nerve fibers which supply the flexor muscles
of the crus. Stated in general it consists of (1) a separation of the
fibers destined for the internal plantar region from the ramus profundus
and their assumption of a more superficial course, a process which occurs
also in the antibrachium; (2) a breaking up of the ramus superficialis
medialis into a number of branches which arise independently; (3) the
union of the internal and external plantar nerves to form the posterior
tibial; and, finally, the association of some fibers of both the ramus
superficialis medialis and the ramus profundus with the posterior tibial.
Taking the reptilian arrangement for a starting point, the rearrange-
ment as it is shown in the opossum, in the mammalia in general as repre-
sented by the cat and the mouse, and in man may be schematized thus :—

Lacertilia. Opossum. Cat and Mouse. Man.
Ramus superf. med- Branches to the super- Branches to the super- Branches to gastrocnemii,
ialis. ficial muscles* ticial muscles* soleus and plantaris.
~S07 or,
~—
Cu 2eg,

Ramus superf. fibu- )

laris.

External plantar____ 77)
> Posterior Tibial. Posterior Tibial.
¢ Internal plantar — ES
ogo
ee 2°
| a <> a we
As 2) Che a co
Ramus profundus. 4 ¥ yor
Branches to the deep Branches to the deep Branches to popliteus,
muscles and perios- muscles and perios- tibialis posticus and the
L teum. f teum.t n. interosseus cruris.

* By the superficial muscles are here meant the gastrocnemii, plantaris, soleus and flexor fibularis.

t+ By the deep muscles are here meant the flexor tibialis, tibialis posticus and pronator tibiz (popliteus).

V. THE HoMOLOGIES oF THE MAMMALIAN CRURAL FLEXORS.

Having now described the arrangement of the muscles in the three
vertebrate groups selected for study and having also elucidated the
modifications presented by the primary nerve stems, we are in a position
to determine the homologies of the mammalian muscles with those of
the lower forms. A comparison of the lacertilian and amphibian muscles
has already been made and the comparison now to be drawn might be
principally with the lacertilia, were it not that it will be necessary in the
following pages to make frequent reference to the conclusions of Hisler,
95, who deduces the mammalian arrangement directly from the amphi-
bian, neglecting altogether the reptilian. It will be convenient to con-
sider the various muscles in succession and to take the arrangement seen
in man as a type.

1. The gastrocnemius of man is formed by the union of two heads,

J. Playfair McMurrich 65

one from the external and the other from the internal condyle, and it
unites with the soleus to form the tendo Achillis, inserting into the
os calcis. Disregarding the soleus for the present, there are two possi-
bilities to be considered with reference to the double origin of the
gastrocnemius; either (1) it represents two originally distinct muscles
which have united below, or (2) it represents the splitting in its upper
part of an originally single muscle. The second of these possibilities
may be dismissed on the ground that in the lower mammals the two
heads, as a rule, remain distinct throughout their entire length. Lisler,
im accepting the view that the two heads are primarily distinct muscles,
takes the ground that one or the other of them has undergone an exten-
sive migration, basing this conclusion upon the crossing of the two
tendons which occurs shortly above their insertion, a peculiarity which
has been considered in detail by Parsons, 94. The crossing, considered
by itself, throws little light upon the question as to which muscle has
undergone the supposed migration and Hisler, turning for evidence to
the nerve supply, finds that Cunningham, 81, has observed in Phal-
angista maculata that the gastrocnemius medialis is supplied from the
external saphenous (sural) nerve, which has a markedly fibular position
and he concludes therefore that it is the gastrocnemius medialis which
has migrated and that primarily it had its origin from the fibula and
lay to the fibular side of the gastrocnemius lateralis, in which case
there would be no crossing of the tendons.

The argument by which such a remarkable migration is deduced is
open to criticism along several lines. In the first place the crossing
of the tendons does not necessarily imply a migration of the muscles.
It may be difficult to give a satisfactory explanation of it on another
basis, and the migration theory, if correct, would certainly explain it,
but it may be pointed out that the same crossing occurs also in the
tendons of the flexor fibularis and the flexor tibialis digitorum in man,
and yet a reversal of the relative position of the two muscles by
migration seems altogether improbable. A theory which explains the
one crossing will probably also explain the other, for, it may be noted,
the tendons of the gastrocnemius and plantaris represent a superficial
layer of the plantar aponeurosis into which both muscles primarily
insert, while those of the two long digital flexors represent a deep layer
of the same aponeurosis. The most probable factor in the production
cf the crossing is a physiological rather than a morphological one, a point
which will be considered Jater on in connection with the discussion of
the flexor tibialis digitorum.

In the second place it would seem that Hisler has placed too much

66 The Phylogeny of the Crural Flexors

stress upon the supply of the gastrocnemius medialis by a branch of the
external saphenous nerve. I have not been able to trace the origin of
the nerve in the opossum, but one must conclude from Cunningham’s
statement, 81, that in the thylacine the nerve arises from the internal
popliteal. In these two forms then, the thylacine and Phalangista, two
different origins of the nerve occur, one of which favors Hisler’s migra-
tion theory while the other is opposed to it. Which is the more primi-
tive origin? I have not been able to find in the literature accessible to
me any sufficiently detailed accounts of the arrangement of the nerves
in other marsupials or in the monotremes, but, since there can be no
question as to the identity of the lacertilian muscle termed above the
plantaris superficialis medialis with the mammalian gastrocnemius inter-
nus, the origin of its nerve fibers may throw some light on the question.
In Scineus it is supphed by a branch from the ramus superficialis medi-
alis, 7. e., from the more medial of the two superficial nerve trunks and
according to Gadow, 82, this is the usual condition in the lacertilia which
he studied, in Ophryoessa only does the branch come from the ramus
superficialis fibularis. In the crocodiles the muscle is supplied by a
branch from the ramus profundus and a weak branch from the super-
ficialis medialis, while in the alligator it receives branches from both
superficial nerves, that from the fibularis being the smaller. It seems,
therefore, that there is a considerable amount of variation in the course
of the nerve fibers in question, a fact which weakens an argument based
solely on the path followed by a group of nerve fibers in a single species
of mammal.

It seems to me that the muscle in question is primarily and finally
a muscle of the tibial side of the crus, and that its homologue in that
position can be found from the urodele amphibia to the highest mam-
malia. lLisler, as has already been pointed out, has failed to recognize
the true plantaris superficialis of the amphibia and has thus been led
widely astray in his attempts to homologize the amphibian and mam-
malian muscles. He finds the amphibian homologue of the gastroc-
nemius medialis in the fibulo-tarsalis (fibulo-plantaris, Eisler) and that
of the gastrocnemius lateralis in the plantaris profundus III minor
(plantaris superficialis minor, Eisler). It may be pointed out that
both these muscles lie beneath the plantaris profundus III (plantaris
superficialis major, Eisler) which Eisler identifies with the mammalian
plantaris. This latter muscle, however, wherever it can be certainly
identified, is in relation with the deeper portion of the gastrocnemius
lateralis and would seem to be a derivative of the deeper portion of
that muscle. Lisler’s identifications would accordingly require an in-

J. Playfair MceMurrich 67

version of the deeper and more superficial muscles, his fibulo-plantaris
and plantaris superficialis minor coming to le on a plane posterior to his
plantaris superficialis major. Such a transposition can only be
accepted on the strongest evidence, and of fare it seems to me, there is a
failure. ;

Finally, as was pointed out in considering the antibrachial flexors,
any theory which requires the migration of a muscle origin over a joint
from below demands the closest scrutiny. Eisler’s homologies make the
gastrocnemius medialis have its origin primarily from the head of the
fibula, and to reach the position it has acquired in the lacertilia and
mammalia it must have migrated upwards over the knee joint as well
as medially. If a plausible homology can be set forth which does not
require this migration, the presumption is in its favor. In the forearm
it was shown that the palmaris superficialis layer was distinguished from
the other fiexor layers by having its origin from the humerus, and that
throughout the whole series of forms studied it retains that origin. The
remarkable similarity which obtains between the amphibian antibrach-
ium and crus leads to the expectation that in all probability the homo-
logue of the superficial palmar layer will have the same relations, and the
identification of the plantares superficiales medialis and lateralis with
the gastrocnemii exactly fulfills the expectation.

The conclusions to which I have been led, then, are that the gastroc-
nemius medialis and lateralis of the mammalia are primarily separate
muscles which insert into the superficial layer of the plantar aponeu-
rosis, and that they represent the greater part of the superficial plantar
layer of the amphibian crus, the gastrocnemius medialis corresponding
to the plantaris superficialis medialis of both amphibia and lacertilia
and the gastrocnemius lateralis to a portion of the amphibian plantaris
superficialis lateralis and to the lacertilian muscle similarly named.

The plantaris —There can be little doubt but that the plantaris is
a derivative of the same muscle mass which gives rise to the gastroc-
nemius lateralis, or, to be more precise, that it represents the deeper
medial portion of that mass. For it is typically associated with the
gastrocnemius lateralis and is frequently united with that muscle in its
upper part, occupying then the position indicated. It is already a dis-
tinct muscle in the reptilia, at least the muscle described above as the
plantaris superficialis accessorius seems to be its homologue, although
the relations which this muscle bears to the plantaris profundus III
seems at first sight to preclude any such homology. But it must be
remembered that after all the association is not directly with the pro-
fundus III, but with the plantar aponeurosis into which the profundus
IIT also inserts.

68 The Phylogeny of the Crural Flexors

As a result of the difference in the views of Eisler and myself regard-
ing the amphibian homologues of the gastrocnemii, a difference also exists
as to the homologue of the plantaris. Eisler finds it in his plantaris
superficialis major, a muscle which, so far as its greater part is concerned,
is fibular in origin and has been termed above the plantaris superficialis
III. Acceptance of Hisler’s homology would again require the migra-
tion of a muscle from below over the knee joint and, furthermore, as
has already been pointed out, a transition of the planes occupied by the
plantaris and the gastrocnemius, both of which phenomena the homo-
logy which I have deduced avoids.

It may be added (1) that the primary connection of the plantaris
below is with the plantar aponeurosis, its insertion into the os calcis in
man being a secondary condition, and (2) that its frequent absence is
probably more correctly to be regarded as a failure to separate from the
gastrocnemius lateralis, in connection with which idea its not unfrequent
union with the tendo Achillis is of significance.

The soleus, the third element in the triceps sure, is a muscle at first
sight apparently peculiar to the mammalia, and among these is possibly
unrepresented as a distinct muscle in the monotremes.” It has been
described as lacking in a number of mammals, in such cases being
probably included in the gastrocnemius lateralis. It has character-
istically an origin from the fibula and this points strongly to its being
a representative of the plantaris profundus group of muscles. The con-
ditions in the lacertilia throw little light upon the question, but it is
io be noted that the two superficial profundus layers are fused together
in these forms. They are, however, clearly distinguishable in the amphi-
bia and it is possible that they again become separated in the mammalia,
a series of modifications similar to those which occur in the antibrachial
flexors taking place. If this supposition be correct then it seems prob-
able that the soleus represents the plantaris profundus III of ‘the
amphibia. The forms which I have studied do not furnish sufficient
data for certainty as to this homology, but it seems to be the only one
consonant with the facts at our disposal. Possibly a renewed study of
the monotreme crus with this idea in mind may yield some light. Kisler,
it may be added, regards the soleus as a derivative of the gastrocnemius
lateralis.

The flexor fibularis and the flexor tibialis are so closely associated that
at first one would have little hesitation in assigning them to a common

°'The erroneous application of the term soleus to the muscle which arises
from the epiphysial process’ of the fibula has already been noted.

J. Playfair MeMurrich 69

origin, the two muscles standing to one another in much the same
relation as the flexor profundus digitorum and the flexor longus hallu-
cis of man. Their nerve supply is altogether different, however, since
the flexor fibularis is supplied by the equivalent of the ramus superfi-
cialis medialis, while the flexor tibialis is supplied from the ramus
profundus. ‘There seems to be no good reason why this should be so if
the two muscles belong to the same original layer, and one is forced to
the conclusion that they have their origin from quite different layers.
In the amphibia and lacertilia it has been shown that the plantares
profundi III and IT are supplied from the ramus superficialis medialis,
while the plantaris profundus I and the interosseus are supplied from
the ramus profundus. ‘The flexor fibularis, accordingly, probably repre-
sents the plantaris profundus II, if the soleus be regarded as equiva-
lent to the plantaris profundus III, while the flexor tibialis probably
represents in part the plantaris profundus I. In other words the flexor
tibialis is a muscle derived from the same primary layer as the tibialis
posticus and is quite distinct from the flexor fibularis.

This view may seem improbable on account of the close relation of the
two muscles in their lower portions and on account of the distinctness
of the tibialis posticus, but it must be remembered that the primary
insertion of a considerable portion of the plantaris profundus I is
probably into the plantar aponeurosis and that in the lacertilia it is in
part united to the sesamoid cartilage developed in the tendon of the
plantaris superficialis III-II. It is this aponeurotiec portion of the pro-
fundus I which becomes the flexor tibialis, while the remainder of it
constitutes the tibialis posticus, and, as will be shown later, the flexor
accessorius digitorum pedis.

Hisler finds the homologue of the flexor tibialis in his plantaris
profundus IT and that of the flexor fibularis in his plantaris profundus I,
thus coinciding with the opinion expressed above that the muscles
belong to different primary layers and also with the identification of the
two muscles, since the muscle he names the plantaris profundus I
is identical with that which I have called the plantaris profundus II and
that which he calls the plantaris profundus II is a part of my plantaris
profundus I.

The remarkable transference of the action of the flexor fibularis from
the fibular digits to the great toe which occurs in the mammalian series
has received its most plausible explanation from Keith, 94, on functional
grounds. It remains to especially emphasize in connection with his
argument the primary insertion of both muscles into the deeper layers
of the plantar aponeurosis, the different arrangements of the tendons

70 The Phylogeny of the Crural Flexors

of the two muscles being but various differentiations, due to differences
of strain, of an originally single aponeurosis and not a secondary fusion
of distinct structures.

The significance of the ¢ibialis posticus has already been indicated ;
from its relations and nerve supply it seems unquestionably a derivative
of the plantaris profundus I, a view not at variance with that of Hisler,
when allowance is made for the differences in our terminologies. Another
derivative of the same layer is the flexor accessorius digitorum pedis
(quadratus plante), which represents a portion of the layer which takes
its origin from the tarsal bones and is inserted into the plantar aponeu-
rosis. ‘The muscle certainly finds no place in the general plan of the
plantar muscles and is clearly represented in the lacertilia, where it is
supplied by a branch from the ramus profundus. Its supply from the
external plantar nerve in the mammalia is readily explained on the basis
of the separation of the plantar fibers from the ramus profundus to form
a special more superficial nerve stem and to subsequently unite with the
external plantar fibers to form the posterior tibial nerve, as has already
been described. Its relations to the tendon of the long digital flexor is
clearly a persistence of its original insertion into the deep layer of the
plantar aponeurosis.

Finally as regards the popliteus, the most usually propounded homology
is with the uppermost portion of the interosseus muscle, and, in truth,
at first sight this seems to be a most plausible suggestion. ‘There are,
however, some difficulties in its way, one of the most important being
its origin from the external condyle of the femur and another that in
some forms it covers in, 7. ¢., lies posterior to the upper portions of the
flexor tibialis and the tibialis posticus. On the other hand, its constant
supply from the ramus profundus seems to imply in almost unmistakable
terms its derivation from either the interosseus or the plantaris pro-
fundus I, and of the two the interosseus seems to be its most likely
origin.

Hisler, though, recognizing a possibility of referring it to the inter-
osseus, finally concludes that it is not properly a crural muscle at all
in the sense in which the term crural is used here, but that it is a femoral
muscle and the equivalent of the brachialis anticus of the arm. It is
difficult to see how such an homology can be worked out in its details.
It would imply that the muscle is a derivative of one of the femoral
flexors, most presumably of the biceps or better of such a muscle as the
eruro-coceygeus of the opossum, which sends a slip obliquely across the
thigh to be inserted into the shaft of the tibia. It is to be noted, how-
ever, that this slip passes superficially to the upper part of the gastroc-
nemius, while the popliteus passes beneath, 7. e., anterior to that muscle.

J. Playfair MeMurrich a

The opossum has no muscle which corresponds exactly to the popliteus.
It has the homologue of the interosseus well developed as the pronator
tibiew, but that muscle is entirely confined in its origin to the fibula,
even its uppermost portion which has been homologized with the popliteus
arising from that bone. It is only in the higher forms that a true
popliteus is found and certain peculiarities in its structure in the
mouse and eat seem to throw some light upon its significance. In both
these forms, as has already been noted, two very distinct portions can
be discerned in the muscle, a more tibial portion whose fibers have a
very oblique direction and a more fibular portion whose fibers are more
nearly vertical. A distinct line of demarcation between the two parts
occurs in any transverse section of the upper part of the crus. Further-
more, the muscle receives two nerves, a fact which in so small a muscle
is in itself noteworthy, and is all the more significant in that one of these
nerves arises, in the mouse for instance, with that for the soleus from
the internal popliteal stem, while the other arises from the branch which
I have identified with the crural portion of the ramus profundus of the
lower vertebrates. And, finally, the internal popliteal branch is suppled
entirely to the more tibial oblique-fibred portion of the muscle, while
that from the profundus passes entirely to the more fibular vertical
portion.

The significance of these facts seems to be evident. The popliteus
is a compound muscle, consisting of a portion derived from the plantaris
superficialis and a portion which represents a part of the pronator
tibie of the marsupials and the interosseus of the lower vertebrates. In
other words the constitution of the mammalian popliteus is exactly equiva-
lent to that of the pronator radii teres in the arm.

The idea that the muscle is a composite one furnishes a simple explana-
tion of the condition occurring in some carnivores. Gruber, 78, has
shown that in the dog, wolf and fox there exists, independently of the
popliteus and lying to a certain extent beneath it, a short muscle extend-
ing between the upper portions of the fibula and tibia. This is the
m. peroneo-tibialis. The same muscle occurs also in Viverra (Dobson,
83), and as an anomaly in man (Gruber, 77 and 78). The fact of the
occurrence of such a muscle in certain carnivores while lacking in others
is certainly reasonably accouted for on the supposition that its absence
in the latter is only an apparent one. That is to say, it seems probable
that the peroneo-tibialis of the dog represents the more vertical portion
of the popliteus of the cat, the dog’s popliteus being, equivalent to the
obliquely fibered portion of the cat’s muscle. And similarly, the appear-

ance of the peroneo-tibialis as an anomaly in man may readily be ex-
6

72 The Phylogeny of the Crural Flexors

plained on the ground of a separation of the profundus portion of the
popliteus from the superficialis portion.

In the opossum the upper partially separated portion of the pronator
tibize is very probably the equivalent of the peroneo-tibialis element, but
what may be the representative of the superficialis portion of the pop-
liteus it is difficult to say. A possible degenerated representative of it
may be found in a strong tendon-like band which extends obliquely across
the knee joint from the external fabellar cartilage to the head of the tibia,
but such an identification can be at present merely a suggestion. More
important, perhaps, are the relations which seem to exist between the
plantaris and the popliteus as shown by anomalies in man, the popliteus
having occasionally an accessory head which often coincides with the
absence of the plantaris.

If the tendon mentioned above as occurring in the opossum really
prove to be the representative in that animal of the superficialis portion
of the popliteus, then there should be some muscular representative
of it in lower forms. This may be found in that portion of the mono-
treme popliteus which arises from the epiphysial process of the fibula,
and in the reptilia it may have a representative in the plantaris super-
ficialis tenuis, although this seems at present very questionable. A study
of a greater number of forms than I have had at my disposal will be
necessary to trace out all the homologies of the popliteus, but I believe
that the observations here recorded make the supposition as to the com-
posite nature of the popliteus exceedingly probable.

In conclusion a few words may be said with regard to the modifica-
tions and homologies of the plantar aponeurosis throughout the series.
In the urodele amphibia it is represented in the crus by the aponeurosis
which covers the posterior surface of the plantaris profundus III and
by the tendons of the plantares superficiales medialis and lateralis. It
receives, therefore the insertions of these three muscles, together with
that of the profundus II and a part of that of the profundus I, and
gives origin to the superficial muscles of the plantar surface of the foot.
With the increase in size of the lateral and medial portions of the plan-
taris superficialis, a portion of the superficial layer of the aponeurosis
becomes separated to form the tendon of those muscles, while the rest
of it is covered in by them and remains included in the tendon of the
plantaris profundus III-II, part of it giving insertion to the plantares
superficiales accessorius and tenuis.

This is the reptilian condition, and the transition from it to that of
the mammalha is comparatively simple. The superficial layer of the
aponeurosis in the mammalia is represented by (1) the tendon or tendons

Plantaris prof. III.
Plantaris prof. IT.

Plantaris prof. I.

Interosseus.

J. Playfair MeMurrich 73

of the triceps sure, (2) the tendon of the plantaris and (3) the plantar
aponeurosis of the foot, a portion of it (4), however, remaining included
with the deep layer in the tendons of the flexores fibularis and tibialis.
It is this fourth portion of the superficial layer which gives origin to
the flexor brevis digitorum in those forms in which that muscle arises
from the tendons of the long flexors.

The homologies of the crural muscles traced out in the preceding
pages may be tabulated thus:

Amphibia. Lacertilia. Opossum, Mammalia.
Plantaris sup. medialis. Plantaris sup. medialis. Gastrocnemius medialis. Gastrocnemius medialis. :
( Plantaris sup. lateralis. Gastrocnemius lateralis Gastrocnemius lateralis.
(less included soleus). i eee
Plantaris sup. lateralis. { Plantaris sup. accessorius. Plantaris. Plantaris.
| Blantarissup. tenuis. oo —.......... Popliteus (superficial
l portion) ?
(Soleus portion of gas- Soleus.
\ plantaris prof. III-Il. troc. lat.
y Flexor fibularis. Flexor fibularis.
; Flexor tibialis. Flexor tibialis.
aes eee Ne | Tibialis posticus. Tibialis posticus.
Plantaris prof. I access. Flexor accessorius. Flexor accessorius.
Interosseus. Pronator tibiz. Popliteus (peroneo-tibial
portion).
SUMMARY.

1. In the crus of the urodelous amphibia the flexor muscles are ar-
ranged in five layers, the superficial one arising in the femoral region,
the others, which have a more or less oblique direction, taking their
origin from the fibula and slightly from the tarsus. They are inserted
for the most part into the plantar aponeurosis, only the deepest layer
inserting into the tibia. Between the second and third layers is a slen-
der longitudinal muscle extending between the fibula and the tarsus.

2. The nerves of the flexor muscles of the amphibian crus are arranged
in two main trunks, a ramus superficialis and a ramus profundus. The
latter is continued into the pes as the internal plantar nerve. The former
divides into rami mediales which are confined to the crus and a ramus
fibularis which is continued into the pes as the external plantar nerve.
The rami superficiales mediales supply the first, second and third layers
of muscles, the ramus superficialis fibularis, the fibulo-tarsalis and the
ramus profundus the fourth and fifth layers.

3. A complete separation of the preaxial and postaxial nerve fibers does
not take place at the knee-joint in the amphibia, but the ramus pro-
fundus for a considerable portion of its course contains fibers which are
distributed to the prxaxial surface of the crus.

4. In the lacertilian crus the same muscle layers that occur in the

74 The Phylogeny of the Crural Flexors

amphibia are readily distinguishable. The superficial layer has increased
greatly in size and shows a differentiation into several muscles. The
second and third layers have fused and the fourth layer has differen-
tiated into two separate muscles. The fibulo-tarsalis has disappeared
and the muscles have in general a more vertical direction than in the
amphibia.

5. The arrangement of the nerve trunks in the lacertilian crus is essen-
tially the same as in the amphibia. The separation of the preeaxial and
postaxial fibers takes place, however, above the knee joint.

6. In the mammalia the same layers of muscles can be distinguished
although they have undergone greater differentiation into individual
muscles than in the lower forms.

%. The plantar fibers of the ramus profundus are separated in the mam- —
malia from the crural fibers and in the opossum form a more superficial
stem, the internal plantar, which traverses the crus without taking part
in its nerve supply. The other rami remain practically unaltered. In the
higher mammalia a further change takes place in that the ramus fibu-
laris (external plantar) and the internal plantar unite to form a single
stem, the posterior tibial, and, in man, some of the fibers belonging to
the ramus superficialis mediales and the ramus profundus become in-
cluded in this.

8. The superficial layer of muscles retains throughout its origin from
the femur and the deep layers theirs from the crural bones, with one
apparent exception. Furthermore the insertion into the plantar apo-
neurosis is largely retained, although some shifting to the bones occurs.

9. The soleus represents the second layer of muscles and its absence
in certain forms is probably, due to its inclusion in the gastrocnemius
lateralis.

10. The flexor fibularis and flexor tibialis belong to different layers,
the former representing the third layer, while the latter is formed from
a portion of the fourth layer, as is also the tibialis posticus.

11. The flexor accessorius digitorum (quadratus plant) is primarily
one of the crural muscles and represents another portion of the fourth
layer of muscles.

12. The popliteus is a compound muscle, being formed of a portion
from the superficial layer, united with a portion of the fifth layer.
The occasional occurrence of a distinct m. peroneo-tibialis in the higher
mammalia is probably due to a failure of the two portions to unite.

ANATOMICAL LABORATORY, UNIVERSITY OF MICHIGAN, July 26, 1904.

J. Playfair McMurrich 15

REFERENCES.

CUNNINGHAM, D. J., 81—Report on some points in the Anatomy of the Thy-
lacine (Thylacinus cynocephalus), Cuscus (Phalangista maculata) and
Phascogale (Phascogale calura), ete. Scient. Results, Voyage of
H.M.S. Challenger. Zool., V, 1881.

Cours, E., 72.—The Osteology and Myology of Didelphys virginiana. Mem.
Boston Soc. Nat. Hist., II, 1872.

Dosson, G. E., 83.—On the Homologies of the Long Flexor Muscles of the Feet
of Mammalia, etc. Journ. of Anat. and Phys., XVII, 1883.

EISLER, P., 95.—Die Homologie der Extremitaten. Abhandl. naturf. Gesellsch.,
Halle, XIX, 1895.

FURBRINGER, M., 7o.—Die Knochen und Muskeln der Extremitaten bei den
schlangenahnlichen Sauriern. Leipzig, 1870.

ELLENBERGER, W., and BAuM, H., 91.—Systematische und topographische Anat-
omie des Hundes. Berlin, 1891.

Gapow, H., 82.—Beitrage zur Myologie der hinteren Extremitaét der Reptilien.
Morphol. Jahrb., VII, 1882.

GRUBER, W., 77.—Ueber den neuen Musculus peroneo-tibialis beim Menschen.

Arch. fur Anat. u. Phys., Anat., Abth., 1877.

78.—Ueber den normalen Musculus peroneo-tjbialis bei den Hundes,
ete. Arch. fiir Anat. u. Phys., Anat. Abth., 1878.
78.—Nachtrage tiber den Musculus peroneo-tibialis. Arch. fiir Anat.
u. Phys., 1878.

HALBERTSMA, H. J., 47,—Ueber einen in der Membrana interossea des Unter-
schenkels verlaufenden Nerven. Miiller’s Arch., 1847.

HoFrFMANN, C. K., 73.—In Bronn’s Klassen u. Ordnungen des Thierreichs.
Bd. VI, Abth. 2, Amphibia, 1873-78.

Humpuey, G. M., 72—The Muscles and Nerves of Cryptobranchus Japonicus.
Journ. of Anat. and Phys., VI, 1872.

KEITH, A., 94.—Notes on a theory to account for the various arrangements of
the Flexor profundus digitorum in the Hand and Foot of Primates.
Journ. of Anat. and Phys., XXVIII, 1894.

LecHE, W., 98.—In Bronn’s Klassen und Ordnungen des Thierreichs. Bd. VI,
Abth. 5, Mammalia, 1874-1900.

McCormick, A., 87.—The Myology of the Limbs of Dasyurus viverrinus.
Journ. of Anat. and Phys., XXI, 1887.

McMourricu, J. P., 03.—The Phylogeny of the Forearm Flexors. Amer. Journ.
of Anat., II, 1903.

Parsons, F. G., 94.—On the Morphology of the Tendo Achillis. Journ. of
Anat. and Phys., XXVIII, 1894.

~

PERRIN, A., 93.—Contribution 4 l’étude de la myologie comparée. Membre
postérieur chez un certain nombre de Batraciens et Sauriens. Bull.
Scient., XXIV, 1894.

76 The Phylogeny of the Crural Flexors

I. Die Sehnenverbindung

Scnuuze, F. E., 66.—Myologische Untersuchungen.
Zeitschr. fiir wiss.

in der Planta des Menschen und der Saugethiere.
Zool., XVII, 1866.

Younea, A. H., 81.—The so-called Movements of Pronation and Supination in
the Hind-Limb of certain Marsupials. Journ. of Anat. and Phys., XV,
1881.

82.—The muscular Anatomy of the Koala (Phascolarctos cinereus).

Journ. of Anat. and Phys., XVI, 1882.

THE FRAMEWORK OF THE GLANDULA PARATHYROIDEA.

BY
JOSEPH MARSHALL FLINT, M.D.

From the Hearst Anatomical Laboratory of the University of California.

WitH 3 TExT FIGURES.

In studying the framework of the thyroid gland in man and the
higher mammals, the author was enabled at the same time to make cer-
tain observations concerning the supporting tissue of the glandula para-
thyroidea. These investigations were carried on by means of the de-
structive digestive methods through which the cytoplasmic elements are
all dissolved, leaving the resistant framework of the organ in the form of
an opaque skeleton, which reveals its original form and relationships,
and demonstrates clearly at the same time, the course of the various
interstitial processes in three dimensions of space. The details of the
method have already been published in another place.’ In both the
monkey and the dog the parathyroid bodies are situated within the general
capsule of the glandula thyroidea. Under normal conditions in both the
living organ and in fixed tissues, the small oval gland is scarcely elevated
from the surface of the larger organ in which it is contained. The
fasciculated capsule of the thyroid practically splits and embraces the
gl. parathyroidea which is oval in both longitudinal and transverse
dimensions. ‘Accordingly the capsule of the thyroid becomes the capsule
of the parathyroidea, with no essential differences in structure. Like the
capsules of most organs it is composed of laminated fasciculi of white
fibrous tissue, with a considerable amount of reticulum in its inner
surface. This capsule contains a small amount of elastic tissue, some of
which may accompany the larger septa that follow the greater vessels
into the substance of the gland. In piece digestions which have been
cut through the thyroid and parathyroid, the organ is clearly shown in
three dimensions. In both dog and monkey, the parathyroid is 24 mm.
broad, about 4 mm. long, and about 2 mm. in thickness. When viewed
with a stereoscopic microscope, the organ is seen just within the capsule

‘Flint: Bulletin of the Johns Hopkins Hospital, 1901; Arch. f. Anat. u.
Ent. Anat. Abth., 1903.

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

78 Framework of Glandula Parathyroidea

of the gl. Thyroidea where its finer structure and limiting envelope bring
it out in sharp contrast to the follicles of the thyroid that embrace it on
three sides. The little organ is oval in both transverse and logitudinal
planes giving it plasticly the form of a prolate spheroid. As shown by
this method the structure of the thyroid has been previously described *
and when the plane of section includes both organs, a glance is sufficient
to separate them owing to the marked differences in their structure. At
first sight in piece digestions, the parathyroid has a homogeneous, ground-
glass appearance without showing any very striking features excepting
the blood vessels that traverse its substance; but if the specimen is care-
fully studied with high oculars and rapid alterations in the quantity and

Fia. 1.—Piece Digestion of the Thyroid and Parathyroid of a Monkey. Extracted with
ether, digested with trypsin and cleared in glycerine. 19. The general form and arrange-
ment of the thyroid follicles are readily made out. Beneath the capsule and embraced by
its split lamineze is the parathyroid body. The large septaand vessels as well as the finer
septa can be seen on the surface, while the vessels are readily followed into the depths.

a= Finer septa of parathyroid. b=Blood-vessel and coarser septa. c=Capsule of parathy-
roid at point of splitting. d=Capsule of thyroid. e=Follicles of thyroid.

variety of light with which it is illuminated, delicate, fine septa on the
surface of the organ come into view. Owing, however, to their extreme
delicacy they are indistinctly shown and the picture accordingly is not as
instructive as one form those organs where the connective tissue is accu-
mulated into larger and more definite processes and septa.

Besides the more delicate septa that embrace the cell complexes of the
parathyroid we see the blood vessels which are always accompanied by
relatively thick connective tissue processes. As a rule these run in the

?Flint: The Johns Hopkins Hospital Bulletin, 1903.

Joseph Marshall Flint 79

central portions of the gland although instances are not uncommon
where they are found either in the capsule or its neighborhood indicating
their points of entrance and exit to and from the substance of the organ.
The quantity of the connective tissue diminishes with the order of rami-
fication until the small arteries, veins and capillaries are reached. These
raturally are found in the finer septa or trabecule about the cell columns.

To Ludwig and his pupils we owe the view that many organs are
divided into a series of similar structural units which have constant and
definite relations to connective tissue processes, blood vessels, nerves and
lymphatics. An organ is composed of a great many of such units which
are repeated again and again in its formation. Glands like the pancreas,
salivary gland, liver, and spleen express their structural relationships
excellently while others as for example the stomach and adrenal cannot

FIa. 2.—Section of block of tissue shown in Fig. 1.about 45 microns thick. X 37. Stained
with Aniline blue. Drawn over ablue print which was subsequently washed out. The sec-
tion shows the capsule of the parathyroid, the larger septa and blood-vessels as well as the
smaller septa limiting the cell columns. c=Capsule. s=Finer septa. d=Coarser septa.
a=Blood-vessel.

be subdivided at all. Accordingly when we know the finer structure of
one unit we know the structure of the whole organ with the exception of
the relation of these units to each other. In this sense, howéver, there
are no structural units in the parathyroid which bear a constant relation-
ship to connective tissue processes. The ultimate structural integers
must be looked upon as the cell columns or cell groups and the adjacent
fibrous tissue which supports them.

In thin digested sections the framework appears as irregular septa
which do not form a continuous network throughout the organ, but are
broken up into smaller processes which support the irregular coiled
columns of cells of which the organ is composed. These septa carry the
arteries, capillaries, veins, and nerves. They are in some places built up
of fasciculi of reticulum fibrils, in others, of a thinner, looser formation

80 Framework of Glandula Parathyroidea

of anastomosing and branching fibrils.) When thick, stained, digested
sections from 50 microns up are studied, these broken septa are obviously
continuous in the third dimension with other processes that turn off and
occupy various planes according to the branching of the anastomosing
cell columns. In this way in these preparations, especially under the
low power, it often appears as though the framework formed a continuum
stretching across the gland. In considering the structure and arrange-
ment of the framework in three dimensions, this is, of course, true, the
broken irregular septa appearing only in thin sections where the con-

Fia. 3.—Section of parathyroid of dog. Stained by Mallory’s method. X 62. c=Cap-
sules. s=Finer septa. a=Cell columns. by=Blood-vessel.

tinuity of the third dimension is broken. Under ordinary circumstances
the septa are comparatively fine and delicate. Relatively speaking, how-
ever, the framework is not abundant. Occasionally large septa project
inwards from the capsule, either in connection with or independent of
the blood vessels. The majority of the large vascular trunks, however,
as is seen in piece digestions also, are found in the center of the organ.
Around the adventitia, the framework is abundant and in these situations
large fasciculi of considerable dimensions are often found. When thin
sections are studied under the immersion lens, the framework can readily
be resolved into the ultimate constituent fibrils of which it is composed.
These fibrils branch and anastomose, and are of extreme delicacy. In

Joseph Marshall Flint 81

sections stained by the ordinary methods and thin sections varying from
3 to 6 microns in thickness, stained by Mallory’s connective tissue stain,
numerous cells with oval nuclei are found embedded in the fibrils. These
are the connective tissue corpuscles, and do not differ in this position
from those found in other parts of the body. In Mallory specimens
(Fig. 3) the relationship of parenchymatous cells to the connective tissue
processes are clearly shown. The cells are polygonal in shape, composed
of granular cytoplasm which stains readily with the acid dyes. These
cells contain spherical nuclei of medium size, possessing a well marked
nuclear membrane with a considerable amount of chromatin along the
linin filaments. They are packed together in irregular coiling and anas-
tomosing columns (Fig. 3, a) of varying size. In some instances as
many as seven or eight cells may be interposed between septa and blood-
vessels while in others only two or three are so placed. No connective
tissue fibrils pass in between the cells to form a finer framework. They
rest against each other, and are supported by the adjacent septa.

RESUME.

(1) In piece digestions, the gl. parathyroidea of the dog and monkey
is seen in the form of a prolate spheroid embraced by a capsule formed
through a splitting of the capsule of the thyroid gland. Within the
gland the larger connective tissue processes accompanying the blood-
vessels are easily seen usually in the central portion of the organ which
under the low powers of the stereoscopic microscope has a homogeneous
ground-glass appearance. Under the higher powers, however, the delicate
septa embracing the cell columns can just be made out.

(2) In thin stained digested specimens, the framework appears as
irregular broken septa composed of anastomosing and branching fibrils
as well as fasciculi or bundles of fibrils. These septa support the irregu-
lar anastomosing cell columns of which the gland is composed. In thick,
stained, digested specimens, however, septa can be followed in three
dimensions where they give almost the appearance of a closed network
owing to the change of direction as they follow the cell complexes of the
gland in the depths of the section.

(5) The relations of the cells to the connective tissue as shown in these
sections, indicates that the cell columns are supported by the septa.
Fibrils from the septa do not run in between the individual cells. The
cell columns are irregular in thickness, and anastomose with each other.
The smaller vessels are found in the smaller septa.

_
lee
?
j

‘

. = =

* p= See
* = r
1
i]
i

>

oes ee
»

a

a

PT

ao 2
a)
e
‘

“

~~ “ee, =:

~

4?
-

THE DEVELOPMENT OF THE CRANIAL AND SPINAL
NERVES IN THE OCCIPITAL REGION OF THE
HUMAN EMBRYO.

BY
GEORGE L. STREETER, M. D.

Instructor in Anatomy, Johns Hopkins University.

WitH + PLATES AND 14 TEXT FIGURES.

The following paper reports the results of a study of the morphology
of the ninth, tenth, eleventh, and twelfth cranial and the upper cervical
nerves, together with their ganglia, in a series of human embryos. It
includes a description of all the stages in the development of these
structures from the time at which they can first be definitely outlined
from the surrounding mesodermal tissue up to the time they have reached
adult conditions. This work was made possible through the kindness
of Prof. O. Hertwig, Prof. His, and Prof. Mall, who gave the writer
access to their valuable embryo collections for the purposes of this study.
For this courtesy the writer takes advantage of the present opportunity
to express his appreciation. Acknowledgment is also to be made to
Prof. Gage, whose Buxton embryo is included in the series studied.

The ultimate histogenesis of the nerve elements, a question which
has recently been thoroughly gone over by Harrison, 01, and Bardeen, 03,
will not be taken up. In the earliest stage where reconstruction was
possible the right and left divisions of the ganglion crest have migrated
ventro-laterally along the side of the neural tube, and are about to form
secondary attachments to it. Fibroblast formation is at this time well
under way, and peripheral fibre paths are beginning to become definite.
It is the consideration of the size, form, and relation of these paths and
the associated ganglion cell masses, in their different stages of growth,
toward which attention has been directed.

The results of this study are tabulated at the end of the paper; but
special mention should be made of the eleventh cranial nerve. In tracing
out its early history it becomes more than ever apparent that it is abso-
lutely similar and continuous with the tenth or vagus nerve. In the
embryo these exist, not as two independent cranial nerves, but rather
as parts of a single structure, each part possessing mixed motor and
AMERICAN JOURNAL or ANATOMY.—VOL. IV.

84 Development of Ovcipital Nerves in Human Embryos

sensory roots with root ganglia derived from the same ganglion crest.
As the development progresses the cranial end of this complex becomes
predominantly sensory and the caudal end predominantly motor, and
also more spread out which gives rise to a difference in the appearance
of the two portions in the adult, and has resulted in their being con-
sidered as two independent structures.

That such a relation between the tenth and eleventh cranial nerves
exists is not a new idea, but was long ago suggested by the work of His,
88, on the human embryo, though this investigator did not work with
sufficiently young stages to make the evidence conclusive. The theory
has since then been supported by the work of Furbringer, 97, and
Lubosch, 99, who believe that phylogenetically the tenth and eleventh
nerves cannot be separated. Chiarugt, 90, however, from the compara-_
tive embryology of these structures, concludes that the eleventh is not a
part of the tenth, but is a nerve for itself which results from the differen-
tiation of the nucleus of origin of the ventral roots into median and Jat-
eral divisions; the latter rootlets losing their segmental distribution take
a new course and depart obliquely through the cranium as an independent
nerve. Another view regarding these nerves is offered by Muinot, 92.
He suggests that a modification may have occurred in the motor fibres
of the dorsal roots of the hypoglossus, by which these motor fibres,
following the abortion of the ganglia, no longer join the ventral roots
of the twelfth, but turn forward to join the vagus thereby forming the
trunk of the accessory nerve. Since the work of His, 88, and Mail, 91,
on the human embryo, further details in the development of the tenth and
eleventh nerves in other mammals have been supplied in the well known
papers of Froriep, 82, 85, and 01, and by the work of Robinson, 92, and
Lewis, 03. The latter two give us a more accurate description of the .
so-called ganghonic commissure, than had before existed; although they
failed to recognize the full significance of these ganglia and their relation
to the precervical ganglia of Froriep.

The comparative morphology of the occipital nerves, particularly with
regard to their bearing on the segmental origin of the head, has been the
subject of much speculation, ever since Gegenbaur, 72, published his
work on the selachian head. The charge may perhaps be justly made
that more space in the literature is given to theories and discussions
concerning these structures than to actual observations on their com-
parative and embryological anatomy. This subject will be briefly treated
under the heading comparative morphology. It will be emphasized there
that the ganglia of the trunks of the ninth and tenth (gang. petrosum
and gang. nodosum) are branchio-meric and largely independent of the

George L. Streeter 85

ganglia of the roots. The latter gangla though not segmental more
closely resemble-the spinal ganglia; the attempt however to reduce the
cranial nerves to a spinal nerve type is deprecated. Some such hypo-
thesis as was long ago suggested by Balfowr, 76, is much easier of appli-
cation. This investigator supposed the head and trunk to have become
differentiated from each other when there was only a mixed motor and
sensory posterior root present and no ventral root, as was then supposed
to be the case in the amphioxus. Since then it has been found (Ransom
and Thompson, 86, Hatscheck, 92, Dogiel, 03) that the amphioxus and
cyclostomes have ventral, purely motor, roots. These do not arise from
the cord at the same level with, nor do they join, the dorsal roots. The
two might be spoken of as ventral and dorsal nerves. If then Balfour’s
hypothesis were modified to fit with our present knowledge, it could be
stated as follows: The head and trunk nerves were differentiated from
each other at a time when there existed mixed motor and sensory dorsal
roots and pure motor ventral roots. These ventral and dorsal roots did
not then arise from the neural tube at corresponding levels, and were inde-
pendent of each other. In the spinal region of higher vertebrates a modi-
fication has occurred, by which the dorsal and ventral roots have become
strictly segmentally arranged, and have joined in pairs, each pair forming
a common nerve, which is situated median to the myotome. In the head
region the nerves have retained the primitive type; the dorsal roots still
, contain a good proportion of motor fibres, and are situated beneath the
epidermis and outside of the myotomes. They are not segmentally
arranged and do not join with the ventral roots to form common nerves,
but form a system of separate ventral and dorsal (lateral) nerves.

MATERIAL AND Mrruops.—The elements of the peripheral nervous
system do not reach a degree of differentiation, which is sufficient for
reconstruction, until toward the end of the third week. From then
changes in the form and relation continue until the third month, when
the structures have practically reached the condition found in the adult,
and development may be considered as completed. The various stages
in their growth were thus found to be covered by the embryos listed in
the following table (page 86). Their ages have been determined by use
of Mall’s rule (Mall, 03), 7. e., the age in days equals the square root of
the product of the length times one hundred.

It was found that the ganglion masses and fibre paths could be satis-
factorily identified and traced by means of profile reconstructions. This
procedure was made use of with all embryos, with the exception of one at
a late stage which was large enough for dissection. The details adopted

86 Development of Occipital Nerves in Human Embryos

in its application consisted in making enlarged drawings of the sections,
usually fifty diameters, with a projection apparatus or camera lucida,
upon separate sheets of transparent paper. Paraffined wrapping paper
is serviceable and inexpensive, better than this, being stronger and more
transparent, is the “ process” paper, used in Germany as “ butter-brodt
papier” and in this country in packing tobacco. When the drawings
were completed, the sheets were piled so that adjacent sections were
accurately fitted over each other. A vertical line for reconstruction was
then established by marking upon‘ each sheet two lines perpendicular to
each other, forming a series of crosses which exactly superimposed
throughout the entire pile. The individual sections were then plotted
off on mm. paper by fitting the crosses to a chosen perpendicular line, the
distance between the sections being determined by the thickness of the.
sections and the enlargement of the drawings in the usual way. Many
of these reconstructions are diagrammatically shown as text figures

TABLE OF EMBRYOS STUDIED.

Length. Age. Source. No. Reconstruction.
4.0mm. 20 days. Hertwig Collection. 137 Profile, one side.
Aidit Ue Mall ss 148 be both sides.
G29 * 26, His Se Br3 oe BY be
TY); 9&2 Gb Se Mall es 2 uf one side.
MeO) <e 26per** Gage se Buxton be cs ne
IU). Se eo Mall eS 114 S es of
OMe 5 His a9 KO es both sides.
10:0" 3° 31 His *e D1 et one side.
IB Hy ts Bim es Hertwig es 67 & both sides.
144.0 “ Bie 5 Mall es 144 .- iS “ , wax plate.
Tey ee ay pss His te FM ss one side,
30.0 ‘ i ee Hertwig sé 161 ss sf a
65:0 Slr Mall st aac Dissection.

(Figs. 1-12), in which fibre masses are represented by lines and ganglion
cell masses by dots. The same enlargement is used in all of these so that
the actual increase in size may be readily seen by comparing them.

In order to reproduce the third dimension a clay model was made of
a 4.3 mm. embryo, and a wax plate reconstruction of a 14.0 mm. embryo.
Drawings of these are reproduced in Plates I and II. ;

Following the suggestion of Dr. Bardeen much assistance was obtained
in studying these structures by making dissections of pig embryos for
comparison. It was possible in this way to control the various stages
from 8.0 mm. upward. In making such a comparison it was found,
however, that the ratio between the length of the pig and the development
of its nervous system does not exactly correspond to that of the human.
The development of the neryous system of the latter is somewhat more

George L. Streeter 87

rapid; for example, a 14.0 mm. human embryo shows a stage found in
the 18.0 to 20.0 mm. pig embryo. The preparation of these dissections
was facilitated by the use of the following method: The entire embryo
is mordanted for several hours in a solution of potassium bichromate
2.5 per cent, and glacial acetic acid 10 per cent, after which it is rinsed
in water and brought into 80 per cent alcohol. The embryo, after a few
minutes dehydration in absolute alcohol, is then attached with a drop of
thick celloidin to an isinglass strip, previously coated with thin celloidin.
Mica or isinglass is used because it can be easily cut to any desired size.
By smoking it, before the celloidin coat, it is possible to write any desired
label on it, and the black serves as a good background for the embryo.
The whole is then hardened for a short time 80 per cent alcohol, and it is
then ready for dissection. In order to hold the embryo during dissection
the isinglass strip on which the embryo is mounted is clamped to a stage
which is made by fastening a glass slide with balsam to one of the facets
of a cut-glass polyhedral paper-weight. Such a stage is steady and can
be placed in any desired plane. With the embryo firmly mounted in this
way the dissection is made under alcohol with a binocular microscope.

DESCRIPTION OF HMBRYOS AT DIFFERENT STAGES.

Embryos of About Three Weeks.

HHeriwicsCollections=No. 1sd4° 3 oo) tess sees = 4.0 mm.
Mall Collleetioma NG, W481i Fo 50 sh ce Bate 4.3 mm.

(See Figs. 1, 2, 3, and Plate T.)

By the twentieth day the structures have become definite enough in
outline to permit of reconstruction. At this stage the ganglion crest of
the after-brain and spinal cord has divided longitudinally into right and
left halves, each of which has migrated ventro-laterally along the neural
tube and forms a flattened cellular band extending caudalwards from the
auditory vesicle along the lateral border of the tube to its extreme tip.

That part of the crest which corresponds to the spinal cord consists
of compactly grouped cells which are so arranged as to present a flattened
continuous portion or dorsal bridge, the dorsal border of which is rather
smooth and sharply outlined, and shows no connection with the neural
tube. here are as yet no dorsal nerve roots. Projecting ventralward
from this bridge is a series of rounded segmental clumps of cells which
form the primitive spinal ganglia. These end diffusely among the devel-
oping fibres of the ventral roots. The ventral border of the ganglionic

crest is ill-defined in contrast to the dorsal edge of the bridge of the crest.
7

88 Development of Occipital Nerves in Human Embryos

At this time the fibres of the ventral roots of the spinal nerves are gath-
ered into loose bundles and can be traced a short distance in the

Fig. 1. Reconstruction of peripheral nerves in three weeks human embryo, 4.0 mm. long,
Hertwig collection No. 137. Enlarged 16.7 diams.

IX.

X XI gang.crest.

Gang. nodos.
Gang petros.

Fiq. 2. Reconstruction of peripheral nerves in three weeks human embryo, 4.8 mm. long,
Mall collection No. 148. Enlarged 16.7 diams.

1X.

x xIgang. ee

if’ores;
ae

Gang petros

Gang. nodos.

Fiq. 3. Reconstruction of right side of same embryo shown in fig. 2.

mesoderm. Among them are many of the so-called sheath cell nuclei.
Owing to the fact that the fibres of the ventral roots run through the

George L. Streeter 89

ventral part of the spinal ganglia it is impossible to determine the
presence or absence of fibroblasts belonging to these ganglia.

The first, or most oral, of the spinal ganglia exhibits special character-
istics. It is separated from the second ganglion by an enlarged interval,
and it is usually smaller and more slender than the other gangha. In
some cases it consists of a very small clump of cells showing no tendency
to unite with its ventral root, or it may be absent completely. On the
other hand it may be accompanied by additional ganglionic cell-masses
representing the occipital ganglia, found by F'roriep, 82, to be constant
in the sheep. In one case (Fig. 2) the ganglion is divided ventrally into
two clumps of equal size, corresponding to a likewise divided ventral root.
The more oral of these two may therefore be considered as a persistent
occipital nerve and ganglion.

That part of the ganglion crest which is situated lateral to the after-
brain differs from the spinal portion of the crest in that it consists of
more loosely arranged cells, and as a whole it is flatter and presents an
irregularly triangular profile, which at this period does not exhibit a seg-
mental character. That part of it adjacent to the auditory vesicle, and
which represents the anlage of the ganglion of the root of the glosso-
pharyngeal nerve, is in most cases completely separated from the rest
of the crest. It is continued ventralward into the third branchial arch
by a looser zone of cells which connects it, just dorsal and caudal to the
second gill cleft, with a rounded compact clump of cells forming the
future ganglion petrosum. In a similar manner the ganglionic crest of
the vagus at its oral end is continuous by means of an ill-defined zone of
cells with the anlage of the ganglion nodosum. ‘This ganglion, as can be
seen in Figs. 1, 2, 3, and Plate I, is larger than the ganglion petrosum.
It is somewhat spindle-shaped, and owing to the branchial arches it takes
a caudal direction ending diffusely beyond the fourth arch. Directly
over it les a patch of thickened epidermis and is apparently adherent to
it. A careful search was made here and in the ganglion petrosum for
evidence of interchange of cells between the ganglia and the epidermis
without success. That these ganglia are not simply ventral projections
of the ganglion crest is in a degree indicated by the zone of less well-
defined tissue, which at this stage separates them from the crest proper.

In the ganglion crest, in addition to the cellular elements, there are
a few fibres found at the junction of the crest with the neural tube, thus
completing the anlages of the glosso-pharyngeal and vagus nerves. A
sreater fibre development is found in the caudal part of the ganglion
crest of the vagus, where inclosed among the cells, as in a sleeve, is found
a well-defined bundle of fibres representing the accessory nerve. This

90 Development of Occipital Nerves in Human Embryos

bundle makes its appearance at the level of the fourth, fifth, or sixth
cervical segment. It consists of fibres which are contributed to it at
irregular intervals by the dorso-lateral border of the neural tube. In-
‘creasing in size it extends forward, running mesially to the ganglion
erest until it reaches the crest of the vagus where the cells of the crest
form a sheath for it. The tip of this nerve reaches to the dorsal border
of the column of cells which extends from the crest to the anlage of the
ganglion nodosum.

The hypoglossal nerve is represented by a row of four or five rootlets,
which are in a direct line with those forming the ventral roots of the cer-
vical ‘nerves, and resemble them with the exception that they successively
taper off smaller as they become more oral. At this stage they have not
grown far enough out for coming together to form a common trunk.
The rootlets are bundled together in three divisions, between which are
situated the first two occipital myotomes. The third myotome les
between the last division of the hypoglossal and the first cervical nerve.

On looking back at these embryos of the third week, an important
feature is observed in the fact that the main motor elements are indicated
at this time by the presence of fibres. We find a few fibres in the roots
of the ninth and tenth nerves; the trunk of the eleventh is definitely laid
down, also the roots of the twelfth and the ventral roots of the spinal
nerves. ‘The sensory elements, however, are only indicated by the cellular
masses of the ganglion crest.

Embryos of About Four Weeks.

FHis- Collections. Embryo Bron. ..0..c15 60s ot nee 6.9 mm.
Mex Collection iNton ois mae oe lenccecs see 7.0 mm.
Gage Collection, Buxton Embryo............. 7.0 mm.

(See Figs. 4, 5 and 6.)

In this group of embryos the ganglionic crest is increased in size, but
aside from a disproportionate increase in fibre elements, it presents much
the same appearance as seen in embryos of three weeks. The flattened
dorsal border of the spinal crest still forms a continuous bridge along the
tops of the primitive ganglia. Cropping out along this bridge numerous
fibroblasts are seen attaching themselves to the spinal cord. These are
the primitive dorsal nerve roots. Attention is directed to the later and
slower development of these as compared with the ventral roots. Aside
from that seen in the dorsal bridge the ganglia show as yet little or no
fibre formation. They consist of cells whose round nuclei, scantly sur-

George L. Streeter wi)

rounded by protoplasm, are compactly clumped together, the whole form-
ing a row of pillow-like bodies whose ventral borders are lost among the
fibres of the ventral roots. It was not possible to determine as to whether
or not the ventral borders of the ganglia are involved at this stage in
fibroblast formation, and contribute fibres to the nerve root, because of

IVs

-xI Gang. GEESE.

Opthal div.

ye q
Nilaryg. SUP.

Gang. petros.

Gang. nodos.

Fig. 4. Reconstruction of peripheral nerves in four weeks human embryo, 6.9 mm. long,
His collection Embryo Br3. Enlarged 16.7 diams.

the difficulty of differentiating them from the mesodermal sheath cells
of the ventral roots.

The fibres of the spinal nerve roots form compact bundles which
branch, as they extend forward, and form intersegmental anastomoses.
The profuse anastomosis of the nerve roots from the fourth cervical to the
first dorsal segment marks a primitive brachial plexus. (See Fig. 4.)

92 Development of Occipital Nerves in Human Embryos

The roots of the caudal end of the cord are somewhat tardier in their
development, and at this stage the lumbar plexus is only slightly
indicated.

x Aa gang crest. Le . ee
1\ a
| A

Opthal. div.

Sup max.div.

N masticatori/us.

Inf max.div.

\w laryn. sup. \»
™~

~ “Gang petros.

Gang. nodos.

Fic. 5. Reconstruction of right side of same embryo shown in fig. 4.

ee ae aes ze
a ae aes
Le So
. [>

t

SA.

x
Cal:
ao \ = 2.
\ 3
v XII lO FB8#=:: 4
\ Vagus // :
p : zn \"
(C vi Ge . \\o
\ Gang.nodos. \
SS with sens: organ. |

|
Pe: N. laryg, sup

Sens.organ on gang. petros.

Fria. 6. Reconstruction of peripheral nerves in four weeks human embryo, 7.0 mm. long,
Mall collection No.2. Enlarged 16.7 diams.

On the right side of Embryo Br3 (see Fig. 5), and on the left side of
the Buxton Embryo, the first cervical ganglion consists of but a small
clump of cells showing no connection with the ventral root, which root,

George L. Streeter 93

however, is well developed. Cases of this kind have in the adult the
appearance of entire absence of this ganglion, a point which will be taken
up later.

The caudal portion of the ganglion crest of the after-brain is longer
and more slender than in the previous stage. The accessory nerve is still
ensheathed by the cells of the crest. As it extends forward it turns the
curve on the back of the trunk of the vagus, and then freeing itself from
the vagus it extends a short distance lateralward and ends abruptly in a
mass of condensed mesoderm, the anlage of the m. sterno-cleido-mastoi-
deus. The oral end of the ganglion crest of the vagus is connected with
the ganglion nodosum (it will be observed that this ganglion is not con-
sidered as simply a part of the vagus crest, which is because of the
apparent independence of the two seen in three week embryos) by a com-
pact mass of cells, among which are found some fibres. From the
ganglion nodosum there arise two distinct fibre bundles, ventrally the
superior laryngeal, and ventro-caudally the main trunk of the vagus,
around which winds the hypoglossal nerve.

The rootlets of the hypoglossal nerve unite and form a stem, as in
the adult, which seems to be joined by fibres from the first and second
cervical nerves at the point where it bends upward to reach the anlage of
the tongue. . The descending branch of the hypoglossal can be identified
on the right side of Embryo Br3 as a short bud at the point where the
nerve crosses the main trunk of the vagus. ‘Thus the descendens hypo-
glossi develops in this case simultaneously with the appearance of
anastomoses between that nerve and the upper cervical nerves.

The root of the glosso-pharyngeal nerve is definitely connected with
the ganglion petrosum by a fibro-cellular mass. Ventral to the ganglion
the trunk of the nerve is represented by a fibre strand extending into the
third branchial arch. In Fig. 6 is represented the area on both the
petrosal and nodosal ganglia which still remains attached to the overlying
thickened epidermis. In Embryo Br3, on both sides of the embryo, the
ganglia nodosum and petrosum appear to fuse.

On review of this group it is seen that at the end of the fourth week
the ganglion crest is not yet entirely differentiated. We find laid out,
however, the roots, the trunks, and the ganglia of the trunks of the ninth
and tenth cranial nerves. Further, all the elements of the eleventh and
twelfth nerves are present, and the dorsal and ventral roots and the
plexuses of the spinal nerves.

Vagus root gang. (jugular)

Accessory root gang.

Se

IX root gang.

Gang. petros:

Nlaryg. sup.

Xl.

Fia. 7. Reconstruction of peripheral nerves in thirty-one days human embryo, 10.2 mm.
long, His collection Embryo KO. Enlarged 16.7 diams.

Vagus root gang.

Accessory root gang

/

X/1.
Vagus. Pt descendens.

Fia. & Reconstruction of right side of same embryo shown in fig. 7.

George L. Streeter 95

Embryos of About Thirty Days.

Malik Covlecizoms INGOs 2Waeee «ts wie oe kts Balas 2 10.0 mm.
Hus Collection; “Buabryo: Ilsa) se bene sks 10.0 mm.
Has: Collection, Hnrbryo KO y40.% 20. 8 econ: 10.2 mm.

(See Figs. 7, 8, 9 and 14.)

On coming to embryos 1.0 cm. long the final steps in the transforma-
tion of the occipital ganglion crest into cranial nerve rootlets and their
ganglia may be seen. When Figs. 7 and 8 are compared with Fig. 4, a
considerable increase in the actual mass of cells is observed in this region,
and also a marked growth of the fibre elements. As this fibre formation

iX* root ganglion.

Vagus root ganglion

Accessory vagus
root ganglia

; Interganglionic
SLA bridge

Lhe => xi
<< / YE =
r descendens and brs — ‘ Ss
of Cia aCm tohyoid m's. Se
Vagus|} 7 (\ DN
5 6 DA
DA ~ SH sRad.dors.
AZ == =
pn aN
Brachial mans &
\ Z

Fig. 9. Reconstruction of peripheral nerves in thirty-one days human embryo, 10.0 mm.
long, His collection Embryo D1. _ Enlarged 16,7 diams,

96 Development of Occipital Nerves in. Human Embryos

continues the cell masses become broken up and separated into ganglionic
clumps. Instead of the uniform cellular crest seen in the occipital
region in Fig. 4 we find in Fig. 9 a chain of ganglia lying among the
rootlets of the ninth, tenth, and eleventh nerves. Just cephalad to the
first cervical ganglion, in Fig. 7, is a cell mass which may be regarded,
either as a fragment which has become separated off from the first cer-
vical ganglion, or what is more likely a persistent occipital ganglion
(precervical ganglion), such as is found by Vroriep, 82, in the sheep.

Accessory root gang.

X root gang.

by

IX root gang.

Ai

ON

NKR

Fria. 10. Reconstruction of peripheral nerves in five weeks human embryo, 13.8 mm. long,
Hertwig collection No. 67. Enlarged 16.7 diams.

Embryos of Fwe to Seven Weeks.

Elertwis Collection No 2670.0 eeieee aoe 13.8 mm.
Mall «Collection, Nos wade 7 par aac ener 14.0 mm.
is Collection, -Himbryo oH veean wae ete 17.5 mm.

(See Figs. 10, 11, 12 and Plate IT.)

At the end of the fifth week the ganglion crest is completely resolved
into a series of more or less segmental cell masses. The dorsal ridge,
which formed an intersegmental bridge across the tops of the spinal
ganglia, has disappeared. Simultaneously with the disappearance of this

George L. Streeter on

structure occurs the outgrowth of the central rootlets of the ganglia. On
comparing Figs. 1, 9, and 11 one gets the impression of an actual conver-
sion of the dorsal bridge into the ganglion rootlets. In Fig. 10 the
dorsal rootlets have attained a considerable length, and show a tendency
toward anastomosis.

Vagus root gang

Accessory root gang.

IX root gang

S

IX
Gang. nodos.
N-laryg. Sup.

XI with r descend’

™ Sympathetic.

Vagus

Fia. 11. Reconstruction of peripheral nerves in five weeks human embryo, 14.0 mm. long,
Mall collection No. 144. Enlarged 16.7 diams. This drawing is reversed right for left.

The eleventh cranial nerve lies median to the dorsal rootlets, and is
attached at irregular intervals to the spinal cord, just ventral to their
attachments to the cord. It may run either mesially or laterally to the
rootlets of the first spinal ganglion, and is usually adherent to the gang-
lion itself. Cell masses are found on the trunk of the nerve in this

98 Development of Occipital Nerves in Human Embryos

region, as though fragments of this ganglion. 'The close relation between
the first cervical ganglion and this nerve serves to explain the conditions
found in the adult. Along the more cranial portion of the nerve there is
a row of ganglia, the accessory root ganglia, which become successively
larger as we go forward, and which form a series with the ganglion
jugulare of the vagus. The number of these accessory ganglia is usually
three or four principal masses, and in addition there are several smaller
clumps scattered among the rootlets. In a series of pig dissections at
the corresponding age it was not possible to determine a true segmental
order in their formation, and there was no correspondence between these
gangha and the number of the hypoglossal roots, and they show no con-
nection with them. Thus they are not to be confused with the occipital
. ganglia of Froriep. The fibre elements of the accessory nerve fuse with
those of the vagus. The occurrence of an actual interchange of fibres
between them cannot, however, be determined. Leaving the vagus at the
ganglion nodosum the accessory can be traced through the m. sterno-
cleido-mastoideus to the m. trapezius.

The relations of the roots, ganglia, and trunks of the ninth and tenth
nerves were seen in the previcus stage (Figs. 7, 8, and 9) to have taken
on the adult type. In Figs. 11 and 12 the resemblance to the adult
conditions is more complete owing to the relative increase of fibre ele-
ments. The glosso-pharyngeal nerve arises by several compactly bundled
rootlets attached to the neural tube median and caudal to the cartilage-
nous mass in which the internal ear is embedded. Among these rootlets
is the ganglion mass which forms the ganglion of the root or Ehrenritter’s
ganglion. Beyond this begins the trunk of the nerve, on which is found
a second ganglion, the ganglion of the trunk. It is to be remembered
that the ganglion of the root and the ganglion of the trunk have
developed separately, and have so far remained discrete structures. From
the ganglion petrosum is given off ventrally the tympanic branch, or
nerve of Jacobson, and caudally the main trunk of the nerve, which hooks
inward and forward toward its terminal distribution. The ninth and
tenth nerves lie closely together and there is ample opportunity for
anastomosis between them, especially between the ganglia of the trunks.
It will be recalled that in a younger embryo (Fig. 4) these ganglia were
apparently continuous.

The vagus presents the same general type as the glosso-pharyngeus ;
the root and trunk gangha are larger, and the trunk itself may be traced
down into the thorax.

In Fig. 11 the chain of cervical sympathetic ganglia is indicated, and
in Fig. 12 is shown their connections with the spinal nerves. The upper

George L. Streeter 99

’

portion of this ganglionic chain fuses with the ganglion nodosum, and
above this it gives off its branches to the carotid plexus.

. Vagus root gang.

Accessory root gang.

Fror/ep

N.laryg. sup.

Gang.nodos

Sympathetic.

Fig. 12. Reconstruction of peripheral nerves in six weeks human embryo, 17.5 mm. long,
His collection Embryo FM. Enlarged 16.7 diams.

100 Development of Occipital Nerves in Human Embryos

The principal branches and communications of the hypoglossal and
cervical nerves may be distinctly traced in embryos of this size. It will
be seen that we thus have in this group the completion of the principal
features of both the fibre and the cellular elements of the nerves under
consideration; but the picture presented by the adult structure varies
somewhat from this, owing to a disproportionate growth of some parts
over others, and instead of what here appears as a ganglion cell pre- —
dominance we meet there with a predominance of the fibre elements.

Embryos of the Second and Third Month.

Hertwag, ‘Collection. INO: iG hate sone eens 30.0 mm.
Matitr Collectors: so.) eee chee taeeene 65.0 mm.
(See Plate IIT.) «

During the second and third months there is a progressive growth of
the fibre elements with a corresponding stretching-out of the nerve trunks
and rootlets, which results in a greater separation of the ganglion masses
from one another. The further growth of the ganglia is not uniform;
while the ganglia of the trunk of the ninth nerve, and of the root and
trunk of the tenth, and the spinal ganglia continue in their development, ,
the ninth root ganglion and the root ganglia of the eleventh reach at
this time a point of development at which they remain stationary.

In a reconstruction of the left side of the Hertwig embryo, No. 161,
which is not here reproduced, the noticeable change from the conditions
shown in Fig. 12 is in the length and sharper definition of the trunk of
the accessory nerve. On the trunk of this nerve, between the first
cervical ganglion and the ganglion jugulare, there are two root ganglia,
and as the accessorius trunk joins the vagus there is a third ganglion
mass, which, however, is partly fused with the ganglion jugulare. There
are also small clumps of cells among the rootlets of the vagus, as well as
on some of the central rootlets of the spinal ganglia.

A dissection of this region in an embryo at the end of the third month
is shown in Plate III. There the ganglia of the root of the ninth and
the eleventh nerves present very little enlargement. They can be dis-
tinguished from the fibre bundles only by a greater opacity, and appear

.as white nodes in the roots of the respective nerves. The first cervical
ganglion is well developed. An arrest in the growth of this ganglion,
similar to that in the ganglia just mentioned, might also be expected.

As the nerve fibre growth continues, the last trace of these rudimentary
ganglia is lost to the naked eye. In order to determine their ultimate fate
a series of sections was made through the structures of this region in an

George L. Streeter 101

adult specimen. This shows the presence of persistent clumps of normal
appearing ganglion cells, situated along the trunk of the eleventh and on
the roots of the ninth and tenth nerves. A diagrammatic reconstruction
of the series is shown in Fig. 13. In this case the first cervical nerve
received a communicating branch from the accessory, but macroscopically
no ganglion was present. In the series, however, this ganglion is repre-
sented by a circumscribed group of cells on the trunk of the accessory.
Among the rootlets in the same region are scattered groups of cells which
may have been separated off from it. On going further forward other
small ganglion groups are met with, either just beneath the connective
tissue sheath of the nerve roots, or among their fibres, and usually near

accessory root ganglia

ee
—secessorit us
y

-—n a
W/
l /

Gang |
Ehrenritter \¥j

c.1 ganglion

Gang nodosum.

Fig. 15. Diagrammatic reconstruction of ganglion cell masses in peripheral nerves of
occipital region in human adult. Compare with Plate IV.

the junction of the roots with the larger trunks. From their posi-
tion these are considered to be the persistent accessory root ganglia.
Although rudimentary in size they are made up of cells which have all
the appearances of functionating ganglion cells. It is possible that it is
these that provide the sensory fibres for that branch of the accessory
which joins with fibres from the vagus to form the pharyngeal branch,
in some such way as is schematized in Plate IV.

DEVELOPMENT OF INDIVIDUAL NERVES.

It has been seen, in tracing the development of the ganglionic crest of
the after-brain, that the ninth nerve stands apart from the more caudal
nerves and develops independently, and apparently uninfluenced by them.
The tenth and eleventh, in contrast, are parts of a single complex, and
cannot be taken up as individual structures without adopting a sepa-

102 Development of Occipital Nerves in Human Embryos

ration that would be artificial; they will therefore be described together.
The hypoglossus and the cervical nerves, where a close relation also exists,
will be likewise treated.

The Glosso-Pharyngeal Nerve from the beginning possesses the char-
acteristics of a mixed nerve. In embryos 4.0 mm. long (Fig. 1) it
consists of a small clump of ganglionic cells, which can be distinguished
in the mesenchyma attached to the neural tube just caudal to the otic
vesicle and extending toward the third branchial arch. This group of
cells represents the anlage of the ganglion of the root, or Ehrenritter’s
ganglion. Among these cells are a few fibroblastic processes, which do
not belong to them, but arise from cells of the neural tube in the dorsal
part of the ventral zone of His, and form the motor elements of the root.
The character of this anlage resembles that of the vagus; and may be
regarded as a part of the ganglion crest of the after-brain, though it is
not continuous with the vagal portion of it. It is evident that this
ganglion is not a part which has become separated off from the ganglion
petrosum as described by Henle, and others (see Thane, 95), but is an
independent structure. Its inconstancy is to be explained by its further
development. It reaches a size early in the embryo at which it ceases to
further develop, in some embryos earlier than others. The fibre elements,
however, continue to grow, and finally overgrow the ganglion and thus
cause 1t to be apparently absent. A similar occurrence will be seen in
ease of the root gangha of the accessory nerve.

Ventral to this group of cells is a somewhat larger clump of cells, the
primitive ganglion petrosum, which is situated directly beneath the epi-
dermis at the caudal and dorsal margin of the second gill cleft (see Figs.
1, 2, and 3). At this time it is separated from the rest of the anlage
of the ninth by a looser zone of cells, and this gives it the appearance of
having developed in situ, rather than of being a subdivision or bud from
the rest of the anlage. It is true that the same appearance might arise
from a migration of cells from the latter followed by a proliferation of
them at this point. The position of the ganglion petrosum here and in
older embryos (Figs. 4, 5, and 6) indicates a close relationship between
it and the branchial arches, and the same is likewise true of the ganglion
nodosum. In this respect the anlages of the ganglia of the trunks differ
from the ganglionic crest proper, or anlage of the root ganglia, which
is well removed from the branchial arches and does not show any trace of
branchio-meric arrangement. This suggests a difference in origin to exist
for the two kinds of ganglia. Another point of difference between the
ganglia of the roots and the ganglia of the trunks of these nerves is the

George L. Streeter 103

fusion of the latter with an overlying patch of thickened epidermis, and
an apparent absence of such a fusion in ease of the ganglia of the roots.

The relation existing between the ganglia of the seventh, ninth, and
tenth nerves and the overlying epidermis has been described in mammals
by Froriep, 85, and it is regarded by him as the anlage of rudiments of
the phylogenetically lost branchial sense organs of Beard, 85, and
van Wiyhe, 82. In elasmobranchs Mroriep, 91, describes later a double
“ne of fusion between ganglia and epidermis, forming the lateral and
epibranchial sense organs, which may perhaps be considered as com-
parable to the ganglia of the roots and ganglia of the trunks. In mam-
mals, however, he had found only a single line of epidermal fusion, that
existing over the ganglia of the trunks. In our series of embryos the
ganglia of the roots do not seem to take part in the formation of epi-
dermal sense organs, and show no sign of fusion. The condition here
resembles that described by Froriep, 85, in his earlier paper. The
adherence between the epidermis and the ganglion petrosum and nodosum
is indicated in Fig. 6. It is found in all the human embryos studied
from 4.5 to 7.0 mm., after which it disappears. This is a somewhat
earlier and briefer period than given by Froriep for other mammals. No
indication of interchange of cells between ganglia and epidermis could be
made out. |

The ganglion petrosum in embryos 7.0 mm. long has become connected
with the ganglion of the root by a definite strand of mixed fibres and cells,
the fibre elements more and more predominating as the embryo becomes
older. At the same time a tapering bundle of fibres sprouts from the
distal end of the ganglion petrosum, and forms the main trunk of the
nerve, the ramus lingualis, and supplies the third arch. Another branch
appears in 14.0 mm. embryos, the ramus tympanicus, and extends for-
ward into the second arch. The ganglion thus gives off a branch both
oral and caudal to the second gill cleft, and this completes the glosso-
pharyngeus as a typical visceral arch nerve. This was pointed out in
mammals by Froriep, 85, who regards the r. lingualis as the post-trematic
and the r. tympanicus as the pre-trematic branch.

Communications exist between the ganglion petrosum and the ganglion
nodosum in an embryo of 7.0 mm. (Figs. 4 and 5), where they seem
almost as a continuous structure; in other embryos of this stage, and
younger, they are completely separated. Later (Figs. 7, 8, and 9), fol-
lowing the relative change in position of the adjacent parts which
succeeds their unequal growth, these structures are gradually brought
close together, and secondary communications are established between

them.
8

104 Development of Occipital Nerves in Human Embryos

The Vagus Complex includes both the vagus and accessory divisions,
the tenth and eleventh cranial nerves, which develop practically as a
single structure; though the complex is more spread out than the tri-
geminal, yet the relation of the accessory to the vagus is embryologically
much the counterpart of that of the motor root of the trigeminus to the
rest of that nerve. To speak of the two divisions as individual cranial
nerves is misleading; perhaps a new terminology should be introduced,
which would express more exactly their comparative and embryological
relations. Onodt, 02, has suggested the entire removal of the name
accessorius as an independent cranial nerve, but does not himself attempt
to carry it out. Eventually such a radical attack upon the nomenclature
may prove advisable; in this paper, however, whenever it is necessary to
distinguish between the different parts of the vagus complex, the original
usuage will be retained which is based on the gross anatomy of adult
specimens: the term “vagus nerve” will be applied to that portion of
the complex represented by the ganglion jugulare with its rootlets, and
the peripheral nerve trunk extending from this on which is found the
ganglion nodosum; the term “ accessory or eleventh cranial nerve,” no
distinction being made between vagal and spinal portions, will refer to
the remainder of the complex situated caudal to this, and includes gang-
honated rootlets and the large motor trunk extending peripherally to the
sterno-cleido-mastoid and trapezius muscles. In their development it
will be seen that both divisions contain motor elements, which spring in
a continuous line from the lateral border of the neural tube as far down
as the third or fourth cervical segment, and sensory elements which are
developed from the cells of the ganglion crest. Later, following its
further growth, the oral or vagus division of the complex becomes pre-
dominantly sensory, and the caudal or accessory division predominantly
motor.

The ganglion crest of the after-brain is apparently directly continuous
with that of the spinal cord, and extends from the first or second cervical
ganglion to the otic vesicle, an interruption indicating the division
between the ninth and tenth nerves. We agree with Dohrn, 01, who
describes the vagus crest as forming a unit with the spinal crest, rather
than with Froriep, 01, who distinguishes between a ganglion crest of the
head and one of the trunk, and states that they do not simply go over
into one another, but overlap and run along adjacent to each other, each
ending for itself. Evidence of such an overlapping could not be made
out.

An embryo of 4.0 mm. represents the youngest stage at which the crest
was sufficiently differentiated from the mesoderm for accurate recon-

George L. Streeter 105

struction. The shape of the crest at this time is represented in Figs. 1,
2, 3, and Plate I. In the figures the presence of developing nerve fibres
are diagrammatically shown among the cells of the crest. Small bundles
of these fibres spring at irregular intervals from the lateral angle of the
neural tube and enter the crest. In the caudal two-thirds these fibres
join to form a definite strand, which is the primitive trunk of the acces-
sory nerve. ‘This trunk reaches down into the region of the spinal gang-
lion crest to the level of the third or fourth cervical segment. The fibres
do not, however, enter this crest, but run along median to it as far as the
first cervical, when they enter the vagus crest as though into a sleeve.
In regard to the first cervical a variability is shown, the trunk may run
median, lateral, or through it. Forward, in a line with these fibres of the
accessory trunk, are found a few others forming small scattered bundles
at the head of the crest. That the fibres which are present at this stage
are motor may be inferred from three facts: firstly, they spring from the
lateral horn region of the neural tube; secondly, there is at this time no
apparent fibroblast development in the cells of the ganglion crest; and
thirdly, some of them can be followed in their further development until
they become known motor elements, as in the case of the main trunk of
the accessorius, and as in the spinal cord where the ventral roots at this
time are well laid out, though there is as yet no trace of dorsal roots.

Ventral to the head of the vagus crest, and partially separated from it
by a looser zone of cells, is found a second ganglionic mass, the primitive
ganglion nodosum, the relation between which and the crest repeats the
condition found between the ganglion of the root and the ganglion of the
trunk of the glosso-pharyngeus, evidence of independence being here
equally strong. As was there pointed out, the ganglion nodosum is
closely associated with the development of the more caudal branchial
arches. The patch of thickened epidermis over the ganglion, as in Fig. 6,
represents an epibranchial sense organ. A complete segmentation of
the ganglion nodosum, which might be expected in considering this anlage
as the morphological equivalent of a series of gill cleft ganglia, is not
found, though the cells show at first a loose irregular grouping, which
represents perhaps a branchio-meric tendency. 'The laryngeal branch of
this ganglion is present in 7.0 mm. embryos, and forms the principal
nerve to the fourth branchial arch. The main vagus trunk is differen-
tiated at about the same time and is seen sprouting out from the distal
end of the ganglion.

At the end of the first month (Figs. 7, 8, and 9) the cellular column
between the ganglion nodosum and the ganglion crest is converted into
a fibrous trunk. At this time the ganglion crest, besides an increase in

106 Development of Occipital Nerves in Human Embryos

size, is modified in form by the development of numerous rootlets attach-
ing it to the neural tube, and by an irregular clumping together of the
cells of the crest, forming ganglion masses along the main trunk of the
accessorius, which now les at the ventral border of the crest. The root-
lets which are developed at this time are in part sensory, as is evidenced
by comparison with the dorsal spinal roots. The division of the crest
into ganglion masses is accompanied by a rapid development of fibres
between its cells, and it is probable that the growth of these fibres is
the cause which spreads the cell masses apart into separate clumps.
Such a separation into clumps radically differs from true segmentation,
the latter does not seem to occur here. As the fibre growth continues
the ganglion masses become more and more separated, and finally the
crest becomes completely converted into a series of discrete ganglia
(Figs, 10 and 11). The most oral one is the largest and forms the vagus
root ganglion, the ganglion jugulare. Caudal to this, successively dimin-
ishing in size, is a chain of three or four accessory root ganglia, which
extend backward along the accessory trunk until they meet the cervical
ganglion series. In the vagus complex the ganglia diminish in
size in the caudal direction, while in the spinal series the reduc-
tion in size is in the oral direction; this fact enables one to distin-
guish between accessorius root ganglia and precervical (Froriep) ganglia.
Those ganglion masses found adherent to the accessorius between the
first and second, and the second and third cervical ganglia, as in Figs. 10,
11, and 12, are developed from the spinal crest, and represent nodules
derived from the spinal ganglia and which have become separated off.
Formed similarly to these, are found isolated masses on the rootlets of the.
jugular ganglion.

The vagus division of the complex at this stage (embryos of 14.0 mm.)
therefore consists of mixed motor and sensory rootlets, the ganglion jugu-
lare, and the nerve trunk on which is situated the ganglion nodosum,
giving off a laryngeal branch as well as communicating branches to the
ganglion petrosum. The accessory division begins at the third or fourth
cervical segment. Its trunk runs median to the dorsal roots, except at the
first cervical where it may be lateral. It is attached to the neural tube by
mixed rootlets; on which are found a varying number of ganglia. The
trunk after an arched course joins the vagus division, but the greater
portion of it soon leaves the vagus and extends to the shoulder region
and supphes the sterno-cleido-mastoid and trapezius muscles.

The essential features of embryos of the fifth and six week, Figs. 11
and 12, will be observed to be the same as in older embryos and in the
adult, Plate III and Fig. 13. The existing difference may be accounted

George L. Streeter LOT

for by the disproportionate growth of the fibre elements over that of
the cellular elements, and of some of the cellular masses over that of
others; most of the cell masses persist, but some of them early reach a
point at which they remain stationary, such as the accessorius root
ganglia, and sometimes the first cervical. Following the increase in fibre
growth they become buried among the rootlets or on the accessorius trunk,
and though not seen by the naked eye they can be seen on section. Fig.
13 represents a case in which the first cervical] ganglion was macroscopi-
cally absent, but microscopically it is present as a large clump of normal
appearing ganglion cells within the sheath of the accessorius trunk.

Anastomoses between the first cervical and the trunk of the accessory
nerve in the adult have excited much interest. Among others they have
been studied by Kazzander, 91, and later by Weigner, 01. A study of
Weigner’s drawings shows that the accessory nerve of one side has no
,constant relation to the accessory nerve of the other side in the same
individual; they bear themselves as independent structures, and his 37
examinations may therefore be considered as 74 individual cases. By
re-analyzing Weigner’s cases in this manner instructive data on our
present subject have been obtained. ‘They show that the relation in the
adult of the first cervical to the accessorius is as follows:

19%—First cervical ganglion and dorsal root are present, and do not
anastomose with the accessorius.

19%—First cervical ganglion and dorsal root are macroscopically

_ absent.

62%—Various kinds of anastomosis between the first cervical dorsal
root and the accessorius. In many of these cases the ganglion is macro-
scopically absent.

These anastomoses are doubtless to be explained on embryological
grounds. The relative position of the two structures at the beginning
of connective tissue formation would determine their permanent relations.
If they lie in contact at that time they become permanently adherent.
Secondarily, when the dorsal roots become thus entangled in the acces-
sory trunk, as they are apt to in case of the first cervical, they are dragged
along out of their original position by later growth and the consequent
relative shifting of all of the structures in that region. Further
irregularities in their course may be caused by the accessory which, being
laid down earlier than they, would have the tendency to guide the imping-
ing dorsal rootlets out of the direct centripetal line to the neural tube,
and along its own trunk, either forward or backward. A diagram showing
some of these variations is reproduced in Plate IV. In the same diagram
is shown the hypothetical course of some of the other fibres of the acces-

108 Development of Occipital Nerves in Human Embryos

sory. No motor fibres are represented as running from the accessory
to the larynx, the absence of such fibres having been well established by
the work of Onodt, 02, and others, and the clinical observations of
Seiffer, 03. Fibres of the accessory doubtless join the trunk of the vagus,
but they are omitted here for sake of simplicity.

The Hypoglossal Nerve can first be made out in embryos at the end of
the third week, at which time it consists of loose fibre strands which can
be traced between the occipital myotomes springing from the ground plate
of the neural tube and extending a short distance in the mesenchyma
(see Fig. 1). These rootlets are formed in three or four segmental
groups and develop in the same line with the ventral roots of the cervical
nerves. During the fourth week they grow forward and fuse in a common
trunk. At the end of the first month this trunk has passed around the
ganglion nodosum, and curves around the sinus cervicalis mesially and
orally to reach the anlage of the floor of the mouth. A week later its
principal branches of distribution are indicated.

As the hypoglossus crosses the ganglion nodosum it gives off the ramus
descendens, which is first definitely seen in embryos 1.0 em. long. Mall,
91, and Piper, 00, report its absence in embryos 7.0 mm. and 6.8 mm.,
respectively. His, 88, pictures a long r. descendens in Br3 (6.9 mm.).
In a reconstruction of the same embryo, made since then by the author
(see Fig. 4), this is not seen. ‘There is, however, on the opposite side
(Fig. 5) a slight indication of a beginning branch. At the time the
descendens is developed the opportunity for communication between the
hypoglossus and the upper two or three cervical nerves already exists;
that is to say, the terminal fibres of the latter end in brush-like tufts in
close contact with the former. The amount of interchange of fibres
cannot be accurately traced, but it is evident that the character of the
descendens is dependent on the nature of the contribution of fibres from
the cervical nerves. The course in the development of this cervical
anastomosis is as follows (compare Figs. 3, 4, 6,9, and 11): The fibres
of the hypoglossal and the upper cervical nerves start out perpendicularly
from the neural tube, and due to the curve of the latter they come together
like spokes in a wheel, and then grow along adjacent to each other into
the premuscle tissue of Froriep’s schulterzungenstrang ; when the forma-
tion of the nerve sheaths begins, adjoining fibres become thereby more or
less bound together, and as the individual tongue and hyoid muscles draw
apart these nerves are led out into an open plexus, the adult arrangement
of which and its variations has been described by Holl, 77. The exact
formation of this anastomosis must depend on the position of the fibres
at the time the sheaths are formed. ‘This introduces a variability which

George L. Streeter 109

might account for the different arrangements found by Holl. A further
source of variation is presented by slight differences in the division line
between the rootlets of the hypoglossus and the first cervical; the fibres
destined for the r. descendens, for instance, may be either picked up with
the more caudal rootlets of the hypoglossal, when there will be little or no
communication between the hypoglossal and the first cervical, or on the
other hand may be picked up with the first cervical and reach their
destination through anastomosis with the hypoglossal. Thus in embryo
No. 144 of the Mall collection (Plate IL) on the right side the first cer-
vical contributes no fibres to the hypoglossal and descendens, while on the
left side a large communicating bundle exists between them.

In the early stages the rootlets of the hypoglossal present a close simi-
larity to the ventral roots of the spinal nerves, and now are generally
considered as a cranial continuation of them; the nerve being thus derived
from the fusion of three or four segmental spinal nerves, which in the
course of phylogenesis have become enclosed in the cranium. In the
hypothetical ancestor the segments of the nerve belonged to the trunk,
and possessed, in addition to the ventral roots, both dorsal roots and
ganglia, the latter becoming subsequently reduced coincidently with the
invasion of the vagus group into this region. Strong support to this
view was given by Froriep, 82, who in the hoofed animals found persist-
ent dorsal roots and ganglia belonging to one or two of the more caudal
divisions of the nerve. Similar precervical ganglion masses and rootlets
were found in the rabbit, cat, and mouse by Martin, 91, and Robinson, 92.
The former describes five hypoglossal ganglia in cat embryos, of which
he finds only the most caudal one to persist. He thus apparently
includes those that in our series of reconstructions are considered as
accessory root ganglia, which we think have a different phylogenetic signi-
ficance. In the human embryo //is, 88, describes an abortive precervical
ganglion, and names it after Froriep. Inasmuch as he considers the
hypoglossus to belong phylogenetically to the vagus rather than to the
spinal nerves, he is inclined to doubt a relation between the Froriep
ganglion and the hypoglossus. In our reconstructions a typical ganglion
may be seen in Figs. 7 and 11. The former is the same embryo
pictured by His, and does not essentially differ. On the other side of this
embryo, Fig. 8, the first cervical ganglion creeps forward a short distance
along the accessorius tract, and thus represents what may be styled as a
precervical tendency. An interesting case is shown in Fig. 2, where the
first cervical ganglion is divided in two equal parts, each having its own
ventral root. With further growth they would have become separated,
as the spinal ganglia do, and then we should have in the more oral one

110 Development of Occipital Nerves in Human Embryos

a typical Froriep ganglion with a ventral root that would have doubtless
joined with the hypoglossus fibres. In Fig. 9 a slight indication of a
ganglion is present, though it is not labelled in the diagram. In such
cases one cannot say whether it belongs to the spinal group or to the
accessorius root ganglia of the cranial group. These two seem to develop
from the same crest, and it could be expected that the oral tendency of
the former and the caudal tendency of the latter might cause in some
cases a fusing of the two; such an instance is seen in Fig. 12. Where
the retrogression of the spinal elements is advanced, the Froriep ganglion
is absent, and the first cervical also then shows abortive tendencies. If
Fig. 5 is compared with Fig. 7, it will be seen that there the spinal
reduction extends an entire segment further caudad; instead of a rootless
Froriep ganglion, as in Fig. 7, there is in Fig. 5 a rootless first cervical
ganglion.

It is evident that there is a great irregularity in the degree of reduction
of the occipito-spinal dorsal roots and ganglia in different individuals.
By comparing individuals of different ages we cannot therefore estimate
the retrogression undergone in the development of a single individual;
one cannot say, for instance, that because a Froriep ganglion is present
in an embryo of 7.0 mm. and is not present in another embryo of 14.0 mm.
that it has in the latter case disappeared. It was found in case of the
accessory root ganglia that ganglion masses once present persist through-
out life, though they may early reach a point beyond which they do not
further develop. The same is doubtless true as regards the Froriep
ganglion.

CoMPARATIVE MorPHOLOGY.

In considering the phylogenetic significance of the nerves of the occipi-
tal region it becomes apparent that we are here dealing with structures of
two different sources; on the one hand, the cranial nerve elements rep-
resented by the glosso-pharyngeus, the vagus, and the accessorius—a
portion of the vagus, and on the other hand the elements of spinal origin,
the upper cervical nerves and the hypoglossus. The literature concern-
ing the comparative anatomy of these structures is voluminous, and par-
ticularly their involvement in the various theories proposing a segmental
origin of the vertebrate skull. A complete review of this literature and
discussion of the morphological bearing of the cranial nerves is given by
His, 8%, and again later by Rabi, 92. Since then has appeared the im-
portant work of Fiirbringer. 97, supplemented by the embryological
investigations of Braus, 99, and Froriep, 02. Mention should also be
made of the work done on the accessory nerve by Lubosch, 99. The
general facts as known may be stated as follows:

George L. Streeter IE

In the lower fishes the cranial and spinal elements are clearly separated
and their territories do not overlap; a line may be drawn oral to which
all the nerves are cranial and caudal to which all are spinal. In the
phylogenesis, owing to a caudal encroachment of the skull into the spinal
region, the more oral of the spinal nerves become included in the head
region and have special foramina of exit. Those that are thus assimi-
lated by the selachii have been styled by Fiirbringer as occipital nerves,
and those assimilated in addition later by the holocephali are called
occipito-spinal nerves. With this assimilation, however, the spinal and
cranial elements are still discretely separated by a transverse line of
demarcation. There is no actual overlapping of the two until we come to
the sauropsida. Here and in all higher vertebrates, accompanying the
conversion of certain vagus. gill muscles into the trapezius and sterno-
cleido-mastoideus, the cranial elements (7. e. vagus complex) make a
caudal invasion into the spinal region, in such a manner that the acces-
sory portion of the vagus is found wedging itself in between the ventral
and dorsal spinal roots mesial to the ganglia, gaining attachments to the
cord just ventral to those of the dorsal spinal roots.

In the human embryo the different stages of this invasion cannot be
demonstrated. Either the early steps are not repeated in the embryo-
logical history of higher types, or it may be, as McMurrich, 03, suggests,
that the derivation of such structures cannot be demonstrated ontogeneti-
cally because the phylogenetic stages occur while the structures are still
in an undifferentiated state. In the embryos studied, as soon as the
nerve elements can be distinguished, they have their final relative posi-
tion, and the accessorius is found extending well down into the cervical
region. Its caudal end is indicated by “EH” in Figs. 1, 2, and 3. The
vagus-accessory anlage is, in all three instances, about of the same size.
Some variation exists in the extent of overlapping of the cranial and
spinal parts, as is evidenced by the variation in distance between the
ganglion jugulare and the first cervical ganglion. It is doubtless a
variation of the individual, and is of the same character as the variation
occurring in the distance over which the accessory nerve extends into the
cervical region of the adult. In Fig. 14 is shown the wedge-like invasion
of the cranial nerve elements into the spinal territory. The figure is a
diagrammatic profile reconstruction of an embryo one month old. The
gill arches, vertebral skeleton and muscular apparatus, and spinal and
cranial nervous systems are plotted out with view to a comparison of their
relative positions. It shows clearly the impossibility of drawing any
transverse line through the body, oral to which everything would be
cranial, and caudal to which everything would belong to the spinal

112 Development of Occipital Nerves in Human Embryos

system. ‘The behavior of the nervous system adds to the irregularity in
the line of junction between the head and trunk.

As the accessorius wedges itself into the spinal territory there occurs
a progressive retrogression of the more oral spinal elements, resulting
in the disappearance of the dorsal roots and ganglia of the occipito-spinal
nerves, these being the first nerves encountered. The ventral roots of
these nerves persist and join to form the hypoglossus, and supply the

Notochord

Occipital myotomes

°C. myotome.

Cranial nervous
system in spinal
territory.

1S*TA myotome.

Nervous System,
Muscular ee
Skeletal a

Fria. 14. Diagrammatic reconstruction of one month human embryo, 1v.0 mm. long, Mall
eollection No. 144. Enlarged 9 diams.

tongue, which in the meantime has been acquired in the floor of the
mouth. In some of the domestic animals (rabbit, pig, cow, and sheep)
one or two of the more caudal of the occipito-spinal dorsal roots and
ganglia persist as was pointed out by Mroriep, 82, whose name they have
received. In man the most caudal ganglion occasionally persists, but it
is usually without any connection with a corresponding ventral root; in
one case, however, a Froriep ganglion with ventral root was present, and
doubtless would have joined with the hypoglossus as its most caudal root.

George L. Streeter 113

Often the connection between the ventral root of the first cervical and its
ganglion is also missing in man, and the ganglhon rudimentary and found
only on section. These rudimentary ganglia during embryonic life
become adherent to the invading cranial member, the accessory nerve, and
though all connection with the ventral root is absent, they still may func-
tionate by sending their fibres forward or backward along the accessorius,
in the latter case joining a more caudal nerve. Although the first cervi-
cal ganglion, and perhaps a precervical or Froriep’s ganglion may thus
lie in the tract of the accessorius, it is to be remembered that embryologi-
cally they are separate structures, the one cranial and the other spinal.
The apparent relation between the two is only due to the fact that in
the early stages they lie closely together, and become adherent in this
position.

In addition to these occipito-spinal (precervical, hypoglossal, or
Froriep’s) ganglia, there are found in the human adult other rudiment-
ary ganglia situated along the accessory nerve, which are of cranial origin,
and similar to the root ganglion of the vagus. ‘These are the accessory
root ganglia; they form, in the six weeks embryo, a series of ganglionic
clumps, which extend caudalward from the ganglion jugulare, successively
diminishing: in size, along the tract of the accessory nerve attached to its
rootlets. A true segmental arrangement of them does not seem to prevail
in the human embryo, and the same is true in dissections of pig embryos.
The ganglion jugulare continues to develop, but these accessory ganglia
early reach a size beyond which they do not further develop. They,
however, do not undergo retrograde metamorphosis, at any rate not
completely, for evidence of them may still be found in the human adult.

The root ganglia of the ninth, tenth, and eleventh nerves develop from
a ganglion crest which has an appearance and history analogous to that
from which the spinal ganglia develop. The ganglia which form on the
trunks of the vagus and glosso-pharyngeus apparently develop independ-
ently from that crest, and they differ from the ganglia of the roots in
being branchio-meric, and in possessing definite traces of rudimentary
sense organs.

The hypoglossus in contrast to the tenth and eleventh nerves, which
show no trace either in rootlets or ganglia that they were ever formed
from a series of segmental nerves, presents a distinct segmental grouping
of its fibres, as may be seen in Plate I and Fig. 1. This fact, added to
its resemblance in its early stages to the ventral roots of the cervical
nerves in point of origin from the same column of cells, its relation to the
myotomes, and the occasional presence of a Froriep ganglion, offer con-
clusive evidence that this nerve is the equivalent of three or four ventral

114 Development of Occipital Nerves in Human Embryos

roots of phylogeneticalty lost occipito-spinal nerves, which have become
fused into a single trunk.
CONCLUSIONS.

1. The tenth and eleventh cranial nerves are parts of the same complex,
both possessing mixed motor and sensory rootlets, together with root
gangla derived from the same ganglionic crest.

2. During the progress of development of this vago-accessory complex
the cephalic end becomes predominantly sensory, and the caudal end -
becomes predominantly motor and also more spread out. This produces
a difference in the appearance of the two portions which has resulted in
their being considered as two independent structures. The cephalic por-
tion forms the vagus or tenth cranial nerve, and the caudal portion the
n. accessorius Willisii or eleventh cranial nerve. The old nomenclature
is retained, and in so doing the term eleventh cranial nerve is used as
synonymous with n. accessorius vagi plus n. accessorius spinalis.

3. The root gangha of the tenth and eleventh cranial nerves do not
present a definite segmental arrangement.

4. The trunk ganglia of the ninth and tenth cranial nerves (gang.
petrosum and gang. nodosum) when first identified are not definitely
connected with the root ganglia of the same nerves, and they differ from
the root ganglia in having an arrangement segmentally related to the
gill arches, and possessing rudimentary sense organs.

5. The ganglia found on the rootlets of the eleventh cranial nerve are
the counterpart of the root ganglion or jugular ganglion of the tenth.
They do not reach the high development of the latter, though traces of
them persist in the adult. They are to be distinguished from the pre-
cervical ganglion of Froriep, which represents an extra spinal ganglion.

6. The eleventh cranial nerve extends caudalward into the spinal region
to the third or fourth cervical segment, in some cases further; the extent
and variation in the embryo is the same as in the adult. The caudalward
invasion of this nerve is phylogenetic, and not ontogenetic.

7. The hypoglossal nerve in young embryos closely resembles the
ventral roots of the adjacent cervical nerves, and is segmentally continu-
ous in the same line with them. That a phylogenetic retrogression has
removed the dorsal roots, which they seem to have at one time possessed,
is evidenced both by the occasional presence of a Froriep ganglion and by
cases in which the retrogression has gone still further caudalward, and
has remoyed the dorsal root of the first cervical nerve.

8. The ramus descendens hypoglossi is developed in some cases before
the hypoglossus has received any connecting branches from the cervical

George L. Streeter 115

nerves; in other cases such connections are formed coincident with or
before the r. descendens appears. <A variable relation exists between
the r. descendens and the communicating cervical branches.

9. The ventral roots of the spinal nerves are developed earlier than
the dorsal roots. Similarly in the cranial nerves those portions generally
recognized as motor are differentiated into fibre paths earler than the
sensory elements.

LIST OF REFERENCES.

Batrour, F. M., 76.—On Development of the Spinal Nerves in Elasmobranch
Fishes. Phil. Trans., Vol. CLXVI. ‘

BARDEEN, C. R., 03.—The Growth and Histogenesis of the Cerebro-Spinal
Nerves in Mammals. Amer. Jour. Anat., Vol. 2.

BEARD, J., 85.—The System of Branchial Sense Organs and their Associated
Ganglia in Ichthyopsida. Quart. Jour. Micr. Science, Vol. 26.

Braus, H., 99.—Beitrage zur Entwicklung der Muskulatur u. des peripheren
Nervensystems der Selachier. Morph. Jahrbuch, Bd. 27.

CHIARUGI, G., 90.—Le Developpement des Nerfs Vague, Accessoire, Hypoglosse
et Premiers Cervicaux chez les Sauropsides et chez les Mammiferés.
Archiv. Ital. d. Biol., Tome XIII.

DoaiEL, A. S8., 03.—Das periphere Nervensystem des Amphioxus. Anat. Hefte,
Arbeiten, Bd. 21. ‘

Dourn, A., or.—Studien zur Urgeschichte des Wirbelthierkorpers. Mitth’l.
Zool. Station zu Neapel, Bd. XV.

FrRoriIEp, A., 82.—Ueber ein Ganglion des Hypoglossus ete. Arch. f. Anat. u.

Phys. Anat. Abth.

85.— Ueber Anlagen von Sinnesorganen am Facialis, Glossopharyngeus

und Vagus. Arch. f. Anat. u. Phys. Anat. Abth.

o1.—Ueber die Ganglienleisten des Kopfs und des Rumpfes und ihre

Kreuzung in der Occipitalregion. Arch. f. Anat. u. Physiol. Anat.

Abth.

o2.—Zur Entwickelungsgeschichte des Wirbeltierkopfes. Verhandl.

d. Anat. Gesell. Halle.

FURBRINGER, M., g7.—Die Spino-Occipitalen Nerven der Selachier u. Holo-
cephalen und ihre vergleichende Morphologie. Festschr. f. Gegen-
baur.

GEGENBAUR, C., 72,—Kopfskelet der Selachier. Leipzig.

HARRISON, R. G., o1.—Ueber die Histogenese des peripheren Nervensystems
bei Salmo salar. Arch. f. Mikr. Anat., Bd. 57.

HATSCHEK, B., g2—Die Metamerie des Amphioxus und des Ammocoetes.
Weiner Anatomen Kongress, Anat. Anz. Verhandl., Bd. 7.

His, W., 87.—Die morphologische Betrachtung der Kopfnerven. Arch. f.

Anat. u. Physiol. Anat. Abth.

88.—Centralen u. peripherischen Nervenbahnen beim menschlichen

Embryo. Abhandl. math.-phys. Cl. Kgl. Sachs. Ges. Wiss., Bd. 14.

Hott, M., 77—Beobachtungen tiber die Anastomosen des Nervushypoglossus.
Zeitschr. f. Anat. u. Entw., Bd. 2.

116 Development of Occipital Nerves in Human Embryos

KAZZANDER, J., g1,—Ueber den N. access. Willisii und seine Beziehungen zu
den oberen Cervicalnerven. Arch f. Anat. u. Physiol. Anat. Abth.

Lewis, F. T., 03.—The Gross Anatomy of a 12. mm. pig. Amer. Jour. Anat.,
Vol. 2.

Lusposcu, W., g9.—Vergleichend-anatomische Untersuchungen tber den
Ursprung und die Phylogenese des N. accessorius Willisii. Arch. f.
Mikr. Anat. u. Entwick. Bd. 54.

MALL, F. P., 91 —A Human Embryo 26 days old. Jour. of Morphol., Vol. 5.

03.—Note on Collection of Human Embryos. Johns Hopkins Hosp.

Bull., Vol. 14.

Martin, P., g1.—Die Entwicklung des neunten bis zwolften Kopfnerven bei
der Katze: Anat. Anz., Bd. 6.

McMovrgeticH, J. P., 03.—The Phylogeny of the Forearm Flexors. Amer. Jour.
Anat.; Vol. 2.

Minot, C. S., 92—Human Embryology. Wm. Wood & Co., N. Y.

Onopt, A., 02.—Der Nervus accessorius und die Kehlkopfinnervation. Arch.
Laryngol. u. Rhinol., Bd. 12.

Preer, H., 00.—EHin menschlicher Embryo von 6.8 mm. Nackenlinie. Arch. f.
Anat. u. Physiol. Anat. Abth.

RABL, C., g2.—Ueber die Metamerie des Wirbeltierkopfes. Verhandl. d. Anat.
Gesell. Wien.

Ransom and THOMPSON, 86.—On the Spinal and Visceral Nerves of Cyclo-
stomata. Zool. Anz., Bd. 9.

Ropinson, A., g2.—Observations on Development of Posterior Cranial and
Anterior Spinal Nerves in Mammals. Report British Assoc. for
Advancement of Sci., Edinburgh. j

SEIFFER, 03.—Die Accessorius-Lahmungen bei Tabes dorsalis. Berlin. Klin.
Wochenschr., 40-41.

THANE, G. D.,-95.—_The Nerves. Quain’s Anatomy, Vol. III, Part II.

VAN WIJHE, 82.—Ueber die Mesodermsegmente u. Entwicklung der Nerven
des Selachierkopfes. Verhandl. koninkl. Akad. Wetenschappen., 22
Deel. Amsterdam.

WEIGNER, K., o1.—Beziehungen des Nervus accessorius zu den proximalen
Spinalnerven. Anat. Hefte, Arbeiten (Merkel-Bonnet), Bd. 17.

DEVELOPMENT OF THE OCCIPITAL NERVES
GEORGE L. STREETER

X-XT gang. crest. LX:

i
,
‘
;
,
‘
/

NT fibres.

Gang. bridge.

PEATEsI

Gang. petrosum.

Gang. nodosum.

PROFILE RECONSTRUCTION OF PERIPHERAL NERVES IN THREE WEEKS HUMAN EMBRYO,
4.3 MM. LONG, MALL COLLECTION No. 148. ENLARGED ABOUT 25 DIAMETERS. THE THIRD
DIMENSION FOR THIS DRAWING WAS OBTAINED FROM A CLAY MODEL, AND THE SHADING IN

~ PART COPIED FROM A PIG EMBRYO OF SAME STAGE.

AMERICAN JOURNAL OF ANATOMY-=-VOL. IV

DEVELOPMENT OF THE OCCIPITAL NERVES
GEORGE L. STREETER

PLATE Il

Accessory root gang,

WAX PLATE RECONSTRUCTION OF PERIPHERAL NERVES
WEEKS HUMAN EMBRYO,

DIAMETERS.

IN OCCIPITAL REGION OF FIVE
14.0 MM. LONG, MALL COLLECTION No. 144. ENLARGED ABOUT 25

AMERICAN JOURNAL OF ANATOMY--VOL, Iv

DEVELOPMENT OF THE OCCIPITAL. NERVES PLATE
GEORGE L. STREETER

X rool ganglion. IX root ganglion.

Corp. quadrigem.

Cerebellum. ee ae

XT root ganglia. Tie ee

A/a S . PR ere creer

A. vertebralis. M. sternocleidomastoideus.

DISSECTION OF NERVES IN OCCIPITAL REGION OF THREE MONTHS HUMAN EMBRYO, 65.0
MM. LONG, MALL COLLECTION. ENLARGED 4 DIAMETERS.

AMERICAN JOURNAL OF ANATOMY--VOL. IV

IV

PLATE

DEVELOPMENT OF THE OCCIPITAL NERVES

GEORGE L. STREETER

“SSNI1 MOVIE AAVSH AG SSYGIS AYOSNSS

3H1L 4O LVH1 GNV ‘SSNI7 G3Y AG GALVOIGNI SI S3USIS YOLOW AHL 40 3SYNOD AHL “S3AYSN TWWLIdIOOO GNV IVOIANSO S3HL
N34SM1L49 ONILSIXS SNOILVOINNWWOO NOWWOO S3HL JO 3WOS GNV ‘SASSVW NOITDNV9 JO NOILVNLIS 3HL ONIMOHS WvuDdvid

‘UolbUuef /BIIAIID 7S414

"e6ueb 3004 Kuossazoy

\ ‘Buksey dns yy

‘wnsopou burg

pp ‘BukKseyd Y

‘8 04320 GueD

‘DINE “ey

407} IMUBIY T
yo) ued

IV

AMERICAN JOURNAL OF ANATOMY--VOL.

A FURTHER STUDY OF THE DEVELOPMENT OF THE
EXCRETORY ORGANS IN BDELLOSTOMA STOUTI.

BY
GEOS Cy PRICK Pres

Associate Professor of Zoology, Leland Stanford Jr. University.

WitH 31 TExt FIGURES.

Much of the following work has been done in the Embryological
Laboratory of the Harvard Medical School, and it gives me great pleasure
to here acknowledge my many obligations to the Director, Professor
Charles Sedgwick Minot, for his unvarying kindness and _ helpfulness.
I wish also to express my thanks to Professor Oscar Hertwig for his
kindness in placing the resources of his laboratory at my disposal during
a four months’ residence in Berlin.

In a previous paper on this subject * the excretory organs were described
in three stages, designated as A, B and C. Of these A showed the sys-
tem in an early though not in the earliest stage of development, B fol-
lowed quite closely upon A, but between B and C there was a wide gap,
the organs in C resembling in many respects those of the adult.

A subsequent study of better material, and of a rather full series,
beginning with specimens younger than A and ending with others older
than C, has revealed certain errors both of observation and interpre-
tation, and has at the same time brought to light some new and inter-
esting facts. It is the object of the present paper briefly to present these
facts and to correct the errors. No attempt will be made to discuss the
inorphology of the excretory organs of the vertebrates in general.

It may be stated at the beginning that additional observations have
abundantly confirmed the main point of the previous paper, namely, that
in Bdellostoma the entire excretory system arises as a pronephros, and
that from this both the pronephros and mesonephros of the adult are
derived.

In the previous work, on account of the difficulty of counting the
myotomes in transverse series, especially at the anterior end, recourse

Price, G. C.: Development of the Excretory Organs of a Myxinoid, Bdel-
lostoma stouti. Zool. Jarb., Vol. 10, Anat., 1897.

AMERICAN JOURNAL oF ANATOMY.—VOL. IV.

118? Exeretory Organs in Bdellostoma Stouti

was had to the spinal ganglia for the purpose of determining relative
positions. It now seems best to employ the myotomes for this purpose,
especially since in the earlier embryos there is an intimate relation
between the myotomes and the excretory organs. It has been found
that the first spinal ganglion is located between the third and fourth
myotomes, so that where trouble is experienced in: counting the myo-
tomes, the ganglia may be counted instead, and the necessary calculations
made.

The anterior end of the excretory system is more or less rudimentary,
and the position of the first tubule is not constant, but varies both in
different individuals and on the two sides in the same individual. In a
few cases where its position could be accurately determined, and where
there was no probability of degeneration having taken place, it occurred
in segments eleven .to thirteen. In one of the specimens previously
studied the first tubule occurred in sections passing through the sixth
ganglion. But here the preservation is not all that could be desired,
and it is possible, though not certain, that this may really be the seventh
ganglion, and the tubule might correspond to the myotome just back of
this, which would be the tenth. But making all possible allowances,
the fact remains that here the first tubule is farther forward than in
any other embryo studied.

The position of the posterior end of the segmental duct has been found
to vary from about the seventy-ninth to the eighty-second segment, and
possibly a larger number of specimens would show still greater variation.

In_ general it may be said that in this animal there is a good deal
of variation, both in the embryo and in the adult, a fact that should
not be lost sight of in attempting to generalize from a small number
of individuals.

So far as my knowledge extends Dean”* is the only other person who
has published observations on the excretory system in the embryos of
Bdellostoma. From surface views alone he has described a series of
segmental structures, about eighty in number, extending from the region
of the neck into that of the tail, and present for a comparatively long
period of the development. They are described and figured as large and
somewhat complicated bodies, each extending outwards from near the
spinal ganglion to the distal region of the somite, and it is suggested
that they may be tubules of a pronephric nature. In sections of embryos

* Dean, Bashford: On the Embryology of Bdellostoma stouti. A General
Account of Myxinoid Development from the Egg and Segmentation to Hatch-
ing. Festschrift zum siebenzigsten Geburstag von Carl von Kupffer. Jena,
1899.

Geo. C. Price 119

of ages corresponding to those in which the above structures are figured
I have been unable to find any such system of tubules, although on
account of their size it would seem impossible to overlook them if present.
In Dean’s figure 134 the parts marked pronephric tubules bear a strik-
ing resemblance, even down to the fine details, to the ventral part of
the muscle segments as seen in sagittal sections of an embryo of about
the same age as the one from which the above figure was taken, and in
the sections there is a comparatively large amount of connective tissue
between the muscles which corresponds well with the space between the
so-called tubules. Moreover, the muscle segments are the only seg-
mental structures that at all correspond in size to the ones in question.

The youngest embryo studied has one hundred and one myotomes on
the one side and one hundred and two on the other, a number fully
equalling that of the muscle segments in the adult. There are seven
or eight pairs of forming gill slits, only five of which could be made
out from the surface view.

Fic. 1.—Section through the fifty-second segment of the youngest embryo studied.
a, aorta; ch, notochord; g, spinal ganglion; Im, lateral mesoblast; my, myotome;
me, nephrocel; nt, nephrotome; scl, sclerotome. There is an artificial break, shown
best on the right of the figure, separating the myotome from the nephrotome and
extending into the sclerotome. The sclerotome is seen to be in connection with the
myotome, and on the right the lateral mesoblast is continuous with the nephrotome.

In this embryo there are neither excretory tubules nor excretory ducts
in the true sense of the word, but there is an extensive system of nephro-
tomes from which both tubules and ducts are derived. These extend
certainly from the thirteenth segment to the seventy-fourth on one side
and to the seventy-fifth on the other, and it is possible they may begin
even farther forward than the thirteenth. From the anterior end back
to the neighborhood of the fiftieth segment they are all practically in
the same stage of development, but from here on they become gradually
less well developed until, in the last few segments, they appear to be
just forming.

An idea of the relations of the nephrotomes to other parts may be

had from Fig. 1, which represents a section passing through the fifty-
9

120 Exeretory Organs in Bdellostoma Stouti

second segment, and therefore very near the middle of the body region
and about five segments back of the middle of the region of the nephro-
tomes. The mesoblast, which alone interests us, is divided on either side
into four parts, the myotome, my, the sclerotome, scl, the nephrotome,
nt, and the lateral mesoblast, /m.

The myotome comes into direct contact with the ectoderm, there being
no mesenchyme between them, and this is true for all the segments
except the first four or five. There is no myocel, but the centre of the
myotome differs from the periphera in being free from nuclei. In the
anterior segments muscle fibres are forming. Nothing of this is seen
in Fig. 1, although in the section next in front, which passes nearer
the middle of the myotome, some of the cells show signs of differentiation,
and this is true also in four or five of the segments further back. .

on
supra ce

Fics. 2-8.—Seven consecutive sections from the middle of one nephrotome to the
middle of the next. mc, the nephroccl shown in Figs. 2 and 8. Fig. 5 represents the
place where the two nephrotomes come together.

The sclerotome, which is formed of compact mesoblast, is still con-
nected with the myotome, and its segmental character is further shown
in running through the series, by the fact that at regular intervals the
tissue here becomes much less compact. In about the last twenty-seven
segments the sclerotomes have not yet been formed.

The lateral mesoblast has the appearance of loose mesenchyme, and
extends outward from the nephrotome with which it is continuous.
(This is shown only on the right in the figure.) In this section there is no
indication of a splitting into somatic and splanchnic layers, although,
as will appear later, the process has begun farther forward.

The nephrotome lies just below the outer end of the myotome, from
which, in this case, it is separated by a narrow, artificial space, caused
presumably by the action of reagents. It is sharply marked off from the
surrounding tissue by the compactness of its walls, and it contains a
rather large cavity, the nephroceel, nc. In running through the series
the segmental character of the nephroceels is very striking, as may be

W:

a

Geo. C. Price P11

gathered from Figs. 2 to 8, representing a continuous series from the
middle of one nephrotome to the middle of the next, but the segmental
character of the solid portion of the nephrotome is not so apparent. This
is because the nephrotomes in adjoining segments come into direct con-
tact with each other, and some part shows in every section, just as the
segmentally arranged myotomes appear in every section. However, the
nephrotome becomes smaller towards the end than it is in the middle,
and the point where two nephrotomes come together (Fig. 5) can quite
easily be detected by the smaller size of the section and the indistinct-
ness of the outline.

In order to gain an idea of the appearance of a sagittal section through
the nephrotomes, camera lucida drawings were made of all the sec-
tions in three segments, and from these Fig. 9 was reconstructed on milli-
meter paper.

The proportions are here fairly accurate, but it was a question whether
the nephrotomes should be represented in
close contact with one another, or as hav-
ing a slight space between them. It is pos-
sible, though not probable, that an actual Fie. S2AY reconsemuchion of a
sagittal section might even show continuity se aa OU HEESe a on
between adjacent nephrotomes. pag ee

The posterior end of one nephrotome and the anterior end of the
next (Figs. 4 to 6 and Fig. 9) may be looked upon as forming a sort
of short, discontinuous rod. This would be more apparent in a frontal
than in a sagittal section. This remark is made because later they form
an actual rod extending from one nephrotome to the next, and giving
rise ultimately to the greater part of the segmental duct between con-
secutive tubules.

What has been said thus far in regard to the nephrotomes applies to
those representing a middle state of development, ‘those toward the an-
terior end being in some respects more advanced, and those toward
the posterior end less advanced. As an example of the latter we may
take the one in the sixty-eighth segment, a section through which is
represented in Fig. 10. Here the nephrotome is attached to the myo-
tome, although a well marked constriction has appeared between them.
There is a very small nephroccel, but the greater part of the centre of
the nephrotome consists of non-nucleated protoplasm which is continuous
with the non-nucleated protoplasm of the myotome. ‘There is no dis-
tinct boundary between the nephrotome and the sclerotome. On this
side the last nine nephrotomes are attached to the myotomes, and a few
just in front have somewhat the appearance of having been torn away.

122 Exeretory Organs in Bdellostoma Stouti

Of the nine that are attached the second, third and sixth have small
nephroceels, while in the rest the center is made up of non-nucleated
protoplasm. The transition between the nephrotomes represented in
Fig. 1 or Figs. 2 to 8 and the one represented in Fig. 10 is gradual.
Toward the posterior end the constriction between the nephrotome and
the myotome becomes less pronounced than in Fig. 10. It is possible
that if this embryo had been allowed to develop nephrotomes would
have appeared still farther back; at least they are found farther back
in older embryos.

The only striking difference between the nephrotomes in the anterior
region and those represented in Fig. 1 is that in the former the nephro-
ceels communicate with the splanchnoceel, as may be seen in Fig. 11.

oy

=
mn aaa

Fra. 10. Fire. 11.

Fic. 10.—Section through the sixty-eighth segment showing a nephrotome connected
with the myotome, though partly separated from it. my, myotome; lm, lateral meso-
blast; ne, a very small nephroce); nt, nephrotome; scl, sclerotome. The myotome,
nephrotome and sclerotome are all continuous with one another, and the lateral meso-
blast is continuous with the nephrotome. :

Pia. 11.—Section showing the nephrocel, nc, opening into the splanchnoceel, sce.
It also shows the way in which the splanchnoceel first appears as small, rounded cavi-
ties in the lateral mesoblast. my, the outline of the end of the myotome.

Here the distinction between the nephroccel and splanchnoccel is at
once apparent by the difference in the character of their bounding walls.
In this region the nephrotomes come into direct contact with the myo-
tomes and in the fifteenth segment on the right side there is actual
continuity between the two.

A few words regarding the distribution of the nephrotomes. In seg-
ments nine to twelve there are cavities having the shape and position
of nephroccels, but with bounding walls less well marked than is the
ease with nephroceels farther back, and it is doubtful whether they
would give rise to any part of the excretory system. Nevertheless the
feeling can hardly be avoided that these are rudimentary nephrotomes.
On the right side in segments thirteen to fifty, and on the left in seg-
ments thirteen to forty-two the nephrotomes are in direct contact with

Geo. C. Price 123

the myotomes, and, in the one case above noted, there is actual con-
tinuity. In segments fifty-one to sixty-seven on the right, and in forty-
three to sixty-five on the left, the nephrotomes and myotomes are sepa-
rated by narrow, artificial spaces. In segments sixty-eight to seventy-
five on the right, and in sixty-six to seventy-four on the left, the nephro-
tomes are joined to the outer end of the myotome. It seems fair to
suppose that in earlier stages all the nephrotomes would be found con-
nected with myotomes, but this could be proved only by the examination
of a younger embryo.

A brief account of the formation of the splanchnoccel will now be given.
This arises in the lateral mesoblast as small rounded spaces (Tig. 11)
which grow together, and thus form a continuous cavity on either side,
and at the same time split the lateral mesoblast into somatic and splanch-
nic layers. (No reference is here made to the pericardial cavity.) The
entire ccelom of the right side, both nephroccels and splanchnoccel, was
reconstructed on millimeter paper, and this shows that the region of
most active formation is in segments nine to
fifty-three. In segments fifty-four to fifty-
seven it is almost absent, while back of this
it is entirely absent. Fig. 12 shows the
ccelom in three segments, the heavy lines indi-
eating the boundary of the nephroccels, and
the light that of the splanchnoceel. No at-
tempt is made to show anything beyond the
outline of the cavities. It will be observed
that the divisions of the splanchnoceel com-
municating with the nephroccels have a sort
of segmental character, and that the cavity

12.—Reconstruction of

Fra.
the coelomin three segwments of

in one segment in no case quite communi-
eates with that in the next. Further forward,
however, such communications do occur, in
one instance the cavities in four consecutive

anembryo in which the splanch-
noccel is in an early stage of
formation, made on millimeter

pean The nephroccel, ne, is
ounded by the heavy line, and
the splanchnocoel by the light.

Nothing is Shown beyond the

ea aie line of the cavities.
segments being joined together. aches ee

In a few cases a splanchnoccelic cavity was found lying just beyond the
nephroceel, but without any communication between them. In others,
as in Figs. 13 to 15, the two are seen to be breaking into each other.
This may not be the only way in which the communication between the
nephroccel and splanchnoceel is established; in some cases (Fig. 1, right
side) a splitting seems to begin at the nephroceel and extend outward into
the lateral mesoblast.

The next embryo to which attention will be called is between the one

124 Exeretory Organs in Bdellostoma Stouti

just described and the youngest of stage A, previously studied, but is
nearer the latter than the former. It was cut in two in the region of
the forty-sixth segment, the anterior part being sectioned sagittally and
the posterior part transversely. ‘There are eleven or twelve pairs of gill
slits. Mesenchyme is present between the myotomes and the external
ectoderm, and the sclerotomes have been converted into very loose mesen-
chyme, which shows no signs of segmentation. The splanchnoccel is ©
found from about the ninth to the eightieth segment, and through the
ereater part of its extent is continuous from segment to segment. In
the posterior region, however, it is poorly developed, and here sections
are occasionally met with in which it is entirely wanting.

The excretory system extends on the one side from the twelfth to the
eighty-second segment, and on the other from the eleventh to the eighty-
first. Through the greater part of its extent it is no longer in contact
with the myotomes, mesenchyme having grown in between, but towards

Fies. 13, 14 and 15.—Three consecutive sections showing one way in which the
nephrocel, nc, and splanchnocel, sc, come together. The section in front of Fig. 13
shows no splanchnocel at all, while Fig. 15 is the only section in which there is a
communication between the two cavities.

the posterior end of the system the ends of the myotomes gradually
approach the nephrotomes, until, in about the last ten segments, the two
are either in contact or are actually joined together. From the anterior
end back to about the twenty-eighth or thirtieth segment the nephroccels
have either disappeared or are disappearing, by being merged with the
splanchnoceel. Beyond this they are all in connection with the splanch-
nocoel, except the very rudimentary ones in the last four segments.

A good idea of the excretory system as found throughout the greater
part of its extent in this embryo may be gained by an examination of
Fig. 16, representing a sagittal section through segments forty-four and
forty-five, Fig. 17, representing a transverse section through the nephro-
tome in segment forty-seven, and Figs. 18 to 20, three consecutive sections
through the segmental duct, back of Fig. 17. The plane of the sagittal
section does not quite coincide with the long axis of the body, hence the
difference in shape of the nephroccels in Fig. 16, the left or posterior one
being nearer the median line than the right.

Geo. C. Price 125

The nephroceel has increased in size, especially in its dorso-ventral
diameter, and the dorsal wall of the nephrotome as seen in transverse
section (Fig. 17), is arched, as if being evaginated to form a tubule.
As will appear later this is true only in part. A glance at the sagittal
section (Fig. 16) proves that we are still dealing with nephrotomes and
that a tubule in the strict sense of the word has not yet appeared.

Extending from one nephrotome to the next is a solid rod of cells
(Fig. 16, d), a part of the segmental duct. At one point, about half
way between the nephrotomes, the duct is sligtly smaller than else-
where, and here the nuclei are much less numerous. This is brought out
still more clearly in the transverse sections (Figs. 18 to 20). Fig. 19
corresponds to the point x in Fig. 16, and Figs. 18 and 20 are the sections

Fic. 16.—Sagittal section through the the nephrotomes in segments forty-four and
forty-five, the one in forty-four being to the right. d, forming segmental duct; ne,
nephrocels ; 2, place where the two nephrotomes have joined together.

Fic. 17.—Transverse section through the nephrotome in the forty-seventh segment of
the same embryo from which Fig. 16 was taken. nc, nephrocel; sc, splanchnocel.

Fics. 18, 19 and 20.—Three consecutive sections through the duct between the
nephrotomes in the forty-seventh and forty-eighth segments. Fig. 19 corresponds to
the point # in Fig. 16 and represents the place where the ends of two nephrotomes have
united.

next on either side. The relation of the duct to the nephrotomes, and
a comparison with the younger embryo suggest very strongly that it has
been formed from the posterior end of one nephrotome and the anterior
end of the next, the point 2 being the place where the two have united.
This point of union is easily made out in almost all of the segments.

The description just given answers in all essential respects for about
three-fourths of the segments, or to be exact, for all except four at the
posterior end and twelve or fifteen at the anterior end. However, as we
approach the posterior end, in the region where in the older embryos the
tubules are rudimentary and early disappear, the nephrotomes become
quite a little smaller.

126 Excretory Organs in Bdellostoma Stouti

In the last four segments the duct is present as a continuous rod, the
outlines of which are not quite so distinct as farther forward. At four
places, however, situated at perfectly regular intervals, the myotomes
approach and are joined to the duct, and here the center of the duct is
filled with non-nucleated protaplasm, in which is a small cavity. These
can be none other than rudimentary nephrotomes. Whether they would
have developed farther cannot be said, but they are interesting as indi-
cating that the entire duct is formed from nephrotomes, for it is not
likely the duct would have extended farther back, at least no case has
been found where it extends beyond this point, the eighty-second segment.
In the previous paper tubules were found to within two segments of
the end of the duct, and as it was then thought that the duct was formed
from the coelomic epithelium, and as there was no ccelom in these seg-
ments it was suggested that here it had grown back independently. The
present observation renders this improbable.

Fic. 21.—Section through the excretory organs in segments twenty-three, twenty-four
and twenty-five, twenty-three being to the right. d, segmental duct; ce, celomic epi-
thelium ; nc, nephrocels. The nephrocels in segments twenty-four and twenty-five are
growing towards each other, while the one in twenty-three has met and united with
the one in twenty-four. The duct does not show where the two ends of the nephrotomes
have united, as in Fig. 16.

Turning now to the anterior end it remains to be shown how the
nephroccels become merged with the splanchnoceel. This is brought about
simply by the nephroccels in adjoining segments growing together, the
process beginning at the anterior end and proceeding posteriorly. Fig. 21
is a sagittal section through the excretory organs in segments twenty-
three, twenty-four and twenty-five. ‘The nephroccel in segment twenty-
five, the one to the left, has not yet united with the one in segment
twenty-four, although the two have approached each other; but the one
ir segment twenty-four has united with the one in twenty-three. From
here forward sections corresponding to the one just given show no signs
of nephroccels, but if the series be followed toward the median line it is
found that in the three or four sections before its entire disappearance
the ccelom presents the appearance of segmentally arranged cavities,
showing that the fusion of the nephroceels is not entirely completed.
Between segments twenty-three and twenty-four the segmental duct, d,
is seen to be separated by a narrow space from the coelomic epithelium.

Geo. C. Price 127

In some of the segments farther forward this is not the case, the two
being in contact or actually fused together. This union is to be looked
upon as secondary.

Stages A and B previously described follow in natural sequence upon
the one we have just been considering, but before proceeding to the gap
between B and Q, it will be well to pause for a time, and describe a certain
phase in the early differentiation of the segmental duct and tubules
and also notice some of the mistakes of the earlier paper.

Beginning with an embryo not a great deal older than the last, and
corresponding well with the youngest of stage A, it is found that a con-
striction has occurred in the nephrotome, the most obvious result of which
is to divide the nephroccel into three parts, a dorsal part, which helps to
form the lumen of the segmental duct, a ventral part, which later forms

Fic. 22.—Transverse section through the tubule and duct in the forty-fifth segment
of an embryo corresponding to stage A of the previous work. cme, what will later be
the cavity of the Malpighian corpuscle; ld, lumen of the segmental duct; 7t, lumen of
the tubule. The line a—b corresponds to the plane of Fig. 23.

Fic. 23.—Oblique longitudinal section of the segmental duct and tubule in the seg-
ment from which Fig. 22 was taken, reconstructed on millimeter paper. The plane of
the section corresponds to the line a—b in Fig. 22. cme, what will later form the cavity
of the Malpighian corpuscle; ld, lumen of the segmental duct; 7¢, lumen of the tubule.

Fic. 24.—Section of a tubule in the anterior region wheré the nephrocels have dis-
appeared, of the same embryo as the one from which Fig. 22 was taken. f, tubule.

the cavity of the Malpighian corpuscle, and a middle part, which forms
the lumen of the tubule proper. This is not apparent in a single trans-
verse section, such as Fig. 22, passing through the forty-fifth segment,
but it is brought out clearly by the study of a series of sections through
a segment, and still better in a longitudinal section such as Fig. 23.
This was reconstructed on. millimeter paper, and represents an oblique
section, the plane of which corresponds to the line a—b in Fig. 22.
The constriction has affected the lower part of the nephrotome, but not
to so great an extent as the middle part. A comparison of Fig. 23 with
the right hand nephrotome in Fig. 16, from the younger embryo, will

128 Exeretory Organs in Bdellostoma Stouti

help to make the matter clear. The posterior half of the excretory
organs in both these embryos were reconstructed on millimeter paper.
It was found that thirty segments of the older embryo occupied the
same space as thirty-one of the younger, so the length of a segment is
very nearly the same in the two embryos. It was found further that
there was no lumen in the segmental duct of the younger embryo, but
in the older embryo in every segment there was such a lumen, extending
in both directions from the tubule, but more in the posterior than the
anterior. The length of each division of the lumen was nearly the same
as the length of the nephroccel in the younger embryo, while the width
of the tubule was less. It was thus clear that the lumen of the segmental
duct had been differentiated from the nephroccel, and had not been
formed in the solid rod of cells forming the segmental duct of the younger
embryo. This accounts for the fact before noted, that the lumen may
extend in both directions from the tubule.

Fia. 25. Fig. 26.

Fig. 25.—Section through a tubule in the region where the nephrocels have dis-
appeared in an embryo corresponding to stage B of the earlier paper. t, tubule; sc,
splanchnocel ; 2, part corresponding to the part # of Fig. 26.

Fic. 26.—Section of a tubule which has been cut off from its connection with the
splanchnocel, and from the same embryo as Fig. 25. eme, cavity of the future Mal-
pighian corpuscle; t, tubule; sc, splanchnoccel; #, part that will form the Bowman's
capsule of the Malpighian corpuscle. .

In an embryo a little older some of the nephroccels, or better tubules,
bave lost their connection with the splanchnoceel, by the somatic mesoblast
at the point where the splanchnoccel and nephroccel come together, grow-
ing down and uniting with the splanchnic mesoblast. In general, the
process occurs first in the posterior region, but it does not begin at any
one particular segment and from there proceed in regular order, as may be
gathered from the fact that on one side in this embfyo the tubules
which are closed off are in segments thirty-nine, forty-three, fifty-eight,
sixty, sixty-four to sixty-six, and sixty-nine to seventy-one. Beyond
this the tubules are degenerate and have almost disappeared. On the
cpposite side the tubules which are closed off are not quite so numerous,
nor do they always occur in the same segments.

In the anterior region where the nephroceels have disappeared the

Geo. C. Price 129

tubules are formed as evaginations of the dorsal walls of the nephro-
tomes. However, it is not possible to tell where the tubules thus formed
end, and the others begin. Figs. 22 and 24 represent transverse sections
through tubules of the two regions in a younger embryo and Figs. 25
and 26 represent corresponding tubules in an older embryo. ‘The resem-
blance in the appearance of the tubules from the two regions is apparent,
notwithstanding the fact. that there are important differences.

In Fig. 26 the thickening marked 2 forms ultimately the membrane
immediately surrounding the glomerulus; the glomerulus itself appearing
in the angle between this and the tubule. In Fig. 25 there is a similar
thickening, x, but here it later entirely disappears, the tubule retains
the characteristics of a pronephric tubule throughout life, and no glomer-
ulus is ever formed. The attention was first called to the similarity of
the two kinds of tubules in the above particular while searching for
indications of the formation of glomeruli in the anterior part of the excre-
tory organs. It is here given for what it is worth, although it might
be interpreted as suggesting that formerly the anterior tubules as well
as the posterior were provided with glomeruli.

By the disappearance of the nephroccels in the anterior part of the body,
and by their persistence and subsequent separation from the splanch-
noccel in the posterior part, the excretory system becomes quite early
divided into two regions; but for a long time it is impossible to deter-
mine the exact boundary between the two, owing to the fact that there
is an intermediate region of a few segments, in which the nephroccels
have neither merged with the splanchnoccel nor have they been cut off
from it, and there is nothing to indicate which their ultimate fate will be.
Moreover, in front of this the nephrocwls do not end abruptly but be-
come gradually shallower and shallower. Later the boundary becomes
definite, but it is not constant for different individuals, nor is it always
constant for the two sides in the same individual. The best that can be
said is that it is in the neighborhood of the thirtieth segment.

Corresponding to the above described regions there are two kinds of
excretory tubules, those in the anterior region retaining the character-
istics of pronephric tubules throughout life, while those in the posterior
region assume the characteristics of mesonephric tubules. These terms
would be employed in speaking of them if it were not that it would lead
later to the contradictory expression, “the mesonephric tubules of the
pronephros.” For this reason the first will be spoken of simply as the
open tubules, and the second as the closed tubules.

Of the closed tubules those in about the last eighteen or twenty seg-
ments degenerate (the number not being constant), while the remainder

130 dxeretory Organs in Bdellostoma Stouti

persist, and all except a small number at the anterior end become
mesonephric tubules of the adult.

Both of these points were noted in the previous paper, but it was
there stated that in the oldest embryo of stage B the tubules had dis-
appeared in the last nineteen segments. A subsequent very careful
examination proved this to be a mistake. ‘They have disappeared in the
last eleven segments, but in the next eight they are still present in a rudi-
mentary condition. ‘Their entire disappearance does not occur in this
region until quite a little later.

In the earlier paper the statement was made that the series was
sufficiently complete to enable one to say that the mesonephric tubules
of the adult were derived from tubules which in the embryo arose as
pronephric tubules. The opinion has been expressed by some authors
that the evidence given was not sufficient to justify the assertion, and it
is not to be denied that, on account of the wide gap between stages B and
C, the objection may have been reasonable. But an examination of
specimens both older and younger than those before studied, as well as
of quite a complete series between B and C, has proved the truth of the
statement. In no case in the region of the mesonephros of the adult
is there the slightest indication of the disappearance of one set of tubules
and the appearance of another, but on the contrary the first tubules to
appear may be traced into what can be none other than the mesonephric
tubules of the adult.

Wheeler * objected further on the ground that in a young specimen of
Myxine Mass found mesonephric tubules forming in the posterior part
of the body independently of the duct. As a matter of fact Mass* de-
scribes and figures two well developed Malpighian corpuscles, one with and
the other without a tubule, neither of which is in connection with the
duct. He does not state, however, whether these occur in segments in
which later mesonephric tubules are invariably present, and until this
is known it seems at least possible that the structures in question may
be degenerating instead of developing tubules. However it may be in
Myxine in Bdellostoma in a few cases very small, degenerating tubules
have been found far back in the body, independent of the duct. But
here a comparison with other tubules in the same embryo, as well as
with embryos both older and younger, shows that these have been

* Wheeler, W. M.: The Development of the Urinogenital Organs of the
Lamprey. Zool. Jahr., Vol. 18, Anat., 1899.

*Mass, Otto: Ueber Entwicklungsstadien der Vorniere und Urniere bei
Myxine. Zool. Jahr., Vol. 10, 1897.

teo. C. Price 131

pinched off from the duct and would soon entirely disappear. As a rule
the degenerating tubules do not become thus separated from the duct.

In the earlier paper the statement was made that the excretory system
developed from behind forward. This was based on two facts; first,
in passing from the anterior end backward through a number of seg-
ments in both stages A and B, both the duct and tubules became gradually
better and better differentiated; and second, in the older embryo the
system extended a little farther forward than in the younger. From the
last it was assumed that if the younger embryo had developed the system
would have extended as far forward as in the older, but this does not
follow, as is proved by the examination of a larger number of embryos.
The above opinion was further strengthened by the observation that the
first tubules to be cut off from the connection with the splanchnoccel were
at the posterior end. From the material at hand it cannot be determined
which are the first nephrotomes to appear although it would seem that
those at the posterior end are the last. The formation of the excretory
system proper seems to begin in a good many segments at about the same
time, but the development lags behind at the anterior end, and this gives
rise to the appearance in certain stages of this being the youngest part.

The mistake of looking upon the anterior end as the youngest part
led to the more serious error of supposing the nephrotomes were pockets
formed from the unsegmented mesoblast; for at the anterior end where
the tubules were apparently just appearing there were no indications
of nephrotomes, the tubules being connected with the unsegmented body-
cavity. Then in passing back, as the tubules became better and better
developed they were connected at first with very shallow pockets and then
with deeper and deeper ones. We now know that this appearance is due
not to the formation of segmentally arranged pockets, but to the dis-
appearance of nephroccels. In a review Felix’ pointed out the proba-
bility of this error.

In what was supposed to be the youngest part of the system the segmental
cuct was found to be in connection with the ecelomic epithelium. This
union is secondary, but it was then thought to be primary, and it was
supposed that the duct in all parts arose as thickening of the coelomic
epithelium. The theory was strengthened by finding the duct in con-
nection with the ccelomie epithelium in one individual in some of the

5 Felix, W.: Die Price’sche Arbeit ‘Development of the excretory organs
of a Myxinoid (Bdellostoma stouti Lockington)”’ und ihre Bedeutung fiir die
Lehre von der Entwickelung des Harnsystems. Anat. Anz., Vol. 18, Nos. 21
and 22, 1897.

132 Iixcretory Organs in Bdellostoma Stouti

posterior segments. But in younger embryos the duct has since been
found in sections where the splanchnoccel had not yet been formed.

In the anterior region the duct is occasionally either partially or
entirely absent between two tubules. This is to be looked upon as due
either to degeneration or to the failure of the duct to develop. In an
extreme case, in an embryo of stage 6, there was a break on one side
between tubules one and two, five and six, seven and eight, eight and nine,
and fourteen and fifteen, and on the other between one and two, five and
six, and nine and ten.

We now come to the differentiation of the pronephros of the adult,
the most important steps of which occur in the stages between B and C,
and were therefore not seen at all in the previous work.

In the oldest embryo of stage B the anterior end of the body was
injured, so that the position of the anterior end of the excretory organs
could not be accurately determined, but it was estimated that the system
had disappeared in about nine segments. It is true that some of the
anterior tubules may degenerate, but in no case in the material in hand,
where the point could be accurately determined, can it be said that so
many as nine have disappeared, so that the above estimate must be
considered erroneous. What was said regarding the pronephros in em-
bryo C was of rather a tentative nature, the preservation not being of
the best, and it being impossible to work out the structure satisfactorily.

In the youngest embryo before studied the gill-shts were found well
forward in the head region, while in the oldest they were well back in
the body region, occupying the same position as in the adult. On en-
tirely insufficient evidence it was assumed that the change had been
brought about by slits degenerating at the anterior end and new ones
being added at the posterior end, and that, in the course of development
a good many more slits appeared than were present in the adult.
Dean * has shown that this is wrong, the first slits formed being perma-
nent, but shifting their position from the head region to the body region.
This change in the position of the gill-slits seems to be the cause of an
important change in the excretory system. At all events, as the gills
become shifted farther and farther backward the anterior tubules be-
come crowded closer and closer together, until finally they form a small,
compact body, the pronephros of the adult, in the region a little posterior
to the thirtieth segment. The crowding affects all of the open tubules,

°* Dean, Bashford: On the Development of the California Hagfish, Bdel-
lostoma Stouti, Lockington. (Preliminary Note) Quart. Journ. Micr. Sci,
No. 158. Vol. 40, 1897.

Geo. C. Price 133

and a small though variable number of the closed tubules as well.
At all times the anterior end of the excretory system is a little back of
the posterior end of the branchial system. For quite a long period the
‘history of the differentiation of the pronephros of the adult is simply
a history of the crowding together of the tubules, as will appear from
the following account of the organ in a series of embryos of different
ages. Before beginning, it may be stated that in all cases every section
of the parts described was drawn, different colors being used for the
different tubules. This was highly advantageous in all cases, and was
absolutely necessary in the older and more complicated specimens. The
arrangement is not usually exactly the same on the two sides, but unless
there is some really important difference only one side will be described.
The first embryo of the series is a little older than the oldest of stage
A, and the process of crowding together of the tubules has just fairly
begun. The first tubule is in the sixteenth segment and is not con-
rected with the rest of the system. In segments sixteen to twenty-four,

Fig. 27.—Sagittal section through the anterior end of the excretory organs in an
embryo in which the tubules are beginning to be crowded together, formed by the
combination of two adjoining sections, There are here five tubules in the space of
three segments. d, segmental duct; t, tubules.

that is in nine segments, there are twelve tubules. Back of this the
position of the tubules has not been affected. If all were arranged seg-
mentally the first tubule would be in the thirteenth segment. The first
three or four tubules are not so well developed, as the following, and look
as if they might be in the process of atrophy, although this is not at all
certain. Fig. 27 is a combination drawing made from two sections,
showing five tubules in the three segments, twenty to twenty-two. There
is quite a little space as yet between the tubules. The segmental duct
is almost without a lumen, but a little further back the lumen begins,
and continues with only a few interruptions to the posterior end of the
duct. However, the duct does not open to the exterior.

In the next embryo the first tubule is in the twenty-second segment,
and in the six segments, twenty-two to twenty-seven, there are fifteen
tubules. They are not evenly distributed, being more crowded in the
anterior part than in the posterior. But in no case are tw» tubules in
actual contact with each other. One or two at the anterior end are
cuite rudimentary, and there was some doubt as to whether they actually

134 Excretory Organs in Bdellostoma Stouti

represented tubules. All are connected with the segmental duct. If
they were arranged segmentally the first would be in the thirteenth seg-
ment. This embryo was sectioned transversely, and it could be deter-
mined accurately that the tubules in the thirtieth segment were the first
to be closed off. However, those in the twenty-ninth segment looked as if
they might be closed off later.

The first tubule in the next embryo is in the twenty-fourth segment.
1t is not connected with the rest of the system and is separated by quite
a wide space from the second, which is in the twenty-fifth segment. In
segments twenty-five to thirty there are sixteen tubules, and of these the
first ten, which are in the space of three segments, are so closely crowded
as to be almost if not quite in contact with one another. If all were
arranged segmentally the first tubule would here be in the fourteenth
segment.

In the next embryo there are seventeen tubules in the four segments
iwenty-seven to thirty. Of these, the first thirteen are in so short a
space that there is not room for them to stand in a row one behind the
other, and some are beginning to be crowded to one side—a process
which is carried much farther in older embryos. A comparison of Fig.
28, taken from this embryo, with Fig. 27, shows how much closer together
the tubules are here than in the younger embryos. The coelomic opening
of most of the tubules shown in Fig. 28 are found in other sections.
The segmental duct is no longer straight, and so does not show in the
full length of the figure.

The next embryo is particularly interesting, because here the crowding
process has affected not only all of the open tubules, but the three
anterior closed tubules as well. The posterior limit of what will be the
pronephros of the adult may be placed, with some degree of probability,
between the third and fourth closed tubules, for at this place the seg-
mental duct shows some slight indications of degeneration. Owing to
an injury in the anterior region of the body, it is not possible to deter-
mine the exact position of the pronephros, but it occupies the space of
a little less than two segments. There are in all twenty-one tubules,
eighteen open and three closed. The first and second tubules are con-
nected with each other, but not with the rest of the system. The three
closed tubules have the same structure as those farther back, and are in
direct line with them, but they occupy the space of only three-fifths ot
a segment, and the third slightly overlaps the second. Just at the point
where the anterior closed tubule joins the duct, the latter bends down-
ward, backward and shghtly outward, and then turns and runs again
forward, thus forming a sort of s-shaped bend. In this way some of the

. Geo. C. Price 135

vpen tubules come to lie ventral to the closed tubules, and also slightly
more lateral, and a transverse section may show parts of one or even of
two closed tubules, and at the same time parts of several open tubules.
Fig. 29 represents such a section. The pronephros now forms a promi-
nent body projecting into the eelom. The segmental duct is cut through
its posterior bend, and four open tubules are seen to be in connection
with it. By running along the series the lumen of these may all be
traced to a connection with the duct on the one hand and with the
coelom on the other. Above, one of the closed tubules is cut through its
full length, but the lumen appears only at the point where it joins the
duct. Farther on in the series it may be traced to a connection with the

Lez, \-
a PZ ral
STASIS
BE ROE
Sorntenaeent
¥:

(
oa
Ns

f coe
Fia. 28. Fria. 29.

Fic, 28.—Sagittal section showing parts of seven tubules, and illustrating the way
in which they become crowded together. In studying the series all the tubules are found
to open to the ewlom. ca, celom; bv, blood-vessels; d, segmental duct; t, tubules.

Fic. 29.—Section of a pronephros showing the two kinds of tubules, four open tubules
below and a closed tubule above. emec, cavity of the Malpighian corpuscle; d, seg-
mental duct, here shown in two places; gl, glomerulus of the closed tubule; t, tubule.

cavity of the Malpighian corpuscle. ‘There are places in the pronephros
where the segmental duct has no lumen. ‘This is the first embryo in
which the segmental duct was observed to open to the exterior at its
posterior end.

In an embryo slightly older than the last, the pronephros was not
thoroughly studied, but it was observed to be more sharply set off from
the mesonephros than in the last, although the segmental duct was still
continuous between them. It occupied parts of the thirty-third and
thirty-fourth segments. The number of closed tubules was not deter-
mined, but two glomeruli were seen to be in very close contact with each
other.

The next embryo is somewhat younger than embryo C of the previous

e

13 Excretory Organs in Bdellostoma Stouti

paper. The pronephros on the right side is in segments thirty-two and
thirty-three. There are in all twenty-one tubules, and of these the first
three haye no connection with the segmental duct, while the last three
can be traced to a connection with Malpighian corpuscles, and must
therefore represent closed tubules. On the left side the pronephros is
in segments thirty-three and thirty-four, and there are here nineteen
tubules instead of twenty-one. The first two are not connected with the
duct, and two instead of three rerresent closed tubules. If the tubules
were arranged segmentally, the first would be in segment thirteen on
the right side and sixteen on the left, while the anterior closed tubule
would be in segment thirty-one on the right and thirty-three on the left.
On both sides the pronephros is connected with the mesonephros by the
segmental duct, and in neither case does it oceupy the full width of two
segments.

Fic. 30.—Sagittal section through the pronephros showing several open tubules, all
but two of which are cut in longitudinal section. In this embryo the limits of the
pronephros are definitely established. be, blood corpuscles; d, segmental duct; t, tubules.

Fic. 31.—Transverse section through the pronephros of the oldest embryo studied,
showing one of the open tubules, t, in the process of branching. bv, blood-vessels; gl,
glomerulus.

Fig. 30 represents a sagittal section through the left pronephros.
Several tubules lie in about the same plane and are cut longitudinally.
Their connection with the segmental duct is shown. ‘Two are cut trans-
versely. Farther on in the series these also join the duct. The section
is not near enough the median line to show any of the glomeruli.

In the last embryo to be considered, which is some little older than
the embryo C, the exact position of the pronephros could not be deter-
mined, owing to an injury farther forward, but it is located in the

—

Geo. C. Price 13

pericardial cavity, in a position corresponding to that of the adult.
There are on the left side twenty tubules, two of which can be traced to
a glomerulus and must therefore represent closed tubules. Some three
or four of the open tubules have begun to branch, as is shown in the
ene in Fig. 31. This is the beginning of a process which results in
the formation of the scores of pronephric funnels found in the adult.
The two closed tubules are small and insignificant, but the glomerulus
to which they go is large and well defined, and has something the
appearance of having been formed by the fusion of two glomeruli. At
one point, shown in Fig. 31, the cavity at the side of the glomerulus,
the cavity of the Malpighian corpuscle, is almost in communication with
the celom. This is interesting as suggesting that in some cases the
communication might actually occur, in which case the glomerulus would
be virtually in the body-cavity. However, this is to be regarded as only
a passing remark, and not as suggesting a homology between this glomer-
ulus and the glomus of the pronephros in other forms. Nothing of
the kind occurs on the other side, nor has it been observed in any other
embryo. There is a complete break in the segmental duct between the
pronephros and mesonephros. In the pronephros itself the duct is dis-
tinct and the lumen well defined through about half its course on one
side and two-thirds on the other. In the rest it is solid, and in places
quite small. It may be remarked that no case has yet been observed
where the lumen of the duct is continuous throughout the entire pro-
nephros.

A point that has not been worked out is the manner in which the
segmental duct becomes shortened as the tubules are crowded together.
The bending of the duct previously described accounts for part of it
though hardly for all.

The account of the pronephros just given does not fit in well with
Mass’ description of this organ in a young specimen of Myxine, where
there is no segmental duct; but the difference is hardly to be accounted
for on the supposition of errors of observation in either the one case or
the other, nor is it likely that later the pronephros of Bdellostoma would
come to be more like that of the young Myxine. An examination of the
pronephros of the adult Bdellostoma reveals the presence of a true
segmental duct, although it may not be continuous through the entire
organ, while, judging from the published accounts of the pronephros of
Myxine, the presence of a true segmental duct here would seem to be
very doubtful. In the light of this the above discrepancy is not sur-
prising.

™Mass, Otto: l.c.

138 Excretory Organs in Bdellostoma Stouti

The presence of the excretory organs in segments where later the gill
slits were found, was formerly thought to constitute a resemblance
between the excretory system of Bdellostoma and that of Amphioxus, in
which the position of the excretory system and branchial system coincide,
but this opinion is no longer held.

WILHELM HIS.

His RELATION TO INSTITUTIONS OF LEARNING.

BY
FRANKLIN P. MALL.

Professor of Anatomy, Johns Hopkins University.

The ancient science of anatomy has been perpetuated and extended
during the many centuries of its existence by great men who have dedi-
cated their lives tc it. The list is a long one for the development of the
science has been slow and progressive from the earliest ages to the present
time; we find in it on the one hand, some of the names of the greatest
who have ever lived—Aristotle, Vesalius—on the other, the names of
those who rank as leaders of a generation, Bichat, His.

Undoubtedly the reason for the continuous progress of anatomy
through so many centuries is that man has always shown an interest in
a knowledge of his own structure and, in turn, that this knowledge has
been of great service in battling for rationalism against mysticism in
many directions. But in order that a science may progress, it must be
in the hands of able men with the highest ideals and with inexhaustible
zeal. The full benefits of a science cannot be obtained from its literature
alone; it is well known that countries without leaders in a science live
almost in ignorance of it. No science will develop to its fullness unless
it is represented by men who are great enough to grasp the science as a
whole and broad enough to understand its relation to cognate sciences
as well as to the needs of a civilized community.

A man of this caliber was Wilhelm His. He was born of a distinguished
family which felt that it owed much to the community and therefore
educated its children for a strenuous life to be dedicated to the com-
munity. He was in no sense a self-made man though he had in him the
qualities to make one; his powers were fortunately developed under the
best possible conditions. His father taught him, by example, simplicity
in living, clearness of thought and seriousness of life. His mother, who
died while he was still a youth, guided his education with the greatest
care, laying much stress upon a good command of German. In a pri-

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

10

140 Wilhelm His

vate school in a country town, he received a training so rigorous that it
makes one shudder to think of it. At the age of thirteen, he returned
to his home in Basel and entered the Gymnasium from which he gradu-
ated five years later, in 1849." During his years in the Gymnasium, His
rated only as an average pupil, for his bent toward physics and natural
history was great, and what spare time he had was devoted to these sub-
jects. He also photographed a great deal, the art then being still in its
infancy; he constructed his own camera and made his own plates. In
those days the professors of the University taught also in the higher
classes of the Gymnasium (the Pedagogium, as it was called), and it was
from them that he received his greatest inspiration. Especially grate-
ful was he to the Professor of German for a thorough training in the
use of the German language in which he had received such a good start
in his father’s house.

Throughout life the good command of language he had gained and the
marked technical ability he had acquired were of the greatest use to him
in furthering the cause of anatomy; it is interesting to note that these
powers were developed long before he entered the University. It was
fortunate for him and for anatomy that he came under the influence of
able men early in his career.

Most of his student friends had decided to study law and naturally
His thought that he must accompany them. His inward feeling was
drawing him towards natural history, but he felt that one’s inclination
is to be viewed as a forbidden fruit and therefore he had nearly decided
to go with the stream of his fellows. Before taking the final step,
however, he consulted his superiors, one of whom was Winscheid, later
one of the foremost jurists of Germany. Winscheid advised him to fol-
low his own bent and this carried him into medicine.

As soon as he became a student of medicine in the university, he found
himself free to follow his own inclinations. For the first few years he
devoted most of his time to the sciences. Anatomy, physiology, natural
history, psychology, chemistry and geology were studied at Basel and
Bern under a variety of men. He soon learned that for the student, the
quality of the teacher is more important than the subject taught; in
order to have a great teacher and the important subject together, he de-
cided to study anatomy with Johannes Miiller, in Berlin.

The lectures of Miiller were a revelation to him and he felt for the
first time the inspiration of a strong personality. His learned from
Miiller and later from Remak, how great an influence a teacher can have
upon a pupil, when, as an authority, he presents his own line of work.
Then the teacher and pupil stand upon the common field of nature. The

Franklin P. Mall 141

power of the successful explorer is most stimulating. It encourages
the pupil to attack open problems, and, if he can, he formulates ques-
tions and takes his position towards them. Thus Miiller’s presence
worked upon His and unconsciously the latter soon found himself in the
library studying Miiller’s monographs.

His also pursued a course in embryology with Remak in Berlin. The
subject was then a new one, but it made a profound and lasting impres-
sion upon the young student for it showed the relation between histology,
embryology and comparative anatomy. To Remak, His owed more
than to Miiller, for Remak’s teaching helped him to formulate problems
which oceupied him during the following half century; his efforts to
solve them, have, in many respects, changed the aspects of anatomy.

How different is the study of medicine in Europe from that in Amer-
ica! ‘There freedom reigns and students wander from place to place
being controlled only by a fairly rational system of examinations in case
they wish to graduate. Weak students fall out, for there is no cram
system to drive them onward; able students select great men as teachers
and thereby develop themselves and become stronger. So His, soon tiring
of the sterile lectures in practical medicine, turned his eye towards a
new star which was beginning to illuminate the medical world and
wended his way to Wurzburg to study with Virchow. During one of
Reinhardt’s lectures, His had happened to see Virchow’s great paper
upon connective tissue which had just appeared and it made so great
an impression upon him, that he at once decided to go to Virchow. He
was then far from prescient of the fact that in a few years he was to
take an important part in the great discussion which Virchow had
opened.” .

In the early fifties, the University of Wiirzburg was at the beginning of
that rapid rise coincident with the appearance there of a brilliant band
of young professors. The change was brought about largely by a re-
former, Professor Reinecker, through whose personal efforts Virchow and
K6lliker were secured. It was the good fortune of His to enter this
atmosphere; a select faculty was chosen by a talented student. He en-
tered Wiirzburg at the beginning of his fourth year of medical study
and began to work in the clinics. He soon found that the laboratories
attracted him more than the clinics and during three semesters much
time was devoted to practical chemistry; the weekly meetings of the
Scientific Society always found him present to hear of the new medicine
from Virchow and Kolliker.

For a time he attended regularly a kind of journal-meeting at Kol-
liker’s house where a variety of scientific subjects were discussed. There

142 Wilhelm His

he became acquainted with Ludwig’s physiology which was then hetero-
dox and called forth severe criticism. Ludwig, very critical towards
others with pupils who were more so, kept the camp stirred up pretty
well; this had the right effect upon his critics (His included) for they
studied the new physiology. Converts were made at a rapid rate and
before His had reached middle age, he found that both Virchow and Lud-
wig were very orthodox; the scientific world had come to them.

In Wiirzburg, a great opportunity came to His—Virchow at once set
him at work in a good field to answer fundamental questions. Probably
the most important step in the life of a scientist is to be started aright
in research after a good preliminary training. The earlier this is done
the better. His was twenty-one years old, mature for his years and he
was given every opportunity to use all of his ingenuity and strength in
answering important questions. Henle and Virchow had crossed swords,
but this did not make the great pathologist try to enslave the mind of
His in order to win a victory, but now as so frequently afterward, a
desire to know and to understand were Virchow’s only considerations
in directing a pupil in his work. The results of His were not what
Virchow had anticipated for they showed that Bowman’s corneal tubes
and Coccius’ serous spaces were artifacts and had nothing whatever to do
with the system of connective-tissue spaces described by Virchow. But
His was encouraged to continue and the monograph upon the cornea
which he published a number of years later has proven to be a standard
until the present day. His had been in the arena and had shown his
prowess; he had originality and strength as well as training from a great
master and henceforth during all his life he was to win victory after
wictory for science.

It was still necessary for His to take his medical degree and as was
customary with many at that time, he proceeded to Prague and Vienna
to study with the great clinicians there. At the end of a year he re-
turned to Basel and passed his examination with the highest mark.

The young student had received the best from his home, the school, the
gymnasium and the university of his native city, had wandered for four
years studying at famous foreign universities receiving information and
inspiration from the greatest masters—Miiller, Virchow, Kolliker and
many others—and had now returned to Basel to receive his degree. How
much longer must we wait for similar privileges in-America?

The following three years of leisure and lack of responsibility, those
in which the real stuff in a scholar is tested, were devoted to a continu-
ance of the work already so well begun.” Part of the time was passed in
his private laboratory, one of the years being broken by a journey to

Franklin P. Mall 143

Paris, another by a visit to Berlin. Throughout this period he associated
only with the best, for there was a never-wavering desire in him to de-
vote his life to science. During this time he became a Privat docent in
the University of Basel and gave regular lectures there on histology.’

In most of the continental universities any person approved by the
faculty may teach and thus new blood is constantly being infused long
before a fogy or a dog in the manger has been removed by a beneficent
Providence. There is thus maintained a constant competition for better
teaching within the walls of the university, and strong young men are
brought to the front. If under these conditions a young scientist takes
deep root without being spoon-fed or coerced, and commands a broad
field, adding to its borders, the greatest assurance has been given that he
will remain active and productive until he is three score and ten. In
America, we frequently find recent graduates who tell us that they would
follow an academic career if their future were assured as far as salary
is concerned. Little do they realize that this attitude of mind alone
should exclude them absolutely from such a career. Unfortunately, we
seem to have some university presidents who are as easily deceived by
such “scientists” as the public is by a Mesmer. Within a year, I have
known of a president who, when seeking a great anatomist for a rich
university, selected a man who had done no scientific work whatever
and had never been tested, simply on his own assurance that he would
“try to do something.”

In 1857 the chair of anatomy at Basel became vacant, and, as is cus-
tomary, the faculty sought the best available man. They found in His
an able earnest young scientist of the best training, tested through free-
dom and research, whose work had been continued with increasing suc-
cessful results through the three years of leisure following his advance-
ment to the doctorate. A man who had studied a subject’ for its own
sake ° and had contributed to it, was more likely to represent it well than
one who had studied it for other rewards. Only too often do we see
scholars whose work is good and imitative, but net profound, shift from
one thing to another in order to keep before the public in an upward
career; they find themselves sterile at forty and have to be shelved in
some good berth as an active “ pensioner,” at fifty. To avoid this dan-
ger, great productive men must sit in faculties, for they alone are able
to recognize genius in a young man.

When the Chancellor of the University informed His of his pee
ment as professor of anatomy and physiology, he said: “ We have thrown
you into the water, learn to swim;” and the remark was appropriate,
for the new professor had never ae an apprenticeship in the teach-

144 Wilhelm His

ing of either anatomy or physiology. Often did he seem to sink but he
always came to the surface again, for he had overcome great difficulties
before—he was independent. After all, research is to the problem of
teaching as manceuvres are to a battle.

The power in His now blossomed and bore fruit. With his great back-
ground he was able to construct strong courses, marked by his individu-
ality, and soon he was to influence the teaching of anatomy the world
over. The attitude of the comparative anatomist he had learned from
Johannes Miller, that of the histologist from Virchow and Wolliker,
but that which made the greatest and most Jasting impression upon him
was the attitude of the embryologist which he learned from Remak. The
histological work begun with Virchow was continued and soon extended
to include the lymphatic system. In 1862 he published a paper on
lymph radicles, advocating a closed lymphatic system, by all odds the best
paper on that side of the question which has ever been published.

During this time, his embryological studies were also actively pros-
ecuted for he had learned of their great value in histology from Remak.
The classic object, the chick, as well as the structure of the ovary was
studied again. That a plan underlay his work was very apparent, but no
one dreamed of its magnitude until the publication of his academic pro-
eram in 1865, entitled “ Die Haute und Hoéhlen des Korpers.” In this
paper he gives the key by which the genetic relation of tissues can be
ascertained and following it faithfully, he made one discovery after an-
other. The “ Program” is now incorporated with our science; it proved
to be really a program for anatomy, and it was fitting that it should have
been reprinted as the last paper during His’s editorship of the Archiv fiir
Anatomie, nearly forty years after its first publication. It certainly
must be gratifying to a scientist to see his early dreams so well realized
before his Work is over. .As His said of Bichat, “It is a mark of genius
to see great truths in a relatively small number of observations.” °

The great contribution to anatomy during the eighteenth century
was the discovery of the tissues, the conception of which received its
full development in the general anatomy of Bichat, published in 1801.
Cellular tissue had been gradually making its way during that century
and among others, Haller was trying to see in it some unit of organiza-
tion, but the cell of Haller proved to be only a connective-tissue space
and the all-important fiber was viewed by most anatomists as an artifi-
cial production. But the doctrine of cellular tissue was the foundation
of general anatomy and this in turn that of histology which, through
embryology, has given us modern anatomy.

During the nineteenth century we see three great steps in anatomy,

Franklin P. Mall 145
general anatomy, associated with the name of Bichat, the cell doctrine
with that of Schwann, and histogenesis with that of His. Compare
the great text-books of anatomy of 1790 with those of 1810, those of 1830
with those of 1850 and those of 1880 with those of 1900 and the rea-
sons for this statement will be apparent.

At the time His began his embryological studies the plan of the de-
velopment of the vertebrate body, as enunciated by Von Baer with
Remak’s classification of tissues, had been accepted generally. But the
work of Remak was incomplete and in some respects unsatisfactory as,
for instance, his conclusion regarding the development of the nervous
system, the central portion of which he believed arose from the ectoderm
and the peripheral portion from the mesoderm. This and other defects
were corrected by His who made a new classification of tissues and germ
layers which differs more from Remak’s classification than this in turn
did from Von Baer’s.

In the “ Program” His also showed that there is an embryological
foundation for Bichat’s classification of membranes since they are related
directly to the germ layers. Further, he extended the conception of the
serous spaces to include the vascular system. All of the serous spaces
arise in the mesoderm, and His showed that they are lined with a special
kind of cell, designated from the time of his paper on, as endothelial.

After this, his greatest work was histogenetic—witness for instance his
studies upon the nervous system and upon the development of the blood-
vessels. Contribution after contribution was published upon histogene-
sis. During the last years of his life he often complained to me that his
time was short and that it was necessary for him to make haste in order
to round up his work.

Thus His continued to be active until he was over seventy years old;
his final papers, “although fragmentary,” are of the highest quality.
In his great paper upon the angioblast he gave his latest classification of
tissues from an embryological standpoint * and stated in conclusion that
the riddle which had interested him so much still remained unsolved.
Similarly, in his last monograph upon the nervous system, he wrote:
“The notes are not to be considered as a report of a finished research,
but they only indicate the line of work which must be followed by united
effort ° in order that a better understanding of the structure of the brain
may be gained.” He saw more and more clearly that many hands are
required to survey these great fields and during the last twenty years of
his life, thought much about the organization of research. He lived to
see his efforts in this direction crowned by the establishment of a com-

146 Wilhelm His

mittee for the study of the brain by the International Association of
Academies, of which more presently.

In 1872, the chair of anatomy at Leipsic became vacant and through
the efforts of Ludwig, His was secured. Professor Ludwig told me that
it was by no means the unanimous will of the faculty to call His, for in
general he was not well known, nor was he considered a good teacher.
But Ludwig knew his man for he had found in His the strongest oppo-
nent of his notions regarding lymph radicles; furthermore, he was fully
able to appreciate the author of the great academic program on “ Hiute
und Héhlen.”* For over thirty years Ludwig and His were colleagues,
consulting each other almost daily, an ideal relation for great scientists.

The work of His had branched in many directions at Basel, but it
grew with increased vigor at Leipsic, for here he had all the material
aid he could desire. Photographer, modeller, mechanic, technician,
artist and others were at his command. A new laboratory, which proved
to be a model, was built according to his ideas. Lach of these factors
was to play a part in the campaign he was conducting and it was soon
‘seen how his hands were thus extended.

Through the better technical assistance, His increased his productivity,
for his plan was broad enough to use it to the greatest advantage. The
members of his staff, however, were never subject to his orders in their
scientific investigations, nor did he ever use “research assistants” for
his ethical standard would not permit such employment of scientists.
That a great man can increase his productivity enormously without the
questionable use of young colleagues is shown by example in the life of
His. Furthermore, his plan of organizing research is one of the best
from both the ethical and the scientific standpoint.

The microtome which His had invented in Basel was now perfected
and better serial sections were’ made of embryos of different classes of
vertebrates than ever before. The great monograph upon the chick had
just been published and smaller but equally important papers appeared
upon fishes; these were soon to be followed in 1880 by the first part of the
monumental work upon human embryology.

Throughout his embryological work His was constantly interested in
the broader problems and he was among the first to view development
from a mechanical standpoint. His views and general plans were
brought together in 1874 in the classic, “ Unsere Korperform” dedi-
cated to his new colleague, Ludwig, appropriately for in it the subject
of development is presented from a physiological standpoint. These
views were immediately antagonized to the utmost; when arguments
failed, his opponents resorted to ridicule, but, in general, His held his

Franklin P. Mall 147

own and continued a forward course. He now became the leading advo-
cate of the theory of mechanics in development and it subsequently
tinctured all of his papers. Although he made no experiments upon the
growth of animals, he must be viewed as one of the pioneers of the new
science of experimental morphology.

Throughout His’s embryological papers we see that he believed that
the form of the animal body, as well as of its organs and tissues, is due
to mechanical influences of the structures upon one another. ‘This con-
ception was often erroneously construed as meaning that there is a me-
chanical cause for growth, but His repeatedly denied this. “'The stimu-
lant which causes cells to multiply cannot be traced to mechanical
influences.” His mechanical conceptions of development are compara-
ble with the study of the mechanics of the circulation rather than with
that of the evolution of the heart. When analyzed, it appears to me
that his mechanical conceptions are related principally to the wandering
of tissues and organs in development. One of the best examples of His’s
ideas is his conception of the growth of nerve fiber from the central cell
to the periphery, where, through secondary connections, it makes itself
fast to its end organ. With this conception, we can understand and
picture to ourselves the formation and the infinite number of variations
of the peripheral nervous system. Other examples may be found in the
wandering of the diaphragm and in the metamorphosis of the branchial
arches. Mechanical influences must guide these structures to their fate.
And finally, great masses of tissue wander in the embryo long before we
can see what is to become of them. The best example of this kind is to
be found in the early development of fishes formulated in His’s brilliant
theory of concrescence. According to His, the word “mechanical” is
to be applied to the movements of these avalanches of tissues and their
influence upon one another in gaining their final position and not to the
cause of growth.

One of the characteristics of His’s embryological work is that he viewed
the embryo as a whole; he always approved of embryological papers which
carried a subject to its logical conclusion.” As he improved the micro-
tome more and more, so that he could cut serial sections fifty microns
thick, he constantly kept in mind the relation of the individual section
to the embryo as a whole; this led to his well-known method of graphic
reconstruction in 1868. At that time haphazard sections of an embryo
were often compared with chance sections of older embryos which led to
all kinds of erroneous conclusions. In the course of time, he perfected
his method by drawing an enlarged picture of the embryo upon ruled
paper into which he projected in their proper positions the sections

148 Wilhelm His

enlarged to the same scale. Photography, as well as a new instrument
which he invented, the embryograph, aided him in his work. In order to
be better able to compare stage with stage, the reconstructions were con-
verted into models which were duplicated by Ziegler; these models now
form the most valuable asset of many of the embryological museums the
world over. More accurate models can be made by drawing the sections
upon wax plates which when placed upon one another reproduce the
embryo; this method, invented by Born, was used to a very great extent
by His in his later years.

Until this time the embryological campaign had been conducted by
His from all sides, but cne great fortress, the human embryo, stil] stood
before him. Human embryos were very difficult to obtain and the litera-
ture upon the subject was meagre and poor, but His was soon able to
select a few important stages from a mass of poor material, normal and
pathological. These he studied with such care that the knowledge of the
anatomy of the human embryo now exceeds that of any other animal.
To be sure the earliest stages were missing, but this mattered but little,
for being master of the whole field of anatomy, he was able to produce a
model work. Each chapter in his monograph is great and original,
giving a mass of information; the work reaches its climax of interest
and value in the part on the nervous system.

During the last dozen. years of his life, His’s attention was taken away
from the study of the nervous system by a variety of subjects; thus, for
example, he was interested constantly, as his letters show, in the
embryology of fishes.” The work on this latter subject he was able to
round out, but it seemed as if his promised work upon the brain would
never appear. When I visited him the summer before his death, I found
a broken man, working away at a large manuscript which by no means
satisfied him. What could be arranged, was published in a monograph
upon the brain, which will serve as a foundation for investigation for
many years to come. He had hoped to write another volume, but his
illness rapidly grew werse, and, unable to work longer, a week before
his death, he wrote that his remaining wish was that the end might
come soon.

The technical ability of His which had been pretty well trained in the
gymnasium, gradually developed farther and proved to be of great value
to him as a teacher. Optical instruments of all kinds, magic lanterns
and microphotographic apparatus were much used by him; demonstra-
tions with the microscope after each lecture aided in illustration. The
pictures which he drew upon the board while lecturing were models of
their kind and he developed them before the students in such a manner

Franklin P. Mall 149

that they could be copied gradually while being evolved. The subject
matter of his lectures was chesen in a very conservative way, the sub-
stance being always sound and free from all kinds of wild theory and
speculation.”

Early in his career, he had made crude models of the mesentery and
the like, for these were subjects the forms of which were difficult to
understand. Toward the end of the seventies, modelling was prosecuted
on an extensive scale with the aid of the modeller, Steger, resulting in a
series of papers on the form and position of the organs which are now
standard. His many models have been duplicated and fill an important
corner in all anatomical musewns; they have made His the founder of a
new school of topographical anatomy. The method which His had in-
troduced in embryology were thus also applied to gross anatomy, for he
was never satisfied until he could see adult forms in the embrvo and the
outlines of the embryo in the adult.

In His the power of visualizing forms—the power which enabled him
to do his work of reconstruction, to make those wonderful blackboard
drawings during his lectures—was developed to an extreme degree. This
power was one of his principal gifts, and one of the chief foundations of
his achievements in science. As he grew older there was not only an
increase in the depth of insight into problems, which is natural in so able
a man, but also what is rarer, a very great improvement in the power of
expounding his results. His last papers are models, characterized by
conciseness of style, great clearness of description and a suppression of
all superfluous details.

While His was teaching and investigating, the question of nomencla-
ture came up. Lach century, each country, each school, each specialty’
and each teacher seemed to have a particular group of terms based upon
an imaginary normal; the result was that there were so many normals
that it was extremely difficult to construct a table of synonyms.

When His and a few others founded the Anatomische Gesellschaft, |
one of the first questions discussed was the formation of a uniform
nomenclature. After much work and expense, His drew up ‘she official
report of the international commission, in 1895; it is another standard
which the leading teachers and authors have agreed to follow. The new
nomenclature is a compromise; it is not radical, but it has reduced the
number of anatomical names, including synonyms, by about eighty per
cent. It will not be difficult for English-speaking anatomists to accept
this terminology, for it differs less from ours than from that of any
other language.

If we consider the great amount of work His did upon the form and

150 Wilhelm His

position of the organs, upon the general morphology and structure of
the brain, upon embryology, histology and histogenesis, and upon
anatomical terminology, it may safely be said that there is barely a page
in the broad field of anatomy from the ovum to the adult in which his
work does not appear.

Early in his career, His showed much interest in physical anthropology
as we see by the great monograph he and Riitemeyer published on Swiss
skulls. But his time was occupied in so many other directions that it was
impossible for him to continue in this kind of work with the inadequate
assistance he had at Basel. However, some thirty years later, he had an
opportunity to open up a new line of research in anthropology. A skele-
ton, presumably that of Bach, had been found and His was asked to give
an opinion regarding it. It was known that Bach had been buried in an
oak coffin near a certain corner of a church-yard. Here among others
was found the skeleton of an elderly man (Bach died at the age of sixty-
five) in an oak coffin. The skull was found to be peculiar and in it the
anatomist could discern the features of the portraits of Bach. His at
ence proceeded to measure the thicknesses of the soft parts over the bony
prominences of the heads of cadavers and found that in average bodies
these thicknesses are constant, varying only with age and sex. He next
drew averages from the measurements taken from cadavers of elderly
men of fair development, and, with these, Seffner, the sculptor, con-
structed a clay bust on the skull in question. It was found that the re-
constructed bust presented all of the characteristics of Bach even more
pronouncedly than do his portraits. The commission that had the matter
in charge decided that the skeleton in question was undoubtedly that of
Bach. As such, it was reinterred. The musical world through His’s
studies now has a bust of the great composer.

These results gave His the greatest esthetic pleasure, for they meant
a new victory.” From this time on he was greatly interested in induc-
tive anatomy and when I began my career at the Johns Hopkins, he
gave me every possible encouragement in this direction.“ He often wrote
and he often talked about statistical work, but little did I realize the
difficulties of it until we began to tabulate a peripheral nervous system
from a very large number of individual records. It became apparent
from our records that variations are more common in certain parts of
the body than in others and this result interested His very much.”

His was never inclined to develop a school nor was he anxious to have
pupils. When I knocked at his door at first I was turned away, but
after appearing a number of times, was finally accepted. When he set
a problem, it was concisely stated; he outlined the general plan by which

Franklin P. Mall 151

it was to be solved. All of the details were left to the pupil and it
annoyed him to be consulted regarding them. He desired that the pupil
should have full freedom to work out his own solution and aided him
mainly through severe criticism. Specimens and drawings of them
which were not analyzed did not appeal to him and he objected much
to pictures which appeared to represent a mass of “baked tissue.”
Through reactions, either with coloring matter or with some destructive
reagents, or by means of reconstructions, tissues must be emphasized
and we see this characteristic in the illustrations of all of his publica- —
tions. A His drawing can always be recognized even if it appears
without his name attached to it.

His was unwilling to give his own problems to pupils and though later
in life he advocated the establishment of research institutes, it is not
altogether clear how he reconciled the one attitude with the other. But
the institute is rather for routine research and for a discussion of work
by leading investigators who consequently formulate problems to be solved
by organized united effort; it is not intended to dwarf individual effort
in any respect. His was unwilling to write papers for his pupils and
the manuscripts they placed before him were improved only through
erasure, for he excluded all doubtful evidence and irrelevant matter.
“Your paper will be read by a few specialists, and they do not want a
treatise on the science,” he would say, and this criticism coming to a
pupil during the same years that silence on such subjects and encourage-
ment came from Ludwig, proved to be of the greatest value.

His did not have much power to extend his own private work through
his assistants and pupils; they were always given the greatest freedom
for it was against his nature to enslave them to the least degree. Nor
did he possess the patience of a Liebig or a Ludwig in training others to
follow in his path. Further, he did not find that successful research can
often be stimulated by exainple, although men, and scientists too, are
imitative. He often told me that the acceptance of a discovery is fre-—
quently postponed by numerous “ confirmatory” publications which
are filled with so much crude and irrelevant matter that the real point
is buried again; and that the desire of authors to attract atten-
tion often induces them to invent names and write much, thereby making
the answer to a question more obscure than it was before.

Several visits to the Zoological Station at Naples made a profound
impression upon His for it showed him what organization can do for
research, and in 1886 he published a paper on the necessity of research
institutes. In this he pointed out that the function of such institutes
is (1) to solve problems which exceed the power of one man to settle,

152 Wilhelm His

and (2) to collect, classify and conserve all the material relating to such
problems. After a whole life-time has been occupied in collecting
material, it seems a great pity that so much work should be lost at the
death of the collector; could such material and such lives be made availa-
ble for the solution of the great problems by an institute or institute
of institutes (as is the case in astronomy) more progress might be made.
In the foundation of such an institute the many details of embryology
and neurology should first be surveyed; were the field divided, this could
be done quickly. Problems would have to be formulated, and a com-
mon nomenclature and standard of measurements agreed upon. In no
case should individual effort be hampered. Conferences would be neces-
sary from time to time to compare, to criticise results and to formulate
new problems and new plans for their solution, thus aiding all with the
best, as should be the case in ideal scientific investigation. The under-
lying thought is to extend the power of able investigators through the
whole science without dwarfing in any way individual effort; in this way
the best qualities of all could be made to serve the progress of science.”
The Anatomische Gesellschaft was founded with this as one of its objects
in view; it naturally resulted as we have seen in the casting of a uniform
nomenclature. Two great steps had thus been taken and His lived to see
the beginning of a third.

The tendency in the world during all of His’s life was more and more
towards specialization and organization in science, a movement which
gradually became international. An index of this tendency may be seen
in the organization of the International Association of Academies and
through its machinery, His was able to launch his favorite scheme.

Unlike many promoters of science, His was not an impressario and
consequently the learned world had full faith in him. What he advocated
always was the advancement of science. The possibility of personal gain
was excluded from his thoughts as is shown by his attitude towards en-
dowments, especially the Nobel Fund.” He maintained all along that
a research fund must fall into the right hands if it is to benefit science
and that it was of positive injury to science when in the hands of im-
pressarios.”

His proposed at the first meeting of the International Association of
Academies, held in Paris in 1901, that a commission be appointed for
the promotion of the studv of human embryology and another for the
study of the anatomy of the brain. The Association agreed to the ap-
pointment of the latter, at the same time recommending that for the
present, human embryology should be taken up by the anatomical socie-
ties. Later, upon the recommendation of the Saxon Academy, the Royal

e

Franklin P. Mall iLays)

Society of London appointed a Neurological Commission of seven mem-
bers with His as chairman. 'The Commission had been established and
three bulletins had been published preparatory to the first general meet-
ing, but His died before the meeting convened. His plan will live, for
it has taken deep root and has the approval of leading anatomists of
the world.”

The life of His was a life of work, and his energy, industry and en-
durance were so great that he hardly knew the meaning of leisure. He
possessed the qualities of a courageous leader, but lacked the magnetism
that compels many admirers and followers. He was a daring and origi--
nal investigator, possessing great technical ability and artistic feeling; he
was fearless and honorable in controversy and knew no compromise. He
was a great character, true to his family, true to his friends and true to
science.”

Through His, another milestone has been set for anatomy. Through
him the great mother science has given birth to a new science, histo-
genesis. His career is marked by a monument of neurological research
which is unique. His’s life was that of the ideal scholar. During youth
he was strengthened through his own efforts, directed by great masters.
During middle age, he won many victories for anatomy, improving the
science in all its parts. In old age, he completed and rounded up his
work, leaving a great legacy to his survivors, no small part of which con-
sists of wise plans for future work.

REFERENCES AND EXTRACTS FROM LETTERS.

1 Wilhelm His was born in Basel on July 9, 1831. His father, Edward
His, was a son of the Swiss statesman, Peter Ochs. In 1818, when Edward
Ochs became engaged to be married to Anna La Roche, he assumed the name
of His, the maiden name of his father’s mother. He did this with the consent
of Peter Ochs in order to remove the ridicule of his name from which he
had undoubtedly suffered. After graduating from the Gymnasium, His
studied medicine in Basel, Bern, Berlin, Wtirzburg, Prague and Vienna, re-
turning to Basel to take his doctor’s degree, in 1854. Later, he studied in
Paris, returning to Basel as “ Privatdocent” in 1856. In the summer of
1857 he was again in Berlin and in the autumn he was appointed Professor
of Anatomy and Physiology in Basel. In 1872 he accepted the call to the
chair of anatomy at Leipzig, where he died May 1, 1904.

A charming account of His’s early life is given by him in his
Lebenserinnerungen (als Manuskript gedruckt), Leipzig, December, 1903.
See also W. SpaLTEHOLZ, Zum siebzigjihrigen Geburtstag von Wilhelm His,
Miinchener Medizinischen Wochenschrift, No. 28, 1901; and Wilhelm His,
ibid, No. 22, 1904. RunotenH Fick, Wilhelm His, Anatomischer Anzeiger, Vol.
26, 1904. B. Rawirz, W. His, Naturwissenschaftliche Rundschau, No. 24, 1904.

154 Wilhelm His

FRANCIS Dixon, Prof. Wilhelm His, Journal of Anatomy and Physiology, Vol.
88, 1904. W. WALDEYER, Wilhelm His, Sein Leben und Wirken, Deutsche Medi-
zinische Wochenschrift, Nos. 39, 40 and 41, 1904. J. KoLLMANN, Wilhelm His,
Worte der Erinnerung. Verhandl. der Natur. Gesell. in Basel, Bd. 15, 1904.
J. MARCHAND, Wilhelm His, Nekrolog, Bericht. d. K. s. Gesell. d. Wiss., Nov.
14, 1904.

*Ich hatte in Berlin bei Johannes Miiller und bei Remak tiefe Anregungen
erfahren, im tibrigen aber einen nur madszigen Schatz von geordneten Kennt-
nissen eingespeichert. Es hatte mir an sicherer Fiihrung gefehlt: unter den
Medizinern hatte ich keine Gleichgesinnten gefunden, und mein eigentlicher
Freundeskreis bestand aus Landsleuten, meistens Theologen und Juristen.
Schone Gelegenheiten, mir griindlichere physikalische und chemische Kennt-
nisse zu erwerben, habe ich verpasst und statt dessen einige recht sterile
medizinische Vorlesungen abgesessen. Auch die historischen Vorlesungen
von Ranke und die geographischen von Ritter, von denen meine Freunde
soviel Interessantes zu erzahlen wussten, hatte ich ohne Opfer an Fachbildung
besuchen koénnen.—Lebenserinnerungen, p. 28.

3 An jene Zeit absoluter Arbeitsfreiheit habe ich seitdem oft mit Sehnsucht
zurlickgedacht. Allerdings bot sie auch die Gefahren des Sichverlierens, so
habe ich einmal acht Tage lang an der Herstellung eines Glasblasertisches
gezimmert, bin aber dann, nach dessen notdtrftiger Vollendung, von einem
_gehorigen Katzenjammer tber die sinnlos vergeudete Zeit heimgesucht worden.
Im Grund habe ich aber in spateren Jahren die Erfahrung gemacht, dass
fiir den Fortgang eigener geistiger Arbeit die Belastung mit einem maszigen
Pflichtenpensum vorteilhafter ist als die absolute Freiheit, und insbesondere
habe ich oftmals beim Beginn ersehnter Ferien gefunden, dass zugleich mit
dem WHintritt freier Zeitverfitigung eine Erschlaffung der geistigen Spannkraft
sich einstellte, die erst allmahlich und durch Zwang sich wieder tiberwinden
liess. Das Gefahrlichste ist hierbei das Abwartenwollen von Arbeitsstim-
mungen; solche wirklich fruchtbare Stimmungen kOnnen ja zeitweise unver-
hofft einbrechen, viel hdufiger aber sind sie nur dadurch erreichbar, dass
man sich erst gewaltsam durch Ode und anscheinend unfruchtbare Anfange
hindurch kampft. Hat man einmal sein Arbeitsziel klar vor Augen, dann
lernt man auch bald die kleinsten Zeitabfalle des sonstigen Tagewerkes
ergiebig zunutze zu ziehen.—Lebenserinnerungen, p. 48.

*Die meisten jungen Manner miissen nach Abschluss ihrer Universitats-
zeit eine Periode des Missbehagens durchmachen, bis es ihnen gelungen
ist ihre idealen Bestrebungen in eine Thatigkeit ftir’s Leben umzusetzen.—
From a letter of March 20, 1887.

> Hs ist ein schweres, dem seiner Natur getreu bleibenden Forscher aufer-
legtes Gestaéndniss, dass die letzten Ziele, fiir deren Verfolgung er seine ganze
Kraft einsetzt, hier, wie auf allen Gebieten der Forschung, in um so entlegen-
ere Ferne riicken, je weiter er auf dem in ihrer Richtung fiihrenden Wege
voranschreitet. In der kraftigenden Arbeit selbst, im Bewusstsein sicheren
Voranschreitens und in den reichen, am Wege ihn erwartenden Friichten
findet er den vollen Ersatz ftir alle getibte Entsagung.—Unsere Kérperform,
1874, p. 215.

® Ist es ja doch die Gabe geistvoller Naturen, dass sie, auch bei beschrankten
Hiilfsmitteln materieller Erkenntniss, Beziehungen zuahnen und in ihrem

Franklin P. Mall 155

Zusammenhang zu durchschauen vermdgen, die Anderen bei weit reicherem
Material nur stiickweise zuganglich sind, und dass sie selbst im Irrthum oft
Gesichtspunkte erdffnen, die der langsam und miihselig vordringenden
EHinzelnforschung als Wegweiser fiir die Richtung ihres Ganges dienen
kénnen.—Die Hiiute und Héhlen des Koérpers (1865), Archiv fiir Anatomie,
1903, p. 369.

7Soll ich zum Schluss noch einmal versuchen, die histologischen Rollen der
Keimschichten zu sondern, so komme ich zu folgender Aufstellung:

Der Epiblast liefert das Nervengewebe und die Horngewebe.

Der Hypoblast gliedert sich in

den embryonalen Mesoblast, die gemeinsame Anlage fiir das quergestreifte
und glatte Muskelgewebe, fiir die Epithelien des Genitalapparates und
flir die embryonalen Bindesubstanzen.

das ausserembryonale Mesenchym,

den Angioblast, die Anlage des Blutes und der Blutcapillaren,

das Endoderm, die Anlage der Epithelien und Dritisen des Hingeweide-
rohres.

Der Lecithoblast, da, wo er zur Entwickelung kommt, bildet einen Theil
des Hypoblast.

Das alte Rathsel erweist sich zur Zeit immer noch ungelést: noch kénnen
wir nicht sagen, weshalb ein Theil der gegebenen Anlagen zu Bindesubstanzen
wird, und was die Blut- und Capillarzellen bestimmt, so friihzeitig und so
scharf sich von ihren scheinbar so nahen Verwandten, den Zellen der Binde-
substanzen, zu scheiden.—Lecithoblast und Angioblast der Wirbelthiere, Ab-
handl. d. K. Sach. Gesellschft. d. Wiss., Bd. 26, 1900, p. 326.

SIch schliesse diesen in jeder Hinsicht fragmentarischen Aufsatz tiber die
intramedullaren Faserbahnen des Gehirns mit der Bemerkung, dass er zur
Zeit nicht viel mehr zu bieten vermag, als ein Arbeitsprogram ftir kommende
detailliertere Forschungen. Noch sind wir eben in Erkenntniss dieser Dinge
in den allerersten Anfangen, und es bedarf hier, wie anderwarts, zaher Arbeit
bis die Entwicklungsgeschichte des Gehirns nach ihren verschiedenen Richt-
ungen hin befriedigend kann klar gelegt werden. Zur Zeit kann ich nur
angeben, wo diese Arbeit einzusetzen hat. Frtiher oder spater wird man auf
diesem Gebiet zum System organisierter gemeinsamer Arbeit tiberzugehen
haben.—Die Entwickelung des Menschlichen Gehirns, Leipzig, 1904, p. 175.

®TIm Leben unsrer Universitaéten macht sich bei aller anscheinenden Fort-
dauer ihrer Leistungen, und auch bei ununterbrochenem Ersatz abgehender
Krafte durch neu eintretende, eine ganz bestimmte Periodicitat der Entwick-
lung geltend. Fiir die Gesammtuniversitat und ftir die Facultaten folgen
auf Perioden geistigen Aufschwunges solche der Ruhe und des Riickgangs.
Aeussere und innere Bedingungen wirken dabei zusammen und es ist nicht
immer leicht, deren Ineinandergreifen zu verstehen. Eine Grundbedingung
muss aber stets erfiillt sein, falls eine Korperschaft bliihen soll. Die Korper-
schaft muss kraftige und zielbewusste Fiihrer besitzen, welche deren Geist
in bestimmte Bahnen zu lenken und unter ihren Gliedern die Gemeinsamkeit
des Strebens zu sichern wissen.

Solch ein fitihrender Geist ist in unsrer Facultat wahrend mancher Jahr-
zehnte Ernst Heinrich Weber gewesen, welcher vom Jahr 1821 ab die Pro-

ili

156 Wilhelm His

fessur der Anatomie und spaterhin (von 1841 ab) noch die der Physiologie
bekleidet hat. Die Spuren seiner machtigen PersOnlichkeit haben sich als
bleibende erhalten nicht nur in den Acten unserer Facultaét, sondern noch
tiefer begriindet in denen der Wissenschaften, die er vertreten und die er
um ausgedehnte neue Gebiete bereichert hat.

Bis zum Jahre 1865 hat Ernst Heinrich Weber, von seinem Bruder Eduard
untersttitzt, die Doppellast der beiden ausgedehnten Facher getragen. Dann
aber, als die Neuschdpfung einer physiologischen Anstalt in Aussicht ge-
nommen wurde, und dadurch neue Verpflichtungen an den Lehrer der Physi-
ologie herantreten sollten, zog sich der alternde Gelehrte auf seine ur-
spriingliche Anatomieprofessur zuriick, und es ist nun auf Ostern 1865 (unter
dem Dekanat Wunderlichs) die Berufung-von Karl Ludwig als Professor der
Physiologie und Director des neu zu begrtindenden physiologischen Instituts
erfolgt.

Die Initiative zu diesen Neuerungen ist von der koniglichen Regierung
ausgegangen. Im Sinn ihres hohen Monarchen, des Konigs Johann, hatten
sich die einsichtigen Leiter des Ministeriums, Hr. Staatsminister v. Falken-
stein und Hr. Geh. Rath Dr. Hiibel, die Aufgabe gestellt, die Universitat
Leipzig mit allen aufwendbaren Mitteln zu neuem Glanze zu erheben. Die
physiologische Anstalt wurde als das erste Glied einer Reihe von Neu-
schopfungen geplant, deren Endziel die Umgestaltung des gesammten natur-
wissenschaftlichen und medizinischen Unterrichts sein sollte. In der Wahl
von Professor Ludwig hat die k. Regierung eine besonders gliickliche Hand
bewiesen, denn sie gewann an ihm ftir ihre ferneren Entscheidungen einen
vermoge seiner Hinsicht und seiner organisatorischen Kraft ganz besonders
befahigten Rathgeber. Ludwig’s Hinfluss hat sich wahrend der v. Falken-
stein’schen Periode weit iiber das medizinische Facultitsgebiet hinaus er-
streckt, und seiner Anregung sind von den bedeutendsten Berufungen jener
Zeit zu verdanken gewesen. Spater, nachdem einmal die Organisation natur-
wissenschaftlichen Unterrichts fiir Leipzig erreicht und nachdem auch das
Cultusministerium in andere Hande tibergegangen war, hat sich Ludwig auf
sein engeres Arbeitsgebiet zurtickgezogen. Was er aber auf diesem Gebiete
geleistet hat, das hat den Ruhm der Leipziger Universitét bald durch alle
Lander verbreitet.—Karit Lupwie und Kart TurerscH, Beilage, Allgemeinen
Zeitung, Nr. 164. 19 Juli, Miinchen, 1895.

”Tch danke Ihnen fiir den inhaltsreichen Aufsatz tiber die Darmentwick-
lung, die in der mir tiberreichten Jubilaumsschrift thatreich hervortritt. Alle
diese Bezeugungen haben mich herzlich gefreut. Dauernd wird die Be-
friedigung liber Ihre Arbeit sein, die ein bis jetzt so wenig klarer Gebiet
endgultig in’s Reine bringt. Was ja bei den meisten unserer bishdrigen ent-
wickelungsgeschichtlichen Vorstellungen fehlt, das ist die Beobachtunggrund-
lage fiir die Uebergangsphasen aus den friih embryonalen in die foetalen und
von da in die ausgebildeten Stufen. Fir den Darm haben Sie nunmehr die
ganze Kette vom Anfang bis zum Ende zusammengefiigt und das halte ich
fiir einen grossen Fortschritt—From a letter of October 29, 1897.

So weit ich tiber solche freie Augenblicke verftige, widme ich sie noch
meiner alten ungliicklichen Liebe den Knochenfischen. Ich habe seit dreissig
Jahren schon unendlich viel Zeit damit verloren, sie sind ein methodisch

Franklin P. Mall 157

sehr schwer zu bearbeitendes und launisches Material, und doch locken mich
die untiberwundenen Schwierigkeiten und 6ffenen Fragen immer wieder zu
neuen Anlaufen.—From a letter of December 25, 1898.

“2 Wie bei der wissenschaftlichen Arbeit, so tritt auch bei unserer heutigen
Lehrweise der Respect vor der Thatsache in den Vordergrund, und wir be-
miihen uns in erster Linie auch unsere Schiiler dazu zu erziehen. Beim natur-
wissenschaftlichen und somit auch beim medizinischen Unterricht ist unsere
Sorge, dem Anfanger die Kunst unbefangener Beobachtung beizubringen.
Wir halten ihn an, die Sinneswahrnehmungen scharf zu trennen von den
daran sich ankntipfenden Schlussfolgerungen, wir warnen ihn vor der Be-
einflissung durch vorgefasste Meinungen und belehren ihn tiber die Tavtisch-
ungsquellen, die in unsern eigenen Sinnen sowie in unsern besten Apparaten
enthalten sind. Vor allem aber suchen wir den Schitiler dazu zu bringen,
dass er sich angewohnt, das Gebiet eigener Erfahrungen selbstandig zu klaren
Begriffen zu verarbeiten. So klein Anfangs das Capital an solch eigenem
Erwerb sein mag, so gewéhrt es dem Besitzer doch bald das Gefiihl einer be-
stimmten geistigen Freiheit und Unabhangigkeit, das Geftihl des tiichtigen
Menschen.

Was hat nun aber diese, vorwiegend auf Scharfung der Kritik hinstrebende
Form der Schulung mit der Spaltung der Lehrfacher zu thun? Der Zusam-
menhang ist leicht nachzuweisen. So lange es sich um blosse Ueberlieferung
systematisch geordneter Begriffe in dogmatischer Form handelt, ist ein
fleissiger Gelehrter mit Hilfe der nothigen Lehrbticher, der duces Arnemann,
Gaubius und Metzgerus im Stande, ein ausgedehntes Gebiet als Lehrer zu
umspannen, ja selbst vom Ueberspringen von einem Fache auf ein anderes,
mehr order minder entlegenes, wird ihn kein inneres Hinderniss abhalten.
Wenn wir horen, dass in einem fritihern Jahrhundert die Lehrficher inner-
halb der philosophischen Facultaét jedes Jahr frisch ausgel6st wurden und
dass auch nach Beseitigung dieses Modus noch die Verpflichtung bestand,
dass ein jedes Facultatsmitglied allen Fachern gerecht sein musste, so ist
diese heutzutage undenkbare Hinrichtung dadurch verstandlich, dass in jenen
Perioden die Bedeutung der allgemeinen Gelehrtenbildung iiber diejenige der
Fachbildung weit tiberwog, wahrend wir nunmehr auf dem entgegengesetzten
Standpunkt stehen. Sowie verlangt wird, dass der Lehrer die wissen-
schaftlichen Ergebnisse seiner Disciplin anstatt blos in dogmatischer Form,
auch nach ihrer Begriindung dem Schiiler mittheile, so fallt eine Hauptseite
des Unterrichts in die wissenschaftliche Methodik.—Ueber Entwickelungsver-
hiltnisse des Akademischen Unterrichts, Rektoratsrede, Leipzig, October 31,
1882, p. 33.

8 Ludwig’s Forscherwaffen waren eine ungemain scharfe Analyse der ihm
vorliegenden Naturerscheinungen, eine stets klare Fragestellung und eine
absolute Sicherheit seiner Methodik. Dabei verftigte er aber auch iiber eine
ausreichende Dosis jenes Findersinnes; ohne den in Erforschung der lebenden
Natur selbst die klarsten Denker oft machtlos bleiben. Die Natur lasst sich
nicht immer mit Logik zwingen, ihre Wege sind nicht selten versteckt, und
sie enthiillen sich nur dem, der sich in ausdauernder und treuer Beobachtung
den Blick auch ftir deren unscheinbare Spuren gescharft hat. Die unmittel-
bare Liebe zur sinnlichen Beobachtung hat aber Ludwig im hohen Maasse

158 Wilhelm His

besessen, und fiir ihn ist ein gelungenes Praparat oder ein schlagender Ver-
such stets Gegenstand eigentlich asthetischen Genusses gewesen.—CarL Lup-
wia, Gedichtnissrede, Bericht, d. K. s. Gesell. d. Wiss., November 14, 1895,
p. 6.

“Thre Bestrebungen eine inductive anatomische Unterrichtsmethode zur
schaffen, interessiren mich sehr lebhaft. Wenn es Ihnen mit fiinfzig Schiilern
gelingt, zum Ziel zu kommen, so ist dies jedenfalls eine anerkennungswerthe
Leistung. Vor Kurzem publicirte der bekannte, Dr. Schweninger, einige Auf-
satze liber die Erziehung von Medizinern, worin er tiberhaubt das Prapariren
verwarf und meinte, man soll die Anatomie gleich am Lebenden vornehmen,
die Studenten durch Percussion, u. s. w. die Organe auf den Korper zeichnen
lassen, u. s. w. Unsere Medizinererziehung ist zwar krank an zu vielem
Auswendiglernen von Bticherweisheit und gewiss konnte auch in der Ana-
tomie dem Studenten manches osteologisches Detail erlassen werden. Aber
abgesehn davon, ist ja der Prapariersaal eine so wichtige Schule der Beobach-
tung und der Handfertigkeit, dass eine grosse Beschranktheit dazu gehort,
das anatomische Prapariren beseitigen zu wollen.—From a letter of December
31, 1896.

16 His’s influence in America has been great, greater than in any other
country, even Germany. He took a lively interest in our whole development,
in the development of our universities, scientific societies and journals. He
was much pleased with the numbers of the American Journal of Anatomy,
and appreciated above all the leading article by Bardeen and Lewis. “ Auch
dartiber habe ich mich gefreut dass Sie mit so viele Andere zusammen ar-
beiten.”’ He always approved of cooperation.

1 Was Sie mir damals von “ Carnegie Institution,’ geschrieben haben, muss
uns, diesseits des atlantischen Oceans Lebende mit innigem Neid erfillen.
Es ist indessen keine Frage, wir sind in eine Periode eingetreten, in der die
zu leistende Arbeitssumme immer grosser und die Ansprtiche an Reichlich-
keit des Materiales und die Pracision seiner Durcharbeitung immer strenger
werden und da hilft eben schliesslich nur ein wissenschaftlicher Grossbetrieb
mit guter Organisation. Noch haben wir in Deutschland bei aller Arbeit ein
zu planloses Durcheinanderrogen, und zu viel Kraft geht in persdnlicher
Reibung verloren. Der Ehrgeiz ist ein wichtiger Antrieb zur Arbeit, aber
anderseits fiihrt er auch vielfach dahin, dass die Arbeiter anstatt sich zu
unterstiitzen, sich gegenseitig herabzumindern suchen. ... Noch vor zehn
Jahren hatte mir die Organisation eines grosseren rein wissenschaftlichen In-
stitutes, die grésste Freude gemacht. Mit zwei und siebenzig Jahren weiss
man aber, dass die Arbeitszeit nur noch knapp zugemessen ist, ganz abgesehen
davon, dass die Arbeit viel langsamer von der Hand geht.—From a letter of
March 17, 1908.

“Tech hatte im vorigen Sommer einen Anlauf genommen, um die Begrtindung
besonderer entwickelungsgeschichtlichen Institute und Gehirninstitute in An-
regung zu bringen, aber bis jetzt habe ich noch nicht Viel erreicht. Es fehlen
uns in Deutschland und in Europa jene Milliardare die bei Ihnen so fix bei der
Hand sind, wenn grosse Schopfungen fundirt werden sollen. Das immense
Capital, dass der Ingenieur Nobel fiir wissenschaftlichen Zwecke vermacht hat,
ist dadurch nutzlos, dass er die Vertheilung der Zinsen in Form von Preisen

Franklin P. Mall 159

bestimmt hat. Da diese Preise jedes Jahr vertheilt werden sollen so wird,
wie ich ftirchte, die Zutheilung bald zu Parteisache werden und viel Un-
frieden herbei flihren.—F'rom a letter of December 31, 1902.

““Wird durch solche Preise die wissenschaftliche Arbeit wirklich gefordert?”
Ich glaube, man kann diese Frage ruhig verneinen: kein aus innerem Antrieb
arbeitender Forscher wird dadurch, dass ihm das Schicksal eine gr6éssere
Summe Geldes in den Schooss wirft, ein Anderer werden. Er wird eben
uber die Ehre des Preises und tiber den empfangenen Betrag sich freuen, im
Uebrigen aber seinen Gang weiter gehen, als ob Nichts geschehen ware. Und
wer den Preis nicht bekommt, wird nicht anders verfahren. Hochstens liegt
fiir den letzteren, wenn er nicht edel veranlagt ist, die Versuchung vor, dem
begtinstigten Collegen, oder denen, die tiber den Preis zu bestimmen hatten,
unfreundliche Geftihle nachzutragen.—Ueber wissenschaftliche Stiftungen,
Bericht. d. K. s. Gesell. d. Wiss., 1901, p. 434.

18 Sie haben an ihren neueren amerikanischen Universitaten einen kraftigen
Nervus rerum, und wenn reiche Hilfsmittel in die richtigen Bahnen kommen,
so lasst sich ja Vieles in verhaltnissmdssig kurzer Zeit erreichen. Die Haupt-
sache bleibt immer dass die Fiihrung solch fortschreitender Bewegungen in
den Handen von Mannern bleibt, die wissen, woraus es bei geistigen Schopf-
ungen ankommt. Es ist immer befriedigender, véllig Neues zu schaffen, als
am Alten herumzuflicken. Letzters Schicksal fallt uns in Europa nun all-
zu oft zu. Augenblicklich soll wieder an unsern Examenreglementen geflickt
werden, eine Arbeit die nur wenig Freude bringt, da der Ballast alter Vor-
urtheile und Wiederstande nicht tiber Bord geworfen werden kann.—From
a letter of April 22, 1899.

19 GENERAL OUTLINE OF THE DEVELOPMENT OF HIS’S INSTITUTE FOR THE STUDY OF
THE BRAIN.

A.—Proposition to the International Association of Academies, Paris, April
20, 1901.

Die Internationale Association der Akademien moge eine Fachcommission
aufstellen zur Berathung der Mittel und Wege, wie auf den Gebieten, eines-
theils der menschlichen und thierischen Entwicklungsgeschichte, anderntheils
der Hirnanatomie eine nach einheitlichen Grundsatzen erfolgende Sammlung,
Verarbeitung und allgemeine Nutzbarmachung von sicherem Beobachtungs-
material erreicht werden kann.

B.—Decision of the Association of Academies.

1. Die Berathung der auf menschliche und thierische Entwicklungsge-
schichte beziiglichen Abschnitte des Antrages ist vorerst den betreffenden
Fachvereinen (den anatomischen Gesellschaften) zu tberlassen.

2. Dagegen setzt die Internationale Association der Akademien eine Special-
commission nieder, die eine nach einheitlichen Grundsatzen erfolgende Durch-
forschung, Sammlung und allgemeine Nutzbarmachung des auf Gehirnanat-
omie beziiglichen Materiales zu berathen hat. Die Commission hat insbe-
sondere die Schaffung eines internationalen Systemes von Centralinstituten
in Erwagung zu ziehen, in denen die Methoden der Forschung entwickelt, das
vorhandene Beopachtungsmaterial aufgespeichert und der allgemeinen Be-
nutzung der dabei interessirten Gelehrten zuganglich gemacht werden.—An-

160 Wilhelm His

trag der Kon. Sach. Ges. an die Royal Society of London, Ber. d. K. s. Ges.
d. Wiss., February, 1902.

C.—Obdjects of the Institute.

1. Die Aufspeicherung und Zugdénglichmachung von wissenschaftlichem
(normalem) Material an Praparaten, Modellen, Photogrammen, Zeichnungen
UL SW

2. Die technische Hilfeleistung bei wissenschaftlichen Untersuchungen.

8. Die Aufbewahrung von wertvollem experimentellphysiologischem und
pathologischem, bereits bearbeitetem oder noch zu bearbeitendem Material.

4, Die Bewaltigung grésserer, tiber die Krafte einzelner hinausgehender
Aufgaben, soweit solche zur Kooperation sich eignen.—Antrag der von der
Internationalen Association der Akademien Niedergesetzter Commission fur
Hirnforschung der Generalversammlung der Association in London zum 25
Mai, 1904, vorgelegt, Leipzig, 1904.

D.—Organisation of the Institute.

1. Arbeitsfeld und Arbeitsweise bleiben jedem einzelnen Institute tber-

lassen. Es sollen jedoch angestrebt werden:
a. Hine einheitliche Nomenklatur.
b. Verwendung eines einheitlichen Masses und Gewichtes.

2. Alljahrlich statten die Institute der Centralkommission einen Bericht
iiber ihre Thatigkeit ab. Dabei sollen der Bestand und die Zuginge an
Druckwerken, Abbildungen, Modellen und Praparaten mitgetheilt werden.

3. Die Institute sind gehalten ihre Arbeitsmaterialien und die Sammlungen
ihrer Praparate einander unter sich, sowie den derselben bediirftigen Forsch-
ern nach Moglichkeit zuginglich zu machen.—Bericht, etc., Bericht. d. K. s.
Gesell. d. a ise.. June 8, 1903.

D.—Sections.

1. Die systematische Anatomie des menschlichen Centralnerven-systems,
einschlieszlich der Anthropologie.

2. Die vergleichende Anatomie.

3. Die histologische Forschung.

4, Die entwickelungsgeschichtliche Forschung.

5. Die Physiologie, einschliesslich der physiologischen Psychologie.

6. Die pathologische Anatomie, experimentelle Pathologie und Teratologie.

7. Die klinische Forschung.—Entwerf. Motiv zu den Antragen, etc., Leipzig,
January 3, 1904.

F.— Special Committees.

1. Waldeyer, Cunningham, Mall, Manouvrier, Zuckerkandl.

. Ehlers, Edinger, Giard, Guldberg, Elliot Smith.

. Golgi, Ramon y Cajal, Dogiel, van Gehuchten, Retzius.
. His, Bechterew, v. Kolliker, v. Lenhossek, Minot.

. H. Munk, Horsley, Luciani, Mosso, Sherrington.

. Obersteiner, Dejerine, Monakow, Langley, Weigert.

7. Flechsig, Hentschen, Ferrier, Lannelongue, Reymond.—Protokoll von der
Internationalen Association der Akademien Niedergesetzten Centralkommis-
sion fiir Gehirnforschung, January 11, 1904. Bericht. d. K. 8S. Gesell. d. Wiss.,
1904.

Oo OF ~ OO DO

Franklin P. Mall 161

* Hs that mir wohl aus Ihrem Brief, wie aus vielen Andern, die ich bekom-
men habe, zu sehen, wie mein Mann nicht nur durch seine Wissenschaft,
sondern noch mehr durch sein Leben und seinen Karakter seinen Schiilern
etwas Gutes erwiesen hat.—From a letter from Frau Professor His of June 8,
1904.

Ich bin mit diesen Aufzeichnung an einen Punkte angelangt, wo ich sie
abschliessen kann. In reichem Wechsel sind mir beim Niederschreiben obiger
Blatter Bilder vor Augen getreten von einer Fille von trefflichen und von
hervorragenden Menschen, mit denen ich im Laufe meiner Entwicklungsjahre
in Beziehung getreten bin. Gar manche Namen hatte ich der Schar noch
beifigen konnen. Von allen diesen Menschen habe ich gelernt oder sonst-
wie Gutes empfangen. Die weit tiberwiegende Mehrzahl derselben sind langst
dahingeschieden, allen aber bewahre ich ein dankbares Andenken. Mégen
andere dereinst auch von mir dasselbe sagen k6nnen.—Lebenserinnerungen.

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN
MAN.

BY
CHARLES R. BARDEEN,
Professor of Anatomy, the University of Wisconsin, Madison, Wisconsin.

WITH 7 PLATES.

There is a somewhat extensive literature dealing with the development
of the spinal column in various vertebrates. The chief stages in its
differentiation are fairly well determined. Special attention has been
given to the early development in the lower vertebrates. The recent liter-
ature on this subject up to 1897 has been reviewed by Gaupp, 97.
Somewhat less attention has been devoted to the mammals. To Fro-
riep, 86, is due a valuable account of the development of the cervical
vertebre in the cow, and to Weiss, or, an important description of the
development of the thoracic and cervical vertebre in the white rat.

We shall not attempt to enter here into a description of the early
stages of differentiation in the spinal axis; that is of the period covering
the formation of the chorda dorsalis and of axial and peripheral meso-
blast, the differentiation of primitive segments, and the origin of the
axial mesenchyme. This period in the human embryo has been well
treated by Kollmann, gi, and some account of it has previously been
given in this JourNAL (Bardeen and Lewis, or). We shall therefore
proceed at once to a consideration of vertebral differentiation in the axial
mesenchyme.

Vertebral development in the embryo may be divided into three over-
lapping periods: a membranous or blastemal, a chondrogenous, and an
osseogenous.”

1 Among more recent papers may be mentioned, those of Baldus, o1; Hay,
97; Kapelkin, 00; Manner, 99; Mannich, 02; Ridewood, or; and Schauins-
land, 03.

2In the text books the first of these periods is usually called the precartilage,
prochondral, or Vorknorpel stage, but the condensed tissue from which
the skeletal parts are derived gives rise not only to cartilage but also to
perichondrium and to ligaments. Recognizing this fact, Schomburg, oo, has
AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

164 Development of Thoracic Vertebree in Man

THE BLASTEMAL PERIOD.

The division of the axial mesenchyme into segments, sclerotomes,
which correspond to the myotomes and spinal ganglia, is marked at an
early stage by intersegmental arteries. Schultze, 96, has shown that
the segmental differentiation of the axial mesenchyme extends into the
region dorsal to the spinal cord. Ventrally it does not, however, extend
quite to the chorda dorsalis. Fig. 1, Plate I, illustrates the conditions
existing in the thoracic region of man at this period.

vy. Ebner, 88, found in the embryos of several vertebrates a fissure
which divides each sclerotome into an anterior and a posterior portion.
Schultze, in 1896, showed, that in selachians and reptiles this fissure is
represented from the time of its formation by a diverticulum which
communicates with the myoccl. In birds the diverticulum arises sec-
ondarily and later becomes fused with the myoccel, and in mammals it
arises after the myoccel has disappeared.

In man the fissure becomes distinct in the thoracic region at about the
end of the third week of development (Fig. 2, Plate 1).* At this period
the median surface of each myotome has become converted into muscle
fibres (Fig. 2, Myo.). At the same time the mesenchyme in the postero-
lateral region of each sclerotome has become condensed so that it ap-
pears, in a stained section, dark when compared with that of the anterior
half (Fig. 2). At the lateral margin of the anterior halves of the
sclerotomes the spinal nerves extend out toward the thoracic wall (Fig.
2, Sp.N.). The division between the sclerotomes is still marked by the
intersegmental arteries (Fig. 2). About the chorda dorsalis the cells of
the axial mesenchyme become densely grouped into a perichordal sheath.
The long axes of the cells lie parallel with the chorda (Fig. 2, Pch. S.).

The condensation of tissue which distinguishes the posterior sclerotome
half begins, as mentioned above, in the posterior lateral area of each

called the condensed-tissue stage, the mesenchymal period, and restricted the
term Vorknorpel to the earlier stages of the formation of cartilage. The
term “blastemal” is now-a-days commonly used to designate a mass of
mesenchymal tissue from which organs are to be differentiated, and is applied
to the tissue of the limb-bud before differentiation has commenced. It seems
to me that it would be well to extend this term to the structures first differ-
entiated in the limb. Thus, “scleroblastema’’ would mean the tissue differ-
entiated from the blastema of the leg and destined to give rise to skeletal
structures; myoblastema, the time differentiated for the muscles, and dermo-
blastema that destined for the skin.

®The figures on this and the following plates are based upon embryos
belonging to the collection of Prof. Mall. I am greatly indebted to him for
the use of these embryos.

Charles R. Bardeen 165

sclerotome. From here the condensation extends dorsally between the
medial surface of the posterior half of the corresponding myotome and
spinal ganglion and gives rise to a dorsal, or neural, process (Figs. 5 and
6, Plate 2, V. Pr.). At the same time it proceeds ventrally along the distal
margin of the corresponding myotome and gives rise to a ventral or
costal process (Figs. 5 and 6, C. Pr.) ; and medially toward the chorda
dorsalis, giving rise to a process which joins about the chorda dorsalis
with a similar one, from the other side of the segment (Figs. 5 and 6,
Disk). These median processes by their fusion form what has been
termed by Weiss, o1, a “ horizontal plate.” “ Primitive disk” seems to
me perhaps a better term.

The whole mass of condensed tissue which gives rise to the primitive
dorsal, ventral and median processes has received various designations,
of which that given by Froriep, 83, “ primitive vertebral arch” seems to
’ be the most widely accepted. Since, however, it represents much more
than a vertebral semi-arch, I have previously, 99, suggested for it the
term “ scleromere.” *

Figs. 8, 9 and 10, Plate III, represent wax-plate reconstructions of
several scleromeres from the thoracic region of Embryo II, length 7 mm,
The outlines of the condensed tissue are not so sharp in nature as it is
necessary to make them in a model of this kind. It is believed, however,
that the general form relations are here fairly accurately shown. In
Fig. A, Plate IT, of the article by Bardeen and Lewis are shown the rela-
tions of the scleromeres to other structures.

During the period of differentiation of the scleromeres the myotomes
undergo a rapid development. The median surface of each myotome
gradually protrudes opposite the fissure of v. Ebner. The dorsal and
ventral processes of each scleromere are then slowly forced into the inter-
val between the belly of the myotome to which it belongs and the one
next posterior, and thus finally they come to occupy an intersegmental
position. It is not, however, correct to call the early processes of the
” as some text-book writers have done. Fig. 4

€

scleromeres “ myosepta,

shows this.

‘By the fissure of v. Ebner each sclerotome is divided into two portions,
of which the posterior in the higher vertebrates plays the chief réle in verte-
bral differentiation. ‘‘ Scleromere” therefore seems an appropriate designa-
tion for the condensed sclerogenous tissue of this half-segment. Goette has
recently, 97, brought forward evidence in favor of the view that primarily in
the digitates there were two vertebre to each body-segment. In the higher
vertebrates, during embryonic development, the posterior skeletal area of
each body-segment alone develops freely. The anterior area becomes fused
with the scleromere in front.

166 Development of Thoracic Vertebre in Man

This figure represents a horizontal section passing through several
spinal ganglia, myotomes and neural processes. The last may be seen
extending gradually into the area opposite the myotomic septa, but they
still cover the whole posterior half of the median surface of the myotome
in the region of the section. The processes are connected by membranous
thickenings of the mesenchyme of the anterior half of each segment. These
membranes may be called interdorsal membranes. They correspond to
the interdorsalia of elasmobranchs. Figs. 12, 13, 15 and 16, Idr. M.,
represent these membranes. A line drawn from A to B in Fig. 12 would
pass through an area corresponding to that of the section represented
in Fig. 4.

In the region where the neural and costal processes spring from the
primitive disks membranous septa are likewise differentiated from the
anterior halves of the sclerotomes. These septa serve to unite the suc-
cessive disks. Each is continuous posterially with a dense tissue which
strengthens the primitive disk and anteriorly it extends into the neural
and costal processes. The relations of these interdiscal membranes are
shown in Figs. 3, 11, 12 and 13, Ids. m. Since at this period structural
outlines are by no means sharp, the figures based upon wax-plate recon-
structions must be taken as semi-diagrammatic. A line drawn from
c to d in Fig. 12 would represent essentially the plane of the section
shown in Fig. 3.

During the development of the interdiscal membranes the primitive
disks become hollowed out on the posterior surface. A comparison of
Fig. 2 with Fig. 3 demonstrates this. The perichordal sheath meanwhile
is developed in a ventrodorsal direction so that the area between the
primitive disks becomes divided into right and left halves. Figs. 11, 13
and 7 all show this. Each lateral area is filled with a lightly staining
mesenchyme which is continuous ventrally and dorsally with the tissue
surrounding the spinal column.

Fig. 17%, Plate V, represents a sagittal section cut slightly obliquely
through an embryo 12 mm. long. In the region where the chorda
(Ch. d.) is cut, the primitive disks may be seen united by a fairly dense
tissue, the perichordal septum (Sptm.). Posterior to this region the
section passes to one side of the chief axis of the embryo. The inter-
vertebral disks may here be seen separated by a lighter tissue and in the
more posterior portion of the section, which passes still more lateral to
the chordal region, the tissue between the disks is seen to be continuous
with that surrounding the spinal column. In this region the interdiscal
membrane (Ids. m.) is seen anterior, the primitive disk posterior to the
fissure of v. Ebner (F. v. F.).

Charles R. Bardeen 167

Meanwhile the ventral processes of the thoracic vertebra extend well
into the thoracic wall, giving rise to primitive ribs, illustrated in Fig. B,
Plate II, in the article of Bardeen and Lewis, o1.

Development proceeds rapidly. In Embryo CIX, length 11 mm., age
about five weeks (Figs. 14-16), the conditions are as follows: The neural
processes are somewhat better developed than those of the preceding stage,
but otherwise are similar in character. The costal processes are consid-
erably farther developed (Bardeen and Lewis, o1, Plate V, Fig. E).
At the angle between the neural and costal processes opposite where they
join the primitive disks a transverse process, but slightly indicated
at the preceding stage, is now fairly clearly marked. Hach primitive
disk has become further hollowed out at its posterior surface, owing, in
all probability, to the conversion of its tissue into that of the area
between the disks. The interdiscal membrane (Jds. m.), on the other
hand, has become thicker and has extended anteriorly and posteriorly
about the area between the disks so that this has become completely
enclosed (Figs. 14, 15 and 16). The tissue of each segment immedi-
ately anterior to the primitive disk has become greatly thickened and the
line between it and the disk indistinct.

The area between each two primitive disks is still divided by the peri-
chordal septum (Fig. 7). Each half represents the anlage of a chon-
drogenous center of the vertebral body. Formation of cartilage has not,
however, begun. The thickening of the ventral margin of the primitive
disk at this stage represents the “ hypochordal Spange,” which Froriep
has shown to play an important part in the development of the vertebra
of birds and of the atlas in mammals. It has merely a transitory exist-
ence in the thoracic region of man.

Summary.—To sum up briefly, we may say that during the blastemal
period each scleromere becomes divided into two portions, an anterior
and a posterior, characterized by a much greater condensation of tissue
in the posterior. From this condensed tissue arises a primitive vertebra
of Remak, or scleromere, with dorsal (neural) and ventral (costal) pro-
cesses and a disk uniting them to the mesenchyme condensed about the
chorda dorsalis. From the tissue of the anterior half of each sclerotome
arise membranes which serve to unite the dorsal processes of the sclero-
meres, interdorsal membranes, and to cover in the areas between the suc-
cessive disks, interdiscal membranes. The primitive disks become hol-
lowed out posteriorly by a loosening up of their tissue and strengthened
anteriorly by a condensation of the tissue immediately bounding the fis-
sure of y. Ebner. The area between each two disks is bilaterally divided
by a membrane springing from the perichordal sheath. The formation
of a cartilagenous skeleton now begins.

168 Development of Thoracic Vertebre in Man

CHONDROGENOUS PERIOD.

The tissue relations during this period have been carefully studied in
representatives of most of the chief groups of vertebrates. The form of
the early structures has been less accurately determined because most
investigators have avoided the somewhat laborious methods of plastic
reconstruction.

On each side of the blastemal vertebra three primary centers of chondri-
fication appear at about the same time, one for neural process, one for
the costal process and one for the vertebral body. Fig. 7, Plate II, shows
these centers as they appear in a cross section at an early period. Figs.
25, 26 and 27, Plate VI, show the early cartilages of an embryo slightly
older, CXLIV, length 14 mm., age 54 weeks.

The cartilages of the vertebral body develop by a transformation of
the tissue lying between the primitive vertebral disks and surrounded by
the interdiscal membrane. A considerable part of this tissue is derived
from the posterior surface of each primitive disk. At first the cartilage
of the left side is separated from that of the right by the perichordal
septum. Soon this is broken through and the two anlages of cartilage
become united about the chorda. In the thoracic region this union seems
to take place at about the same time dorsally that it does ventrally. A
sagittal section of an embryo at this stage is shown in Fig. 18. The
chorda dorsalis is surrounded by a perichordal sheath. The latter is
encircled by dense intervertebral disks which alternate with light cartil-
agenous rings. ‘The latter are surrounded by perichondrium which is less
condensed than the tissue of the disks, but more so than that of the
bodies and about the same as that of the perichordal sheath. Ventrally
and dorsally a longitudinal hgament has been differentiated from the
surrounding mesenchyme.

It is probable that the disks seen in this section are formed in part
from the primitive disks, in part from the posterior layer of the anterior
sclerotome halves; in other words, that each is formed about the rudiment
of the fissure of v. Ebner. Compare Figs. 17 and 18. The tissue is con-
centrically arranged in a way somewhat resembling that of the interver-
tebral disks of the adult.

The perichordal tissue rapidly decreases in thickness. At the same
time the cartilage of the vertebral bodies grows also at the expense of the
intervertebral disks (Figs. 19, 20, 21 and 22). According to Schultze,
96, the cartilages of the bodies finally fuse to form a continuous car-
tilagenous column. This does not seem to be the case in man. In all of
the embryos belonging to the collection of Prof. Mall some mem-
branous tissue may be seen separating completely the successive bodies,

Charles R. Bardeen 169

but in embryos between 20 and 40 mm. in length this membrane in the
vicinity of the chorda dorsalis is very thin. At the periphery of the
disks the annulus fibrosus is meanwhile differentiated more and more
into a condition resembling the adult (Figs. 17-23, Plate V).

The chorda dorsalis at the period shown in Fig. 18 is of about the
same size at the level of the disks and between them, but as the bodies
increase in size at the expense of the disks the chordal canal becomes en-
larged in the intervertebral areas and constricted at the centers of the
bodies (Figs. 19, 20 and 21). The chorda loses its continuity and the
chordal cells become clumped in the vicinity of the disks (Figs. 21 and 22)
and finally spread out there in the form of a flat disk (Fig. 23). At this
last period the perichondrium of the bodies is again becoming well
marked and the portion of each intervertebral disk in the vicinity of the
chorda dorsalis is better developed than during the stages immediately
preceding. The chordal canal long remains in the vertebral body (Figs.
23 and 24).

The cartilage of the bodies in Embryo CXLIV (Fig. 18) is of an early
embryonic hyaline type. Ata slightly later stage (Fig. 19) two regions
may be distinguished, a central and a peripheral. The central cartilage
is denser than that of the preceding stage, while the peripheral cartilage
resembles it. Gradually the cartilage at the center of the body undergoes
further changes. The cells enlarge and become sharply set off against the
intercellular substance (Figs. 22, 23 and 24), and finally an invasion of
blood vessels takes place, chiefly from the posterior surface (Fig. 23).
These changes in the cartilage, represented also in Fig. 41, Plate VII,
are preliminary to ossification.

Deposit of calcium salts and actual ossification begins in the distal
thoracic and proximal lumbar vertebre of embryos about 5 to 7 em. long
and three months of age. Fig. 42 shows a center of ossification in an
embryo of 70 mm. /

During the development of the vertebral bodies changes have been
active in the neural cartilages. At the period represented in Fig. 7, Plate
II, the neural cartilage is a small, flat-body situated in the dorsal process
of the scleromere; from this as a center, pedicular, transverse, anterior
(superior) and posterior (inferior) articular, and laminar processes are
rapidly developed. This structural differentiation is best followed in the
figures representing the models (Figs. 25-36). The pedicular processes
are at first slender rods (Fig. 26), each of which grows out towards and
finally fuses with its corresponding vertebral body. Froriep has shown
(83) that in the chick this process forms a more essential element of the
body than in mammals. In the atlas it forms a lateral half of the

170 Development of Thoracic Vertebre in Man

ventral arch, but in the thoracic region of mammals it fuses with the
antero-lateral portion of the corresponding vertebral body. After its
junction with this the pedicle increases in size but otherwise shows no
marked alteration of form.

The transverse process is at first a short projection which lies at some
distance from its corresponding rib (Fig. 26). The cartilagenous rib
rapidly increases in size and at the same time the transverse process grows
outward and forward to meet it (Figs. 29, 32 and 34). At first the
developing cartilage of the rib and that of the transverse process are
embedded in a continuous blastema, but before chondrification has pro-
ceeded far, branches from successive intervertebral arteries become anas-
tomosed in the area between the neck of the rib and the transverse
process and separation is effected (Figs. 36, 38 B and 39).

Between the extremity of the transverse process and the rib a joint is
developed (Figs. 39, 40, 41 and 42), and the surrounding blastema con-
verted into costo-transverse ligaments.

The articular processes develop slowly from the cartilage. Extension
takes place anteriorly, A. A. Pr., and posteriorly, P. A. Pr., in the inter-
dorsal membrane. In an embryo of 14 mm. (Figs. 25, 26 and 27) these
articular plates are separated by a distinct interval. In one of 17 mm.
they have approached each other very closely (Fig. 37); and in one of
20 mm. not only do the articular processes show distinctly more form
(Figs. 28, 29 and 30), but in addition the superior articular process
slightly overlaps the inferior (Fig. 38). This overlap of the superior
articular processes is distinctly more advanced in an embryo of 28 mm.
(Fig. 39), and still more so in one of 33 mm. (Figs. 31-33). In an
embryo of 50 mm. (Figs. 34, 35 and 40) conditions essentially like the
adult have been reached.

The laminar processes scarcely exist in Embryo CXLIV (Fig. 26).
In Embryo XXII (Fig. 29) they have begun to project posteriorly to the
region of the articular processes (Fig. 29). The dense embryonic con-
nective tissue covering the laminar processes at this stage gives attach-
ment to a membrane covering, the dorsal musculature, F. D. M., and to
a membrane surrounding the spinal cord, M. Rk. D. This accounts for
the two projections seen dorsally on the side of the model representing
the membranous tissue. In Embryo CXLV, length 33 mm., the laminar
processes extend well toward the dorsal line (Figs. 32 and 33); in Em-
bryo LXXXIV, length 50 mm. (Figs. 34, 35 and 40), they completely
encircle the spinal canal and from the region of fusion of each pair a
spinous process extends distally, though not so far as in the adult.

Alterations in the cartilage of the neural processes preliminary to

Charles R. Bardeen iva
ossification begin at about the time they take place in the vertebral bodies.
They are first seen in an area which corresponds to that in which the
neural cartilage begins. The earliest calcification appears in Embryo
CLXXXIV, length 50 mm., in the arches of the first cervical to the sixth
thoracic vertebre.

The development of the ribs I shall not attempt in this place to describe
in detail. Figs. 25-34 and 37-42 show sufficiently well the relations of
the proximal ends of the ribs to the vertebra. They are developed oppc-
site the intervertebral disks. The blastemal tissue which surrounds the
developing heads of the ribs becomes converted into costo-vertebral liga-
ments. Differentiation in the cartilage preliminary to ossification takes
place in the shafts of the ribs even earlier than in the vertebral bodies
and in the neural processes. Ossification is well under way in the shafts
of the ribs of Embryo LX XIX, length 33 mm.; XCVI, length 44 mm.;
XCYV, length, 46 mm.; and LXXXIV, length 50 mm.

Summary of the Chondrogenous Period of Vertebral Development.—
Each cartilagenous vertebra is developed from four centers of chon-
drification. In addition, a separate center appears for each rib. In com-
paring these centers with the blastemal formative centers, we find that
each primative center of blastemal condensation enters into union with
tissue derived from the anterior half of the body-segment next posterior
and then gives rise to three centers of chondrification, one for the neural
arch, one for the rib and one for half a vertebra. When ossification first
takes place the centers for the ossification of the neural arches and the
ribs correspond to the original chondrification centers in the blastema,
but the centers for ossification of the bodies show little trace of the
bilateral condition which marks the cartilagenous fundaments. '

The processes of chondrogenous form differentiation are shown in
the drawings of the models. The period of ossification of the vertebrze
has been so often and so well described that no attempt will be made to
enter upon a further acount of it in this paper. I have, however, not
found two primary ossification centers, such as Renault and Rambaud
have described, for each neural arch.

LITERATURE.

BADE, P.—Die Entwicklung des menschl. Skelets bis zum Geburt. Archiv
mikr. Anat., LV, 245-290, 1900.

Batpus.—Die Intervertebral Spalte v. Ebners und die Querteilung der
Schwanzwirbel bei Hemidactylus mabuia. Dissertation, Leipzig, 1901.

BARDEEN.—Development of the Musculature in the Body-wall of the Pig.
Johns Hopkins Hospital Reports, IX, 367, 1899.

12

172 Development of Thoracic Vertebre in Man

and Lrewis.—Development of the Back, Body-wall and Limbs in Man.
American Journal of Anatomy, I, 1, 1901.

vy. EBNER.—Urwirbel und Neugliederung der Wirbelsaiile. Wiener Sitzungs-
berichte, XCVII, 3 Abtheil, 1888.

Ueber die Beziehungen des Wirbels zu den Urwirbeln. Wiener Sit-
zungsberichte,, CI, Abth. 8, 1892.

Frorrerp.—Zur Entwicklungsgesch. der Wirbelsatile. Archiv f. Anatomie und
Physiologie, Anat. Abth., 177-184; 1883, 69-150, 1886.

Gaupp.—Die Entwicklung der Wirbelsatile. Zool. Centralblatt, IV, 533-546,
849-853, 889-901, 1897.

GorrTtr.—Ueber den Wirbelbau bei den Reptilien und einigen andern Wir-
belthieren. Zeitschrift. f. wiss. Zoologie, LXII, 343, 1897.

Hacen.—Die Bildung des Knorpelskelets beim menschl. Embryo. Archiv f.
Anatomie und Physiologie, Anat. Abth., 1900.

HassE.—Die Entwicklung des Atlas und Epistropheus des Menschen und der
Saugethiere. Anat. Studien, I, 1873.

Hay.—On the Structure and Development of the Vertebral Column of Amfa.
Field Columbian Museum Publications, Zool. Series V, 11, 1897.
American Naturalist, XX XI, 397-406, 1897.

Hott, M.—Ueber die richtige Deutung der Querfortsatze der Lendenwirbel,
ete. Wiener Sitzungsb., 3 Abth., LXXXV, 181-232, 1882.

KAPELKIN.—Zur Frage tiber die Entwicklung des axialen Skelets der Amphi-
bien. Bull. Soc. Imper. d. Naturalisten, Moscow, 1900, 433-440.

KoLLMANN.—Die Rumpfsegmente menschlicher Embryonen von 13-35 Urwir-
beln. Archiv f. Anatomie und Physiologie, Anat. Abth., 1891.

MACALISTER.—The Development and Varieties of the 2d Cervical Vertebra.

MANNER.—Zeitschrift f. wiss. Zoologie, LXVI, 43, 1899.

MANNICH.—Beitrage zur Entwicklung der Wirbelsatile von Endypleschry-
scome. Jenaische Zeitschr., XXXVII, 1-40, 1902.

Moser, E.—Ueber das Wachsthum der menschlichen Wirbelsatile. Disserta-
tion, Strassburg, 1889.

RAMBAUD et RENAULT.—Origine et developpement des os. Paris, 1864.

RIDEwoop.—On the Development of the Vertebral Column in Pipa and
Xenopus. Anat. Anzeiger, XIII, 1901.

SCHAUINSLAND.—Uebersicht tiber die Entwicklung der Wirbelsdéule in der
Reihe der Vertebraten. Verhandl. Deutsch. Zool. Gesellsch., Wiirz-
burg, 112-113, 1903.

ScHomBure, H.—Entwicklung des Muskeln und Knochen des menschlichen
Fusses. Dissertation, Gottingen, 1900.

ScHULTzE.—Ueber embryonale und bleibende Segmentirung. Verhandl. der
Anat. Gesellschaft, 10 Vers., Berlin, 87-92, 1896.

WEIss, A.—Die Entwicklung der Wirbelsaiile der weissen Ratte, besonders

der vordersten Halswirbel. Zeitschr. f. wiss. Zoologie, LXVI, 492,

L901:

WELCKER.—Ueber Bau und Entwicklung der Wirbelsaiile. Zoolog. Anzeiger,
1878.

Charles R. Bardeen 173

ABBREVIATIONS USED TO DESIGNATE STRUCTURES ILLUSTRATED
IN THE FIGURES.

A. A. Pr., anterior articular process. WN. Pr., neural process.

C. V., cardinal vein. Pch. S., perichordal sheath.

OC. Pr., costal process. P. A. Pr., posterior articular process.
Cel., celom. Pd., pedicle.

Ch. d., chora dorsalis. Rib, rib.

Der., dermis. Scl., sclerotome.

Disk., intervertebral disk. Sptm., perichordal septum.

D. L., dorsal ligament. Sp. C., spinal cord.

D. M., dorsal musculature. Sp. G., spinal ganglion.

F. v. E., fissure of v. Ebner. Sp. N., spinal nerve.

F. D. M., fascia of dorsal musculature.Sp. Pr., spinous process.

Ids. M., interdiscal membrane. Trap., trapezius.

Idr. M., interdorsal membrane. Tr. Pr., transverse process.

Is. A., intersegmental artery. V. L., ventral ligament.

L., lamina, V. B., vertebral body.

Myo., myotome. 5, 6, 7, 5th, 6th and 7th thoracic vertebre.

WM. R. D., membrana reuniens dorsalis.

EXPLANATION OF FIGURES.

PAVE le

Fics. 1, 2,3 and 4. Frontal sections through the thoracic region of several
embryos during the blastemal period of vertebral development. 47.5 diam.
(1) Embryo CLXXXVI, length 3.5 mm. (2) Embryo LXXX, length 5 mm.
(8) and (4) Embryo CCXLI, length 6 mm. (3) Through the region of the
chorda dorsalis, (4) through a more dorsal plane. Figures 1, 3 and 4 repre-
sent sections cut somewhat obliquely so that the right side of the sections
is ventral to the left. In Figs. 2 and 4 on the right side the bodies of several
embryonic vertebre are represented in outline. In Figs. 2 and 3 owing to
artefacts the myotomes are pulled away from the sclerotomes.

PLATE II.

Fics. 5, 6 and 7. Cross-sections through midthoracic segments during the
blastemal period of vertebral development. 55 diam. (5) Embryo LXXVI,
length 4.5 mm. The right side of the section passes through the middle, the
left side through the posterior third of the 5th segment. (6) Embryo II,
length 7 mm. 5th thoracic segment. The right side of the drawing repre-
sents a section anterior to that shown at the left. (7) Embryo CLXXYV,
length 13 mm. The left half of the 6th vertebral body, neural process and rib
are drawn in detail, the body-wall, spinal cord and spinal ganglion are shown
in outline.

PLATES III AND IV.

Fies. 8, 9, 10, 11, 12 and 13. Views of models representing the blastemal
stage of vertebral development. (8-10) Embryo II, length 7 mm., 3314 diam.
(11-13) Embryo CLXIII, length 9 mm., 25 diam. (14-16) Embryo CIX, length
11 mm., 25 diam. 8, 11 and 14 views from in front; 9, 12, 15, views from the
side; 10, 13, 16, views from behind.

174 Development of Thoracic Vertebree in Man

PLATE V.

Fies. 17-24. Sagittal sections in the mid-line through the 6th, 7th and 8th
thoracic segments of a series of embryos from 12 to 50 mm. long. (17) Em-
bryo CCXXI, length 12 mm. This section includes several segments anterior
and posterior to the three above mentioned, 6th, 7th and 8th. (18) Embryo
CXLIV, length 14 mm. (19) Embryo CVIII, length 22 mm. (20) Embryo
LXXXVI, length 30 mm. (21) Embryo CXLY, length 33 mm. (22) Embryo
LXXIX, length 33 mm. (23) Embryo XCVI, length 44 mm. (24) Embryo
CLXXXIV, length 50 mm.

PLATE VI.

Fies. 25-35. Dorsal, lateral and ventral views of models made by the Born
method to illustrate vertebral form-differentiation in the 6th, 7th and 8th thor-
acic vertebre during the chondrogenous period. On the left side the carti-
lagenous, on the right the enveloping fibrous tissue is shown. The latter is
also shown on the eighth vertebra in Figures 29 and 35. (25-27) Embryo
CXLIV, length 14 mm., 20 diam. (28-30) Embryo XXII, length 20 mm.,
13 diam. (31-33) Embryo CXLYV, length 33 mm., 10 diam. (34, 35) Embryo
LXXXIV, length 50 mm., 10 diam. (34) Dorsal view, left half; (35) median
view.

PLATH VII.

Fics. 36-42. Transverse sections through mid-thoracic vertebre of a series
of embryos. 5 diam. (36) Embryo CVI, length 17 mm. (37) Embryo
CCXVI, length 17 mm. (38) Embryo XXII, length 20 mm. (39) Embryo
XLV, length 20 mm. (40) Embryo LXXXIV, length 50 mm. (41) Embryo
XLIV, length 70 mm. (42) Embryo XXIII, length 70 mm.

The models from which the illustrations in this article were drawn
have been reproduced by Dr. B. EF. Dahlgren at the American Museum
of Natural History, New York, N. Y., and arrangements may be made
for securing copies by applying to the Director of the Museum. |

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN
C. R. BARDEEN

PLATE |

Fig.2

AMERICAN JOURNAL OF ANATOMY--VOL. Iv

: ‘ i ie
an i -
Hy :
' Fur ;
uy 7 a 9
* :
_
we > \,
7
¢
i
;
*
,) 4
* §
%
4

PLATE Il

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN

C. R. BARDEEN

IV

AMERICAN JOURNAL OF ANATOMY--VOL.

_.

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN PLATE Ill
Cc. R. BARDEEN

N.Pr.

Is.A.

C.Pr.

Is.A.

Fig.10

AMERICAN JOURNAL OF ANATOMY--VOL. IV

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN
Cc. R. BARDEEN

Ch.d.

: OTe |

. Mi mn
\ i Ni H

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE IV

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN PLATE V
Cc. R.. BARDEEN

Fig. 18

Fig.22

Fig. 24

AMERICAN JOURNAL OF ANATOMY--VOL. IV

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN PEAKED)
c. R. BARDEEN

< Fig.33 :

AMERICAN JOURNAL OF ANATOMY=-VOL. IV

ae Pee see

LAA
, [a * 2

¢ aren

.' aie : pay n°

THE DEVELOPMENT OF THE THORACIC VERTEBRA IN MAN PLATE VII
Can BARDEEN

AMERICAN JOURNAL OF ANATOMY--VOL. IV

13

r

THE ELASTIC TISSUE OF THE HUMAN LARYNX.

BY
DEAN D. LEWIS, M. D.
From the Hull Laboratory of Anatomy, University of Chicago.

WitH 5 PLATES.

The elastic tissue of the larynx, since it was first described by Lauth,’
in 1835, has been the object of frequent studies by anatomists and laryn-
geologists. The functional importance of this tissue in the production
and modification of voice has aroused interest in its distribution and
arrangement, and while the earler descriptions, which are based upon the
study of specimens prepared by methods which do not reveal the finer
elastic fibrils, are correct, a knowledge of the more minute relations exist-
ing between the elastic fibers, muscle, cartilage and epithelium is desir-
able.

The introduction of specific stains for elastic tissue has given a new
impetus to its study from a developmental and pathological view-point.
and has drawn attention to the elastic tissue of the larynx, as is evidenced
by the number of articles appearing upon the subject. Friedrich’s
work followed closely upon the introduction of the Taenzer-Unna orcein
stain, and Katzenstein, using the Weigert resorcin-fuchsin method, has
repeated recently the former’s work. The Weigert. method certainly
differentiates the elastic fibers more distinctly than the Taenzer-Unna
method, and stains the finer fibrils, which may escape the latter. These
two investigators agree in general, but differ in so many points that the
publication of this article, which was about completed when Katzenstein’s
article appeared, seems justified.

The specimens from which this study is made were prepared from the
larynges of the new-born. They were hardened in alcohol and Zenker’s
fiuid. The former fixative, recommended by Weigert, gives excellent
results, but equally good results are obtained after fixation in Zenker’s
fluid. The ordinary celloidin technique was used, with the addition of

1Lauth: Mem. de l’acad. royale de med., 1835, t. 4, p. 98.

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

176 The Elastic Tissue of the Human Larynx

the slow method of celloidin infiltration, as recently recommended by
Miller.’

Sections, fifteen micra in thickness, were made and mounted serially.
The sections were passed through a mixture of alcohol and glycerine,
from which they were transferred to a paper corresponding in size to
the cover-glass to be used. The paper was then inverted upon a slide,
which had previously been coated with a thin layer of albumin fixative.
Between two slides thus prepared a piece of filter paper was inserted, and
the two slides were tied together. They were then placed in a thermostat
until dry, when they were removed and placed in equal parts of absolute
alcohol and ether to remove the celloidin. The sections are always firmly
attached, and no care need be exercised to prevent their floating off. This
method was devised by Prof. EH. C. Jeffrey. Weigert’s resorcin-fuchsin
method was used to stain the elastic fibers. Orange G. as a counter-
stain offers a sharp contrast to the blue-black of these fibers. Van
Gieson’s picro-fuchsin was used in studying the collagenic fibers and the
nodules found in the anterior extremities of the vocal cords.

Henle * describes beneath the mucous membrane of the larynx an elastic
fiber layer, which in some regions is poorly, and in others well developed,
and in some closely and in others loosely connected with the epithelium.
Where this layer is thickened after removal of the mucous membrane or
the tissue which covers them externally, there remain ligaments. These
ligaments are attached at definite points to the perichondrium of the
laryngeal cartilages, and such points of attachment may be regarded as
the points of origin of the ligaments. It is not to be disregarded, how-
ever, that the elastic fibers of these ligaments are in direct continuity
with the elastic elements of the whole mucous tract, and that, therefore,
their limits are not sharp and are arbitrarily made.

Following the classical description of Luschka* the elastic tissue of
the larynx may be divided for descriptive purposes into three zones,
corresponding to the three compartments of this organ. The inferior
zone includes all the elastic tissue within and below the ligamenta vocalia ;
the middle zone includes the elastic tissue surrounding the ventriculus
laryngis; the superior zone includes the elastic tissue of the membrana
quadrangularis and epiglottis. The discussion of the arrangement of
the elastic fibers will be proceeded with in the order given above.

*Miller: J. of Applied Microscopy and Laboratory Methods. Rochester,
INS Ye; Vol: 6; No: 4.

3 Henle: Handb. der Eingeweidelehre des Menschen. Braunschweig, 1873,
p. 254.

*Luschka: Der Kehlkopf des Menschen. Tiibingen, 1874.

Dean D. Lewis aida

ELASTIC TISSUE OF INFERIOR ZONE.
Conus Exasticus (LuscHKa): Lice. CRICO-THYREO-ARYTANOIDEA
OF KRAUSE” AND Lic. CRICOTHYREOIDEUM MeEpIuM (HENLE).

If a lamina of the thyroid cartilage and the subjacent muscle be re-
moved, a fan-shaped mass of elastic fibers will be seen, which passes from
the angle of the thyroid cartilage downward, backward and _ laterad
to be attached to the ascending upper border of the cricoid cartilage
and the inferior surface of the vocal process of the arytenoid cartilage.
This elastic membrane is the corus elasticus of Luschka. Sections at
various levels through it will be described.

In frontal sections, made through the anterior part of the conus elasti-
cus, a dense network of elastic fibers arising from the upper border of
the cricoid cartilage, and passing cephalad and laterad to be attached to
the lower border of the thyroid cartilage, will be seen. ‘The fibers com-
posing this network tend to pass vertically, anastomose freely, and some
of the fibers arise on each side of an indefinite median raphé. -On both
sides these fibers are continuous with elastic fibers passing from the upper
border of the cricoid cartilage. The mass of elastic fibers occupying the
median line form the ligamentum cricothyreoideum medium, and
are seen to be merely the anterior continuation of the conus elasticus,
as previously shown by His. This ligament is pierecd by the crico-
thyroid artery, and the arrangement of some of its component fibers
about a median raphé suggests that functionally it is divided into sym-
metrical parts. (See Fig. 1:)

In frontal sections made through the middle of the conus elasticus,
elastic fibers, few in number, and small in size, are seen arising from the
upper border of the cricoid cartilage, and passing cephalad and mediad
to reach the ligamentum vocale. These fibers, increasing constantly in
number and size as they ascend, form a gentle curve, the convexity of
which is directed mediad, being separated in the subglottic region from
the subepithelial elastic layer, by numerous glands, and loose connective
tissue, which favors the development of cedema at this point. The con-
cavity is occupied by the musculus vocalis. The fibers below are obliquely
arranged, and only as the plica vocalis is approached do they tend to
become sagittally directed and parallel. (See Fig. 2.)

Sections made through the posterior part show that the conus becomes
shorter and approaches nearer the median line. The fibers are more
uearly vertical in arrangement, passing up to the inferior surface of the

5 Krause: Handb. der menschlichen Anatomie. Hannover, 1879.

178 The Elastic Tissue of the Human Larynx

vocal process of the arytenoid cartilage. While these sections are instruc-
tive, much more may be learned by tracing transverse sections serially.

In transverse sections made through the ligamentum cricothyreoideum
medium, the elastic fibers are found grouped in well-defined vertical
bundles, which are separated from each other by horizontal fibers passing
inward toward the indefinite median raphé. Fine elastic fibers, verti-
cally directed, which surround numerous groups of mucous glands, pass
from the subepithelial layer to the ligament. (See Fig. 4.)

More posteriorly the fibers of the conus pass obliquely, upward, for-
ward and mediad, but they are intersected by other fibers passing
almost at right angles to them. The oblique fibers predominate, both in
size and number. ‘Traced backward, the elastic fibers of the conus are
found to be attached to the cricoid cartilage. Anastomosing fibers, most
numerous anteriorly, connect the fibers of the conus with the subepithe-
lial layer, where the two systems are not separated by the glands pre-
viously mentioned. (See Fig. 5.)

In transverse sections just below the ligamentum vocale, the arrange-
ment of elastic fibers anteriorly has been accurately described by Fried-
rich. The elastic tissue, reduced in amount, is replaced by collagenic
fibers, which separate the fibers of the conus from the hyaline substance
of the thyroid cartilage. This collagenic tissue is rich in glands. The
nearer the ventricle is approached, the more nearly horizontal the fibers
become to form the

LIGAMENTUM VocALE. (See Fig. 7.)

Henle, in describing the ligamentum vocale, states that some of its
elastic fibers fuse posteriorly with the elastic cartilage, forming the vocal
process of the arytenoid. Other fibers are attached about the spina
inferior, above the vocal process, and from this attachment fibers course
upward posterior to the ventricle of the larynx. Still other fibers are
inserted below the vocal process upon the medial surface of the arytenoid
cartilage, or upon the anterior surface of the cricoid cartilage.

Kanthack * describes posteriorly a sesamoid cartilage, which marks the
point of transition of the elastic fibers of the ligamentum vocale into the
vocal process of the arytenoid cartilage.

Reinke’s * description of the arrangement of the elastic fibers at the
posterior extremity of the hgamentum vocale, and their relation to the

®Henle: Handb. der Eingeweidelehre des Menschen; p. 255.
*Kanthack: Arch. f. path. Anat., etc., Berl., Bd. exvii, p. 533.
SReinke: Anat. Hefte. Bd. ix, pp. 108-110.

Dean D. Lewis . 179

processus vocalis of the arytenoid cartilage, is the most accurate that has
yet been given. His findings can merely be verified; nothing can be
added. He states that in macroscopical preparations it can be seen that
the ligamentum vocale has a wide area of attachment to the vocal process
of the cartilage, covering its upper and medial surfaces, leaving the
lower and lateral surfaces free for the attachment of the fibers of the
musculus vocalis. The study of microscopical preparations made in
frontal and horizontal planes shows that the middle fibers of the ligament,
only, are the direct continuation of the fibers of the elastic cartilage,
which forms the apex and anterior border of the vocal process. The
greater number of fibers, occupying the lateral part of the hgament, are
derived from the perichondrium, which consists here almost wholly of
elastic fibers, which cover the elastic cartilage above, laterally and
medially.

In the beginning of the elastic cartilage the fibers intersect each other
at various angles, but soon pass parallel in a sagittal plane. The fibers
of the perichondrium upon the medial and lateral surface of the cartilage
pass in front of the apex and anterior border of the processus vocalis in
curves, which intersect each other at right angles. All these fibers form
anterior to the vocal process a dense network, probably the sesamoid car-
tilage of Kanthack, out of which parallel fibers emerge, to pass forward.
The fibers of the hgament receive from the side additional fibers, which
he between the bundles of the musculus vocalis, and by means of its
perimysium they are attached to the vocal process of the arytenoid
cartilage.

In the middle part of the hgament the fibers usually course parallel
to each other, and in a sagittal direction. When the processus vocalis is
in the position of rest, two great divisions of the ligament may be differ-
entiated. The one adjacent to the muscle is dense and only in thin sec-
tions can the separate fibers be differentiated; the other, equally wide,
borders upon the preceding, above and medially, and has the same form.
In sections, it is recognizable as a lighter zone. The fibers of the denser
part of the ligament form a curve, the concavity of which is directed
toward the free border of the labium vocale, while the fibers of the less
compact part are straight.

Upon superficial examination, the fibers of the ligament seem to pass
parallel to each other, without anastomosing. C. L. Merkel has stated
that the elastic fibers of the ligamentum vocale differ from the elastic
fibers in other parts of the body, in that they do not anastomose. Reinke
has shown, however, by the use of specific stains and higher lenses that
there are anastomosing fibers, but that the principal fibers are so well

180 The Elastic Tissue of the Human Larynx

developed that it is difficult to see the finer anastomosing ones. Reinke’s
description is most accurate, and he has demonstrated that the elastic
fibers of the ligament have a definite structure, relative to their function.
His conclusions may be summed up as follows: The elastic fibers of the
ligamentum vocale attain their greatest development in planes parallel
to constant tension, and at right angles to constant pressure, while the
fibers passing obliquely to anastomose with either of the above systems
remain atrophic or have disappeared.

THe ANTERIOR ATTACHMENTS OF THE LIGAMENTA VOCALIA.

There is so much difference of opinion among anatomists and laryn-
geologists concerning the anterior attachment of the hgamentum vocale,
that some account of the views held by the different investigators may
be expedient.

C. Mayer’ was the first to describe in the anterior extremity of each
ligament a small cartilaginous nodule, which was found by him in man,
and some of the higher apes. These nodules are sometimes designated
as the “ cartilagines sesamoidez anteriores.” (See Fig. 8.)

Gerhardt * noted in sections made at the level of the insertion of the
ligamenta vocalia into the thyroid cartilage a small firm median process,
occupying the angle of the thyroid cartilage, which he considered
tc be formed from its hyaline substance. This median process is pro-
longed on each side by yellowish flexible bands into the anterior extremi-
ties of the hgamenta vocalia. The yellowish color of the anterior com-
missure and its thickening is produced by these processes. The intimacy
of the relation between the median process and its lateral prolongations
is variable. However, Gerhardt macerated sections many days in water,
but was still unable to separate them without the use of a knife. He
considered that microscopically there is repeated at the anterior com-
missure the same histological structure that Rheiner had previously
described for the processus vocalis of the arytenoid cartilage, and sug-
gested that this median process, with its lateral pr sloneee be known
as the poe vocalis of the thyroid cartilage.

Verson “ notes that the ligamentum vocale is thickened to form a small
nodule immediately behind its attachment to the angle of the
thyroid cartilage. Upon sectioning this nodule, it is seen to be composed

®Mayer, C.: J. F. Meckel’s Arch. f. Anat. u. Physiol.; 1826, p. 194.
Gerhardt: Arch. f. path. Anat., etc., Berl., Bd. xix, pp. 436-437.

™ Verson: Stricker’s Handb. der Lehre von den Geweben des Menschen u.
der Thiere. Leipzig, 1871, p. 460.

Dean D. Lewis. 181

of elastic fibers, round and spindle cells. It is found in the new-born.
Chondrification never occurs in it.

Sappey ~ states that the ligamentum vocale is attached anteriorly by
means of a nodule composed of elastic tissue, “nodule glottique
anterieur.”

Frankel * describes in the anterior extremity of the hgamentum vocale
a small firm nodule, which is attached to the hyaline thyroid cartilage
by loose fibrous tissue. He states that the nodule undoubtedly contains
cartilage cells.

Nicolas “ speaks of the “ cartilagines sesamoidex anteriores” as small
white or yellowish-white nodules, which can be isolated without difficulty
from the fibro-elastic tissue in which they are lodged. They do not
exceed a millet seed in size; often they are smaller. They are attached
to the angle of the thyroid cartilage by a dense fibrous tissue.
In the majority of cases, if not in all, among the elastic and collagenic
fibers forming the nodules, cells are found which are undoubtedly carti-
laginous in character.

Reinke” describes the nodule as being composed of a deeply stainable
substance, resembling histologically the tissue just anterior to the pro-
cessus vocalis of the arytenoid cartilage. Some cartilage cells are found
in the nodule. The elastic fibers passing off from the anterior extremity
of this nodule are attached to the perichondrium of the thyroid cartilage.

None of the investigators had paid much attention to the finer histo-
logical structure of the median process described by Gerhardt, and it re-
mained for Friedrich” to study it in detail, and to question the con-
clusions of the former. Friedrich describes accurately the relation
existing between the collagenic fibers, occupying the angle of
the thyroid cartilage, and the anterior fibers of the conus elasticus. This .
fibrous tissue increases in amount as sections passing cephalad are ex-
amined; it attains its greatest development opposite the anterior attach-
ments of the ligamenta vocalia. The fibrous tissue will be found
erouped in well-defined vertical bundles immediately adjacent to the
thyroid cartilage. These vertical bundles are surrounded by horizontal
fibers. Posteriorly, as the anterior extremities of the hgamenta vocalia
are approached, the bands of fibrous tissue become horizontal, corre-
sponding in direction to a horizontal group of elastic fibers, passing off

p)

2 Sappey; quoted by Friedrich.

3 Frankel: Arch. f. Laryngol. u. Rhinol. Bd. i.

™ Nicolas: Poirier et Charpy’s Traite d’anatomie humaine., t. iv, p. 435.
“Reinke: Anat. Hefte. Bd. ix, p. 110. .

1% Friedrich: Arch. f. Laryngol. u. Rhinol. Bd. iv, pp. 192-193.

—
io 6)
©

The Elastic Tissue of the Human Larynx

from the anterior extremities of the nodules described by Mayer. Inter-
spersed throughout this fibrous tissue are some cartilage cells and fine
elastic fibrils, which are vertically directed. (See Fig. 8.)

Friedrich states that the perichondrium passes between the hyaline
substance of the thyroid cartilage and the median process described by
Gerhardt, and that it forms a distinct line of separation between the two.
He concludes, therefore, that there is no gradual transition from a hya-
line to an elastic cartilage, as is the case in the arytenoid cartilage, and
that Gerhardt is not justified in speaking of the process described by
him, as the processus vocalis of the thyroid cartilage. Friedrich found
no cartilage cells in the nodules occupying the anterior extremities of
the ligamenta vocalia.

Katzenstein,” while agreeing with Friedrich in many points, takes
issue with him concerning the anatomical significance of the median
process. He states that the perichondrium is reflected upon the sides
cf the process, and that it does not form a line of separation between the
fibrous median process and the hyaline substance of the thyroid cartilage.
In establishing the perichondrial relation, he makes use of the law first
suggested by Rawitz, that the perichondrium exerts a directive influence
upon the orientation of cartilage cells. Upon the external surface of the
thyroid cartilage the cells are arranged parallel to the fibers of the peri-
chondrium. In the center of the cartilage they he at right angles to its
long axis; upon the inner surface of the cartilage the cells are arranged
parallel to the fibers of the perichondrium, until the median process is
reached, where they are gathered into irregular clusters and some of the
cartilage cells are displaced posteriorly. Katzenstein agrees with Ger-
hardt regarding the anatomical significance of this process, and considers
it as quite comparable to the processus vocalis of the arytenoid cartilage.
Like Reinke, in his work upon the ligamentum vocale, Katzenstein has
shown that the fibers are arranged in definite tension planes.

In some of the lower animals, white rat, cat, ete., Katzenstein has de-
scribed between the laminze of the thyroid cartilage a wedge-shaped
cartilage, which is covered by elastic fibers anteriorly, and receives poster-
iorly the attachment of the anterior extremities of the ligamenta vocalia.
According to him, this wedge-shaped cartilage is the homologue of the
median process described by Gerhardt.

While Katzenstein has accurately described the histological findings,
he has misinterpreted their anatomical significance. If the develop-
mental history of the thyroid cartilage be reviewed, some facts will be

7 Katzenstein: Arch. f. Laryngol. u. Rhinol. Bd. xiii, pp. 336-337.

Dean D. Lewis . 183

met with which will explain the different orientation of cartilage cells
adjacent to the process, and the significance of the wedge-shaped carti-
lage described by the investigator.

Rambaud and Renault,” in discussing the development of the thyroid
cartilage, say that the lamine of the thyroid cartilage are united by means
of a circumscribed median cartilage—“ le cartilage vocal.” It may be
distinguished by its transparency. This cartilage is well-marked in young
subjects. In the adult, however, the cartilage may not be present, but
is represented by an indistinct point of ossification. It is lozenge-shaped,
and its borders unite with the lamine of the thyroid cartilage.

Henle™ states that horizontal sectivns through the thyroid cartilage
show that its lamine are separated more or less distinctly from a middle
piece by a condensed layer of interstitial substance, curved so that the
convexity is directed mediad. ‘This middle piece of the thyroid carti-
lage is the lamina mediana cartilaginis thyreoidee of Halbertsma. In
this piece the cartilage nests are smaller and more closely grouped than
in the lamine. The ligamenta thyreo-arytenoidea inferioria and their
corresponding muscles arise from this part, or a connective tissue mass,
connected with it. The fibers from the mass pass a short way into the
middle piece, so that this tissue immediately posterior to the hyaline
cartilage resembles in structure fibro-cartilage.

Nicolas” states that immediately after birth there is found in the
median line of the thyroid cartilage, at the level of the vocal cords, a
special arrangement of the cartilage cells, corresponding to the position
of the lamina mediana. In the adult this middle portion of the thyroid
cartilage can be distinguished from its lamin only by the different
orientation of its cells.

The embryology of the thyroid cartilage explains the different orienta-
tion of the cartilage cells adjacent to the median process, and in it. I
have sectioned the larynges of young dogs, and have found, occupying
the space between the lamine of the thyroid cartilage, the wedge-
shaped cartilage which Katzenstein states is comparable to the median
process described by Gerhardt in man. I consider it to be the lamina
mediana which fuses later in the lower animals than in man.

Friedrich’s description of the median process is correct, with the
exception of that of the perichondrial relation. The perichondrium does
not pass between the hyaline substance of the thyroid cartilage and the

1’ Rambaud et Renault: Origin et devellopement des os. Paris, 1864.
” Henle: Handb. der Eingeweidelehre des Menschen; p. 243.
* Nicolas: Poirier et Charpy’s Traite d’Anatomie humaine. T. 4, p. 447.

18+ The Klastie Tissue of the Human Larynx

fibrous tissue composing the process, nor is it reflected upon the sides
of it, as described by atzenstein. ‘The whole process is formed by a thick-
ening of the perichondrium, which passes directly backward to receive the
attachment of the elastic fibers of the ligamenta voealia. This is indi-
‘ated by the arrangement of the elastic fibers of the perichondrium at
this point, which, adjacent to the thyroid cartilage, are directed either
antero-posteriorly or obliquely. The attachment of the elastic fibers of
the ligamenta vocalia at this point is comparable to their attachment
by means of the perichondrium to the laryngeal cartilage at other points.
The great increase in the amount of elastic tissue in the ligamenta vocalia
Gemands an increase in the number and size of the perichondrial fibers
to which they are attached. This increase, gradual from below up-
wards, corresponds to the increase in the number of the fibers composing
the conus elasticus, as it passes upward.

Adjacent to the thyroid cartilage, the fibers are grouped in vertical
bundles, which are separated from each other by horizontal fibers. Pos-
teriorly, the perichondrial fibers are collected into horizontal bundles,
which are separated from each other by blood vessels, and ducts of glands,
vertically directed. Into these parallel bands of fibrous tissue are attached
parallel bundles of elastic fibers, which pass off from the nodules occupy-
ing the anterior extremities of the lgamenta vocalia. Scattered
throughout the perichondrium at this point are fine elastic fibers, verti-
cally directed. I have been unable to find cartilage cells as far posterior
as described’ by Friedrich and Katzenstein. These cells seem to be
restricted to a narrow zone immediately adjacent to the thyroid carti-
lage.

In specimens stained by Van Gieson’s method, fibrous bands are seen
to pass off posteriorly from perichondrium to surround a nodule, which
oceupies the anterior extremity of each ligamentum vocale, and which
is composed of round and spindle cells. While histologically these
nodules resemble somewhat those found anterior to the processus vocales
of the arytenoid cartilages, I have been unable to find cartilage cells
in them. (See Fig. 9.)

In specimens stained by Weigert’s method, the elastic fibers of the
ligamentum vocale will be seen to pass into the nodule, parallel to each
other, posteriorly, while from its anterior extremity and medial surface
several heavy bundles of anastomosing elastic fibers pass off, to be attached
to the parallel fibers of the perichondrium. Some elastic fibers appar-
ently originate in the nodule, for the fibers passing off from its anterior
extremity exceed in number those passing into it posteriorly.

In conclusion, I agree with Friedrich in not considering the process

Dean D. Lewis 185

described by Gerhardt as the processus vocalis of the thyroid cartilage.
It may be compared to the other perichondrial processes by which the
elastic tissue of the larynx is attached to the laryngeal cartilages at differ-
ent points. It is impossible to explain why the thyroid cartilage develops
as it does, but the relation existing between the fibrous tissue occupying
the concavity of the lamina mediana, and the elastic tissue of the liga-
mentum vocale does not resemble in the least the histological relations
existing between the hyaline substance of the arytenoid cartilages, and
their vocal processes.

I have not found cartilage cells in the nodules occupying the anterior
extremities of the ligamenta vocaha. I would suggest that these
nodules be known as the noduli vocales. It is difficult to assign to
these nodules their physiological function, but the increase in the number
of elastic fibers and their arrangement at this point, strengthened, as
they are, by numerous round and spindle cells, would suggest that the
ligamenta vocalia are here subjected to their greatest tension, and are
therefore re-inforced.

THE RELATION OF THE MuscuLus VOCALIS TO THE
LIGAMENTUM VOCALE.

The relation of the musculus vocalis to the elastic fibers of the hga-
mentum vocale is highly important, and although it has been studied by
many anatomists and laryngologists in recent years, there is at the
present time no uniformity of opinion concerning it, or its functional
significance in the production or modification of higher tones.

Ludwig *

‘ describes the musculus thyreo-arytenoideus as being divided
into a portio aryvocalis and arythyreoidea. The former division begins
upon the lower extremity of the anterior surface of the arytenoid carti-
lage, and passes in parallel bundles by the side of the ligamentum vocale,
to end in it. The shorter fibers end directly anterior to the apex of the
vocal process; the longer near the thyroid cartilage. These fibers, acting
simultaneously, draw the ligamentum vocale downward and outward. If
they act independently of each other, different segments of the ligament
will be affected in different ways. Fibers anterior to the insertion of the
muscle are rendered tense, while the fibers posterior to it are relaxed.
Ludwig regards the ligamentum vocale as the tendon of the musculus
thyreo-aryteenoideus.

Verson ~ denies that any fibers of the musculus thyreo-arytenoideus
are inserted into the elastic fibers of the ligamentum vocale.

“1 TLudwig: Lehrbuch der Physiologie der Menschen. Bd. i, pp. 567-570.
“Verson: Beitrage z. Kenntniss des Kehlkopfes u. der Trachea. Wien,
1868, p. 3. :

186 The Elastic Tissue of the Human Larynx

Luschka,” after making a study of larynges in which the musculature
was well developed, came to the conclusion that the muscle fibers belong-
ing to the free border of the ligamentum vocale pass along the whole
length of the ligament, retaining their muscular characteristics from the
arytenoid to the thyroid cartilage. His findings in the larynges of
children verified his conclusions as to the condition in the adult.

Henle™ states that the fibers of the musculus thyreo-artyenoideus
internus, adjacent to the ligamentum vocale, are small. The fibers lying
nearest to the ligament pass in between the elastic fibers composing it,
and are closely connected with them. A number of the muscle fibers
either arise from or end among the elastic fibers of the ligament. Regard-
ing the functional significance of this relation, he says that the fibers
ending in the elastic‘tissue must have some infiuence upon the movements
of the ligamentum vocale, and suggests that the short fibers acting upon
segments of the ligament may account for the production of falsetto
tones.

Jacobson ~ describes the musculus thyreo-arytenoideus as having a very
complicated structure. He finds, in horizontal sections, muscle fibers
arising from the processus vocalis, and the lateral surface of the lower
part of the arytenoid cartilage, which pass inward to the free border of
the ligamentum vocale, and end in bundles of parallel elastic fibers, which
eventually pass into the hgamentum vocale. He sums up his conclusions
concerning these muscle fibers by saying that there can be no doubt that
the musculus aryvocalis of Ludwig may be so developed in some cases,
that the ligamentum vocale may be rendered tense, while the arytenoid
cartilage remains stationary or in the position of adduction. Thus, the
short fibers of the muscle may oppose the long fibers, which act as
adductors.

Kanthack™ states that. the medial fibers of the musculus thyro-
arytznoideus pass between the elastic fibers and appear to end in them.
In sections, which are made exactly parallel to the course of the muscle
fibers, it can be seen, however, that they pass uninterruptedly from ary-
tenoid to thyroid cartilage, without ending among the elastic fibers of
the ligament. The ligament is not to be regarded as the tendon of the
muscle.

Friedrich notes that there is no definite arrangement of the elastic

° Tuschka: Der Kehlkopf des Menschen. Tiibingen, 1871, p. 121.
**Henle: Handb. der Hingeweidelehre des Menschen; p. 266.

*> Jacobson: Arch. f. Mikr. Anat., Bonn, Bd. xxix, pp. 624-627.

* Kanthack: Arch. f. path. Anat., etc., Berl., Bd. cxvii, p. 542.

7 Wriedrich: Archiv. f. Laryngol. u. Rhinol. Bd. iv, p. 207.

Dean D. Lewis 187

fibers about the end of the muscle fiber. He is inclined to believe, how-
ever, that there is a close relation between the muscle fibers and the
elastic elements of the cord, and emphasizes the fact that by muscle fibers
leaving the body of the muscle and running for a short distance in the
ligament, as fine an influence could be exerted upon the elastic elements
of the ligament as could be explained by the idea of the insertion of
muscle fibers directly into the elastic fibers. He does not regard the
ligament as the tendon of the musculus vocalis.

Katzenstein ~ does not regard the ligamentum vocale as the tendon of
the musculus vocalis. He has never seen the direct transition of muscle
fiber into elastic fiber. ,

In horizontal and frontal sections, muscle fibers may be found, which
are closely related to the elastic fibers of the ligamentum vocale. I have
found these to be most numerous posteriorly, in front of the vocal process
of the arytenoid cartilage. These muscle fibers have no such complicated
arrangement as Jacobson depicts. They seem to pass in between the
elastic fibers of the ligament, and to be surrounded by these fibers, but it
is probable that they do not end among the elastic fibers of the ligament.
Smirnow’s investigation ” as to the mode of insertion of striated muscle
into soft tissue will aid in settling this question. He says that in all
cases in which striated muscle is not in direct relation to the bony or
cartilaginous skeleton, in which the fibers are attached to the softer varie-
ties of connective tissues, these tendons consist, wholly or almost wholly,
of elastic tissue. In attempting to establish this relation, I have been
unable to find in any case a transition of muscle fiber into elastic tissue.

The laryngoscopic findings in the production of falsetto tones as given
by Stork, quoted by Jacobson, would suggest that the ligamentum vocale
may act in segments. I am inclined to believe that these fibers, which are
so closely related to the elastic tissue of the ligamentum vocale, but still
cannot !:e considered as inserting into it, may by their contraction make
tense the vocal ligaments, while the arytenoid cartilage remains stationary,
and may by their contraction render the production of falsetto tones pos-
sible. There is still another possibility, however. The fibers of the liga-
mentum vocale, as they pass forward to their anterior attachment, are
re-inforced by additional elastic fibers, which are derived from the
perimysium of the musculus vocalis, and through it are attached to the
arytenoid cartilage. It is possible that by the contraction of muscle
fibers related to these elastic elements different segments of the cord

* Katzenstein: Arch. f. Laryngol. u. Rhinol. Bd. xiii, p. 346.
7° Smirnow: Anat. Anz., Jena. Bd. xv, p. 488.

188 The Elastic Tissue of the Human Larynx

could be acted upon, and the cord abducted and rendered tense, while the
arytenoid cartilage remained in a position of adduction.

.

MIDDLE ZONE.

THE ELAstic TISSUE OF THE VENTRICULUS LARYNGIS.

The continuation upward of the subepithelial elastic layer of the
labium vocale forms the delicate elastic membrane surrounding the ven-
triculus laryngis. This membrane lies just beneath the epithelium of
the ventricle, and is poorly developed. Above, it becomes continuous
with the elastic tissue occupying the plica ventricularis. ‘Transverse
sections through the lower part of the ventricle and through the vocal
ligament show well the relation existing between the elastic fibers belong-
ing to each at this level (Fig. 7). The elastic fibers upon the medial
surface of the ventricle are continuous with the lateral fibers of the
ligament. The elastic fibers upon the lateral surface of the ventricle
become continuous posteriorly with the fibers of the hgament, or are
attached to the antero-lateral surface of the arytenoid cartilage. Anter-
iorly, as they approach the nodulus vocalis, they divide, part to be inserted
into the nodule and part to be reflected laterad to become continuous
with the elastic fibers of the perichondrium, covering the inner surface
of the thyroid cartilage.

SUPERIOR ZONE.
ARYTENO-EPIGLOTTIDEAN JAIGAMENTS: MEMBRANA QUADRANGULARIS.

The elastic fibers of this zone are situated in the aryteno-epiglottidean
folds. Their general direction is from above and anteriorly, downward
and backward. Posteriorly, these fibers are attached to the medial sur-
face of the arytenoid eartilages, and anteriorly, to the lateral borders
of the epiglottis. Above, they pass upward to the free borders of the folds,
and are related to the cartilages of Santorini and Wrisberg. Below, the
fibers become thickened to form the ligaments occupying the labia ven-
tricularia.

LIGAMENTUM VENTRICULARE: LIGAMENTUM ‘'THYREOARTHNOIDEUM
SUPERIUS.

Luschka describes the elastic fibers forming the ligamentum ventri-
culare as being grouped into well-defined bundles anteriorly and poster-
iorly. In the middle of their course the elastic fibers are separated from
each other by numerous glands.

Dean D. Lewis . 189

Verson” denies the existence of a proper ligamentum ventriculare.
He states that the elastic fibers occupying the labium ventriculare have
no definite direction. A section made at right angles to the labium
reveals some elastic fibers which are separated from each other by
numerous glands. Interspersed among the elastic fibers are numerous
eollagenie fibers. |

Henle™ describes the ligamenta ventricularia as arising on each side
of the ligamentum thyreo-epiglotticum from a connective tissue mass,
which fills the angle of the thyroid cartilage at this level.
Anteriorly, the ligament is an independent band. Posteriorly, the fibers
separate to enclose spaces, which lodge glands and fat. In the vicinity
of the arytenoid cartilage, between the spina superior and inferior, a band
of elastic tissue passes downward posterior to the ventricle. This is the
ligamentum arcuatum of Tortual.

Henle’s description of this ligament seems to be the most accurate.
The elastic fibers arising from the angle of the thyroid cartilage
unite to form a distinct ligament in their anterior third. (See Fig. 10.)
The lhgament is not dense, and the single fibers composing it can be
readily recognized. It arises from a connective tissue mass occupying
the angle of the thyroid cartilage, which is much smaller than
the one at the level of the ligamenta vocalia. In the posterior two-thirds
of their course the fibers composing the ligament separate from each
other and enclose spaces, in which are lodged numerous glands and fat.
The fibers of the hgament anastomose frequently with each other. Nu-
merous collagenic fibers are scattered among the fibers composing the
ligament.

The elastic cartilage forming the epiglottis is broken up by numerous
glands. (See Fig. 11.) I have attempted to find in this elastic cartilage
@ definite functional arrangement of the fibers, but this has been impos-
sible because of their number and frequent anastomoses.

The discussion as to the relation existing between the epithelium and
elastic tissue of the larynx has been definitely settled by the use of specific
stains. The elastic fibers are differently directed in different divisions of
the larynx, and bear a different relation to the epithelium in various
regions. At the level of the ligamenta vocalia the epithelial cells rest
directly upon an elastic fiber layer, which is only arbitrarily separated
from the elastic fibers of the ligament. The subepithelial fibers are

*® Verson: Stricker’s Handb. der Lehre von den Geweben des Menschen
u. der Thiere. Leipzig, 1871, p. 459.
31 Henle: Handb. der EHingeweidelehre; p. 254.
14

190 The Elastic Tissue of the Human Larynx

parallel to the fibers of the ligament, and to the mucous folds found at
this level. The elastic fibers enter the bases of the latter, but do not pass
vertically into them. An exception to the latter statement is found upon
the medial surface of the vocal process of the arytenoid cartilage, where
the fibers of the subepithelial layer pass vertically or obliquely into the
base of a small mucous fold. This change in direction of the elastic
fibers corresponds approximately to the level at which the linea arcuata
inferior of Reinke * crosses the vocal process. This line limits the extent
of an cedema of the labium vocale posteriorly; the vertical arrangement
of the elastic fibers probably acts as a barrier at this point. This
arrangement may have some further functional import. In this region
the epithelium is subject to the greatest stress, owing to the frequent
and wide range of movement of the processus vocalis. The arrangement
of the elastic fibers here undoubtedly anchors more securely epithelium to
vocal process. Reinke states that occasionally a similar relation between
epithelium and subepithelial elastic layer is found at the level of the
nodulus vocalis. I have not found this relation in my specimens. In the
subglottic region the subepithelial elastic fibers are separated from the
epithelium by a thin connective tissue layer. The elastic fibers are here
vertically directed. The same relation is found in the inter-arytenoid
space, and in the superior laryngeal zone.

I am indebted to Mr. Leonard H. Wilder, artist to the laboratory, for
the accompanying drawings.

EXPLANATION OF PLATES.
PLATE 1:

Fic. 1.—Frontal section through the anterior part of the conus elasticus.
1. Thyroid cartilage. 2. Elastic fibers of the conus passing mediad on each
side to form the ligamentum cricothyreoideum medium, which is pierced by
the crico-thyroid vessels. Most of the fibers pass cephalad and laterad from
the cricoid cartilage to attach to the thyroid cartilage. Some arise from a
raphé formed by the union of the horizontal fibers of the ligament. 3. Cri-
coid cartilage.

Fic. 2.—Frontal section through larynx, just posterior to the anterior com-
missure. 1. Thyroid cartilage. 2. Cricoid cartilage. 3. Fibers of the conus
elasticus. 4. Nodulus vocalis, showing the convergence of the elastic fibers of
the ligamentum vocale upon its medial surface. 5. Musculus vocalis. 6. Ven-
triculus laryngis. 7. Ligamentum thyreo-epiglotticum. 8. Subepithelial elas-
tic layer separated from the elastic fibers of the conus by glands and colla-
genic tissue.

Fic. 3.—Transverse section through the ligamentum cricothyreoideum
medium, showing its relation to the fibers of the conus and the subepithelial
elastic layer.

® Reinke: Fortschritte der Medicin. 1895, p. 476.

Dean D. Lewis : 191

PLATE II.

Fic. 4.—Transverse section through the ligamentum cricothyreoideum
medium, showing the general direction of the elastic fibers composing it.
Bundles of vertical fibers are surrounded by horizontal fibers, which pass
toward the median line to meet corresponding fibers of the opposite side. In
this section the crico-thyroid artery is seen piercing the ligament. The direc-
tion of the fibers of the subepithelial elastic layer and their relation to the
ligament is shown.

Fig. 5.—Transverse section through the larynx at the lower part of the
thyroid cartilage. 1. Thyroid cartilage. 2. Conus elasticus. 38. Cricoid car-
tilage. 4. Posterior crico-arytenoid muscle.

PLATE III.

Fig. 6.—Anterior part of preceding figure as seen under low power, showing
the mode of attachment of the elastic fibers of the conus to the thyroid carti-
lage, and the general direction of the fibers. 1. Thyroid cartilage. 2. Peri-
chondrial process by which the elastic fibers of the conus are attached to the
cartilage. 3. Anterior part of the conus elasticus. 4. Subepithelial elastic
layer.

Fic. 7.—Transverse section through the larynx at the level of the ligamenta
vocalia. 1. Thyroid cartilage. 2. Perichondrial process by which the elastic
fibers attach. This is the median process described by Gerhardt. 38. Nodulus
vocalis (cartilaginous nodule described by Mayer). 4. Ligamentum vocale.
5. Processus vocalis of the arytenoid cartilage. 6. Arytenoid cartilage. 7.
Ventricle of larynx.

PLATE IV.

Fic. 8.—Transverse section through the larynx at the level of the attach-
ment of the anterior extremities of the ligamenta vocalia—low power. 1.
Thyroid cartilage. 2. Thickened perichondrium described by Gerhardt as the
processus vocalis of the thyroid cartilage. 3. Nodulus vocalis (cartilago sesa-
moidea anterior); anastomosing bundles of elastic fibers pass off from it
anteriorly to be attached to the horizontal fibers of the perichondrium. 4.
Parallel elastic fibers of the ligamentum vocale passing into the posterior
extremity of the nodulus vocalis.

Fic. 9.—Transverse section through the anterior attachment of the ligamen-
tum vocale. The larynx is divided in the median line. Low power: Van
Gieson’s picro-fuchsin. 1. Thyroid cartilage. 2. Perichondrial fibers. 3.
Nodulus vocalis, showing numerous round and spindle cells. .4. Musculus
vocalis.

PLATE V.

Fig. 10. Transverse section through the larynx at the level of the ligamen-
tum ventriculare. 1. Thyroid cartilage. 2. Cricoid cartilage. 3. Ligamen-
tum ventriculare. 4. Fibers of the ligamentum thyreoepiglotticum. 5. Ven-
triculus laryngis. 6. Group of glands.

Fic. 11.—Transverse section through the epiglottis.

THE ELASTIC TISSUE OF THE HUMAN LARYNX
DEAN D. LEWIS

PLATE |

AMERICAN JOURNAL OF ANATOMY--VOL. IV

THE ELASTIC TISSUE OF THE HUMAN LARYNX
. DEAN D. LEWIS

PLATE II

AMERICAN JOURNAL OF ANATOMY--VOL. Iv

THE ELASTIC TISSUE OF THE HUMAN LARYNX
DEAN D. LEWIS

PLATE III

FIG. 6.

AMERICAN JOURNAL OF ANATOMY--VOL.

IV.

THE ELASTIC TISSUE OF THE HUMAN LARYNX
DEAN D. LEWIS

= >

—<=,
ame « A

vi
Mis
RVi-g

He

t

AN
AS

A
a

Kae
Kes

ete oR

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE IV

PEATE

THE ELASTIC TISSUE OF THE HUMAN LARYNX

DEAN D. LEWIS

FIG. 11

AMERICAN JOURNAL OF ANATOMY--VOL, IV

a

STUDIES OF THE INTERSTITIAL CELLS OF LEYDIG.

No. 2.—THEIR POSTEMBRYONIC DEVELOPMENT IN THE PIG.

BY

R. H. WHITEHEAD, M. D.
Medical Department, University of North Carolina.

WITH 5 TEXT FIGURES.

In a recent article * I presented the results of a study of the embryonic
development of the interstitial cells of Leydig in the pig. In the present
article I wish to give a brief account of the findings in a study of their
postembryonic development in the same animal. The methods employed
were the same as those described in the first article, to which I may refer
also for the more important facts in the literature of the subject.

The youngest pig in my series was one month old. In sections of the
testis at this age, compared with one from the embryo pig near term, the
cross-sections of seminal tubules are somewhat more numerous and closer
together, and the masses of interstitial cells are proportionally smaller.
Beneath the albuginea the cells are much reduced in size, and are ar-
ranged in a few more or less parallel rows, separated by small bundles
of connective-tissue fibres. In the deeper portions of the gland, however,
the cells are still of about the same size. Three main types of cells can
be observed, as follows (Fig. 1):

1. Cells with cytoplasm condensed around an eccentric nucleus, while
the periphery is extensively vacuolated. The vacuoles have more or less
uneven, ragged margins. Some of them hold inclusions which vary in
size; the largest of these have the granular appearance and staining
reactions of the cytoplasm, while others are more hyaline in appearance.
Occasionally structures entirely similar to these inclusions are found
between the cells.

2. Cells whose cytoplasm is condensed around an eccentric nucleus,
while their periphery is much clearer, containing only a few scattered
cytoplasmic threads.

1Amer. Jour. Anat., Vol. 3, No. 2, 1904.

AMERICAN JOURNAL OF ANATOMY.—-VOL. IV.

194 Studies of the Interstitial Cells of Leydig

3. This variety is similar to the preceding, but is characterized by the
presence of large acidophile granules, which have about the same size
as those noted in the germinal epithelium of the early embryo. They are
more granular in appearance, however, than the latter, and in prepara-
tions stained by Mallory’s method they take the acid fuchsin, whereas
the granules of the epithelium are stained by the aniline blue. With
Mann’s mixture of methyl blue and eosin they are stained by eosin. Al-
though acidophile, they do not stain with as much intensity as the
granules of eosinophile leucocytes. The granules are situated, for the
most part, in the peripheral portion of the cytoplasm; occasionally a
cell is found which seems loaded with them throughout, but, as a
general rule, they are largest and most numerous near the periphery. No
such granules were seen within the seminal tubules; in the spaces be-
tween the Leydig’s cells, however, small collections of them were rather
frequently encountered, which in most cases were undoubtedly small por-

Fia. 1. Fig. 2.

Fie. 1. Types of Leydig’s cells in pig one month old. x S800.
Fie. 2. Pig three months old. A small group of Leydig’s cells. x 800.

tions sectioned from the periphery of granule-bearing cells, but in some
instances seemed to be free. All the cells of these three varieties have
rather coarse, well defined cell-boundaries, especially marked in the case
of the vacuolated cells. These boundaries frequently stain differently
from the cytoplasm proper; in preparations stained by Mann’s solution
of methyl blue and eosin the cell-boundaries quite commonly are blue,
while the cytoplasm takes the eosin. Many of the cells have two nuclei,
but no mitotic figures were observed. A review of my preparations
of the testis from the embryo just before term shows that all these varie-
ties of cells are present there; indeed the principal difference between
the two glands, so far as Leydig’s cells are concerned, is the atrophy of the
subalbugineal layer in the pig one month old. The granules, however,

R. H. Whitehead - 195

in the granule-bearing cells of the embryonic gland are smaller and not
so limited to the periphery of the cells.

In preparations stained with Sudan III or osmic acid numerous
globules of fat, oftentimes very large, are found constantly in the
seminal tubules; but the interstitial cells contain at the most only a
few fine droplets—many of them contain no fat whatever.

The collections of cells have a rich blood-supply through a network
of thin-walled capillaries. A rather striking feature in the testis at
this age is the large number of eosinophile leucocytes, both in capillaries
and free among the interstitial cells.

In the pig two months old Leydig’s cells, in general, are smaller than
in the preceding specimen. The varieties of cells described above are
still present, but with some differences.

Fia. 3. Fia. 4.

Fig. 3. A small area of testis in pig one month old. x 50.

Fie. 4. Small area of testis in pig five months old. x 50.

The granule-bearing cells are scanty, while the cells with pale periph-
ery are quite abundant, as are also the vacuolated cells, the two to-
gether forming the great majority of the interstitial cells. In the case
of the vacuolated cells the number of inclusions is noticeable. In many
of these cells the septa between the vacuoles are breaking down.

At three months the convoluting of the seminal tubules has increased
considerably, so that many more cross-sections of them are seen, and
the interstitial cells are divided up into smaller collections. The break-
ing down of the septa between vacuoles and the concentration of cyto-
plasm around the nucleus have also progressed, so that now the indi-
vidual cells are much smaller than in the preceding stages, and most
of them present the pale periphery (Fig. 2). A very few containing
acidophile granules may be seen.

196 Studies of the Interstitial Cells of Leydig

These two processes, growth of the tubules and atrophy of the inter-
stitial cells, continue at such a rapid rate that in the five-months pig the
tubules greatly predominate over the interstitial cells (compare Figs.
3and4). The latter are now so reduced in size as to almost be identical
in appearance with the subalbugineal cells of the pig at one month. In-
dividual cells are shown in Fig. 5. Many of them are like the central
one in the figure, others entirely lack the distinct cell-boundaries and
are little more than naked nuclei, and others show distinct cell-boundaries
only at intervals, especially at the margin of a vacuole.

Of the three adult testes at my disposal two were evidently pathological,
as the tubules in one case contained no sexual cells, and in the other
only a few spermatogonia; probably they were ectopic testes, and need
not be considered here. The third one, however, was normal, and sper-
matogenesis was quite active. Sections show that the growth and con-
voluting of the seminal tubules have progressed still further, with the
‘result that there are very few Leydig’s cells between the albuginea and

Fie. 5. Types of Leydig’s cells in pig five months old. x 800.

the bases of the tubules, but they have been crowded against the lines of .
attachment of the septa to the albuginea. In the deeper portions of the
sections the general appearance of the cells is quite similar to that found
in the five-months pig, with the exception that they are somewhat larger.
They do not contain any acidophile granules, nor could Reinke’s crys-
talloids be demonstrated in them. The subdivision of the groups of
Leydig’s cells has increased, and in many situations there are none be-
tween adjacent tubules.

While, of course, this study does not warrant conclusions as to the
function of Leydig’s cells in the adult, we may at least inquire if it furn-
ishes any data in support of any of the hypotheses which have been
advanced as the result of histological investigation of adult conditions. In
favor of the theory of v. Bardeleben, that Leydig’s cells replace Sertoli
cells as the latter are worn out in the performance of their function, I can
find no evidence in any of the preparations. After the basement membrane
of the tubules is laid down it forms a barrier which completely prevents
the passage of interstitial cells into the tubules. So also as to the view

R. H. Whitehead 197

of Plato, that the function of Leydig’s cells is to store up fat and pass
it on through the walls of the tubules to be used as pabulum in sper-
matogenesis, the evidence is negative. The Leydig’s cells of the pig’s
testis contain little or no fat, while the tubules show large quantities of
that substance; nor could I detect the minute canals described by him in
the walls of the tubules. Moreover, if recent investigations upon fat
metabolism are to be accepted, fat entering the tubules from the outside
would probably pass through their walls, not as such, but rather as its
two liquid components. Some support, however, might be derived for
an extension of Plato’s theory as suggested by v. Lenhossek, according to
which the function of the interstitial cells is to store up, not merely fat,
but other material as well, to be used as pabulum by the tubules. The
most important facts in the development of Leydig’s cells, it seems to me,
are the alternating periods of hypertrophy and atrophy, and the struc-
tural characters of these cells during the stage of hypertrophy. The
periods of hypertrophy precede, while those of atrophy are synchronous
with, periods of rapid growth by the seminal tubules. Moreover the
changes in the interstitial cells, though occupying much more time, are
comparable, to some extent, with those which occur in secreting cells.
So that the appearances described might be interpreted as possibly in-
dicating that the Leydig’s cells elaborate a specific pabulum for the
tubules during the development of the testis.

I wish here to thank Professor F. P. Mall for the courtesy of a seat
in his laboratory while this article was in preparation.

PROPHASES AND METAPHASE OF THE FIRST MATURA-
TION SPINDLE OF ALLOLOBOPHORA FCTIDA.

BY
KATHARINE FOOT AND E. C. STROBELL, Woods Holl, Mass.

WITH 9 PLATES.

In the most mature eggs found in the ovaries of Allolobophora
fetida, the germinal vesicle is intact, the large nucleolus is present and
the chromosomes are not yet formed.

In the eggs of the freshly deposited cocoon, the first maturation spin-
dle is at the metaphase." From the time, therefore, that the eggs leave
the ovaries, until they reach the cocoon, the prophases of the first matura-
tion spindle take place, i. e., the forming of the chromosomes, the
breaking down of the germinal vesicle, and the disappearance of the
large nucleolus. In the summer of 1901, we found that these stages of
development occur, while the eggs are in the receptacula ovorum *
and we have thus far observed no exception to this rule.

In an earlier paper (Foot and Strobell, ’02), we demonstrated that these
worms deposit their cocoons about every third day, and it is therefore
probable that the eggs accumulate in the receptacula during this
period, indicating that the development of the eggs progresses very slowly
in the receptaculum ovorum. The earliest stages shown by these eggs,
have the germinal vesicle intact, with the nucleoplasm still undiffer-
entiated into distinct chromatic and achromatic substances, a large
nucleolus with one or more vacuoles, and no indication of any centriole
or rays in the cytoplasm, while the most mature eggs have the first
maturation spindle at the metaphase.

1 We have found only one egg in which the chromosomes are not oriented
in the equator of the spindle. The normal stage of development appears to
be the metaphase of the first spindle, though there are many degenerated
and disintegrated eggs, some showing a vestige of a germinal vesicle but
with no differentiation of the nucleoplasm, the chromatin being disintegrated
and structureless, and in only one or two cases have we found even an indica-
tion of a persisting nucleolus.

2Marshall and Hurst state in their text book, ‘“ Practical Zoology” (1888),
that immature eggs may be found in the receptacula ovorum of Lumbricus
at certain seasons of the year.

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

200 First Maturation Spindle of Allolobophora Feetida

Metron.

In the summer of 1901, we fixed and sectioned a large number of
receptacula ovorum, but we found this a very unprofitable method;
accurate study of the normal egg being very much hampered by the
number of abnormal eggs found in the receptacula, in some cases the
entire receptaculum being filled with eggs in various stages of degenera-
tion. We found in one receptaculum ovorum as many as forty eggs and
of these only four were normal. A more: serious difficulty was the un-
favorable action of the fixative on the receptaculum as a whole. The
swelling or shrinking, or the combination of both, produced by some fixa-
tives, acts with intensified effect on the mass of eggs crowded into the
receptacula, often distorting the normal eggs in a way to render them
valueless for cytological study.

In the summer of 1902, we tried removing the eggs from the receptac-
ula after they had been cut from the worm and placed in a watch glass
in distilled water. By carefully teasing the walls of a receptaculum,
under a dissecting microscope, the eggs can be pressed out, and only those
that appear normal selected for study, the subsequent technique being
the same used for eggs collected from the cocoons (Foot, ’98). The fol-
lowing fixatives were used, chromo-acetic, corrosive sublimate, corrosive
acetic, Rabl’s picro-sublimate, Boveri's picro-acetic, picro-sublimate, Flem-
ming’s chromo-aceto-osmic, and the same proportion of chromic and osmic
omitting acetic acid. Hermann’s platino-aceto-osmic and the same pro-
portion of osmic and platinum chloride, omitting the acetic acid. A
comparison of the photographs will show that the platino-osmic has
proved the least injurious to both nuclear and cytoplasmic structures.
The sections were stained with iron-hematoxylin followed by dilute
Bismarck brown, or with Bismarck brown alone. In some cases, un-
stained sections give the most satisfactory photographs. This is especially
true for archoplasm, where in stained preparations the dense stain
taken by the archoplasm produces such a strong contrast to the faintly
stained cytoplasm, that it is impossible to get an accurate reproduction
of one without sacrificing the other. All the preparations were stained
with the end in view of securing satisfactory photographs, as we aim to
present only such cytological phenomena as can be clearly demonstrated
by photography.

The impossibility of securing a clear demonstration of the prophases
in fixed and sectioned eggs led us to devise a new method which is a
modification of the smear method. Instead of smearing a mass of e
on the slide, we handle each egg separately. After isolating a living egg

ggs

Katharine Foot and E. C. Strobell 201

in a very small drop of water on the slide, the membrane is carefully
pricked with a very fine needle and as the cytoplasm flows out, the ege
membrane is gently dragged away with the needle, allowing the contents
of the egg to spread and dry immediately. By this method the germinal
vesicle, and sometimes even the spindle, flow out of the egg membrane
intact, and dry so quickly that the structures are remarkably well pre-
served. The vesicle when drying flattens out over a larger area, leaving
the individual chromosomes or threads sufficiently separated to be
clearly identified, and when the entire spindle passes out of the egg
membrane intact, all the eleven chromosomes are beautifully demonstrated
on one plane. In this manner we arrange from 20 to 30 eggs on a slide,
which has been previously ruled with a diamond into definite areas, so
that the position of each egg is known and any egg can be studied later
in connection with data taken before it was killed.’

Cytoplasm.—During the development of the egg when the prophases
of the first maturation spindle occur there is a marked change in the
structure of the cytoplasm and during this period there is a decrease in
the size of the egg and an increase in the amount of the substance between
the egg membrane and the outer membrane, greatly increasing the width
of the latter area. Compare the space between the two membranes
shown in Photos. 13, Plate I, and 68, Plate IV, eggs at the germinal
vesicle stage, and that of Photos. 99 and 100, Plate V1, where the first
maturation spindle is at the metaphase.“ We have demonstrated these
two membranes in earlier papers (1897, Fig. 2) (1901, Photos. 57 and
59), and shown that Allolobophora possesses in common with many
Oligochetes, a delicate outer membrane and an equally delicate membrane
in contact with the egg itself, the space between the two being filled with a
relatively non-stainable substance. Vejdovsky and Mrazek have observed
these two membranes with coagulated substance between them in many
Lumbricide and they criticise Gathy’s interpretation of the three layers
in Tubifex as one thick membrane.

The difference of size and density between the eggs at the germinal
vesicle stage and those containing the first maturation spindle is very
evident in living eggs, the latter being smaller, more dense and more
opaque and these features are equally evident in the dried preparations.
At the germinal vesicle stage the typical cytoplasm shows a distinct

3 Such slides do not resemble smear preparations, they suggest rather slides
with a few thick sections far removed from one another.

*This contrast is really more marked than shown in the photographs, as
Photos. 13 and 68 are magnified nearly 300 diameters more than Photo. 99.

202 First. Maturation Spindle of Allolobophora Feetida

honey-comb structure and at the spindle stage there are fewer alveoli and
those present are much smaller. The increase in the amount of substance
between the inner and outer membranes of the egg and decrease in the
number and size of the alveoli, suggest that the former may be increased
at the expense of the latter, but we have no proof of this.

Osmophile Granules——Typical osmophile granules in the cytoplasm of
normal eggs found in the receptaculum ovorum are demonstrated in the
unstained sections of Photos. 68 and 69, Plate IV. A comparison of
these sections with the ovarian eggs of Plate 42 of an earlier paper
(1901) shows a diminution in the amount of osmophile substance in
these older odcytes. In the above mentioned paper we noted a decrease
in the amount of osmophile substance between the ovarian and the
cocoon eggs and suggested that the storing.of the osmophile substance in
the ovarian egg must be for the use of the egg from the time it leaves
the ovary until it is supplied with the nutritive albumen of the cocoon,
i. e., during the prophases of the first maturation spindle. This supposi-
tion seems confirmed now that we know these processes go on very
slowly in the receptacula ovorum and that the osmophile substance
gradually diminishes in amount during this period.

We first demonstrated these osmophile (deutoplasmic) granules in
1898, and in subsequent papers have presented additional data. We have
shown that they are not dissolved out of the cell by either turpentine or
xylol, but that after staining they are as a rule invisible, having lost all
trace of the blackening caused by the osmic acid. This also fades in
unstained eggs that have been kept for a long time in paraffine or
mounted in balsam. If a section is photographed before staining, Plate
IV, Photos. 68 and 69, and then stained in iron hematoxylin the granules
are indistinguishable even with a 2 mm. lens, but if the section is com-
pared with a photograph of the unstained preparation that shows exactly
where to look for each granule, they can be identified as clear, colorless
bodies that would entirely escape observation without the photograph
as a guide.

Centriole and Spindle-—In this paper we shall adopt Boveri’s term
centriole for the small central granule of the asters for which in an
earlier paper we retained the old term centrosome. We drop the term
centrosome for this granule not because we find a second body in the
asters of Allolobophora which answers to the centrosome of Boveri and
others, but to avoid the confusion of retaining an old term which implies
a structure and accompanying complicated changes which we have not
found in this egg.

The centrioles destined for the two poles of the first maturation

Katharine Foot and E. C. Strobell 203

spindle are first seen at opposite poles of the germinal vesicle, indicating
that they arise independently and not by division of a primary centriole.
This stage is shown in Photos. 81, 82, 84 and 86, Plate V. The eggs
of Photos. 81 and 82 are greatly marred by poor fixation, chromo-acetic
being as harmful to cytoplasm as it is to nuclear structure (see p. 215).
The centriole of each aster is so small that only the presence of the rays
justifies our interpreting the central microsome as the centriole. In Photo.
81, Plate V, the centriole is in contact with the membrane of the germinal
vesicle, and in Photo. 82, Plate V, only slightly removed from it. The two
asters are at opposite sides of the germinal vesicle, being separated by
fifteen sections, the entire germinal vesicle being cut into ninteen sections.
In all these sections the membrane of the germinal vesicle is intact and at
the points where the rays of the asters focus the membrane slightly
protrudes.’

It might be claimed that the theory of the nuclear origin of the cen-
triole is supported by the presence of these two centrioles almost within
the germinal vesicle, but they certainly do not support the particular
phase of the theory that holds the nucleolus responsible for the centriole,
for in this egg the nucleolus is still intact and is found two sections
removed from one aster and thirteen sections from the other. The
injurious action of chromo-acetic on the nucleoplasm of the germinal
vesicle does not always affect its membrane, although it breaks the con-
nection between the two, leaving the membrane in contact with the sur-
rounding cytoplasm, this being favorable for the identification of the
earliest appearance of the centriole. On the contrary other photographs
show that many fixatives favorable for the study of nuclear constituents
are of no value for studying the early appearance of the centrioles, the
cytoplasm being torn away from the germinal vesicle, destroying all trace
of the centrioles and asters (Photos. 22 to 25, 29 and 30, Plate II, 46 to
49, 51 and 52, 59 to 63, and 65, Plate III). In a few cases we have
found eggs killed in these same fixatives showing an equal shrinkage of
cytoplasm and nucleus leaving the membrane in continuity with cytoplasm
and nuclear reticulum, but the exceptions are very rare and have not been
at the stage to throw any light on the origin of the centrioles. Platino-
osmic, as stated above (p. 200), appears to be the least injurious for all
constituents, and we hope its further use will enable us to collect more

° The blurred effect of the part of the membrane seen in the photographs is
due to its being on a different plane from the centrioles. The two sections
shown in Photos. 81 and 82 are cut so close to the periphery of the germinal
vesicle that part of its membrane is seen en face. In the sections beyond
those photographed, the membrane is the only part of the germinal vesicle left.

204 First Maturation Spindle of Allolobophora Feetida

satisfactory data as to the origin of the centrioles of the first maturation
spindle.

Meves’, ’02, observations as to the constancy in size of the centrioles,
regardless of the size of the cells are supported by this egg. There is
an insignificant difference in size between the centrioles at the metaphase
of the first maturation spindle (Photos. 26, Plate II, 91, 92a and b,
Plate V, 99, 100, 101, 102 and 106, Plate VI), the second matura-
tion spindle (Photos. 103, 104, 105, 108 and 109), the first, second
(Photo. 110), and third cleavage spindles (Photo. 107, Plate VI).”
There are often exceptions to this rule, but in many of these cases the
cause is obviously overstaining (Photo. 83, Plate V), for centrioles
that we have seen or photographed in unstained preparations, as a rule,
show no more variation in size than can be accounted for by different
fixation, or individual variations. Photos. 92a and b, Plate V, show
also a similarity in size of the centrioles of the peripheral and inner
poles of the first maturation spindle. There is, however, a dissimilarity
in size between the centrioles of the prophases and metaphases of the
first maturation spindle, the former being smaller, indicating that the
centriole passes through stages of growth (compare the centrioles of
Photos. 81, 82, 84 and 86, Plate V, with those of the metaphase, Plate
VI). That centrioles are sometimes found out of the center of the
sphere is probably due to fixation, for if fixation can produce such marked
variations of cytoplasmic and nuclear structures as demonstrated in these
photographs, it is indeed remarkable that the central position of the
centriole is maintained as constantly as we find it in these eggs.

The first maturation spindle can be readily identified in the living egg.
In Photos. 125, 128, 129 and 130, Plate IX, we see spindles that have
retained their form after the contents of a living egg has been pressed
out of its membrane and allowed to dry quickly on a slide. In such
preparations we have found no trace of a centriole, but we cannot give
much weight to this evidence for in all fixed material, both stained and
unstained, a centriole is invariably present at each pole. Unstained cen-
trioles are demonstrated in the spindles of Photos. 91 and 99, Plates V
and VI, and stained centrioles in the spindles of Photos. 26, Plate II,
92a and b, Plate V, and Photos. 100 to 110, Plate VI.

Photos. 84 to 89, Plate V, suggest that a large part of the spindle is
formed of achromatic nucleoplasm. In Photos. 86 to 89 a part of the
membrane of the germinal vesicle is still seen and the nucleoplasm within

*The magnification of Photos. 99, 102, 108 and 110 is 710 diameters, and
Photos. 26 and 109, 1100 diameters. All the others are 1000 diameters.

Katharine Foot and E. C. Strobell 205

this area is unmistakably assuming the form characteristic of spindle
fibers. Photos. 81, 82, 83, 84 and 86, Plate V, indicate with equal
clearness that the cytoplasm contributes to the polar rays. These
photographs (84 and 86) suggest a very close if not causal relation be-
tween centriole and spindle and may be called in evidence to support the
theory that the spindle is formed under the influence of the centrioles.
In these preparations the chromosomes certainly do not influence the
form of the spindle, for although they are all massed on one side,
yet the spindle remains symmetrical in relation to the centrioles.
Even if this massing of the chromosomes in one-half of the spindle is
not the normal condition it should produce an abnormal and distorted
spindle, if the rays were formed only in relation to the chromosomes.
We are forced to conclude, therefore, that the spindle is a life expression
of the nucleoplasm and polar cytoplasm, or is formed under the in-
fluence of the centrioles.

Nucleolus.—In a comprehensive history of the nucleolus, Montgomery,
’98, shows what a bewildering number of conflicting interpretations have
accumulated around this structure since it was first figured by Fontana,
in 1781, and what little progress has been made in solving the problems
of its significance, even its morphology being enveloped in a mass of con-
tradictory evidence. We hope to be able to establish the morphology of
the nucleolus of the egg of Allolobophora fatida by a careful compara-
tive study of its form after killing in a variety of fixatives, by a study of
the living egg, and of the dried germinal vesicles, as obtained by the
method described on p. 200. We are convinced that its variety in form
can be best appreciated when demonstrated by a number of photo-
graphs and with sufficient data for each fixative it may be possible to
arrive at a correct decision between the artificial and the normal struc-
tures.

The germinal vesicle of Allolobophora in common with many other
forms contains two distinct kinds of nucleolar formations, and for these
we shall adopt Flemming’s term, principal and accessory nucleoh. The
former is the relatively large nucleolus which has persisted and increased
in size since its first appearance in the smallest odcyte of the ovary
and is peculiarly the nucleolus of the odcyte first order. The accessory
nucleoli are first seen in the large ovarian eggs, and the earliest stage at
which we find them they are very small and in close proximity to the
chromatin (Photos. 9, 11, Plate I, 23, Plate II, and 49, Plate III), and
after many fixatives stain like the chromatin. The two structures, the
principal nucleolus and the accessory nucleolus, differ in several respects,
as a rule only one of the former is present in a germinal vesicle (Photos.

15

206 First Maturation Spindle of Allolobophora Feetida

4,7, 12, Plate I, 22, 27, 29, 31, Plate II, 49, 50, 58, 60, Plate III, 71,
Plate IV), whereas the accessory nucleoli vary in number from one to
six or more.’ The principal nucleolus is composed of two substances dif-
fering in density and producing, though not invariably, the so-called
vacuolated appearance, whereas the accessory nucleoli are dense homoge-
neous bodies and are, as a rule, not vacuolated. The principal nucleolus
and the accessory nucleoli are clearly seen in unstained sections, even the
smallest of the accessory nucleoli appearing homogeneous and refractive.

In Photos. 23, 29, 39, 40, Plate IJ, 56, Plate III, both kinds of
nucleoli are shown in the same section of a germinal vesicle. Photos.
3 and 4 show two sections of the same germinal vesicle, the former
containing the accessory nucleolus and the latter the principal nucleolus.
Photos. 6, 7, 8 and 9 are sections of one germinal vesicle, Photos. 7 and
8 showing the principal nucleolus and Photos. 6, 8, 9 showing each one
accessory nucleolus. After certain fixatives, e. g., platino-osmic, the
principal nucleolus stains very faintly with iron hematoxylin and can be
readily decolored, whereas the accessory nucleoli retain the hematoxylin
with as much tenacity as do the chromosomes.* The principal nucleolus
disappears before the first maturation spindle is formed, whereas the
accessory nucleoli often persist as late as the metaphase of the first
maturation spindle. They do not always persist, however, until this
period, and when they are present, their position is most inconstant, some-

times being within the area of the spindle, even at the equator, but more

often in the cytoplasm at different distances from the spindle. The
presence of the persisting nucleoli was noted and figured in an earlier
paper (Foot, ’94), but at that time we had not recognized them as
accessory nucleoli but supposed them to be fragments of the large egg
nucleolus.

In unstained sections the so-called vacuoles of the principal nucleolus
are entirely transparent and sharply differentiated from the rest of the
nucleolus which is dense and refractive (Photos. 14b, 15 to 21, Plate I,
35, 39, 40, Plate II, and 74, Plate IV). Stained preparations, however,

7This applies to sections. In all dried germinal vesicles that appear to be
normal, we find as a rule only one accessory nucleolus, though we have some-
times found two (Photo. 115, Plate VII), rarely three, and in one or two cases
five. Sections of fixed eggs indicate that the single accessory nucleolus
probably owes its origin to the fusing of several small ones.

’That the principal nucleolus fails to stain readily at these stages when
the accessory nucleolus stains very intensely may be due to the disintegra-
tion of the former, for at an earlier stage of development the principal nucle-
olus stains intensely.

Katharine Foot and E. C. Strobell 207

show that many of the vacuoles contain a substance which after some fixa-
tives can be so densely stained that the entire nucleolus appears homogen-
eous, these results supporting the observations of investigators who claim
that the so-called vacuoles of the nucleolus contain a fluid substance. In
dried germinal vesicles we have not been able to demonstrate any vacuoles
in the nucleolus, but in the living egg we have sometimes seen a single
vacuole in the principal nucleolus. In one case the nucleolus at first
showed no vacuole, one appearing about five minutes after the egg was
under observation and persisting until the egg died. In other eggs the
large nucleolus at first contained a vacuole which disappeared in about
five minutes, this condition persisting until the egg died. It would
seem, therefore, that vacuoles in the principal nucleolus of living eggs
appear and disappear as maintained by some authors.

Many of Montgomery’s figures show the vacuole stained in differential
colors, but his interpretation that they “are derived from the small
fluid globules which first appear in the nuclear sap” is not supported by
Allolobophora. Photo. 57, Plate III, might be forced to such an inter-
pretation, but in the hght of more than fifty other negative examples
it must be interpreted with them as merely another expression of the
effect of fixation and dehydration on the more fluid portion of the
nucleolus. Even unstained preparations show that the refractive parts
of the principal nucleolus and its vacuoles represent two substances of
very different degrees of density and this must produce an unstable
condition which is very readily disturbed by fixatives, and must be,
therefore, largely responsible for the great variety of forms seen in fixed
material.

The vacuoles in the principal nucleolus of Allolobophora sometimes
appear to be true vacuoles and again they appear to be a thin fluid sub-
stance which can be stained so as to completely obliterate the vacuoles
(see Photos. 22 and 27, Plate II), which will reappear after decoloring
as shown in Photos. 37, 42, Plate II, 49, 53, 54, 58, 60, Plate III, etc.,
and again the contents of many ring-like nucleoli closely resemble
the surrounding nucleoplasm suggesting that it has been artificially
forced into the nucleolus. This condition is shown in Photos. 36, 38,
43, 44, 45, Plate II, 50, Plate III, 79, Plate IV, 98, 94, 95, 96, and 98b,
Plate V, and many of these resemble the nucleoli figured by Coe, ’gg, in
Cerebratulus. In some cases, e. g., Photo. 50, Plate ITI, it is clearly seen
that a substance is massed in the nucleolus at the expense of the surround-
ing nucleoplasm, and Photo. 98a and b, Plate V, show a distinct break in
the nucleolar ring indicating how the fluid portions of the nucleolus and
the nucleoplasm may be brought into contact. Many nucleoli which are

208 First Maturation Spindle of Allolobophora Feetida

not vacuolated, contain small dark specks (see Photos. 5 and 14a) which
can be clearly seen and photographed in unstained preparations, for they
are quite as black as the osmophile granules (see Photos. 68 and 69,
Plate IV, for these granules). They differ from osmophile granules,
however, in retaining their original color after staining with iron
hematoxylin. Photo. 14a shows them in an unstained, and Photo. 5
in a stained preparation, those of the latter were clearly seen and photo-
graphed before the nucleolus was stained. In 1888 Vejdovsky figured
and described two nucleoli containing granules in the germinal vesicle
of Rhynchelmis (Taf IIL) and Montgomery in ’98, figured dark granules
in several nucleoli, e. g., Figs. 267 to 269, but they cannot be the same as
those shown in our Photos. 5 and 14a, for Montgomery finds by change
of focus these dark granules can be transposed into small, clear vacuoles.
Photo. 14a demonstrates that this is not the case for Allolobophora, in this
photograph some of the granules are out of focus, not all being on the
same plane, yet none of them appear as vacuoles. Photo. 14b is a nucleo-
lus from the same preparation as 14a and shows some of the vacuoles
quite as small as the black granules of 14a, but a change of focus on these
small vacuoles does not transpose them into granules. The granules
in 14a may represent an early stage of degeneration, a later stage being
shown in Photo. 75, Plate IV.

The sixty photographs of nucleoli, shown in our plates, represent forms
figured by investigators for widely different material. ‘These photo-
graphs show not only the varying forms of the so-called vacuoles, but
Photo. 4 shows a nucleolus sharply differentiated into the so-called chrom-
atic and achromatic portions, which can be differentially stained, and have
been described under a variety of names. Among later papers, such a dif-
ferentiation of the nucleolus has been demonstrated in Helix by Ancel,
’02, and in Teleosteans by Stephan, ’o02. Photos. 20, Plate I, 54, 55,
60, Plate III and 77, Plate IV, show the so-called nucleololus, or endo-
nucleolus, which Montgomery and others pronounce an artefact.

It is impossible to determine how many of the fantastic forms assumed
by the nucleolus of Allolobophora are artefacts, but the fact that definite
forms appear more.or less constantly after certain fixatives creates a well-
founded suspicion of every form that cannot be verified by comparison
with the living egg.

In Allolobophora there appears to be no fundamental difference be-
tween the principal nucleolus and the accessory nucleoli, and may not
the individualities of the former be due merely to its adaptation to
special needs of the egg during its growth period?

In many points the accessory nucleolus corresponds to the nucleoli of

Katharine Foot and E. C. Strobell 209

the male and female pronucleus. In the vesicles formed at the telophase
of the second maturation spindle’® a small dense homogeneous nucleolus
is first seen in close proximity to each chromosome (Foot and Strobell,
’o0). These increase in size by growth and by irregular and inconstant
fusing with one another. Thus in the resting female pronucleus we
find nucleoli, which like the accessory nucleoli are inconstant in size and
number, and this inconstancy is true also for the nucleoli of the male
pronucleus. One or several of these may persist until the metaphase of
the first cleavage spindle and like the accessory nucleoli may be in the
spindle or in the surrounding cytoplasm. These, like the accessory
nucleoli, are relatively dense homogeneous structures as compared with
the large nucleolus of the odcyte first order, and these points of agree-
ment suggest the possibility of a closer relationship—may not the acces-
sory nucleoli of the germinal vesicle arise in connection with the chromo-
somes of the first spindle before instead of after their division? If, as
held by a number of investigators, the chromosomes of one division are in
some manner related to the nucleolar substance of the following rest
stage, may not this be established at an earlier period and the accessory
nucleoli of the germinal vesicle be a precocious appearance of the nucleoli
which are so conspicuously absent between the first and second spindles ?—
the processes involved in the rest stage occurring before instead of after
the first division, the origin and growth of the accessory nucleolus being
part of them. The second division precociously foreshadowed in the four
part chromosomes of the germinal vesicle suggests a precedent for this
interpretation.

If our interpretation of the accessory nucleolus is correct, a like struc-
ture should be present in spermatocytes, and there should be two nucleoli
in the spermatocyte first order in all cases where a resting stage is omitted
between the first and second division. Such a condition is figured by
Vom Rath, ’92, in Gryllotalpha. His Figs. 10 and 11 show two nucleoli
in the spermatocyte first order at the spireme stage in which they are
conspicuous also in Allolobophora, and it is significant and interesting
that the two nucleoli in the spermatocyte are nearly equal in size. See
also Schreiner’s, ’04, Figs. 23 and 24.

*The transformation of the chromosomes into vesicles at the telophase of
the second maturation spindle was first seen in Rhynchelmis and described
and figured by Vejdovsky in 1887. Our Photo. 32 shows one of these vesicles
in a second maturation spindle and indicates that their formation is not
necessarily dependent upon a definite form or position of the chromosomes,
as this vesicle is formed before the telophase, and the chromosomes have not
assumed the shape they usually show, when at the lower pole of the spindle,
prior to the formation of the vesicles (Foot & Strobell, 00, Photo. 24).

210 First Maturation Spindle of Allolobophora Feetida

We hesitate to interpret the accessory chromosome of some authors
as the equivalent of the accessory nucleolus, but it is a suggestive fact
that many investigators have interpreted this structure as a nucleolus
and there is a significant disagreement as to whether it divides in the
first or second spindles, or in fact whether it divides at all. Our inter-
pretation that the accessory nucleolus of Allolobophora is the true
nucleolus of the odcyte second order supports Wilson’s, ’96, surmise that
the accessory nucleoli of egg cells “perhaps correspond to the true
nucleoli of tissue cells ” (p. 93), though he bases this conclusion on his
interpretation that the principal nucleolus does not correspond to the
“true nucleoli of tissue cells.” He mentions two kinds of nucleoli
in egg cells, the “ principal nucleolus,’ or net knot, and the “ accessory
nucleoli,” which are of the second (smaller) type, and although they do
not agree in their affinity for stains with the accessory nucleoli of
Allolobophora (which at this stage stain with greater intensity than the
principal nucleolus and retain the color with more tenacity even than
the chromosomes), they do agree in other more essential points, 1. e.,
their relation to the single large nucleolus as to size, number, advent
and persistence. Ina later edition of “ The Cell” (’00), Wilson’s conclu-
sions are greatly modified and he states that the principal and accessory
nucleoli “ differ widely in staining reactions, but it does not yet clearly
appear whether they definitely correspond to the plasmosomes and karyo-
somes of tissue cells.” He further says that the principal nucleolus
“cannot be directly compared to the net knots or karyosomes of tissue
cells,’ leaving the implication that they resemble the true nucleoli
(plasmosomes) of tissue cells, although he adds that in color reaction the
accessory nucleoli are comparable to these, pp. 127 and 128.

€

In Chetopterus, Mead, ’98, figures a ring-shaped nucleolus (Fig. 6)
closely resembling those of Allolobophora, and although he says nothing —
of accessory nucleoli he has demonstrated in several of his figures two or
three nucleoli which appear to answer to the accessory nucleoli of Allolo-
bophora. He says the nucleolus “ breaks up into a number of pieces
which remain for a time in the vicinity of the spindle, but gradually
degenerate and disappear,” p. 196. In Allolobophora it is the acces-
sory nucleoli which often persist until after the first maturation spindle
is formed, the principal nucleolus disappearing at an earlier period.

10“ rom its staining-reaction this type of nucleolus appears to correspond,
in a chemical sense, not with the ‘true nucleoli’ of tissue cells, but with
the net knots or karyosomes, such as the nucleoli of nerve cells and of many
gland cells and epithelial cells,” p. 92.

Katharine Foot and E. C. Strobell 211

Wheeler in Myzostoma, ’97, finds that the nucleolus persists in some
cases later than the second cleavage but he does not identify any accessory
nucleoli.

The accessory nucleolus of Allolobophora probably corresponds to the
second nucleolus, Gathy, ’oo, describes in Tubifea as arising indepen-
dently and disappearing later than the first. Gathy, however, does not
interpret the nucleolus he sometimes finds persisting until the metaphase
of the first spindle as the above mentioned second nucleolus. His de-
scription of the gradual disappearance of the nucleo without fragmenta-
tion is supported by our observations on the principal nucleolus of dried
germinal vesicles (Photos. 121, 122, 123 and 124, Plate VIII), though
after fixatives the nucleolus is sometimes seen breaking up into frag-
ments (Photos. 20, Plate I, 31, 39, 41, Plate IT, 54, 55, 57, 67, Plate III,
75, Plate IV and 97, Plate V). Dried preparations clearly demon-
strate that the principal nucleolus gradually loses its capacity to stain,
decreases in size and -finally disappears while the membrane of the
germinal vesicle is still intact (Photos. 114, Plate VII, 121 to 124, Plate
VIII). It does not pass out into the cytoplasm there to degenerate as
observed in odcytes in many other forms. This suggests that its func-
tional value is confined to the nucleus, and if our interpretation is cor-
rect that there is no fundamental difference between the principal
nucleolus and the accessory nucleolus we cannot accord any special
significance to the fact that the accessory nucleolus, unlike the principal
nucleolus, persists in the cytoplasm before its final disappearance. Its
cytoplasmic destiny may be due merely to the fact of its later origin and
consequently later disappearance, 1. e., after the germinal vesicle has
been replaced by the spindle. Dried germinal vesicles indicate that a
single accessory nucleolus is typical of normal odcytes. Sections of fixed
egos indicate that the single accessory nucleolus owes its origin to the
fusing of several smaller ones.

Many authors have recognized a more or less radical difference between
the large nucleolus of the germinal vesicles and the nucleoli of the cleav-
age stages. Allolobophora does not show the difference between the two
nucleoli that Korschelt, ’95, indicates for Ophryotrocha. In both Anne-
lids the large nucleolus contributes nothing to the formation of the
chromosomes, but in Ophryotrocha the cleavage nucleoli “ vielleicht ein
Theil des vorher im Kernkorper niedergelegten Chromatins dem Kern-
faden beigefiigt wird.” He adds “Was die erwihnten Verschieden-
heiten des verhaltens der nucleolen in dem Ei- und Embryonalzellen
betrifft so liessen sich diese vielleicht durch die recht verschiedenartige

212 First Maturation Spindle of Allolobophora Feetida

Ausbildung und Funktion der Kerne in den Beiderlei Zellen erkliren,”’
p. d79.

There has been a recent revival of interest in the theory that the
chromatin destined for the chromosomes of the first maturation spindle
is stored at an earlier stage in one or several nucleoli. In a recent
paper Blackman, ’03, sums up the weight of authority for this view,”
and to his list of authors may be added recent papers by Goldschmidt,
’°o2, Hartmann, ’o2, Bryce, ’or, and especially Le Brun, ’or, ’o2,
whose extensive publications on the maturation stages of Batrachians
are illustrated by many figures demonstrating this point. The evidence
furnished by the egg of Allolobophora cannot be interpreted as a sup-
port for the above theory. It must unquestionably be classed as support-
ing the interpretation of the many authors who claim that the chro-
matin destined for the chromosomes of the spindle is at no time aggregated
into a large nucleolus. In Allolobophora the chromosomes are formed
by a gradual segregation of the chromatin which is dispersed throughout
the germinal vesicle, and in order to maintain the theory that the
nucleolus is the storehouse of the chromatin there should be a definite
and constant relation between the formation of the chromosomes and the
breaking down and disappearance of the nucleolus. This, however, is
not the case in Allolobophora, the two processes do not invariably occur
in unison. If the chromosomes have their origin in the nucleolus we
should never find the chromosomes formed while the nucleolus is intact—
before it shows any evidence of breaking down. Photos. 27 and 28,
Plate II, 51 and 52, Plate III, and 68 to 73, Plate IV, demonstrate that
the chromosomes can be formed in the germinal vesicle without any
marked morphological disturbance of the nucleolus, these nucleoli not
differing essentially from those of the ovarian eggs and from the nucleoli
of those eggs in the receptaculum ovorum in which the chromosomes are
still unformed. ‘This is demonstrated also in the dried germinal vesi-
cles. In Photos. 111 and 113 to 115, Plate VII, the chromatic spireme is
formed and the principal nucleolus is still intact, showing no evidence
of having contributed to the chromatin of the spireme; and in those
cases in which the large nucleolus loses in staining capacity, while the
chromosomes increase in staining capacity, the phenomenon is probably
due to the normal disintegration of the former, and not to a contribu-
tion of its substance to the chromosomes. As a rule the principal
nucleolus has disappeared, or is in process of disappearing when the

u“ Blachmann, Carnoy, Davidhoff, Hermann, Holl, Sobotta, R. Hertwig,
Wilson and others,’ p. 197.

Katharine Foot and E. C. Strobell 213
chromosomes are formed, e. g., Photos. 121 to 124, Plate VIII, but
Photos. 116 to 118, Plates VII and VIII demonstrate that the disap-
pearance of the principal nucleolus may be retarded until after the com-
plete formation of the chromosomes. Such cases demonstrate that the
chromosomes in Allolobophora are not formed at the expense of the
nucleolus. JKorschelt, ’95, has reached a like conclusion for the Annelid
Ophryotrocha and in its morphology the large nucleolus of Ophryo-
trocha is strikingly like the large nucleolus of Allolobophora (compare
Korschelt’s figures of the germinal vesicle 72, 74, 75 and 79 with our
Photos. 29, Plate FI, 49, 58, Plate III). On this point observations on
a number of Annelids are in accord (compare Myzostoma, Wheeler’s
Figs. 3 and 4, ’97, Thalassema, Griffin’s Figs. 3, 6 and 8, ’99). In
Batrachians, Lubosh, ’02, has supported his criticism of the nucleolar
origin of the chromosomes by a reproduction of many interesting photo-
graphs of the nucleoli in Triton eggs. A few of them resemble those
of Allolobophora, cf. his Photo. 8 with our Photos. 5, 14a, and 75, Plate
IV, also his Photos. 5 and 18 with our 20, Plate I, 54 and 55, Plate ITI.

CHROMATIN, A COMPARATIVE Stupy oF SECTIONS AND OF ENTIRE
GERMINAL VESICLES DRIED ON THE. SLIDE.

The photographs of the germinal vesicles of Plates I to IV show the
average results obtained by sectioning these eggs after killing them in the
variety of fixatives given on p. 200.

The photographs of Plates VII, VIII and IX show the average results
obtained by drying individual germinal vesicles on the slide, according
to the method described on p. 200. A comparison of these two sets of
photographs demonstrates the advantage of the latter method, the former
proving inadequate almost to the point of being misleading, and it is
here evident that we have not yet found a fixative for this egg that shows
the development of the chromosomes as clearly as they are demonstrated
in dried preparations.

In sections, the chromatin of the most mature eggs of the ovary and of
the youngest eggs of the receptacula ovorum, is at about the same
stage of development, the relatively achromatic reticulum of the germinal
vesicle showing only indefinite chromatic aggregations which appear
to be the first steps towards forming the filaments out of which’ the
eleven bivalent chromosomes are finally formed. The first indications
of chromatic filaments are shown in Photos. 1 to 9, and 31, Plate IT, and
66, Plate ITT.

In dried germinal vesicles an early stage of the aggregation of chro-

214 First Maturation Spindle of Allolobophora Feetida

matin is shown in Photo. 111, Plate VII, where the germinal vesicle
is traversed by an extremely delicate chromatic thread or threads. Prior
to this stage we have found no structure in the germinal vesicles other
{han the nucleoli, the entire nucleoplasm being uniformly chromatic
and obscuring all differentiation. The only indication of the presence
of the diffused chromatin being expressed by the fact that the nucleo-
plasm reacts to nuclear stains with much more intensity than it does
after its chromatin constituent has formed the spireme or chromosomes.

This is true of eggs taken from the free end of the ovary, and the
earliest stages of those found in the receptacula. A comparison of
results obtained from dried germinal vesicles with sections creates the
suspicion that much of the structure seen in sections of these eggs in
the earlier stage may be due to fixation. Such artificial demonstration
of the chromatin is, however, instructive in showing how the accessory
nucleoli arise in close connection with the chromatin in widely separated
areas of the germinal vesicles, later fusing into the one accessory nucle-
olus typical of later stages.

In sections, a more marked aggregation of the chromatin into pro-
nounced chromatic filaments is shown in Photos. 22 to 25, Plate II, 46 to
49 and 59 to 62, Plate III, these filaments undoubtedly corresponding
to portions of the skeins shown in the dried germinal vesicles of Photos.
13 to 115, Plate-V iL.

At the upper, right-hand periphery of the germinal vesicle of Photo.
30, Plate II, there are two isolated, loosely granular filaments, the
granules perhaps representing individual chromomeres, which are ob-
scured in the more dense chromosomes of Photos. 28, Plate II, 51, 52 and
64, Plate IIT. Later stages of chromosome formations are shown in
Photos. 68 to 72, Plate IV, and further stages of development in Photos.
85 to 89, Plate V.

In dried germinal vesicles we see a much more intelligible evolution
of the chromosomes,—in Photos. 113 to 115, Plate VII, the delicate
chromatic thread or threads of Photo. 111 have contracted or fused into a
relatively thick spireme. This spireme divides transversely into bivalent
chromosomes, the character of each chromosome being clearly expressed
by transverse constrictions in the center, demonstrating that each bivalent
chromosome is composed of two equal parts. Photos. 114 and 115, Plate
VII, show a distinct longitudinal split of the spireme, and Photos. 116,
117, 119 to 130, Plates VII to IX, show the persistence of this split to
the chromosome stage, producing typical tetrads.

In all our sections showing the stages represented in Photos. 68 to 70,

Katharine Foot and E. C. Strobell 215

Plate IV, 86 to 89, Plate V, we find the chromosomes in a tangled condi-
tion, and we are unable to identify in such confused masses of chromo-
somes the rings or other forms of an earlier stage (Photos. 51 and 52,
Plate IIT). Presumably the small ring of Photo. 52 answers to the small
four part chromosome of Photo. 72, Plate IV, and perhaps to the small
chromosome of Photos. 33, Plate II, 91, Plate V, and 99, Plate VI, but
our sections give no proof of this, chromosomes being so rarely isolated
in fixed material that no trustworthy comparison of progressive stages
can be made.

In dried germinal vesicles, however, the chromosomes of each stage
are clearly isolated and can be readily compared and a few of the chromo-
somes of the first maturation spindle can be identified in the prophases,
though their individual differences are not sufficiently marked to make
such identification conclusive.

In the majority of cases the photographs of sections show that the
chromatic filaments are partly formed, more or less at the expense of the
reticulum in which they are imbedded (Photos. 22 to 25, Plate IT, 46 to
49 and 59 to 62, Plate III). We are convinced that this is not com-
parable to the living condition; but due to fixation, for its degree varies
with different fixatives. Eggs fixed in chromo-acetic show the extent
to which this process can be carried, the achromatic substance” and
’ chromatin being coagulated into thick, loose coils that are so dense they
hold the stains with as much tenacity as the chromosomes (see Photo.
34, Plate Il). This photograph also shows the nuclear membrane dis-
torted and torn, whereas the membrane of the germinal vesicles of eggs
killed in platino-osmic remains unbroken and in perfect contact with the
nucleoplasm as well as the cytoplasm. Compare for example, Photo. 34,
Plate II, with Photos, 68 and 69, Plate IV, the latter egg showing the
smooth contour of the membrane typical of a living egg, and this is also
shown in germinal vesicles dried on the slide.

All the sections of germinal vesicles are photographed at the same
magnification (1000) and a comparison of any of them with Photo. 34,
Plate II, demonstrates the swelling of the germinal vesicles in the
chromo-acetic preparation. Watching the effect of chromo-acetic on the
living egg under the microscope shows that this fixative first swells the
ego before the usual shrinkage of dehydration commences—the final
shrinkage being less than that of many other fixatives. The actual size

2 We use the expression achromatic substance because it is so well estab-
lished as a definite part of the nucleoplasm as distinct from chromatin. It
is often, however, a misnomer, for in some cases it stains intensely (Photos.
27 and 28, etc.).

216 First Maturation Spindle of Allolobophora Feetida

of the germinal vesicle of Photo. 3+ more nearly approaches that of the
living egg; but the distortion of the nuclear constituents indicates the
injurious effects produced by the chromo-acetic, probably due largely to
the initial swelling. Such definite and varying response to fixatives.
serves to support the present skeptical attitude towards all cellular struc-
ture seen in fixed material. The reticulum which must represent the
residuum of the achromatic substance after its dehydration, presents a
varied aspect, definite fixatives being responsible for definite forms.
These form differences are clearly seen by a comparison of the reticulum
in Photos. 68 to 73, Plate IV (fixed in platino-osmic) with those of
Photos. 51 and 52, Plate III, showing an egg of nearly the same stage of
development, but fixed in corrosive sublimate. ‘The achromatic substance
of the former (Photos. 68 to 73, Plate IV), is a relatively transparent
homogeneous substance in which the chromosomes are imbedded, while
that of Photos. 51 and 52, Plate III, is a distinct network. The char-
acteristics of the achromatin of the first egg (Photos. 68 to 73) are as
pronounced in the unstained sections (Photos. 68 and 69) as in those
stained with iron hematoxylin (Photos. 70 to 73, Plate IV), and this
ege shows that dehydration and shrinkage have taken place with much
less distortion of the nuclear and cytoplasmic elements. The nuclear
membrane is intact and the cytoplasm is not torn away from the nucleus
as is the case in corrosive sublimate preparations, and in all these eggs
killed in fixatives containing acetic acid “s—-compare those of Plates II
and III with Platesl and IV. The extent to which a fixative may distort
the nucleoplasm is seen in Photo. 34.

The photographs demonstrate that apart from the nucleoli only two
constitutents are clearly differentiated in the germinal vesicle, the rela-
tively achromatic substance and the sharply chronaatic filaments, which,
as stated above, appear to be not pure chromatin but chromatin plus a
part of the achromatic substance. Photos. 22 to 25, Plate II, 46, 47
and 59 to 62, Plate III, show in many cases clear areas around the
filament indicating that the chromatic filaments are partly formed at the
expense of the surrounding reticulum, and the fact that after some fix-
atives, (e. g., chromo-acetic, Photo. 34, Plate IT), all the achromatic sub-
stance and chromatin are welded together, suggest that the fixative may
be responsible even in those cases where only a small part of the achro-
matic substance contributes to the chromatic filament. ‘This has a dis-
tinct bearing on the theory that only a small part of the chromatin of

* Tellyesniczky’s, ’o02, criticism of the use of acetic acid in the study of
nuclear constituents is supported by its effect on the egg of Allolobophora. It
unquestionably produces artefacts in both nucleolus and nucleoplasm.

Katharine Foot and E. C. Strobell 217

the germinal vesicle takes part in forming the chromosomes of the
first spindle. The evidence points rather to the conclusion that the
apparent surplus of chromatin in this egg is due to its earlier artificial
combination with the achromatic substance causing a misconception as
to the actual amount, and that all the chromatin, excepting that possibly
contributed to the accessory nucleoli, is finally consigned to the chromo-
somes.

We have not attempted to analyze the reticulum by differential anilin
staining, because the experimenting we have done with anilins on later
stages of the egg, has convinced us of the justice of the criticism of those
investigatcrs who question all results obtained by this method. Our aim
is to place in evidence only such phenomena as can be seen and photo-
graphed without the aid of any complcated method of staining, and in
nearly all cases we control the evidence of the stained preparations by
photographs of unstained sections, e. g., Photos. 68 and 69, Plate IV,
and forty other unstained sections showing nucleoli, centrosomes, archo-
plasm, ete.

With thin unstained sections much can be seen and photographed at a
thousand diameters—the centriole and even individual microsomes can
be clearly registered by photographs and such evidence as this method
furnishes is at least relatively reliable. It may be an objection that this
simple method throws out of court a number of so-called nuclear struc-
tures, for we are indebted to the anilin stains for several analytical sub-
divisions of the reticulum, e. g., Heidenhains’ lanthanin or oxychro-
matin granules which, according to Tellyesniczky, ’02, are the same as
Schwartz’s paralinin and Pflitzner’s parachromatin—the non-staining
linin, and Reinke’s cedematin spheres, or cyanophilous granules. It may
be justly asked, whether this is a question of indebtedness to the anilins
or a score to settle.

Chromosomes.—The development of the 11 tetrads and their subse-
quent division in the first maturation spindle are so clearly demonstrated
by our new method described on p. 200 that the successive steps of the
process can be illustrated by a few photographs “ of these preparations
(Plates VII, VIII and IX).

In the dried germinal vesicles of eggs from the distal end of the ovary
and of the youngest eggs from the receptacula ovorum we have been
unable to identify any differentiation of the nucleoplasm into the rela-

1 We have more than two hundred preparations demonstrating these stages
with equal clearness and many of these we have already photographed. In
a future paper we shall reproduce some of them in connection with photo-
graphs of later stages demonstrated by the. same method.

218 First Maturation Spindle of Allolobophora Feetida

tively achromatic and chromatic segregation which appears later. Photo.
111, Plate VII, shows an early stage of the segregation of the diffused
chromatin into a delicate thread or threads which later form the pro-
nounced spireme of Photo. 113. At this early stage of the spireme the
entire germinal vesicle is traversed by a delicate thread so closely entwined
that it gives the appearance of a network and it has been impossible to
determine whether this is composed of one continuous thread. Photo.
111 represents a typical distribution of the chromatin at this stage. We
have a large number of similar preparations and many photographs, but
lack of space prevents our reproducing more than one.

Photo. 112, Plate VII, shows a very different segregation of the chro-
matin, the chromatic granules of the nucleoplasm are collecting directly
into a coil-hke structure without passing through the stage shown in
Photo. 111. We have only one such preparation and we interpret it as
abnormal, but have reproduced it because it shows so clearly that the
chromatin is distributed throughout the entire germinal vesicle, and be-
cause that part of the nucleoplasm which is not yet differentiated into
chromatin and achromatin gives a very faithful picture of the entire
nucleoplasm of odcytes in an earlier stage of development. The granular
nucleoplasm as shown on the right periphery of the germinal vesicle of
Photo. 112 gradually segregates (in normal eggs) into an extremely deli-
eate chromatic network, which is at first as indistinct as that shown at
the left side of the germinal vesicle of Photo. 111.

Photo. 113, Plate VII, is a germinal vesicle showing a typical early
stage of the spireme. A study of this photograph in the hght of Photo.
111 suggests that the spireme of Photo. 113 has been formed by a con-
traction and thickening of the delicate thread or threads of Photo. 111
or by the fusing of parallel strands.

A study of Photo. 114 in the light of Photo. 113, Plate VII, suggests
that each part of the double thread of Photo. 114 may be the single thread
of Photo. 113, or, as we are inclined to think, that the single thread has
increased in thickness by contraction and growth and has subsequently
split. The longitudinal split of the spireme seen in Photos. 114 and 115
persists throughout the prophase and can be clearly seen in many of the
chromosomes at the metaphase (cf. Plates VIII and IX). We interpret
Photos. 114 and 115 as a later stage than Photo. 113 because the thread
has commenced to break apart transversely to form the eleven bivalent
chromosomes.

In Photo. 116, Plate VII, the entire spireme has divided into bivalent
chromosomes with the exception of the two bivalent chromosomes which
are close to the accessory nucleolus. 'These are still attached end to end,

Katharine Foot and E. C. Strobell 219

forming a part of the original coil including four univalent chromosomes
or two bivalent chromosomes. The bivalent character is shown in the
lower of the two chromosomes by a clear space in the center, and a similar
clear space is shown in the bivalent chromosome just northeast of the
accessory nucleolus.

We have several intermediate stages between Photos. 115 and 116, Plate
VII, where the chromosomes are in the form of long thread-like loops, but
lack of space prevents our reproducing them. The bivalent character
of the chromosomes is clearly shown in many of the photographs. The
three rings in the upper part of Photo. 117, Plate VIII, show not only that
each ring is composed of two univalent chromosomes attached end to end,
but the longitudinal spht of each is indicated, completing the transverse
and longitudinal markings typical of the tetrad. We may say that the
chromosomes of the prophase and metaphase are typical tetrads, for in
every preparation in which the eleven chromosomes are shown, one or
more of them show beyond question both the longitudinal and transverse
markings.

In Photo. 118, Plate VIII, at least five of the eleven chromosomes
show the transverse constriction, though in all these chromosomes the
longitudinal split is obscured. In Photo. 119 the tetrad character of at
least five of the chromosomes is almost schematically shown, the large
figure 8 shows not only the longitudinal spht but a marked constriction
in each loop indicating the point of contact of the two univalent chromo-
somes. The smali chromosome southwest of the figure 8 shows with
equal clearness its bivalent character and the longitudinal split, and the
three bivalent chromosomes north of the figure 8 admit of only one inter-
pretation. The fact that eleven bivalent chromosomes are typical of
the prophases of the first maturation spindle of Allolobophora is demon-
strated by the photographs of Plate VIII and the tetrad character of
these chromosomes is clearly shown. A good deal of scepticism has re-
cently been expressed as to the constancy of the number of the chromo-
somes in the first spindle, discrediting the great significance that has been
attached to this point. We would therefore accentuate the fact that in
every case where the chromosomes are so distributed as to admit of an
accurate count, we have not found a single exception to the number eleven
in the prophases and metaphase. Rods, rings and figures 8 are the
most common forms, though there are examples of the cross-shaped
chromosome which several investigators have demonstrated in other
forms. In Allolobophora an interpretation of their origin appears to
present no difficulties, they undoubtedly arise by a simple contraction
of a bivalent chromosome, i. e., two rod-shaped univalent chromosomes

220 First Maturation Spindle of Allolobophora Foetida

placed end to end. As they contract and are pressed together each splits
open along the line of the longitudinal furrow, the ends are thus pressed
out at right angles forming the two arms of the cross. As our prepar-
ations show the cross type of chromosome in all stages of its development,
no other explanation of its origin for this egg seems possible. The be-
ginning of the formation of a cross is seen in Photo. 124, Plate VIII,
in contact with the asymmetrical figure 8, and northeast of it a cross in a
further stage of development. Varying forms of the cross chromosomes
are seen in Photos. 117, 120, 123, Plate VIII, and 116, Plate VII. The
last photograph shows also the first stage of a cross formation in ‘the
bivalent small chromosome at the lower periphery of the germinal vesicle,
and in Photo. 126, Plate IX, the method of forming a cross is almost
schematically shown in the fourth chromosome from the left periphery of
the photograph.

In the preparations reproduced on Plate IX the membrane of the
germinal vesicle as well as the principal and accessory nucleoli have dis-
appeared. The eleven bivalent chromosomes in all cases are present and
~in Photos. 129 and 130 are symmetrically arranged in the equator of tne
spindle ready to divide. These preparations appear to us to demonstrate
conclusively that the first division separates two univalent chromosomes,
but we do not yet know that these two univalent chromosomes are two of
the somatic chromosomes of the odgonia, so we cannot assert that the first
division is a reducing division in Weismann’s sense. We can only say
that the prophases and metaphase of the first maturation spindle of
Allolobophora support the observations of Korschelt, Montgomery and
others, who do claim that the first division is reducing. But in Allolo-
bophora several questions still remain unanswered. Does each bivalent
chromosome represent two somatic chromosomes which are exactly similar
in size and form, or does this exact similarity only indicate a foreshadow-
ing of the first division? Do the two represent the paternal and maternal
inheritance as held by Montgomery, ’or, Sutton, ’o2, and others, or
does the longitudinal furrow indicate this double line of inheritance? We
must delay an attempt to answer these questions until we can determine
whether the pairs of chromosomes, represented by the bivalent chromo-
somes of the prophase, are present in the odgonia as Montgomery and Sut-
ton find them in certain insects, and whether the longitudinal furrow of
the prophase can be explained as a foreshadowing of the second division.

The photographs of Plates VII and VIII demonstrate that the ring
chromosomes are formed by the uniting of the free ends of two wnivalent
chromosomes and the photographs of Plate IX show that such rings are
divided at the metaphase at the points of contact of these two chromo-

Katharine Foot and E. C. Strobell 221

somes. In most cases this point of contact is expressed by a clear space
or by a knob-like thickening. The clear space is shown in the three rings
of Photo. 117, Plate VIII, and in one or more of the chromosomes in
Photos. 116 to 130, Plates VII, VIII and IX. The knob-like thickening
at the point of contact of two univalent chromosomes is shown in one of
the chromosomes of Photos. 116 and 130.

That the spindles must have some tenacity of form in the living egg
is demonstrated by the characteristic spindle formation with the two polar
spheres often remaining undisturbed by the process of pricking the
membrane of the egg and allowing the cytoplasm to flow out upon the
slide. An indication of the spindle form is shown in Photos. 125, 128,
129 and 130, Plate IX. The fact that the egg is dried so rapidly that
the form of the spindle is not distorted argues that some confidence may
be placed in the form of the chromosomes as well.

In the equator of the spindle of Photo. 125, Plate LX, a ring chromo-
some is seen showing a distinct longitudinal spht and the clear transverse
space which indicates one point of contact of the two univalent chromo-
somes which form thering. ‘This space is in the equator and unquestion-
ably indicates one of the points of separation of this chromosome. The
chromosome on the extreme left shows a like clear space, the other half
of the ring, having already separated and contracted, resulting in one of
the forms typical of the metaphase (the lower arm of this chromosome
is in contact with the upper arm of a like chromosome). This form of
division is seen in two of the chromosomes of Photo. 128, in at least five
of the chromosomes of Photo. 129 and five of Photo. 130, Plate IX.
Rings with a longitudinal furrow and characteristic indications at the
points of contact of the univalent chromosomes of which they are formed
are shown in Photo. 126, Plate IX, and three similar rings dividing
transversely are shown in Photo. 127, the one near the center of the
photograph being especially instructive. These examples, with the ring of
Photo. 129 and the two rings of Photo. 130, Plate IX, appear to dispel
all doubt as to the manner in which the bivalent chromosome of A/lo-
lobophora is divided in the first maturation division. In each of these
six photographs there are examples also of the simple transverse separ-
ation of the two rods attached end to end which represent the simplest
form of these bivalent chromosomes. Many of them still show the
longitudinal furrow which has persisted from the spireme stage and
leave no doubt that this division is not along the lines of this longitudinal
spht.

Many of the chromosomes demonstrated in the spindles of our Plate

IX closely resemble those figured by Nekrassoff, ’03, in the first matur-
16

Cas)
oo
Co

First Maturation Spindle of Allolobophora Feetida

ation spindle of Cymbulia, Fig. 7, though he interprets their division as
longitudinal.

This egg agrees with the observations of the many investigators who
have demonstrated for both odcytes and spermatocytes a marked differ-
ence in the size of the chromosomes of the first spindle. In Crepidula,
Conklin, ’o02, has shown this inequality to have reached the ratio of one
to fifteen in volume. In Allolobophora the inequality in the size of the
chromosomes is distinctly seen at the prophases and metaphase; compare
the two chromosomes in the same germinal vesicle of Photos. 51 and 52,
Plate III, and chromosomes in the germinal vesicle of Photo. 116, Plate
VII, and those of Plate VIII. Compare the size of the chromosomes in
the first spindle (metaphase) Photo. 33, Plate II, and those of Plate IX.

It is more difficult, however, to demonstrate a persistent and individual
form for each chromosome and in fixed and sectioned eggs we have found
this quite impossible. For example, in a collection of thirty-four photo-
graphs showing every chromosome in four first spindles at the meta-
phase it was impossible to identify any one chromosome in the four
spindles as the same individual. This is undoubtedly due in part to
distortion of normal forms by fixation and as a rule the chromosomes are
so closely massed in fixed material that their individuality is obscured
(e. g., Photos. 69 and 70, Plate IV, 85 to 89, Plate V). In dried germ-
inal vesicles this massing of the chromosomes is avoided and individuals
can be distinctly differentiated. Although the individuality of these
chromosomes is not sufficiently pronounced to admit of a definite identifi-
cation of the individual at each stage, a comparison of the chromosomes
of the prophases of Plate VIII with those of the metaphase on Plate IX
will demonstrate that a few of the individual chromosomes of the pro-
phases can be identified in the metaphase with some accuracy, and this
argues strongly for the individuality of all, and supports the theory which
has been so frequently and ably defended by Boveri.”

The prophases of the first spindle of Allolobophora as above demon-
strated confirm Vom Rath’s interpretation of the prophases of the first
spindle of the spermatocytes of Gryllotalpha published in 1892. When

& After our paper had gone to press Baumgartner’s interesting article
appeared, giving “Some new Evidence for the Individuality of the Chro-
mosomes,” Bio. Bull., Vol. VIII, 1904. Our Photos. 116 to 130, Plates VII,
VIII and IX demonstrate that Baumgartner’s suggestive conclusions are not
supported by the egg of Allolobophora. We find no constant form differences
of the chromosomes, the simplest form of the bivalent chromosome is two rods
attached end to end, and these present a variety of shapes, rings, figures 8,
crosses, ete., without any regularity or constancy. The free ends of the

Katharine Foot and E. C. Strobell 223

the “Samenmuttenzelle” (spermatocytes Ist order of authors) has
attained its growth the chromatin is distributed as a delicate “ Maschen-
werk ” (cf. his Figs. 10 and 11 with our Photo. 111, Plate VII). He
next figures and describes the chromatin as a coil with a single longitud-
inal split (Fig. 12), this coil dividing transversely into half the number
of somatic chromosomes, each of the bivalent segments representing two
somatic chromosomes attached end to end, later their free ends uniting
to form rings, these rings showing the same longitudinal spit which he
demonstrated in the coil. He is uncertain, however, whether this longi-
tudinal split foreshadows the first or second division. He says: “Hs
kann folglich die eine der beiden Trennungen der Chromosomen auf diese
vorseitige Spaltung des Chromatinfadens zuriickgefiihrt werden; ob dies
nun aber die erste oder die zweite Theilung ist, kann nach den Prap-
araten nicht mit Sicherheit entschieden werden, ich méchte eher an die
zweite Theilung denken,” p. 113.

Montgomery’s, ’or, ’04, interpretation of the prophases studied in a
variety of forms, is supported by Allolobophora in the longitudinal split
of the coil (cf. Montgomery’s ’04, Figs. 10 and 11 with our Photos. 114
and 115, Plate VII), and the separation of the univalent chromosomes at
the first division, but this egg does not support Montgomery in certain
points in which his observations differ from those of Vom Rath, Riickert
and Hicker. These points are clearly stated by Montgomery, ’o4:
“ Riickert, ’94a, and Hacker and others after him, concluded that there
was a continuous chromatin spireme preceding the first maturation mitosis,
and that the apparent reduction in number of the chromosomes is effected
by this chromatin spireme segmenting into half the normal number of
chromosomes. JI showed for Peripatus, ’00, on the contrary, that a
continuous linin spireme is present at this stage but not a continuous
chromatiu spireme, and that the bivalent chromosomes are produced by a
later conjugation without the formation of a continuous chromatin loop.
According to Riickert it is a case of chromosomes already closely con-
nected remaining so; according to me, of chromosomes not in contact at
first, becoming so secondarily. Hence I spoke of this act as the con-
jugation of the chromosomes, and argued that this is the important

bivalent chromosomes show a tendency to unite into a ring and in some cases
nearly all the eleven chromosomes are rings (Photo. 122), and sometimes not
a single ring is formed, Photos. 116 and 118. This by no means disproves
Baumgartner’s conclusions, for the variety of shapes of the chromosomes
of Allolobophora may be due to mechanical disturbance of the living form
incident to the technique. This point can be determined only by the study
of living chromosomes.

224 First Maturation Spindle of Allolobophora Feetida

eriterion of the synapsis stage.” The photographs of our Plates VII and
VIII demonstrate that in Allolobophora “it is a case of chromosomes
already closely connected remaining so.”

Many of our photographs confirm A. and Ix. E. Schreiner’s, ’04, obser-
vations on the spermatocytes of Myaine glutinosa and Spinax niger. The
delicate thread-like reticulum in the early prophase (cf. Fig. 2 with our
Photo. 111), the coarse spireme of the later stage (cf. their Figs. 3, 5 and
23 with our Photos. 114 and 115, Plate VII), and finally the form of the
bivalent chromosomes. ‘The rings, figures 8, ete., of Schreiner’s Fig. 15
are reproduced in Allolobophora, though in their mode of formation and
subsequent division there are fundamental differences. ‘They interpret
the first furrow of the spireme as due to a union of two of the delicate
threads of the earlier stage, and at a later stage they identify a second
longitudinal split of the spireme, these two longitudinal divisions indi-
cating ‘the method of separation of the chromosomes for the first and
second divisions. The individuals of each bivalent chromosome are
paired longitudinally in the spireme, whereas in Allolobophora they are
placed end to end, thus though the rings, figures 8, etc., of Schreiner’s
Fig. 15 and those of Allolobophora are formed alike by the uniting of
the ends of two univalent chromosomes, they have attained this final
arrangement by an entirely different method. In both forms, therefore,
the first division separates univalent chromosomes, though in one case
the division is longitudinal and in the other transverse.

This transverse division of the chromosomes supports Lillie’s, ’o1,
observations on Unio. He says, “The first division is certainly at right
angles to the long axis of the chromosomes, as these lie in the equatorial
plate,” p. 236.

The spireme demonstrated in our photographs of Plate VII and the
“heterotypic”’ chromosomes of Plate IX confirm Flemming’s obser-
vations on the spermatocytes of Salamandra maculosa published in 1887.
The “heterotyyic” chromosomes of his Figs. 22, 23 and 25 are accurately
reproduced in several of our photographs of Plate IX. Since the ring
chromosome was demonstrated by Flemming many investigators have
identified them in a variety of forms. The ring chromosomes of our pho-
tographs of Plates VIII and IX are similar to those demonstrated by
Henking,” ’91, in Pyrrhocoris ; Moore, ’95, 1n Hlasmobranchs; Bolles Lee,
*97, in Helix; von Klinckowstrém, ’97, in Prostheceraeus, de Sinéty, ’or,
in Orthoptera, Schockaert, ’04, in Thysanozoon and Montgomery, ’o4, in

* Henking interprets the first division as a reducing division and the second
as an equation division.

Katharine Foot and E. C. Strobell 220

Plethodon. McClung, ’oo-’02, has shown rings, figures 8 and crosses
in certain insects clearly demonstrated by excellent photographs. Helen
Dean King, ’o1, has demonstrated the ring in the egg of Bufo, and her
interpretation that the knob-lke thickenings represent “the place of
union of the two chromosomes that fused to form the ring ” is confirmed
by the photographs of our Plates VII and VIII; in Allolobophora, how-
ever, the rings do not divide longitudinally in the first division as in
Bufo.

Many forms of the chromosomes in Allolobophora are similar to those
demonstrated by Korschelt, ’95, in Ophryotrocha and his interpretation of
the first division separating two univalent chromosomes is confirmed by
the photographs of our Plates VII, VIII and IX. Further details in
which Korschelt’s observations are supported by this egg are stated under
the heading ‘“‘ Comparisons with other Annelids.”

Archoplasm.—In earlier papers we have used Boveri’s, ’88, term
archoplasm for that substance in the cytoplasm which in the youngest
odcytes is massed in different shapes close to the germinal vesicle (yolk
nucleus of authors). As the egg grows this substance increases in
amount, becomes distributed throughout the cytoplasm; after fertiliza-
tion much of it gradually segregating to the periphery, and finally a large
part of it contributing to the formation of the polar rings. We use the
term archoplasm because at definite phases of the development of the
egg the substance appears to contribute to the formation of asters and
spindles and may thus be the homologue of Boveri's archoplasm. - Pro-
gressive steps in the development of the substance from yolk nucleus to
polar rings were illustrated by a series of photographs (Foot and |
Strobell, ’or), but in the present paper we have reproduced only three
sections of the earlier stages, merely to illustrate our interpretation. At
the upper periphery of Photo. 12 about half of a very small odcyte is
shown, with a mass of archoplasm (yolk nucleus) at the opposite poles of
its nucleus.. The next stage represented in this paper is shown in
Photo. 76, Plate 1V, the archoplasm being somewhat removed from the
nuclear membrane. A later stage is seen in Photo. 78, Plate IV, in
which the substance is distributed throughout the cytoplasm, and Photo.
12 shows a lke distribution in an older odcyte. In recent investigations
of this polar-ring substance we have attempted no analysis by compli-
cated methods of staining, studying it only in the light of comparative
fixation and aiming thus to demonstrate its presence in eggs in which,
after some fixatives, its identity is questionable.

It is interesting to compare the effects of fixation on the two con-
stituents of the cytoplasm, chromatic archoplasm and relatively achro-

226 First Maturation Spindle of Allolobophora Feetida

matic cytoplasm, with the effects of fixation on the two nuclear constitu-
ents, chromatin and achromatic nucleoplasm, for in both cases the reac-
tion to fixation is strikingly alike. Such a crude comparison is instruc-
tive only to illustrate the fact that both chromatin and archoplasm can
be identified after some fixatives and obliterated after others, yet the
specific character of chromatin is universally admitted, whereas the spe-
cific character of archoplasm has been very generally doubted. We do
not mean to imply that the two substances, archoplasm and achromatic
cytoplasm, are the sole constituents of the cytoplasm, any more than the
chromatin and achromatic nucleoplasm may be the sole constituents of
the nucleus. Archoplasm and achromatic cytoplasm are as much at the
mercy of the fixatives as are the chromatin and achromutic nucleoplasm,
and, like them, can be fused together into a network or into masses in
which the constituents are indistinguishable, or they may be so separated
that the two are readily differentiated. Such a differentiation is seen
in the unstained sections of Photos. 68 and 69, Plate IV, the chromatic
archoplasm and relatively achromatic cytoplasm being segregated into
quite definite areas. We do not claim that such pronounced cases of seg-
regations more nearly approximate the living condition, but they may be
instructive as an aid to demonstrating the individuality and continuity
of the substances during these stages, and the same may be said of some
forms of segregated chromatin shown in many of our photographs of sec-
tions. The achromatic cytoplasm of Photos. 10, Plate I, 68 and 69, Plate
IV, like the achromatic nucleoplasm, is a relatively homogeneous sub-
stance, but it becomes granular and chromatic when combined with the
archoplasm (Photo. 13, Plate I, just as does the achromatic nucleoplasm
when combined with chromatin (cf. the nucleoplasm of Photo. 34, Plate
II, with that of the germinal vesicles of Plates II, III, IV). Aggrega-
tions of archoplasm like the three shown in Photo. 13 are distributed
throughout the cytoplasm of the entire egg (cf. Foot and Strobell, ’98,
Photo. -9). Such aggregations are typical of chromo-acetie preparations
and in the hght of our recent study of the germinal vesicle we are
convinced that the entire cytoplasm of such preparations represents an
artificial combination of archoplasm and achromatic cytoplasm compar-
able to the fusing of the chromatin and achromatic nucleoplasm in the
germinal vesicles of the same preparations (Photo. 54), and they support
the observations of the investigators who question the specific nature of
archoplasm. Such an interpretation would be supported also by Photo.
100, Plate VI, but a comparison of these two photographs (13 and 100)
with 68 and 69, Plate IV, suggests that the individuality of the archo-
plasm, so clearly shown in the last two photographs, is only obscured

Katharine Foot and K. C. Strobell ipa

in Photos. 13 and 100. We believe the chromatic granular substance
of the prophase (Photos. 68 and 69) is the same as the chromatic gran-
ular substance of the metaphase (Photo. 99, Plate VI). These eggs were
killed in the same fixative, platino-osmic, and the substance can be recog-
nized at the two stages. In Photo. 100 the substance has a very different
distribution from that of Photo. 99, Plate VI, though the two eggs are at
exactly the same stage of development, i. e., metaphase of the first spindle.
In the chromo-acetic preparation (Photo. 100) the homogeneous achro-
matic cytoplasm is in the form of pronounced rays, combined in
such a way with the archoplasm that the latter may be interpreted
as cyto-microsomes. After some fixatives it certainly does assume
the form of cyto-microsomes and in these cases its identification as
a specific substance is possible only where it is accumulated into
dense masses. Its interpretation as a specific substance or as an
integral part of the cytoplasm depends upon its special .manifesta-
tion after a given fixative and suggests that the opposing interpretations
are largely a question of terms. In this egg we claim its individuality
only on the ground that we think we can trace the substance with un-
broken continuity from its earliest aggregation as yolk nucleus in the
youngest odcytes to the cleavage stages—a large part of it contributing to
the formation of the polar rings. Aggregations of archoplasm not alone
in chromo-acetic preparations, but in corrosive sublimate and many
others are readily differentiated by double staining (Foot, ’96), but this
method obscures its presence when it is most evenly distributed through-
out the egg, and for this reason study of comparative fixation has seemed
the more profitable method to follow (Foot and Strobell, ’00). When the
odcyte first order has reached its maximum growth it is especially diffi-
cult to differentiate the archoplasm. Its presence at this stage is demon-
strated in Photos. 68 and 69, Plate IV, and Photo. 10, Plate I, shows
an interesting segregation of the substance in the form of a “ polar ring”
which is not normally due until the pronuclear stage. This is a section
of an odcyte with the germinal vesicle intact and the chromosomes not
yet formed, a stage earlier than that shown in Photos. 68 and 69. There
is a similar aggregation of archoplasm at the opposite pole of this egg
and these two polar aggregations present a striking resemblance to many
polar rings of the pronuclear stages which are not invariably in the form
of aring. This precocious polar segregation of the substance in Photo.
10 appears to us to demonstrate the presence of this definite substance
in the egy during these early stages and the granular appearance of the
archoplasm in this photograph is typical of all fixed material. The
chromatic centers of asters fail to show this granular effect (Photos. 84,

228 First Maturation Spindle of Allolobophora Feetida

86, 91, 92, Plate V, 99 to 110, Plate VI), but we do not think this
necessarily means that these centers are devoid of archoplasm; it may
indicate rather a definite chemical combination with the achromatic
cytoplasm that causes a different morphological reaction to fixation
(Foot, ’96). An obvious contribution of archoplasm to the spheres is
largely dependent on fixation, and in some cases it is aggregated into
granular masses or heavy rays around the mark-zone (Foot and Strobell,
700), and again the mark-zone itself is granular and stains intensely.

Differentiation of a special chromatic substance in the cytoplasm of
young odcytes or spermatocytes is very common, but more rarely is this
substance traced to the later stages of development. Among recent
papers Voinoy, ’03, has traced a substance to the first spindle in Cybister,
and finally to the Nebenkern. He has designated it as “zone interne”
as distinguished from his “ zone externe” (Figs. 29 and 30), and a com-
parison of his figures with our photographs leads to the impression that
the archoplasm of Allolobophora is synonymous with his “ Mitochondria”
(Benda’s) and “ zone interne ” combined (see his Figs. 35 and 39).

Data as to the specific nature of polar-ring substance have been pre-
sented by Wilson in his interesting paper on Dentalium, ’04. He iden-
tifies an upper and lower polar area in the odcyte and of these he says:
.“T believe it is probable that at least the lower protoplasmic area and
probably also the upper disc are in a general way comparable to, if not
identical with, the polar rings observed in the eggs of certain leeches and
Oligochetes.” Of the lower polar area he says: “It is evident
that material from the interior of the egg must flow into the
lobe as it forms,” and of the upper polar area he adds: “ It is here again
evident that an extensive flow of this material must take place from the
interior cf the egg” (pp. 12-15). These facts have a special bearing
on our interpretation that the polar ring substance is distributed through-
out the egg and later aggregates at the poles. (Foot and Strobell, ’98-)
Wilson interprets both areas as “ specific cytoplasmic material.”

In this connection Wheeler’s work on Myzostoma, ’97, is of special
interest. He interprets certain phenomena at the upper pole of the egg
as homologous to the polar rings of Annelids and he identifies in the
odcyte (Fig. 1) a denser area of protoplasm which strikingly resembles
the yolk nucleus of Allolobophora and this he traces to the yolk-lobe
(opposite pole), though he does not interpret the substance as yolk-
nucleus, nor the yolk-lobe as homologous to a polar ring of Annelids.
Conklin, however, in 1897, identifies at the vegetal pole of the egg of
Crepidula a mass of hyaline substance which he homologizes to the
yolk-lobe described by Mead in Chetopterus and to the polar rings of

Katharine Foot and EH. C. Strobell 229

Annelids. He says: “I am convinced that this peculiar body is homolo-
gous with the problematical lobe which is described by Mead, ’95, in the
egg of Chetopterus and further it is probably identical with the polar
rings observed by Whitman, ’78, in Clepsine and since then by various
authors in different Annelids.”

COMPARISONS WITH OTHER ANNELIDS.

In the odgenesis of Annelids there has been very little work done on
the prophases of the first maturation spindle.

Among the Lumbricidae we have not found any record of observations
on these stages, but in the spermatogenesis of Lumbricus terrestris
Calkins, ’95, has studied the prophases of the first division and the two
species, Allolobophora and Lumbricus, are in accord in showing a spireme
with a longitudinal furrow. In Lumbricus, however, the spireme divides
transversely into the full number of somatic chromosomes, and there is,
therefore, no numerical reduction of the chromosomes by two univalent
chromosomes remaining attached end to end as in Allolobophora.

In the first spindle of Luwmbricus there are sixteen tetrads, the
spermatocytes second order receiving each sixteen double chromosomes.
Calkins adds: “Whether this is a reducing division in Weismann’s
sense cannot be ascertained.”

Vejdovsky’s and Mrazek’s recent valuable work, ’03, on Rhynchelmis
is confined to later stages, the material being unfavorable for the study
of the prephases of ‘the first maturation spindle. Of these stages they
say: “ Das Studium der ersten Vorgiinge der Hireifung ist schon tech-
nisch sehr zeitraubend, wenn man auf den endlosen Schnittserien durch
die einige Zentimeter langen von Hiern prall angefiillten vorderen
Abschnitte des Wurmleibes stets nur entweder noch ruhenden Kernen
oder den bereits fertigen Reifungsspindeln begegnet ” (pp. 454 and 455).
They have, however, supplemented their work on Rhynchelmis by a
study of the prophases of the first spindle in Tubifex, Limnodrilus and
Ilyodrilus, but their results are demonstrated by a single text figure
showing a germinal vesicle with tetrads. In explanation of this figure
they say: “ Den ganzen Vorgang der Chromosomenbildung konnten wir
nicht verfolgen. rst in spiteren Stadien fanden wir die in Textfigur 3
abgebildeten Formen der Chromosomen. Es sind dies Gebilde, die wie
aus zwei dicht an einander gelegten sichel- oder biskuitformgen Teilen
zusammengesetzt erscheinen. Ein Vergleich mit den an anderen Ob-
jekten gewonnenen Resultaten fiihrt zu dem Ergebnisse, das wir hier
langsgespaltene doppelwertige Elemente vor uns haben, die den typischen

230 First Maturation Spindle of Allolobophora Feetida

Vierergruppen entsprechen. Wie die Abbildung zeigt, kann die Form
der einzelnen Gruppen etwas variieren, doch muss ausdriicklich bemerkt
werden, dass wir Ringbildungen niemals beobachten kénnten. Dagegen
sind kreuz- oder x-férmige Figuren die haufigsten.”

In Text Figs. 4 and 7 they show 37 different forms of the chromosomes
of the first maturation spindle of Rhynchelmis. In our Photo. 72, Plate
IV, there is an exact duplicate of one of the chromosomes in their Text
Fig. 4, and a comparison of their other figures with our photographs of
the first spindle of Allolobophora (Plate IX and our Text Fig. 4 of an
earlier paper, ’98), show a suggestive similarity in form. Vejdovsky
and Mrazek state, and their Text Fig. 5 demonstrates, that only the
central part of the first maturation spindle of Ilyodrilus is of nuclear
origin, but in Allolobophora a much larger proportion of the spindle is
derived from the achromatic nucleoplasm of the germinal vesicle (see
Photos. 84 to 89, Plate V). On this point Allolobophora is more in
accord with Gathy’s, ’oo, observations on T'ubifex, though the membrane
of the germinal vesicle persists longer in Tubifex than in Allolobophora.
Gathy’s Fig. 11 shows the first spindle nearly at the metaphase and yet
the membrane of the germinal vesicle is almost intact, whereas in
Allolobophora the nuclear membrane entirely disappears before the
spindle reaches the metaphase. Our Photos. 84 to 89, Plate V, show
part of the membrane of the germinal vesicle persisting until both cen-
trioles and asters are present at opposite poles, though not developed to
the stage shown in Gathy’s Figs. 10 or 11; these photographs indicate,
however, that the achromatic nucleoplasm of the germinal vesicle of
Allolobophora, ike Tubifex, contributes to a large part of the first mat-
uration spindle. Allolobophora further supports Gathy’s observations as
to the independent origin of the two centrioles, their first appearance
close to the nuclear membrane and the indication that the spindle is
formed under their influence. Gathy omits other important details in
the formation of the spindle and his Figs. 8, 9 and 10 demonstrate that
he has not observed the successive steps of the development of the chro-
matin of the germinal vesicle into the chromosomes.

Among the Annelids these stages have been most thoroughly investi-
gated by Korschelt in Ophryotrocha, ’95, and Allolobophora corroborates
almost every detail of the process Korschelt describes. In both Oli-
gochetes the chromatin forms a skein, though in Ophryotrocha the longi-
tudinal furrow does not appear until after the chromosomes are formed.
The skein divides transversely into chromosomes, in Ophryotrocha these
being univalent, whereas in Allolobophora they are bivalent, as a rule
remaining bivalent until separated at the anaphase of the first spindle.

Katharine Foot and KE. C. Strobell 201

In both Oligochetes the chromosomes have a distinct longitudinal furrow,
which has persisted in Allolobophora from the skein stage, and in both
forms the first division separates two univalent chromosomes. In these
two Annelids the achromatic nucleoplasm contributes to the formation
of the spindle fibers, the fibers forming within the germinal vesicle
while its membrane is partly intact, and the centriole in both cases is first
seen outside the germinal vesicle close to its membrane, though in Ophryo-
trocha the two arise by division of one, while in Allolobophora they are
first seen at opposite poles of the vesicle.

Thalassema.—Griftin’s interesting paper on the maturation and fer-
tilization of the egg of Thalassema, ’99, gives a clear demonstration of
the prophases of the first maturation spindle. These are reproduced in
his Figs. 1 to 12, but Griffin demonstrates no spireme in the germinal
vesicle, and he neither figures nor describes stages answering to the
stages shown in our Photos. 111 to 115, Plate VII, and in Korschelt’s, ’95,
Figs. 67 to 74. He figures a spireme only in the nuclei of “ minute ova,”
the size of these nuclei in relation to the germinal vesicles of later stages
showing them to be the young nuclei emerging from the telophase of the
last odgonial division, and the spireme of these nuclei is not comparable
to the spireme demonstrated in our photographs of Plate VII. They can
be compared only to similar minute cells in the ovary of Allolobo-
phora, stages with which we are not concerned in the present paper. In
the text Griffin describes the spireme of the nuclei of these minute cells
as showing an occasional longitudinal split and dividing transversely
into bivalent chromosomes at the beginning of the growth period (Fig.
2), these chromosomes persisting “as double rods throughout the entire
growth period” (p. 605). In Allolobophora the chromosomes do not
persist through the growth period nor can any indication of the aggre-
gation of the diffused chromatin into the spireme of Photos. 111 to 115 be
demonstrated until some time after the germinal vesicle has attained its
maximum size (see p. 217 for details). The two Annelids agree, how-
ever, as to the form of the final tetrads. Griffin’s Text Figs. 1 and 2
show chromosomes in the form of rings, figures 8, crosses, etc., which
are strikingly like those in many of our photographs, though their origin
is apparently very different. In Z’halassema the spireme divides trans-
versely into half the number of somatic segments, these bivalent chromo-
somes differing, however, from Allolobophora in the important point
that their bivalent character is expressed by a longitudinal division of
each, instead of two univalent chromosomes being attached end to end.
as in Allolobophora. 'Thus the rings, figures 8, ete., which are com-
mon for the two Annelids have a different origin, necessitating a different

232 First Maturation Spindle of Allolobophora Feetida

interpretation for the first division, in T'halassema the first division
being longitudinal (giving to each cell one-half of every wnivalent
chromosome), and in Allolobophora transverse (giving to each cell entire
univalent chromosomes, each receiving one-half the somatic number).
Griffin has clearly stated his results in the following summary :

“1. By longitudinal fission and transverse segmentation of the
spireme thread, there arise 12 (reduced number) ellipse-shaped chromatin
masses.

“2. These persist throughout the growth period of the egg.

“3. During prophase they concentrate into crosses, the arms of which
are tight loops.

“4. In the first polar division these are drawn out again into ellipses
which divide to form daughter-V’s (equation division).

“5. The V’s break apart at the angle in the second polar division (re-
ducing division),” p. 612.

The persistence of the chromosomes throughout the entire growth
period, during the time that the nuclear reticulum is gradually develop-
ing, led Griffin to the conclusion that “its development is independent of
the chromosomes which are passive during its growth” (p. 604), and
Griffin’s conclusions as to the independence of these two substances are
supported by our observations on Allolobophora. The two Annelids are
further in accord as to the first appearance of the maturation centrioles.
In both types they are first seen as minute asters close to the germinal
vesicle, though in Thalassema they are closer to each other than we have
yet found them in Allolobophora. This independence of the centrioles
accords with Mead’s observation on Chetopterus, ’98 (cf. our Photos.
81 and 82, Plate V, with Mead’s Figs. 8 and 9). The two extremely
small centrioles of the above-mentioned photographs show a more marked
independence of origin than those figured by Mead, for they are at oppo-
site poles of the germinal vesicle, while closer to its membrane and at an
earlier stage of development than those figured for Chetopterus. ‘These
primary asters of Chetopterus “arise at some distance from the wall
of the germinal vesicle,” and Mead adds: “J am not prepared to say at
present whether the primary asters are formed by the further growth
and specialization of two of the secondary asters or by the union and
coalescence of several.” These secondary multiple asters which Mead
has demonstrated in his Fig. 7, he has shown to be normal in Chetop-
terus, having watched the phenomena in living eggs, which continued to
develop after the multiple asters had disappeared. In Allolobophora we
have found only one egg showing structures that could be interpreted as
multiple asters, but the egg was unquestionably pathological, the germinal

Katharine Foot and li. C. Strobell

ras)
Go
vo

vesicle had broken down and its contents scattered throughout the cyto-
plasm. These structures in Allolobophora resemble some of the asters in
Cerebratulus which Kostanecki interprets as expressions of a pathological
condition. They show, however, no evidence that they have originated
by division of the normal aster as Kostanecki, ’02, interprets those of
Cerebratulus, they indicate rather that their irregular dense centers are
small aggregations of dispersed nuclear substance, around which cyto-
plasmic rays focus.

Mead gives no account of the development of the nine chromosomes®
which he finds in the first maturation spindle. These stages are omitted
also in Wheeler’s work on Myzostoma, ’97, where he first figures the
chromosomes as 12 tetrads suspended in the chromatic network of the
germinal vesicle. These are composed of two rods swollen at their
ends, and of these Wheeler says: “I have not studied the origin of the
double rods in the germinal vesicle so that I am unfortunately unable
to pass an opinion on the nature of the division in the first spindle. In
the case of the second spindle, however, I feel confident that there is a
longitudinal splitting of each of the single chromatin rods remaining in
the egg after the formation of the first polar body” (p. 51). Allolobo-
phora supports Wheeler’s observations as to the occasional persistence of
part of the membrane of the germinal vesicle until the spindle is formed
(cf. Photos. 84 to 87, Plate V, with Wheeler’s Fig. 63). In Myzostoma
the two centrioles are first seen near the membrane of the germinal
vesicle and in close relation to each other, being “ connected by a delicate
achromatic bridge.” Wheeler’s Figs. 3 to 5 demonstrate the gradual
separation of these centrioles to form the poles of the first spindle.

DECEMBER 1, 1904.

PAPERS REFERRED TO.

ANCEL, P., ’02.—Histogénése et structure de la glande hermaphrodite d’Helix
pomatia (Linn.) Arch. de biologie, Tome XIX.

BLACKMAN, M. W., ’03.—The Spermatogenesis of the Myriopods. On the Chrom-
atin in the Spermatocytes of Scolopendra heros. Bio. Bull., Vol. V.,
No. IV.

Bover!, TH., ’88.—Zellen-Studien. Jena.

o2.—Ueber mehrpolige Mitosen als Mittel zur Analyse des Zellkerns.

Vehr. d. phys. Med. Ges. Wutirzburg., N. F., Bd. XXXV.

Bryce, THos. H., ’o1—Maturation of the Ovum in Echinus. Quart. Journ.
Mic. Sci., Vol. XLVI.

CALKINS, Gary N., ’g5.—The Spermatogenesis of Lumbricus. Jour.
Morph., XI.

Cor, WESLEY, R., ’99.—The Maturation and Fertilization of the Egg of Cere-
bratulus. Zool. Jahrb., Bd. XII.

234 First Maturation Spindle of Allolobophora Feetida

ConkKLIN, E. G., ’97—The Embryology of Crepidula. Journ. Morph., Vol. XIII,
No. 1.
’o2.—Karyokinesis and Cytokinesis. Journ. Acad. Nat. Sci., Phila.,
Vol. XII.
FLEMMING, W., ’87.—Neue Beitrage zur Kenntniss der Zelle. Arch. f. mik.
Anat., Bd. X XIX.
Foor, KATHARINE, ’96—Yolk nucleus and Polar rings. Journ. Morph.,
Vol. XII.
’97.—The origin of the Cleavage Centrosomes. Journ. Morph., Vol. XII.
’98.—The Cocoons and Eggs of Allolobophora fceetida. Journ. Morph.,
Vol. XIV.
Foor & STROBELL, ’98.—Further Notes on the Egg of Allolobophora feetida.
Zoological Bulletin, Vol. II, No. 3.
700 and ’o1.—Photographs of the Egg of Allolobophora feetida, I and II.
Journ. Morph., Vols. XVI and XVII.
?902.— Further Notes on the Cocoons of Allolobophora feetida. Bio. Bull.,
Viole Eki
Gatuy, E., ’o00.—Contributions a l’étude du développement ce l’oeuf et de la
fécondation chez les Annelides. La Cellule, T. XVII.
GoLpscHMIDT, R., ’9o2.—Untersuchungen iiber die Hireifung, Befruchtung und
Zelltheilung bei Polystomum integerrimum Rud. Zeit. f. wiss. Zool.,
Bd. LXXI.
GRIFFIN, Br. B., ’96.—The History of the Achromatic Structures in the Matur-
ation and Fertilization of Thalassema. Trans. N. Y. Acad. Sci.
’99.—Studies on the Maturation, Fertilization and Cleavage of Thal-
assema and Zirphea. Jour. Morph., Vol. XV.
HARTMANN, MAX., ’02.—Ovarialei und Eireifung von Asterias glacialis. Zool.
Janrb. Abt. 1, Anat. Bd. xv.
HENKING, H., ’91.—Ueber Spermatogenese und deren Beziehung zur Wient-
wicklung bei Pyrrhocoris apterus. Zeit. f. wissen. Zool., Bd. LI.
Kine, HELEN D., ’o01.—The Maturation and Fertilization of the Egg of Bufo
lentiginosus. Journ. Morph., Vol. XVII.
KLINCKOWSTROM, A. VON, ’97,—Beitrage zur Kenntniss der Eireifung und Be-
fruchtung bei Prosthecereus vittatus. Arch. f. mik. Anat., Bd. XLVIII.
KORSCHELT, E., ’95.—Ueber Kernteilung, Eireifung und Befruchtung bei Ophry-
otrocha puerilis. Zeit. f. wiss. Zool., Bd. LX.
KOSTANECKI, K. v., ’02 Ueber abnorme Richtungskorpermitosen in befruchte-
ten Hiern yon Cerebratulus marginatus. Abhandl. u. Bull. Akad.
Wiss., Krakau.
LEBRUN, H., ’o1.—La vésicule germinative et les globules polaires chez les
Anoures. La Cellule, T. XIX.
%o2.,—La vésicule germinative et les globules polaires chez les Batra-
ciens. La Cellule, T. XX.
LeEE, A. BOLLES, ’97—Les cinéses spermatogénétiques chez |’ Helix pomatia.
La Cellule, T. XII.
LILLif, FRANK R., ’o01.—The organization of the Egg of Unio, based on a study
of its Maturation, Fertilization, and Cleavage. Journ. Morph., Vol.
XVII, No. 2.

ra)
SS)
Ot

Katharine Foot and E. C. Strobell

LuposH, WILHELM, ’o2.—Ueber die Nucleolar-substanz des reifenden Tri-
toneneies nebst Betrachtungen tiber das Wesen der Hireifung. Jen.
Zeit. f. Naturwissenschaft, Bd. XXXVI.

. MARSHALL and Hurst, ’88,—Practical Zoology. Smith, Elder & Co., London.

McCuung, M. E., ’00.—The Spermatocyte Divisions of the Acrididze. Bull. Uni.

Kansas, Vol. 1, No. 2.

’o2.—The Spermatocyte Divisions of the Locustide. Kansas Uni. Sci.

Bull., Vol. I, No. 8.

MEAD, A. D., ’98.—The Origin and Behavior of the Centrosomes in the Annelid
Egg. Journ. Morph., Vol. XIV, No. 2.

MEVES, FR., ’02.—Ueber die Frage, ob die Centrosomen Boveri’s als allgemeine
und dauernde Zellorgane aufzufassen sind. Verhand. Anat. Gesell.,
Bd. XXI.

MontTGoMErY, THOs. H., JR., ’98.—Comparative Cytological Studies with especial
regard to the Morphology of the Nucleolus. Journ. Morph., Vol. XV,
No. 2.

Montcomery, T. H., ’o91.—A Study of the Chromosomes of the Germ Cells of

the Metazoa. Trans. Amer. Phil. Soc., Vol. XX.

’04.—Some Observations and Considerations upon the Maturation Phe-

nomena of the Germ Cells. Bio. Bull., Vol. VI, No. 3.

’o4.—Prof. Valentin Haecker’s Critical Review on Bastardization and

Formation of the Sex Cells. Zool. Anz., Bd. XXVII, No. 20, 21.

Moore, J. E. S., ’95.—On the Structural Changes in the Reproductive Cells
during the Spermatogenesis of Elasmobranchs. Quart. Journ. Micr.
Sci., N. S. 38.

NEKRASSOFF, A., ’93.—Untersuchungen tiber die Reifung und Befruchtung des
Eies von Cymbulia Peronii. Anat. Anz., Bd. XXIV.

ScHOCKAERT, RUFIN, ’04.—La Fécondation et la Segmentation chez le Thysano-
zoon broschi. La Cellule, t. XXII.

ScHREINER, A. and K. E., ’o04.—Die Reifungsteilungen bei den Wirbeltieren.
Ein Beitrage zur Frage nach der Chromatinreduktion. Anat. Anz.,
Bd. XXIV, No. 22.

DE SINETY, ’o1.—Recherches sur la biologie et l’anatomie des Phasmes. La
Cellule, t. XI.

STEPHAN, P., ’o2.—Sur quelques points relatifs 4 l’évolution de la vésicule
germinative des Téléostéens. Arch. d’anat. Microscopique., T. V.

SuTTon, WALTER S., ’o2,—On the Morphology of the Chromosome Group in
Brachystola Magna. Bio. Bull., Vol. IV.

TELLYESNICZKY, K., ’o2.—Zur Kritik der Kernstrukturen. Arch. f. mik. Anat.
u. Entw., Bd. LX.

VEJDOVSKY, FR., ’88-’92,—Entwicklungsgeschichtliche Untersuchungen. Prag.

VEJDOVSKY, FR., UND MRAZEK, ’03.—Umbildung des Cytoplasma wahrend der
Befruchtung und Zellteilung. Nach den Untersuchungen am Rhynchel-
mis-Hie. Arch. f. mik. Anat., Bd. LXII.

Vorinoy, D. M., ’03.—La spermatogénése d’été chez le Cybister Roeselii. Arch.
Zool. exper. et générale, Vol. I.

Vom Ratu, O., ’92.—Zur Kenntniss der Spermatogenese yon Gryllotalpha vul-
garis Latr. Arch. f. mik. Anat., Bd. XL.

236 First Maturation Spindle of Allolobophora Feetida

WHEELER, WILLIAM MorTON, ’97,—The Maturation, Fecondation and Early
Cleavage of Myzostoma Glabrum Leuckart. Arch. de Biol., T. XV.

Witson, E. B., ’96’00.—The Cell in Development and Inheritance.

’o4.—The Germ-regions in the Egg of Dentalium. The Journ. Exper.

Zool., Vol. I.

EXPLANATION OF PLATHS.

All the photographs are of the egg of Allolobophora fetida and were taken
by the method described in Zeit. f. wiss. mik., Bd. XVIII, 1901, “A new method
of focusing in photo-micrography,”’ Foot and Strobell.

The Zeiss apo. 2 mm. immers. lens 140 apr. and compensating ocular 4
were used, and for the one thousand magnification a camera draw of 25%
inches.

Many of the photographs were taken on the Seed No. 27 and the Agfa Isolar
plates. With these rapid plates an exposure of only one to three minutes was —
required even on cloudy days. In the photographs of Plates VII, VIII and IX
the wider area of accurate focus demanded by the subjects made it necessary
to close the substage diaphragm to a point requiring doubling the time of
exposure.

Only a small part of each section is reproduced, except in photos. 68 and 69.

The reproductions of Plates I, IV and VI are by the half-tone process, and
Plates II, III, V, VII, VIII and IX are by the Rotograph process.

PLATE I.

The oocytes shown in Photos. 1 to 11 inclusive were fixed in platino-osmic
(Hermann’s fluid without acetic acid) and stained with iron hematoxylin fol-
lowed by dilute Bismarck brown. All the sections are 242

PHoto. 1. 1000. First of four consecutive sections of a germinal vesi-
cle of an oocyte first order from a receptaculum ovorum, showing chromatic
filaments stained black, and faintly stained nucleoplasm.

PuHoto. 2. * 1000. Next section to above.

Puoro. 3. 1000. Next section to Photo. 2 showing an accessory nucleo-
lus stained black like the chromatic filaments. This is the only accessory
nucleolus in this germinal vesicle.

Puoto. 4. 1000. Next section to above showing the principal nucleolus
which stained yellow by the Bismarck brown. This nucleolus is also present
in the two following sections.

Puoro. 5. 1000. Section of a germinal vesicle of an odécyte first order
from the receptaculum ovorum, showing black chromatic filaments, and
large yellow nucleolus with small black granules (these granules are lost
in some of the reproductions). This section was first photographed unstained
and the granules in the nucleolus were as black and sharply differentiated as
in the stained preparation.

Puoro. 6. 1000. First of four serial sections of a germinal vesicle of
an oocyte first order from receptaculum ovorum showing black chromatic fila-
ments and one accessory nucleolus, also stained black.

Puoro. 7. 1000. Second section from above showing the principal
nucleolus. This nucleolus stained brownish yellow and is present in the two
following sections.

“~

Cos)
(SG)
~+

Katharine Foot and E. C. Strobell

PuHoro. 8. x 1000. Next section to Photo. 7 showing the same large nucleo-
lus and a second accessory nucleolus stained black, at the upper left hand
periphery of the germinal vesicle.

Puoro. 9. x 1000. Second section from Photo. 8 showing a third accessory
nucleolus which in this section is attached to one of the chromatic filaments.

Puoro. 10. x 1000. Section of an odcyte first order from a receptaculum
ovorum, at the same stage of development as those shown in Photos. 1 to 9.
The archoplasmic granules are aggregated at two poles of the odcyte bearing
a striking resemblance to the polar rings of the ripe egg. Only one pole is
shown in the photograph.

PuHoro. 11. X* 1000. Section of germinal vesicle of an o6écyte first order
from a receptaculum ovorum, showing black chromatic filaments and one
accessory nucleolus attached to the longest filament. This was the only acces-
sory nucleolus in this germinal vesicle, and the principal nucleolus stained
brownish yellow as in the other o6écytes. ;

Puoro. 12. x 710. Young odcyte from ovary showing the large vacuolated
nucleolus and well defined areas of archoplasm in the cytoplasm. On the
periphery of the photograph is shown part of a very small odcyte with aggre-
gations of archoplasm (yolk nucleus) at opposite poles of the germinal vesicle.
Fixative, corrosive sublimate (saturate). Stain, iron hematoxylin.

Puoro. 138. 1000. Part of periphery of odcyte first order from a receptac-
ulum ovorum. The two membranes, with an indication of the substance
between, are distinctly shown. The archoplasm is fused with the cytoplasm
in a form characteristic of chromo-acetic fixation, c. f. Photo. 10 for platino-
osmic fixation. Fixative, chromo-acetic. Stain, iron hematoxylin followed
by dilute Bismarck brown. .

Puoro. 14a. x 1000. Nucleolus with black granules in large unstained
oocyte from ovary. The granules resemble the black granules found in the
cytoplasm of the same preparation. Fixative, chromo-acetie followed by osmic
acid, which blackened the granules of the nucleolus and cytoplasm.

Puoro. 14b. 1000. Nucleolus in a large unstained odcyte from distal
end of ovary. Fixative, Hermann’s fluid.

Puotos. 15, 16 and 17. x 700. Vacuolated nucleoli from large unstained
oocytes from ovary. Fixative, Hermann’s fluid.

Proto. 18. x 700. Vacuolated nucleolus in medium sized unstained odcyte
from ovary. Fixative, Hermann’s fluid.

Puoros. 19, 20 and 21. 700. Vacuolated nucleoli in large unstained
oocytes from ovary. Photo. 21 shows both a principal and accessory nucleo-
lus. Fixative, picro-nitric followed by osmic acid.

PLATE II.

The section of Photo. 26 is 2. All the others are 2% up.

PuHotos. 22-25. 1000. Four sections selected from fourteen of a germinal
vesicle of an oocyte first order from a receptaculum ovorum. The principal
nucleolus is present in three consecutive sections, two of them being shown in
Photos. 22 and 23. Photo. 23 shows one accessory nucleolus attached to the
largest chromatic filament. Fixative, Rabl’s picro-sublimate. Stain, iron
hematoxylin, followed by dilute Bismarck brown.

17

238 First Maturation Spindle of Allolobophora Feetida

PuHoro. 26. X about 1100. Section of upper pole of a first maturation
spindle with sphere and centriole in fertilized odcyte from freshly deposited
cocoon. There is a centriole at the lower pole of this spindle, but we selected
this photograph as more interesting for reproduction, because it is much more
difficult to demonstrate a centriole at the upper pole, when it is close to the
periphery. Fixative, Flemming’s fluid without acetic acid. Stain, iron hem-
atoxylin, followed by dilute Bismarck brown.

PuHotos. 27 and 28. xX 1000. Two sections of the same germinal vesicle of
an oocyte first order from a receptaculum ovorum. There are four intervening
sections between the two reproduced here and the principal nucleolus of
Photo. 27 shows no connection with the chromatic filaments. Fixative,
platino-osmic. Stain, iron hematoxylin followed by dilute Bismarck brown.

Puoro. 29. 1000. Section of a germinal vesicle of an odcyte first order
from a receptaculum ovorum, showing the large vacuolated nucleolus and
the smaller accessory nucleolus. Fixative, Picro-sublimate. .Stain, iron
hematoxylin, followed by dilute Bismarck brown.

PuHotTo. 30. 1000. Section of a germinal vesicle of an odcyte first order
from a receptaculum ovorum showing chromatic filaments in transverse
section and two short granular filaments crossing each other. The large
vacuolated nucleolus is in the 3rd, 4th and 5th sections from that of Photo. 30.
Fixative, corrosive-acetic (5 per cent acetic). Stain, iron hematoxylin fol-
lowed by dilute Bismarck brown.

PuHoto. 31. 1000. Section of a germinal vesicle of an oocyte first order
from a receptaculum ovorum showing the large vacuolated nucleolus and an
early stage of the segregation of the chromatin into chromatic filaments.
Fixative, corrosive sublimate. Stain, iron hematoxylin.

PuHotTo. 32. 1000. Section of lower pole of second maturation spindle
showing four of the eleven chromosomes approaching the lower pole. This
section was photographed to show a precocious formation of one of the
eleven vesicles not due normally until the telophase. The centriole is in the
next section. Fixative, chromo-acetic. Stain, iron hematoxylin.

PuHoro. 33. 1000. Section of first maturation spindle, showing five of
the eleven chromosomes. This egg is one of eight found in an immature
cocoon encircling the clitellum of a worm. The four chromosomes which are
on the same plane are clearly defined and show a decided difference in size.
We have six photographs of this spindle showing all the eleven choromosomes
but cannot spare space for more than one reproduction. Fixative, corrosive
acetic (10 per cent acetic.) Stain, iron hematoxylin, followed by dilute
Bismarck brown.

Puoto. 34. 1000. Section of a germinal vesicle of an odcyte first order
from a receptaculum ovorum, showing the typically injurious effect of
chromo-acetic on the nucleoplasm. This photograph shows a distinct net knot
resembling a large nucleolus with two vacuoles, but it is unmistakably an
artefact, for the principal nucleolus of this egg is present in the seventh sec-
tion from the one reproduced in this photograph. Stain, iron hematoxylin
followed by dilute Bismarck brown.

PuHoro. 35. 710. Section of a vacuolated nucleolus of an oocyte from the
Ovary. Fixative, corrosive sublimate followed by osmic acid. Unstained.

Katharine Foot and E. C. Strobell 239

Puoro. 36. 710. Section of a ring nucleolus of an oédcyte from the
ovary. Fixative, Lindsay Johnson’s fluid. Stain, acid fuchsin.

Puoro. 37. X 710. Section of a vacuolated nucleolus. A small accessory
nucleolus is also present. Fixative and stain same as Photo. 36.

PuHoro. 38. 710. Section of a ring shaped nucleolus of an odcyte from
the ovary. Fixative, Merkel’s fluid followed by osmic acid. Stain, iron
hematoxylin.

Puoro. 39. 710. Section of a vacuolated nucleolus from an ovarian
oocyte and near it an accessory nucleolus. WFWixative, picro-sulphuric followed
by osmic acid. Unstained.

Puoto. 40. 710. Section of a vacuolated nucleolus from an ovarian
odcyte and near it a small accessory nucleolus. Fixative, picro-sulphuric
followed by osmic acid. Unstained.

PuHoto. 41. 710. Section of a vacuolated nucleolus from an ovarian
oocyte. This nucleolus appears to be surrounded by a definite membrane and
this is shown in three consecutive sections. Fixative, picro-sulphuric. Stain,
iron hematoxylin.

Puoto. 42. 710. Section of a vacuolated nucleolus from medium sized
ovarian odcyte. Fixative, corrosive sublimate. Stain, iron hematoxylin.

Puoto. 43. 710. Section of a ring shaped nucleolus in odcyte from a
receptaculum ovorum. Fixative, corrosive acetic (10 per cent acetic). Stain,
iron hematoxylin followed by dilute Bismarck brown.

Puoto. 44. 710. Section of aring shaped nucleolus in an ovarian odcyte.
Fixative, Graf’s picro-formalin followed by osmic acid. Stain, iron hema-
toxylin.

Puoro. 45. 710. Seetion of a crescent shaped nucleolus in ovarian oocyte.
Fixative, Graf’s picro-formalin followed by osmic acid. Unstained.

PLATE III.

All sections 2% wy.

Puortos. 46 to 49. 1000. Four sections selected from eighteen of a
germinal vesicle of an oocyte first order from a receptaculum ovorum.
Photos. 46 to 48 are consecutive sections, and Photo. 49 is the third section
beyond Photo. 48 and is one of three consecutive sections in which the large
vacuolated nugleolus is present. There are two accessory nucleoli also present
in this section. All these sections show the chromatin segregated into more
or less definite filaments and Photo. 47 shows a filament with what appears to
be a longitudinal split. Fixative, saturate corrosive sublimate. Stain, iron
hematoxylin.

Puoro. 50. * 1000. Peripheral section of a germinal vesicle of an oocyte
first order from a receptaculum ovorum showing a ring shaped nucleolus.
Fixative, platino-osmic. Stain, iron hematoxylin followed by dilute Bismarck
brown.

PuHotos. 51 and 52. * 1000. Two consecutive sections of a germinal
vesicle of an oocyte first order from a receptaculum ovorum. Each section
shows a ring chromosome, the two differing greatly in size. The prnicipal
nucleolus is present in this germinal vesicle and intact. Fixative, saturate
corrosive sublimate. Stain, Bismarck brown. |

240 First Maturation Spindle of Allolobophora Foetida

Proros. 58 and 54. ™ 1000. Each of these photographs shows one of three
consecutive sections of a vacuolated nucleolus in an oocyte first order from a
receptaculum ovorum. Fixative, saturate corrosive sublimate. Stain, iron
hematoxylin.

Puoro. 55. ™* 710. Section of a nucleolus in an ovarian oocyte. Fixative,
bichromate-cupric sulphate. Stain, iron haematoxylin.

PuHoro. 56. 1000. Part of a section of a germinal vesicle of an oocyte
first order from a receptaculum ovorum showing the principal and the
accessory nucleoli. Fixative, platino-osmic. Stain, iron hematoxylin fol-
lowed by dilute Bismarck brown.

Provo. 57. * 710. Section of a vacuolated nucleolus in an ovarian oocyte.
Fixative, bichromate-cupric sulphate. Stain, iron hematoxylin.

Puoro. 58. X 1000. Section of a germinal vesicle of an oocyte first order
from a receptaculum ovorum. This is one of four consecutive sections of the
principal nucleolus. Fixative, corrosive acetic (5 per cent acetic). Stain,
iron hematoxylin followed by Bismarck brown.

Puortos. 59 to 62. 1000. Four sections selected from sixteen of a germ-
inal vesicle of an o6dcyte first order from a receptaculum ovorum. The first
three sections are consecutive. The principal nucleolus of Photos. 59 and 60
shows no connection with the chromatic filaments. Fixative, Rabl’s picro-
sublimate. Stain, iron hematoxylin followed by Bismarck brown.

Puoros. 63 to 65. * 1000. Three sections of a germinal vesicle of an
oocyte first order from a receptaculum ovorum. Photo. 63 shows an accessory
nucleolus with transverse sections of a chromosome at opposite sides of it.
The principal nucleolus is in the sixth section from the accessory nucleolus.
In Photo. 64 there is a chromosome in the form of a figure eight and in Photo.
65 a rod shaped chromosome and transverse sections of two more. Fixative,
saturate corrosive sublimate. Stain, iron hematoxylin. ;

Puoro. 66. 1000. Section of a germinal vesicle of an o6dcyte first order
from a receptaculum ovorum showing chromatic filaments and an accessory
nucleolus attached to a chromatic filament. Fixative, saturate corrosive subli-
mate. Stain, iron hematoxylin.

Puoro. 67. 710. Section of a nucleolus of an odcyte from the ovary.
Fixative, bichromate-cupric sulphate. Stain, iron hmwmatoxylin.

PLATE IV.

The section of Photo. 76 is 3 7, of Photo. 77 2 ~, and all others on this plate
are 2% wm.

Puoros. 68 to 73. 1000. Six consecutive sections of an oocyte first
order, from the receptaculum ovorum. Photos. 68 to 69 are entire sections
unstained. They demonstrate very clearly the presence of the polar-ring
substance (yolk-nucleus, archoplasm) and the deuto-plasmic granules black-
ened by the osmic in the fixative (platino-osmic). In Photos. 70 to 73 only
the germinal vesicle and a small part of the surrounding cytoplasm are
reproduced. Photo. 70 shows a group of chromosomes, Photo. 71 the principal
nucleolus and a small accessory nucleolus, Photo. 72 a large accessory nucleo-
lus and a cross shaped chromosome, and Photo. 73 a second section of the
accessory nucleolus of Photo. 72. Photos. 70 to 73 are stained with iron
hematoxylin followed by dilute Bismarck brown.

Katharine Foot and E. C. Strobell 241

PHoro. 74. 710. Section of a vacuolated nucleolus of an odcyte from
the ovary. Fixative, corrosive-acetic (20 per cent acetic) followed by osmic
acid. Unstained.

Proto. 75. X 710. Section of a nucleolus with dark granules. From an
oocyte in the ovary. Fixative, picro-acetic followed by osmic acid. Unstained.

Puoto. 76. 710. Section of a small odcyte in the ovary showing the
nucleolus and yolk nucleus. Fixative, saturate corrosive sublimate followed
by osmic acid. Unstained.

PuHotTo. 77. %* about 1100. Section of a nucleolus of an oocyte from the
ovary. Fixative, Flemming’s fluid (strong). Stain, iron hematoxylin fol-
lowed by dilute Bismarck brown.

PHoro. 78. x 710. Section of a young odcyte in the ovary showing a ring
nucleolus and archoplasm (yolk-nucleus) in the cytoplasm. Fixative, saturate
corrosive sublimate. Stain, iron hematoxylin.

Puotro. 79. 1000. Section of a ring nucleolus of an oocyte from the
ovary. Fixative, Merkel’s fluid followed by osmic acid. Stain, iron hema-
toxylin. (Photo. 8C was omitted by mistake.)

PLATE V.

All sections 2% uv

Puotos. 81 and 82. x 1000. Two sections of a germinal vesicle of an
oocyte first order from the receptaculum ovorum. This vesicle was cut into
nineteen sections and the two reproduced here were fifteen sections apart and
both very close to the periphery of the germinal vesicle; the next section on
each side being the last to show the germinal vesicle. These photographs
show the earliest appearance of the centrioles and asters at opposite poles of
the germinal vesicle. The centriole in Photo. 81 is still in contact with the
membrane of the vesicle and the centriole in 82 is very close to the membrane.
Fixative, chromo-acetic. Stain, iron hematoxylin followed by dilute Bismarck
brown.

Puoro. 83. 1000. Section of a germinal vesicle of an odcyte first order
from the receptaculum ovorum. The centriole and aster are further developed
than those in Photos. 81 and 82 and not so close to the membrane of the
germinal vesicle. The other centriole is eleven sections from the one in this
photograph. Fixative, Boveri’s picro-acetic. Stain, iron hematoxylin fol-
lowed by dilute Bismarck brown.

Puortos. 84 to 89. x 1000. Consecutive sections of a germinal vesicle of
an oocyte first order from a receptaculum ovorum. 86 and 87 are photographs
of the same section, 86 being taken on the plane of the small centriole, and 87
on a lower plane to show the rest of the chromosomes. The two centrioles are
seen at opposite poles in Photos. 84 and 86. There are two accessory nucleoli
in Photo. 89. Traces of the persisting membrane of the germinal vesicle are
seen in nearly all the sections. Fixative, Rabl’s picro-sublimate. Stain, iron
hematoxylin followed by dilute Bismarck brown.

Puotos. 90 and 91. * 1000. Two planes of the same section of a first
maturation spindle in an odcyte from a receptaculum ovorum, showing four of
the eleven chromosomes. Both centrioles are in this section but not on the
same plane as the two chromosomes of Photo. 90. One centriole is shown in
Photo. 91, which was taken before staining and focused for the unstained

242 First Maturation Spindle of Allolobophora Feetida

centriole at the upper pole, the centriole of the lower pole was not on the same
plane. Photo. 91 shows two unstained chromosomes on a different plane from
the two in Photo. 90. Fixative, platino-osmic. Stain, Photo. 90 iron hema-
toxylin followed by dilute Bismarck brown. Photo. 91, unstained.

Puotos. 92a and b. 1000. Two consecutive sections of a first matur-
ation spindle of a fertilized oocyte in a freshly deposited cocoon. We have
twelve photographs of this spindle showing all the eleven chromosomes and
both centrioles. In the two selected for reproduction both centrioles are
shown and a few of the chromosomes. 92a was focused for the centriole,
sacrificing a sharp definition of the chromosome, and 92b was focused for the
peripheral centriole, only one of the four chromosomes being on the same
plane. Fixative, chromo-acetic. Stain, iron hematoxylin.

PHotos. 93 to 97. 710. Sections of nucleoli of odcytes from ovary. Fix-
ative, Rabl’s picro-sublimate. Stain, iron haematoxylin.

Puotos. 98a and b. X 710 and 1000. Section of a nucleolus of an odcyte
from the ovary. Fixative, Platino-osmic. 98a unstained. 98b stained with
iron hematoxylin.

PLATE VI.

The sections in Photos. 99 and 105 are 3, those of 102 and 106 are 5 y, all
the others are 244 nu.

Puoto. 99. 710. Unstained section of an unfertilized odcyte from a
freshly deposited cocoon showing the first maturation spindle at the meta-
phase, with two of the eleven chromosomes. The peripheral centriole is not
visible in this unstained section, but the centriole at the lower pole of the
spindle could be clearly seen. Fixative, platino-osmic.

Puoro. 100. x 1000. Section of an unfertilized odcyte from an immature
cocoon still encircling the clitellum of the worm. This photograph shows the
peripheral pole of the first maturation spindle with the centriole and thread-
like rays typical of chromo-acetic preparations. The two membranes of the
ege with the substance between are clearly defined, and this is shown also
in Photo. 99. Fixative, chromo-acetic. Stain, iron hematoxylin followed by
dilute Bismarck brown.

Puoro. 101. x 1000. Section showing the centriole at lower pole of a first
maturation spindle of an odcyte from the receptaculum ovorum. Fixative,
Boyeri’s picro-acetic. Stain, iron hematoxylin followed by dilute Bismarck
brown.

Puoro. 102. X 710. Section showing the centriole at peripheral pole of
a first maturation spindle of an odcyte from the receptaculum ovorum. Fix-
ative, chromo-acetic. Stain, iron hematoxylin.

PuHoto. 103. »X 1000. Section showing the centriole at the lower pole of a
second maturation spindle of a fertilized odcyte from a cocoon. Fixative,
chromo-acetic. Stain, iron hematoxylin.

Puoto. 104. 1000. Section of the second maturation spindle of a
fertilized odcyte from a cocoon, showing the centriole at the lower pole, but
not in the centre of the sphere. Fixative, chromo-acetic. Stain, iron hema-
toxylin.

Puoto. 105. X about 1000. Section showing the centriole at the lower pole
of a second maturation spindle of a fertilized odcyte from a cocoon. Fixative,
saturate corrosive sublimate. Stain, iron h#matoxylin.

Katharine Foot and E. GC. Strobell 243

Puoro. 106. x 1000. Section showing the centriole at the peripheral pole of
a first maturation spindle, of an odcyte from the receptaculum ovorum. Fix-
ative corrosive acetic (10 per cent). Stain, iron hematoxylin followed by
dilute Bismarck brown.

Puoto. 107. x 1000. Section of the spindle of one of two small cells of a
three celled egg from the cocoon, showing one of the two centrioles. Fixative,
chromo-acetic. Stain, iron hematoxylin.

Puoro. 108. 710. Section showing the centriole at lower pole of the
second maturation spindle of a fertilized odcyte from a cocoon. Fixative,
Merkel’s fluid. Stain, iron hematoxylin.

Puoto. 109. X about 1100. Section showing centriole at the lower pole of
a second maturation spindle of a fertilized odcyte from a cocoon. Fixative
chromo-acetic. Stain, iron hematoxylin.

PuHoro. 110. X 710. Section of a second cleavage spindle of an egg from
a cocoon. Both centrioles are shown though they are not on exactly the same
plane. Fixative, chromo-acetic followed by osmic acid. Stain, iron hema-
toxylin.

PLATES VII, VIII and IX.

The preparations shown in the photographs of these plates are from odcytes
from the receptaculum ovorum (except Photo. 128), and were obtained by the
method described on p. 200.

The preparations were stained with Bismarck brown.

Magnification of all the photographs, 1000 diameters.

Puotos. 111 and 118. Two germinal vesicles each showing a principal nu-
cleolus, one accessory nucleolus and a fine chromatin thread.

Puoro. 112. Germinal vesicle showing a pathological aggregation of the
chromatin.

PuHotos. 114 and 115. Two germinal vesicles showing each a principal nucle-
olus and a longitudinally split spireme. In Photo. 114 there is one, and in
Photo. 115 there are two accessory nucleoli.

PuHoTOS. 116 to 124. Germinal vesicles showing different forms of the eleven
bivalent chromosomes immediately after the transverse division of the spi-
_ reme, in Photo. 116, two of the bivalent chromosomes are still attached end to
end. Nearly all these preparations show a longitudinal split in some of the
chromosomes, and in Photos. 116, 118, 121, 122, 123 and 124 both the principal
and the accessory nucleoli are present.

PuHotos. 125, 126, 127, 129 and 130 show the eleven bivalent chromosomes
arranged with more or less regularity in the equator of the first maturation
spindle. For detailed description of these photographs, see p. 217.

PuHorTo. 128. The eleven chromosomes nearly in the equator of the first
maturation spindle of an odcyte from an immature cocoon.

_

pe

PROPHASES OF FIRST MATURATION SPINDLE OF
ALLOLOBOPHORA FOETIDA.

~ eg <2

a

AMERICAN JOURNAL OF ANATOMY--VOL. IV.

FOOT & STROBELL PHOTOS.

PLATE |.

PROPHASES OF FIRST MATURATION SPINDLE OF PLATE
ALLOLOBOPHORA FOETIDA. FOOT & STROBELL PHOTOS.

ERICAN JOURNAL OF ANATOMY--VOL

PROPHASES OF FIRST MATURATION SPINDLE OF PEATE Stil.
ALLOLOBOPHORA FOETIDA. FOOT & STROBELL PHOTOS

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PROPHASES OF FIRST MATURATION SPINDLE OF

ALLOLOBOPHORA FOETIDA.

AMERICAN JOURNAL OF ANATOMY--VOL. Iv.

FOOT & STROBELL PHOTOS:

a

PLATE IV.

=

PROPHASES OF FIRST MATURATION SPINDLE OF
ALLOLOBOPHORA FOETIDA.

PLATE V.
FOOT & STROBELL PHOTOS-

7
RY

we 1’
re

AMERICAN JOURNAL OF ANATOMY--VOL. tv

¥ aae Lee*

oe oe

| ee) Avi vy oe.
Pe Tha:

PROPHASES OF FIRST MATURATION SPINDLE OF PLATE VI.

ALLOLOBOPHORA FOETIDA.

FOOT & STROBELL PHOTOS.

*

<2

ea

7 ‘

~.

at ie RN Sg,
Cah

AMERICAN JOURNAL OF ANATOMY--VOL. IV.

a a

ROPHASES.OF FIRST MATURATION SPINDLE OF 5 PLATE VII
ALLOLOBOPHORA FOETIDA FOOT & STROBELL PHOTOS

AMERICAN JOURNAL OF ANATOMY--VOL

PROPHASES OF FIRST MATURATION SPINDLE OF

PATE Witt
ALLOLOBOPHORA FOETIDA.

FOOT & STROBELL PHOTOS

AMERICAN JOURNAL OF ANATOMY VOL. IV

PEATE Ix

PROPHASES OF FIRST MATURATION SPINDLE OF

FOOT & STROBELL PHOTOS

ALLOLOBOPHORA FOETIDA.

Pe |
-. Gee, j 126

NAL OF ANAT

GENETIC INTERPRETATIONS IN THE DOMAIN OF
ANATOMY.*

PRESIDENTIAL ADDRESS BEFORE THE ASSOCIATION OF AMERICAN
ANATOMISTS.
BY

CHARLES-SEDGWICK MINOT, LL. D., D. Sc.

The science of anatomy, although one of the oldest of all sciences, was
long neglected in’ America, and taught only in a routine fashion by
professors who had little or no thought for the promotion of the science
or any aim higher than teaching a certain number of established facts
in gross anatomy to the maximum possible number of students. Within
the last generation the few pioneers of anatomy have been succeeded by
teachers, many of whom share the highest ideals of anatomical science,
and have contributed important discoveries by which it has been really
advanced. Our Society is at once the symbol and the outcome of, these
comparatively new conditions in America, and we have as our duty not
only actively to encourage research, to spread anatomical knowledge, and
to earn appreciation of anatomy as a living science, but also to exert a
missionary influence by which the dignity and vitality of our science
shall be brought to recognition at all our universities.

* The following recent or new technical terms are used in the course of the
address and are recommended for adoption.

Cytogenic glands, false glands which produce cells, as for example, the
lymph and genital glands.

Cytomorphosis, to designate comprehensively all the structural modifica-
tions which cells or successive generations of cells may undergo from the
earliest undifferentiated stage to their final destruction.

False glands, all glands, which develop without ducts.

Lympheum, a more or less definitely circumscribed area consisting of cellu-
lar reticulum, the meshes of which are charged with leucocytes and are in
direct communication with lymph-vessels or more rarely with blood-vessels.
It is a site for the multiplication of leucocytes.

Mesepatium, the membrane (French, méso) extending from the stomach
and duodenum to the median line of the ventral abdominal wall, and in
which the liver develops. It comprises a dorsal mesepatium (lesser omen-
tum) and ventral mesepatium (falciform or suspensory ligament).

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

246 Genetic Interpretations in the Domain of Anatomy

It is sometimes said, and perhaps more often thought, that anatomy
is a completed science. This assertion is based upon the thoroughness
and exhaustive character of the descriptions to be found in our text-
books of the anatomical conditions in the human adult; yet even as
regards the organization of the adult we have still much to learn, espe-
cially concerning the microscopic structure with which we are still very
imperfectly acquainted.

But anatomy is not alone a descriptive science. It is also comparative
and genetic. In both these directions its development is very far from
complete, and a vast amount of original research must still be completed
before comparative anatomy and embryology shall have approached any-
where near even the present perfection of descriptive human anatomy.

To embryological research must be attributed a large part of the
extraordinary progress which anatomy has made during the last twenty-
five years. By embryology we have gained a far deeper understanding
of all anatomical forms, we have acquired new interpretations for path-
ological facts, and we have secured for the first time some clear insight
into the essential structure of the brain. I need not do more than
allude to these achievements, since they are familiar to us all, and have
most profoundly affected our anatomical conceptions. Our point of view
has changed, and we interpret the anatomy of the adult in terms of the
genesis of the organs and tissues during their embryonic development.

Perhaps no man has contributed so much towards this result as the
great Leipzig anatomist, Wilhelm His, whose death this year we have
to lament. He was a great master. He had full command over the prob-
lems of anatomy and contributed in the richest measure to their solution.
His influence in America has been especially strong and widespread, and
has certainly had much to do in bringing about the progress of anatomy
in this country, which we are seeking to maintain, and if possible, in-
crease. In what I have to say to you on this occasion, you will perceive

Phrenic area, the area on the superior or cephalad surface of the liver, by
which the liver is attached permanently to the diaphragm. It includes
the whole of the territory of the coronary and triangular “ligaments,” so-
called in current text-books.

Sinusoid, an irregular blood space, produced by the subdivision of a larger
blood-vessel by the ingrowth of the parenchyma of an adjacent organ.

Structural unit, the territory of an organ supplied by a single terminal
branch of an afferent vessel (artery or vein); the volume of such a unit
is often only 10-20 cubic millimeters.

Trophoderm, the special layer of cells formed on the exterior of the
young mammalian blastocyst, and serving to secure the implantation of the
ovum in the uterus.

Charles-Sedgwick Minot 247

the influence of His clearly, and I cannot let the opportunity pass of
expressing publicly my gratitude and admiration for the greatest anat-
omist of his time.

Although embryology has already contributed in so ample measure to
the promotion of our science, we are still far from having accepted all the
enlightenment which she offers us. With your permission I will try
to present to you certain embryological aspects of anatomy, the character
of which I have sought to indicate in the title of this address, by the
words “ genetic interpretations.”

First of all, let us consider the subject of cytomorphosis. This word
I proposed in 1901* “to designate comprehensively all the structural
modifications which cells or successive generations of cells may undergo
from the earliest undifferentiated stage to their final destruction.” As
stated on that occasion it is convenient, though somewhat arbitrary, to
distinguish four fundamental successive stages of cytomorphosis. These
stages are (1) the undifferentiated stage: (2) the stage of progressive
differentiation, which itself often comprises many successive stages; (3)
the regressive stage or that during which degeneration or necrobiosis
occurs; and (4) the stage of the removal of the dead material. The gen-
eral data on which the conception of cytomorphosis is based have been
briefly put together also in my text-book of embryology, and it seems
therefore superfluous to dwell upon them at length in addressing you.

I cannot of course claim any greater originality in the establishment
of the conception of cytomorphosis than is implied by the definite formu-
lation of the ideas upon which it is based. These ideas have been
gradually gathered as the fruit of numerous investigations in histogenesis.
The mentioned investigations have made us all familiar with the con-
ception of undifferentiated embryonic cells, with the gradual progress of
differentiation in the cells during the embryonic, foetal, and even post-
natal periods; and have also made us acquainted with various examples
of degeneration and atrophy occurring in the course of development, both
before and after birth. Up to the time when I proposed the term there
had been, so far as I know, no attempt to survey all this array of facts
from a single unifying point of view. But such a point of view is, I
believe, well calculated to render our notions more precise as to many
processes of development, and to afford us at the same time the practical
benefit of being able to present the facts of histogenesis in our teaching
in a way, which is very advantageous, because it facilitates in the stu-
dent’s mind the establishment of a real insight into the general course of
development by emphasizing principles of very wide application. To me,
at least, it seems that the conception of cytomorphosis should be made the

248 Genetic Interpretations in the Domain of Anatomy

foundation of all our instruction in anatomy, and that its importance
should be constantly emphasized in our class-rooms and that when good
illustrations of cytomorphosis are encountered by the student, his atten-
tion should be especially directed to them, so that he may become familiar
with the conception. Let me mention a few illustrations which I have
found serviceable in teaching.

But first I must call your attention to an aspect of cytomorphosis, which
has not hitherto, so far as my knowledge goes, been sufficiently empha-
zised. We may distinguish two fundamental phases. During the first,
cell division occurs, during the second, cell division does not occur. Dur-
ing the first phase we may find a progressive alteration, which gradually
takes place in successive generations of cells, but apparently the amount of
differentiation which can occur while cells retain the power of active
division is comparatively slight. During the second phase, since the cell
no longer divides, the alteration takes place in the single cell, and the
alteration, which occurs under these conditions, is typically great and may
be best designated by the term final differentiation, differentiation being
here held in our minds clearly distinct from degeneration. By final
differentiation we mean the establishment of that special organization
of a cell, which brings to perfection the specialized physiological function
for which the cell is destined. Thus the alteration of a mesenchymal cell
into a muscle fibre is its final differentiation, and establishes the physi-
ological perfection of that cell as a contractile element. Beyond the
final differentiation of the cell comes the series of degenerative changes.
A comprehensive study of cell degeneration is yet to be made, nevertheless
we can already say that, although cell degeneration is chiefly character-
istic of the second phase of cytomorphosis, which is also characterized
by the cessation of cell division, yet the degeneration may be initiated
before the power of cell division is lost and the degenerative change in the
cells may go on while they are still proliferating; but typically it seems
rather that degeneration belongs to the second phase of cytomorphosis,
and this seems to be alike true for necrosis and atrophy, that is to say,
simple cell death, and for necrobiosis, that is to say, cell death preceded
by structural changes, which we know commonly under the name of
hypertrophic degeneration.

Let us pass on now to a few illustrations of cytomorphosis:

The first to which I would direct your attention is afforded by the
formation of the “ trophoderm.” This is a new term which I have
recently brought forward to designate the special layer of cells formed
apparently from the ectoderm (or according to Assheton’s theory, from

Charles-Sedgwick Minot 249
the entoderm) which serves to secure the implantation of the mammalian
ovum in the walls of the uterus. In my “ Text-Book of Embryology ”
I have figured these cells from the human ovum and applied to them
the term trophoblast, but as Professor Hubrecht, who introduced this
last term into science, has objected to this application of it, it has been
necessary to introduce a new term, hence the designation trophoderm.
It corresponds in large part, perhaps wholly, to that which Duval desig-
nated as the ectoplacenta. It is the first tissue in the mammals to be
distinctly differentiated. The cells by their large size, distinct bound-
aries, and characteristic nuclei, are readily distinguished from any other
cells existing in the embryo at the time the trophoderm is differentiated.
Very soon after the development of the trophodermie cells, a large part of
them begin to complete their cytomorphosis by undergoing degeneration
and resorption. By their disappearance, as I have elsewhere pointed out,
the intervillous spaces arise. The trophoderm therefore is not only the
earliest tissue to be specialized in the development of mammals, but also
the earliest tissue to absolutely complete its cytomorphosis.

Another striking illustration of the eytomorphic cycle with its phases
of differentiation, degeneration, and disappearance of cells is offered
to us by the blood corpuscles. The first blood corpuscles are cells with a
minimum amount of protoplasm. The cells then proceed to grow, and
as they grow, differentiate themselves in part at least, into red blood
globules. In mammals there follows the stage, degenerative in character,
by which the nucleus of these red blood corpuscles disappears. ‘The man-
ner of its disappearance is, to be sure, still perhaps a matter of debate,
but for us for the moment is of minor importance. After the degen-
eration or disappearance of the nucleus, the blood corpuscles are destroyed
and, having completed their cytomorphosis, are replaced by fresh ones.

A third admirable illustration is offered us by cartilage, and a fourth
by bone. In cartilage we see at first a differentiation of simple mesen-
chymal cells which then enlarge, becoming the characteristic cartilage
cells. When ossification of the cartilage occurs we can easily follow the
hypertrophic degeneration of these same cartilage cells, which has been
so much studied that good accounts of the enlargement and breaking down
of these cells preliminary to the ingrowth of the osteogenetic tissue can be
found in all the better text-books of histology; but I regret to say I do
not recall any text-book either of anatomy, histology, or embryology,
which points out the fact that this succession of changes in cartilage cells
is a typical and almost perfect illustration of cytomorphosis. Almost the
same can be said of bone, for in the formation of this tissue also we have
first, the differentiation of the mesenchymal cells into osteoblasts, which

250 Genetic Interpretations in the Domain of Anatomy

are always of larger dimensions than the cells from which they arise;
and after these osteoblasts have become bone cells they cease their devel-
opment and apparently degenerate. I have to say apparently, because, so
far as I know, the fate of the bone corpuscle has not been ascertained
with certainty. We risk but little, however, in asserting that the bone
cells also offer an instance of a normal, complete cytomorphosis.

As a fifth and last illustration, let us choose the epidermis, in which
we have a distinct type of differentiation. In the basal layer are the cells,
which divide and produce, according to our present notions, all of the cells
of the epidermis. When the basal layer cells divide, however, some of
them only, pass immediately through further cytomorphic changes in
order to make first, the cells of the mucous layer, and later, by under-
going cornification, to constitute the horny layer. Others of the basal
cells remain members of the basal layer and continue to proliferate. We
thus see the progeny of the original basal cells divided into two classes:
the cells of one class pass on in their development, the others retain their
ancestral type. In the epidermal cells we observe as in other instances
of eytomorphosis, first the enlargement and differentiation of the cells,
here occurring in the mucous layer, and later their degeneration or corni-
fication followed by their necrosis and destruction.

It would be easy to multiply these illustrations. All of you could
supply more. That which I would urge upon your consideration is the
value of the cytomorphic interpretation in explaining the origin and
differentiation of tissues in the light of the broadest principle of cellular
development which we have up to the present time been able to establish.

I will now ask you to consider certain possible genetic classifications.
The most fundamental and important of these seems to me to be that of
tissues and of organs according to the germ layers from which they arise.
This classification was made the basis of his entire course of lectures upon
animal morphology by Professor Carl Semper, the Wiirzburg zoologist,
under whom I had the pleasure of studying in 1875-76. It is not merely
very practical and advantageous alike to teacher and pupil, but is also
the only thoroughly scientific classification of structures and organs which
we can adopt. No other classification should, in my judgment, be seri-
ously considered. So firmly do I hold this conviction that I greatly
deplore the fact that our text-books of histology are not written upon
an embryological basis, the lack of which deprives them of much of the
scientific character and value they ought to have.

As our knowledge of the development from the germ layers has grown,
we have learned with ever-increasing certainty that each germ layer
has its specific réle to play. Each germ layer produces its own specific

Charles-Sedgwick Minot 251

set of tissues, which are not duplicated by the tissues of any other germ
layer. I have already pointed out on another occasion that the import-
ance of the germ layers is as absolute and unvarying in the domain of
pathology as in normal differentiation; I need not dwell on that aspect
of the question now, but will only repeat the declaration of my belief that
the entire teaching of the pathologist as well as of the histologist and
anatomist should be based on the doctrine of the germ layers and their
specific réles in histogenesis.

Almost any group of tissues would offer a favorable opportunity for the
discussion of genetic classification. We may select those which are
differentiated from the embryonic mesenchyma and which are commonly
grouped in the adult under the names of the connective and supporting
tissues. It is almost superfluous, so much is the genetic point of view
neglected, to call your attention to the fact that in our current text-books
of histology there is often little or nothing which would enable the student
to grasp the relations of these tissues to one another or to understand
their genetic relationships. It is true that our knowledge in spite of the
great advances of recent years is still too incomplete to justify our assert-
ing that the classification which we can now make is final. Nevertheless
we can already perhaps attain approximate finality. A very great step
in advance was made when the character of the cellular reticulum was
established and it was shown that this tissue is different from ordinary
connective tissue. It has two principal characteristics: first, the matrix
or intercellular substance is nearly or absolutely fluid, so that leucocytes
can wander freely in the intercellular spaces of the reticulum; second,
the network of original protoplasmic filaments has become directly con-
verted into a network preserving more or less the original form, but con-
sisting not of protoplasm, but of a new chemical substance, reticulin.
Where cellular reticulum is developed, as for instance in the so-called ade-
noid tissue, there may be formed from the cells a minimum amount of
connective tissue fibrils and of elastic substance, but if we may judge
from our present knowledge the cells, which have produced reticulin,
preserve but a very small capacity for the production of other elements
of connective tissue. Hence, it seems to me that we may well put cellular
reticulum in a class by itself, quite apart from the true connective tissues
in which the intercellular substance is not mainly fluid and in which there
is an abundant development of fibrillar or elastic substance, or of both,
and in which, further, reticulin is nearly if not wholly absent. We shall
thus come to place all the connective tissues, properly so-called, in a second
genetic group. When we follow in the embryo the history of young con-
nective tissue, we learn that it undergoes two principal kinds of modifica-

252 Genetic Interpretations in the Domain of Anatomy

tions, those affecting the matrix, and those affecting the cells. On these
differences the classification in the following table is based.

We also know that connective tissue can be directly transformed. into
cartilage, which, therefore, unquestionably belongs in the same second
genetic group as the true connective tissues. As regards bone, I find it
somewhat difficult to reach a decision, but incline to the conclusion that
bone should be regarded as distinct from the true connective tissue, thus
making a third genetic division of the tissues derived from the primitive
mesenchyma. This conclusion appeals to me partly as a protest against
the absurd, though long established and honored, custom of separating
cartilage from connective tissue, and putting cartilage and bone together
in a common group under the head of supporting tissues. The following
table presents the proposed classification in a form which you can easily
follow:

TABLE I.—THE MESENCHYMAL TISSUES.

Cellular reticulum ( Mucous tissue
Matrix specialized.+ Adult connective tissue
Cartilage

GESS ULC ie aie eves erzvasaiei ol Syereietets

Embryonic connective
MESENCHYMA.

( Fat cells

Pigment cells

Smooth muscles

Cells specialized ...~ Basement membranes
Pseudo-endothelium

| Genital interstitial cells,
Bone L ete.

Let me refer briefly to a third and more special example of the
genetic classification of tissues, namely, that of the blood-vessels. As you
probably all know, recent embryological investigations have compelled us
to recognize not only the three familiar and long-known classes of blood-
vessels, arteries, veins, and capillaries, but also a fourth class, that of the
sinusoids. Capillaries arise as small vascular sprouts from pre-existing
vessels, and these sprouts grow in the mesenchyma. A sinusoid, on the
contrary, has an entirely different developmental history, for it is pro-
duced by the subdivision of a pre-existing and relatively large vessel.
The subdivision is accomplished by the proliferating tubules (or trabec-
ule) of an organ, which encounter a large vessel and invade its lumen,
pushing the endothelium before them.’ The endothelium of the vessel,
on the other hand, expands and spreads over the tubules (or trabecul).
By the convolutions and anastomoses thus produced, a large vessel is sub-
divided into small ones. It follows that a sinusoidal circulation is
purely venous or purely arterial. It may suffice, upon this occasion, to
point out again that the structure of many important organs, as for

Charles-Sedgwick Minot 253

example, the liver and Wolffian body, cannot be understood or even
described correctly without taking into consideration the sinusoidal
character of their circulation. In this case also, the adoption of the
genetic interpretation is much needed. We shall apply presently the
concept “ sinusoid ” to aid us in the interpretation of glands.

We may pass now from a consideration of tissues to that of organs,
and begin with the glands. The classification, which at the present time
prevails widely, is one based upon certain incidental peculiarities in the
shape of the secretory portion of the glands, so that they are put into two
main divisions::the tubular and alveolar; then under each of these we
have three main parallel groups:

the simple tubular or alveolar glands;
the simple branched tubular or alveolar glands;
the compound branched tubular or alveolar glands.

But in this classification there is no place for such a gland as the liver
and the thyroid. In text-books of histology, we find the liver tucked in
under the tubular glands and designated as a reticulated tubular gland,
and the thyroid placed as a follicular gland under the head of the alveolar.
But in another way the system also fails, for there are tubulo-alveolar
glands which again must be classified as simple, simple branched, or
compound branched. LEssentially this classification’ is adopted by the
authors of the manuals of histology which I have examined. To classify
glands thus, seems to me about on a par with classifying organs by their
being solid or hollow, a principle, which would put the spinal cord in the
same class with the intestine, and nerves in the same class with the
tendons. The peculiarities of shape of the secretary portions of glands
are entirely secondary, and do not indicate anything fundamental in
regard to the structure of the gland itself. We cannot call a system good
which, if apphed in accordance with its own definitions, would put some
of the mucous glands of the stomach in one division, others in another
division, because, although these glands are alike in their histological
structure, some of them are branched and some are not. Must we not
condemn a view, which excludes the ovary from the glands and makes
the testis a compound tubular gland, although ovary and testis are strictly
homologous organs, even in the details of their structure? These are
only samples of the innumerable difficulties which the system encounters,
-because it is essentially pedantic, admirable as an orderly arrangement of
names, but impossible as a presentation of anatomical facts.

It appears to me not difficult to make an entirely new classification of
18 :

254 Genetic Interpretations in the Domain of Anatomy

glands, which shall be based upon their genesis and upon the morpho-
logical distinctions, which exist between them. To begin with, we may
put the unicellular glands, of which the goblet-cells serve as the most
familiar type known in man, in the Class A, v. p. 256; next we may have
the true multicellular glands of the epithelial type, which always develop
with ducts by which their secretion is discharged; these form Class B;
while a third class would include the false glands which never develop
with ducts, which produce either merely an internal secretion so-called,
or are adapted to the development of cells of special kinds, as, for ex-
ample, the lymph- and genital-glands; such structures constitute Class C.

We must first attempt a classification more in detail of the true epi-
thelial glands (Class B). In my opinion we can best make two funda-
mental divisions. The glands of the first division have often been called
single or simple follicular glands; I propose for them the term “ simple
glands.” 'The glands in question are always small and have one or several
centers of growth according as they are simple tubes or slightly branching.
Those of each kind are always very numerous and they occur more or less
near together over considerable areas. There are two types of these
known. ‘The first are simple invaginated areas with scattered unicellular
glands, as for instance the glands of the large intestine, the so-called
Lieberkiihn’s follicles; they might be called simple follicles. Glands of
the second type are invaginated areas with specialization of the cells, as,
for example, the sweat, gastric, and sebaceous glands; they might be called
glandular follicles. In the accompanying table the principal glands of
this division are enumerated.

The glands of the second division are of greater bulk and are often
referred to as organic or branching glands. I propose to name them
“compound glands.” They are provided with a single main duct,
which is more or less freely branched, each branch connecting finally with
the secretory portion proper of the organ, which portion may itself also
be branched or not. Each gland falling in this division is a more or less
complete organ by itself, receiving its special blood supply, and its special
innervation—is, in short, a clearly marked physiological entity. Such a
gland differs profoundly in its plan of organization from the glands of
our first division. Of the second division there are clearly three main
types to be distinguished. In the first type the branches of the glands
are found to be supported by mesenchyma or its derivative, connective
tissue, which is more or less abundant between the ducts and secretory
elements of the organ, and in the mesenchyma there is a capillary circu-
lation, which is often brought, however, into.intimate proximity with
the epithelial elements of the organ. These organs are further character-

Charles-Sedgwick Minot 255

ized by the fact that their branches remain distinct. In the second type,
on the other hand, the branches unite together and form an anastomosing
gland structure, and when this anastomosing condition is found it is
associated, not with the development of connective tissue and capillaries,
between the epithelial elements of the organ, but, on the contrary, with
the presence of a sinusoidal circulation. The branching glands with
capillary circulation are numerous, and they may arise, as is noted in the
table, from either the ectoderm, the entoderm, or the mesothelium.
Glands of the second type, anastomosing and furnished with sinusoids,
are few in number. The liver is, of course, the most typical and the
most important. With the liver we ought perhaps to associate the para-
physis, for the recent and still unpublished investigations of Dr. John
Warren show that when this gland is highly developed it is of an anastom-
osing type, and make it probable that its blood supply is sinusoidal.

There remains still a third type, which is necessary, because the ducts
become obliterated in a certain number of true epithelial glands, which
develop primarily with ducts. There results in each case a group of hol-
low epithelial follicles, which are characteristic. For this type I propose
the name, “ ductless epithelial glands.” ‘The thyroid gland and the
hypophysis are probably the best-known illustrations of this group of
glands. Although the morphology of the pineal gland (epiphysis) is
obscure, the organ seems at present to belong to our third type.

Our third Class, C, comprises the false glands, which never develop with
ducts. So far as I am aware this statement may be made absolute for
all glands of this class. It is true beyond any possible question for most
of the glands, which are here to be considered, but it is perhaps as well
to note that possibly some of the glands of the first division may be found
in some vertebrates to have been primitively provided with ducts. This
seems to me possible, but not probable. The first division of the false
glands are the epithelioid. They are perhaps exclusively, so far as the
essential gland elements are concerned, of entodermal origin, and it has
become probable that their circulation is typically sinusoidal. In the
present state of our knowledge it would be venturesome to make positive
assertions on these two points. In the epithelioid glands we have groups
of cells of epithelial origin separated, in the adult at least, from the layer
which produced them, and brought into intimate relation with blood-ves-
sels. A second division comprises the mesenchymal ductless glands,
which are similar to the epithelioid glands in their general appearance,
but their specific elements are derived from the middle germ layer. As
an illustration of the duct!ss false gland of the first division, | may men-
tion the parathyroid; of the second division, the suprarenal cortex. As

256 Genetic Interpretations in the Domain of Anatomy

to the position of the thymus, I feel quite uncertain and hardly dare to
say whether it should be placed among the epithelioid glands or among
the cell-producing glands. Similarly, how to place the interstitial cells
of the genital glands in our system is not yet quite clear to me. The
third division is that of the cytogenic glands, and of these we may
readily distinguish three important types: the first, those in which lymph
cells arise; second, those which produce red blood corpuscles; and third,
those which yield the genital elements. The glands of the first type may
be called lympheal structures. “ Lymphxal” is a new term derived
from “ Lymphzeum,” itself a new technical expression, which I have used
for several years in my lectures on histology and have found advantageous.
A lympheum may be defined as follows: it is a site for the multiplication
‘ of leucocytes and is a more or less definitely circumscribed area consisting
of cellular reticulum, the meshes of which are charged with leucocytes
and are in direct communication with lymph-vessels, or more rarely with
blood-vessels. The following offer examples of lymphza: solitary fol-
licles, tonsils, thymus, lymph glands, hemolymph glands and spleen.
As stated above, whether the thymus should belong in the first or third
division, I cannot say. Of the second type in this division, the bone
marrow is the most important example. Of the third type, that of the
genital clands, we have of course to distinguish two forms, the ovary and
the testis.

With these explanations, I hope the accompanying table will be clear
and I trust that the proposed new classification of glands will seem to you
both more scientific and more available than the classification now pre-
valent, which I should lke to see displaced.

TABLE II. CLASSIFICATION OF GLANDS?

Crass A. Unicellular.
CLASS B, True Glands, always developed with ducts.
Division 1. Simple Glands, (unifollicular or single glands).
a. Ectodermal.
1. Sweat glands.
2. Sebaceous glands.
3. Buccal glands.

b. Entodermal.
1. Gsophageal.
2. Gastric.
3. Intestinal.

c. Mesothelial.
1. Uterine.

Charles-Sedgwick Minot 257

Division 2. Compound Glands (organic or true compound glands).
Type a, ductless epithelial branching (with capillary circulation).
1. Ectodermal. -
Salivaries, tear gland, Harderian.
Mammary glands.
2. Entodermal.
Pancreas.
3. Mesothelial.
Appendicular glands of the urogenital system.
Type b, anastomosing (with sinusoidal circulation).
1. Liver.
2. Paraphysis (in Necturus).
Type c, ductless epithelial (with secondary obliteration of duct).
1. Thyroid.
2. Hypophysal gland.
8. Infundibular gland.
4, Pineal (epiphysis).

Crass C. False Glands, never developed with ducts.

Division 1. Epithelioid glands (exclusively entodermal?)
1. Parathyroid.
2. Carotid.
3. Thymus (cf. below) (?).
Division 2. Mesenchymal ductless glands.
1. Suprarenal cortex.
2. Coccygeal gland and other chromaffinic cell organs.
3. Interstitial cells of genital glands (?).
Division 38. Cytogenic glands.
a. Lympheal structures.

1. Lymph glands and follicles (tonsil?).
2. Hemolymph glands.
3. Spleen.
4. Thymus (?).
b. Sanguifactive organs.
1. Bone marrow.

ce. Genital glands.
1. Ovary.
2. Testis.

I should like to include, in passing, reference to another general
anatomical conception which, though not based strictly on embryological
results, may be appropriately mentioned. I mean that unit of adult or-
ganization, which is sometimes referred to as the “ lobule,” but, as this
term is somewhat confusing owing to the manifold meanings assigned to
it, I venture to express the hope that the term “ structural unit” will be

258 Genetic Interpretations in the Domain of Anatomy

used instead, as has already been done by a few writers. We can then
continue to employ the term lobule for the lung and the liver in the senses
tradition gives to the term, as used for these two organs, and avoid
confusion. The slructural unit* may be defined as the territory of an
organ supplied by a single terminal branch of an afferent vessel (artery
or vein). The volume of such a unit is often only 10-20 cubic milli-
meters. In the case of the liver, the structural unit comprises parts of
several adjacent so-called lobules. It is a pleasure to recall that the
recognition of the anatomical importance of these units is due to one
of our most distinguished American investigators, Dr. Mall.

Finaliy, I should like to apply the principle of genetic interpretation
to descriptive anatomy. It will, I think, sufficiently expound the point
of view I am advocating to consider the application of the principle to
a single organ, and for this purpose we may conveniently select the liver.
In order to show that what I propose is practically a real and great
innovation, let me indicate to you briefly the character of the anatomical
descriptions of the liver to be found in some of the leading text-books of
human anatomy.

In Cunningham’s Anatomy (1902), the account of the liver is written
by Professor Birmingham. He describes, 1, the general form of the sur-
face ; 2, the topographical relations and surfaces in detail; 3, the fissures,
without giving their morphological relations; 4, the division into right
and left chief lobes; 5, the peritoneal relations and ligaments; 6, the phy-
sical characteristics.

In the tenth edition of Quain’s Anatomy (1896), the description opens,
1, with the dimensions and weight; 2, the surfaces; 3, the fissures; 4,
the ligaments and the omentum; 5, the topographical relations; 6,
vessels and nerves; 7, the ducts.

Testut in the third volume of his Anatomy (1894) gives, 1, the situ-
ation; 2, fixation; 3, volume and weight; 4, general confirmation, includ-
ing the two chief right and left lobes; 5, the surfaces in detail.

The account of the liver in Poirier’s Anatomy, Volume IV (1895),
is written by Charpy, who begins with 1, the definition, and continues with
2, situation; 3, fixation; 4, data as to weight, consistency, ete.; 5, the form
and surfaces. Under the head of fixation Charpy says:

“La foie est suspendu A la voute du diaphragme par deux moyens d’attache:
par des replis péritonéaux et par la veine cave inferieure.”’

This misleading statement is the more deplorable because he mentions
only incidentally that the liver adheres directly to the diaphragm. Quite
at the end, the division into the right and left lobes is mentioned.

Charles-Sedgwick Minot 259

In the fourth edition (1901) of Merkel-Henle’s Grundriss, there is, 1,
a general account, which is distinctly not morphological in character ;
2, detailed description of the surfaces and topography ; 38, of the histology.

Gegenbaur in the seventh edition of his Anatomy (1899) proceeds very
differently, for he has strong morphological inclinations. He gives, 1,
the general account of the development of the liver; 2, general account
of the surfaces, including the division into the chief lobes; 3, the relation
of the veins to the omentum and the falciform hgament. Gegenbaur is
the only author of a text-book of human anatomy, known to me, who gives
a distinctly morphological account of the human liver, but even his pre-
sentation of the subject leaves much to be desired, chiefly because his
knowledge of embryology was meagre, and quite insufficient for an ade-
quate interpretation.

It would be easy to analyze descriptions in other text-books, but
enough has been presented to show that they are usually characterized
by certain common tendencies. The authors dwell upon the position and
shape of the liver, seeking to emphasize its exact form, but not endeavor-
ing at all to emphasize the essential characteristics of the organ, er to
bring out the significance of its parts in a manner satisfactory to either
an embryologist, a physiologist, a morphologist, or a pathologist. With
the exception of Gegenbaur, none of the accounts rises above the level of
sheer description.” They simply perpetuate the tradition inherited from
the time when human anatomy was only the description of what was
actually found in the human adult. That tradition has undoubtedly been
in part maintained by the demands of surgeons, whose interest is necessa-
rily chiefly given to the exact determination of the topographical divisions
in the body, hence the influence of the surgeons, when dominant in the
anatomical laboratory, has often exerted an influence unfavorable to the
becoming maintenance of a scientific spirit, such as we ought to insist upon
for the sake alike of anatomy and medieine.

If we review collectively the brief analyses just given of the actual
descriptions in the text-books, we realize at once that those points, which
the genesis of the liver reveals to us as fundamental, are scarcely heeded
by the authors whom we have reviewed. ‘This is not a fitting occasion to
attempt a new description of the liver, and I can merely indicate to you
the principal points upon which a scientific description ought, in my
opinion, to be based. No little study and care would be necessary to work
out practically the suggestions, embodied in the following schedule. In-
deed, the schedule can doubtless be improved by others.

_ In order to prepare an adequate description of the liver, we must begin
by laying aside certain bad habits which we have inherited and have

260 Genetic Interpretations in the Domain of Anatomy

allowed ourselves to perpetuate. I mean the habit of applying the term
ligaments, and the habit of applying the term fissures, to the liver; also
the habit of describing the hepatic segment of the vena cava inferior as
a vessel distinct from the liver, it being in reality, strictly, in every sense
of the word, a portion of the organ. It may be further suggested that the
introduction of a new term, mesepatiwum, may assist in clarifying the
relations. The “ mesepatium” is the membrane (French méso), which
stretches from the ventral border of the stomach and duodenum to the
median line of the ventral abdominal wall. It is in this membrane that
the liver develops. Above the liver, between it and the stomach, is the
dorsal mesepatium (lesser omentum). Between the liver and the body
wall is the ventral mesepatium (falciform or suspensory ligament). In-
stead of speaking of the ligaments, we should speak of the insertion of the
dorsal and ventral mesepatium into the liver; and instead of coronary and
triangular ligaments, we should speak of the attachment of the liver to the
diaphragm. This area of attachment might be called, as regards the dia-
phragm, the hepatic area, as regards the liver, the phrenic area.

With these preliminary explanations in mind, it may be suggested that
a description of the liver must begin, as many authors have begun it
already, with a general statement in regard to the position, size, color,
and general form of the organ, and explaining that it is a gland, with a
duct opening into the duodenum, and having the gall bladder appended
to it, and that the circulation is sinusoidal, and not capillary.

Next, I should place a careful statement of the fundamental relations,
as follows, first, of the broad connection of the liver with the diaphragm.
This connection is primitive embryologically, is maintained throughout
life and constitutes the phrenic area. It is not by the so-called ligaments
or peritoneal folds, nor is it by the vena cava inferior that the liver is
attached ic the diaphragm. On the contrary it is by a large and charac-
teristically shaped phrenic area of the organ that the connection is
established. Second, the relation of the liver to the mesepatium, pointing
out especially that the insertions of the dorsal and ventral mesepatia
mark the division of the liver into right and left lobes and that the inser-
tion is enlarged at one point towards the right to form the so-called porta
of the liver, which admits from the dorsal mesepatium the hepatic artery,
bile duct, and portal vein. Third, the relation of the veins to the organ,
emphasizing that the portal vein marks the border of the dorsal mesepa-
tium, and that its branches within the organ mark the so-called portal
canals; emphasizing also that the umbilical vein or venous hgament
marks the free edge of the ventral mesepatium, and explains the position,

Charles-Sedgwick Minot 261

origin and adult state of the ductus venosus. Jourth, the entrance and
exit of the vena cava inferior. In this connection there should be made
clear the role of the caval mesentery in furnishing a path for the cava
inferior, and at the same time shutting off the lesser peritoneal space,
and keeping the surface of the Spigelian lobe as part of the boundary of
this space.

Next, again, might be presented the secondary features, especially the
marking off of the caudate lobe from the chief right lobe by the vena
cava inferior, and the marking off, similarly, of the quadrate lobe by the
porta and the gall bladder.

Finally, according to this schedule, the description of the finer surface
modelling and the contact with various adjacent organs, such as the stom-
ach, colon, duodenum and kidney. Not one of these topographical re-
lations is indispensable for a comprehension of the general character of
the organ. Even from the standpoint of the surgeon and physician they
are of minor importance. If they are put, as has been customary, in
the forefront of text-book descriptions, attention is distracted from more
essential things. Surely one need not argue to prove that a general
comprehension of each organ is, first and last, the most important goal,
to be striven for in the study of it.

In regard to almost every organ in the body it may be said, I think
without injustice, that the current anatomical text-books offer bare and
barren form-descriptions, seldom giving much, and often giving no, con-
sideration to the essential morphological features of the parts. ‘Take, for
example, the urogenital system. We all know that the internal female
genitalia are formed of two urogenital ridges, which fuse in the median
line, making the so-called genital cord. There is in each ridge a longi-
tudinal epithelial duct, which becomes the Fallopian tube, and by fusing
with its fellow in the genital cord, produces the cavity of the uterus and
vagina. A projection on one side of each ridge forms the ovary. Where
the ridges have not united, rudiments of the Wolffian body of the embryo
occur. The surface of the ridges, both where they are separate and where
they are united, is covered by mesothelium. Around the duct (Fallopian
tube), there is developed a muscular layer, and around the uterine portion
of the fused ducts in the female a very powerful musculature is developed.
By the union of the two ridges a partition is formed across the pelvic
end of the abdomen, so that the abdominal cavity forms a pocket on the
dorsal, and another on the ventral, side of the genitalia. Now the
anatomical way of describing these organs is not to mention the ridges
at all, but in the case of the female to speak of the uterus and its liga-

262 yenetic Interpretations in the Domain of Anatomy

ments. It seems sometimes as if a deliberate effort were made by the
descriptive anatomist to exclude all liberal use of the understanding and
of the intelligence from the study of anatomy, and to reduce it almost to
mere memorization of shapes and proportions, exceedingly difficult to fix
in the mind by that method.

Is one not justified in condemning with great severity the perpetuation
of this old type of anatomy? Is it nota grave mistake to fail to take ad-
vantage of the progress of anatomical science, and to utilize the best
results of anatomical investigation to aid us in forming for ourselves, and
still more, perhaps, for our students, clear notions of the essential charac-
teristics of human organization? There has been within the last twenty-
five years a very great progress in our knowledge of the topographical
anatomy of the viscera, both thoracic and abdominal. When I plead
for the presentation of the subject from the genetic standpoint, I do not
mean to imply- that this superior topographical knowledge should be
slighted, but, on the contrary, I believe that if the student can first master
the essential morphological relations of the body, it will be easier for him
to master subsequently the finer, and often practically important, topo-
graphical details. Let our motto be, not “to memorize,” but “to com-
prehend” the facts of anatomy.

Embryology illuminates anatomy. Its teachings give us the intel-
lectual mastery of anatomical science, because embryology analyzes de-
tails, discriminates the essential from the secondary facts, and establishes
the genetic interpretation, in the solvent light of which the obscurities
of ancient.anatomy vanish, and we see, where before was a dead sea of
innumerable facts, new vital laws arising and guiding principles.

NOTES.

1. P. 247. Cytonforphosis was first used in the Middleton Goldsmith Lec-
ture for 1901, entitled “The Embryological Basis of Pathology,’ Science,
XIII, 481.

2. P. 256. Perhaps all or some of the salivary glands are entodermal. The
submaxillary gland belongs among the organs, when it is a single large
compound gland with a Bartholini’s duct. When the submaxillary is repre-
sented by a group of small glands, they belong with the other simple buccal
glands.

The position of the mammary gland must remain uncertain, until we can
decide whether it is merely a group of glands, or morphologically a true
compound gland. The significance of its peculiar development is still un-
settled.

The hypophysis will perhaps, with more accurate study, be found to be an
anastomosing gland with a sinusoidal circulation.

Charles-Sedgwick Minot — 263

3. P. 258. The morphological characteristics of the structural (or histo-
logical) unit have been pointed out by Mall, so that the brief inadequate
definition seems sufficient for the occasion.

4. P. 252. The account of the formation of sinusoids is somewhat sche-
matic. We now know that the intercrescence of the vessels and parenchyma
offers variations especially in its mode of beginning.

5. P. 259. Huntington’s Anatomy of the Peritoneum, ete. (1903), is written
entirely from the genetic and comparative standpoint. This excellent work,
however, is not a general text-book, and in no sense belongs in the class of
manuals criticized in the text. Even Huntington’s account of the liver
seems to me not to take sufficient advantage of our morphological knowledge,
especially as regards the primary connection of the liver with the diaphragm
and also as regards the sinusoidal circulation.

STUDIES OF THE DEVELOPMENT OF THE HUMAN
SKELETON.

(A). THE DEVELOPMENT OF THE LUMBAR, SACRAL AND CoCcCYGEAL VERTEBRA.

(B). THE CURVES AND THE PROPORTIONATE REGIONAL LENGTHS OF THE SPINAL
COLUMN DURING THE First THREE MONTHS OF EMBRYONIC DEVELOPMENT.

(c). THE DEVELOPMENT OF THE SKELETON OF THE POSTERIOR LIMB.

BY
CHARLES R. BARDEEN.
Professor of Anatomy, The University of Wisconsin.

WitH 13 PLAtTEs.

The following studies on skeletal development are based upon embryos
belonging to the collection of Prof. Mall, at the Johns Hopkins Uni-
versity, Baltimore. I am greatly indebted to Prof. Mall for their use.

A.

THE DEVELOPMENT OF THE LUMBAR, SACRAL AND COCCYGEAL
VERTEBRA. '

Recently I have given an account of the development of the thoracic
vertebre in man (This Journal, Vol. IV, No. 2, pp. 163-175). In the
present paper it is my purpose to describe the special features which
characterize the differentiation of the more distal vertebra.

During the early stages of vertebral development the skeletal apparatus
of the various spinal segments is strikingly similar. This is shown in
Fig. 1, Plate I, which illustrates the spinal column of Embryo II,
length 7 mm. Yet even during the blastemal stage some regional differ-
entiation becomes visible. ‘The costal processes of the thoracic vertebre,
for instance, develop much more freely than those of other regions. It
is, however, during the chondrogenous period that the chief regional
features appear.

To what extent morphological similarity in the sclerotomes and sclero-
meres indicates equal formative potentiality experiment alone can deter-
AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

21

ras)

Studies of the Development of the Human Skeleton

lor)

6

mine. While it is unlikely that experimental studies of the required
nature can ever be made on mammalian embryos it is quite possible that
they may on embryos of some of the lower vertebrates. From the evi-
dence at hand it seems probable, however, that the primitive vertebrae are
to a considerable extent potentially equivalent and that their subsequent
development depends upon the demands of their regional environment.
The strongest arguments in favor of this view come from a study of
variation in the adult. It is well known that at the regional boundaries
vertebral variation is frequent. Thus the 7th vertebra often carries a
short “cervical” rib (Gruber, 69), and rarely it has two cervical ribs
which run to the sternum (Bolk, or). On the other hand the 8th ver-
tebra, usually the first thoracic, may assume all the characteristics of a
cervical vertebra (Leboucq, 98, Low, or).

At the thoracico-lumbar margin variation is more frequent than at the
cervico-thoracic. Thus out of 1059 instances described statistically in
the literature I found, 1904, that the 19th vertebra, commonly the last
thoracic, had no free ribs and was hence of the lumbar type in 30
instances, 2.8%, and that on the other hand the 20th vertebra, commonly
the first lumbar, had free costal processes in 23 instances, 2.2%. Cases
have also been reported where the 21st vertebra, usually the 2d lumbar,
has carried free ribs (Rosenberg, 99). Variation takes place in the
articular processes as well as in the costal elements of the vertebra at the
thoracico-lumbar border (Topinard, 77).

Variation of the lumbo-sacral boundary is likewise frequent. Thus out
of the 1059 instances above mentioned in 28 instances, 2.7%, the 24th
vertebra, commonly the 5th lumbar, was the first sacral and in 47
instances, 4.4%, the 25th vertebra was of the lumbar type.. The 25th
vertebra may exhibit one of many transitions from the sacral to the
lumbar type. This subject has been well treated by Paterson, 93.
Papillault, oo, has contributed an interesting paper on lumbar variations
and Cunningham, 89, on the proportion of bone and cartilage in the
lumbar region.

At the sacro-coceygeal border variation is even more frequent than in
the regions more anterior. Paterson, 93, found diminution in the number
of sacral vertebrae in 2.62%, and increase in their number in 35.46% of
the 265 sacra he examined; and Bianchi, 95, in 17.5% of the female,
23.3% of the male, and 21.23% of the total number (146) of sacra
examined. In this count he excluded sacra in which compensation for
lumbo-sacral alterations was to be seen. Bianchi thinks that the 1st
coceygeal vertebra belongs properly to the sacrum. In the 1059 instances
mentioned above I find that the 30th vertebra, usually the 1st coccygeal,

Charles R. Bardeen _ 267

was reported sacral in nature in 91 instances, 8.6%, and the 29th, com-
monly the Jast sacral, coccygeal in 27 instances, 2.5%. It is possible
that variations of this nature were sometimes overlooked by those making
up the tables from which the above data were obtained.

Variation other than border variation has been reported most fre-
quently in the cervical region. Ribs have been found not only on the 7th
and 6th vertebre but also on those more anterior (Szawlowski, or).

It seems fair, however, to assume that the primitive vertebrae become
differentiated according to the demands of their environment. ‘Thus
the factors commonly exerted on the 8th to 19th costo-vertebral funda-
ments causing them to develop into thoracic vertebre with free ribs,
may be so exerted as to call into similar development the 7th to 18th, the
7th to 19th (20th), the 8th to the 18th, the 8th to the 20th (21st), or
the 9th to the 19th (20th). While the thorax may be segmentally
extended or reduced at either end, extended at both ends, or extended at
one end and reduced at the other, a simultaneous reduction at both ends
has not been reported (Rosenberg, 99).

Differentiation in the post-thoracic region depends apparently in the
main upon the position of the posterior hmb (Bardeen, oo, Ancel and
Sencert, 02). When the developing ilium becomes attached to the costal
processes of the 25th, 26th (and 27th) vertebrae the conditions of the
lumbar, sacral and coccygeal regions are commonly normal. But the
developing ilium may become attached further anterior than usual,
either directly to the costal process of the 24th vertebra or so far forward
that a close hgamentous union is established with it. In such instances
the 12th rib is usually either very rudimentary or absent and often the
29th vertebra is of the coccygeal type. In rare instances the thorax
may at the same time advance a segment into the cervical region. On
the other hand the developing ilium may take a position more posterior
than usual, leaving the 25th vertebra either free to develop into the
lumbar type or but incompletely united to the sacrum (Paterson, 93).
When this occurs the 20th vertebra is very apt to have ribs developed in
connection with it and the 30th vertebra usually becomes an integral part
of the sacrum.

The coccygeal vertebre, with the exception of the first, which is more
directly than the others subjected to the differentiating influences of the
developing limb, are relatively more rudimentary in the adult than in the
embryo.

Rosenberg, 76, advanced the opinion that the ilium is attached more
distally in the embryo than in the adult. I have recently, 1904, shown
that this is not the case. On the contrary, as might have been inferred

268 Studies of the Development of the Human Skeleton

from the distribution of the nerves to the posterior limb, the ilium is
differentiated in a region anterior to the site of its permanent attachment
and the differential activities which it stimulates in the sacral vertebra
are exerted first on the more anterior of these vertebra.

The two limbs do not always call forth a similar response on each side
of the body. Thus Paterson, 93, found asymmetry of the sacrum in 8.3%
of the instances he studied.

Assuming the specific differentiation in the lumbar, sacral and coccy-
geal regions of the spinal column represents a response to stimuli arising
in part from the developing limb, we may turn to a consideration of the
differentiation thus brought about in each of these regions. Attention
will here be directed chiefly to the more salient differences between de-
velopment in the distal half of the vertebral column and that recently
described for the thoracic region.

LUMBAR VERTEBRZ.

Rosenberg, 76, seems to have been the first to take up a detailed study
of the early development of the lumbar vertebre in man. He described
the costal rudiment of these vertebre and found in several embryos that
this rudiment of the 20th vertebra had given rise to a cartilagenous
13th rib. A careful study of a large number of human embryos
has led me, however, to the conclusion (1904) that a 13th rib is no
more frequent in the embryo than in the adult and that in the series
studied by Rosenberg it must have been unusually frequent. Holl, 82,
found no 13th rib in the embryos which he examined. He also came to
the conclusion that the transverse processes of the lumbar vertebra do not
represent ribs. Most investigators rightly disagree with him on this
point.

The development of the external form of the lumbar vertebre in a
series of embryos belonging to the Mall collection is shown in Figs. 1-13,
Plates I-V.

The bodies of the lumbar vertebre during the earlier periods of differ-
entiation are essentially like those of the thoracic vertebre. In embryos
over 12 mm. long, however, the former become progressively thinner,
broader and longer than the latter. The intervertebral disks and the
enveloping ligamentous tissue are similar in both regions. *In the
thoracic region the canal of the chorda dorsalis lies nearer the ventral
surface of the vertebral column than it does in the lumbo-sacral region.
The marked alterations in the curvature of the spinal column which occur
during embryonic development seem especially associated with changes
in the intervertebral disks.

Charles R. Bardeen 269

The neuro-costal processes of the lumbar vertebrae are also at first
similar in form to those of the thoracic vertebra. This is the case in
Embryo II, length 7 mm., Fig. 1, Plate 1; CLXIII, length 9 mm., Fig..
2, Plate II; and CIX, length 11 mm., Figs. 3 and 4, Plate II.

Marked differences in the costal processes are to be seen when. chondro-
fication begins. Thus while the costal process of the 12th thoracic ver-
tebra has early a separate center of chondrofication (Embryo CLXXV,
length 18 mm., Fig. 14, Plate VI and Embryo CXLIV, length 14 mm.,
Fig. 5, Plate III), the processes of the lumbar vertebree remain for a
considerable period dense masses of mesenchyme (Embryo CLXXV,
length 13 mm., Figs. 15 and 16, Plate VI, and Embryo CCXVI, length
17 mm., Figs. 18 and 19, Plate VI). Finally, however, they undergo
chondrofication at the base (Embryo XXII, length 20 mm., Figs. 21 and
22). I have been unable to determine whether this chondrofication
always takes place from a separate center, as it certainly often does, or
sometimes represents merely an extension into the costal mesenchyme of
cartilage from the transverse process. I incline to the former view.

Sometimes the costal element of the 1st lumbar vertebra may remain
for a considerable period separate from the cartilage of the transverse
process. his is true of the right side (left in the figure) of Embryo
XLV, length 28 mm., Fig. 24. But usually at an early period the costal
and transverse processes become intimately fused (Embryo XXII, length
20 mm., Fig. 21; XLV, length 28 mm., Fig. 24, right side of figure;
and Fig. 25; Embryo LXXXIV, length 50 mm., Figs. 27 and 28). The
“transverse ” process of the adult lumbar vertebra represents in the main
an ossification of a membranous, not cartilagenous, extension of the fused
costal element (C. Pr., Fig. 28).

At first the neural processes of the lumbar vertebre are essentially
hike those of the thoracic (Embryo CXLIV, length 14 mm., Fig. 5,
Plate III). Union of the pedicles with the cartilage of the body takes
place and the laminz extend out dorsally in a similar manner in each.
It is in the transverse and articular processes that the chief character-
istic differentiation takes place.

The lumbar transverse processes are broader and much shorter than
the thoracic. At an early period, as mentioned above, they become inti-
mately united to the costal processes. In the dense mesenchyme between
the region of the transverse process and the costal element there is com-
monly developed no loose vascular area such as serves to separate the
neck of the developing rib from the transverse process in the thoracic
region. ‘The occasional appearance of a foramen in a transverse process
of a lumbar vertebra has led to the supposition that they may occur

270 Studies of the Development of the Human Skeleton

regularly in the embryo (Szawlowski, 02, Dwight, 02). The differences
between the transverse processes of the 12th thoracic and the first two
‘lumbar vertebrae in several embryos are shown in Plates VI and VII,
Figs. 14 to 28.

The articular processes in the lumbar, as in the thoracic, region are
at first flat plates connected by membranous tissue,” Fig. 5, Plate III.
But in the lumbar region the superior articular process develops faster
than the inferior so that each superior process comes partly to enfold the
inferior process of the vertebra next anterior. These conditions may be
readily followed in Figs. 9, 12 and 13, Plates IV and V, and in Figs.
20-28, Plates VI and VII. The mammillary and accessory processes of
the adult lumbar vertebre probably represent an ossification of muscle
tendons attached to the transverse and articular processes.

During the development of the vertebree in embryos from 30 to 50 mm.
in length alterations preliminary to ossification, similar to those in the
thoracic, occur in the lumbar vertebre. No actual ossification occurs in
any of the centers in the latter in embryos less than 5 cm. long in Prof.
Mall’s collection. It begins in the bodies of the more anterior of the
lumbar vertebre in embryos between 5 and 7 em. long, and in those 8 em.
long it has usually extended to the more distal. Meanwhile, ossification
of the neural processes has extended from the thoracic into the lumbar
region and soon may be seen throughout the latter. See Bade, oo.

SACRAL VERTEBRA.

Although much has been written about the ossification of the sacral
vertebre comparatively little attention has been devoted to their early
differentiation. Rosenberg, 76, contributed several important facts,
although the general conclusions which he drew concerning the trans-
formation of lumbar into sacral, and sacral into coccygeal vertebre are
unwarranted. Holl, 82, studied more especially the relation of the ilium
to the sacrum in adults and embryos and added materially to the knowl-
edge of the sacro-iliac articulation. Petersen, 93, has given an incom-
plete description of the sacrum in several early human embryos, and
Hagen, 00, of one 17 mm. long. j

Figs. 1 to 13, Plates I to V, show the general external form of the
sacral vertebrae in embryos of the Mall collection. In Embryo II, length
7 mm., Fig. 1; CLXIII, length 9 mm., Fig. 2; and CIX, length 11 mm.,
Figs. 3 and 4, the sacral appear to resemble the lumbar in all essential
particulars, although there is a progressive decrease in size from the mid-
lumbar region distally. No line of demarcation can be drawn in these
embryos between the sacral and lumbar vertebre on the one side and the
sacral and coccygeal on the other.

4

Charles R. Bardeen Para!

Soon after the stage shown in Figs. 3 and 4 the iliac blastema
approaches more closely the vertebral column, usually in the region
opposite the costal processes of the 25th and 26th vertebrae. These, then,
are stimulated to more active growth and extend in their turn out toward
the ilium. The costal processes of the 27th, 28th and 29th vertebra
are likewise stimulated into more active growth. Lateral to the ventral
branches of the spinal nerves the tissue derived from the costal elements
of these five vertebra becomes fused into a continuous mass of condensed
tissue (Fig. 37, Plate IX; and Figs. 5 and 6, Plate III). Against the
anterior and better developed portion of this the iliac blastema comes to
rest (Figs. 5 and 37). From the time of the fusion of the costal ele-
ments of the sacral vertebre into a continuous lateral mass of tissue these
vertebree may be distinguished from the lumbar and coccygeal. Vari-
ation in the vertebre entering into the sacrum occurs in the embryo as in
the adult (Bardeen, 04). .

At the period when the iliac blastema comes into contact with the
costal mass of the sacrum centers of chondrofication have appeared: in
the bodies of the sacral vertebrae. The bodies, compared with the inter-
vertebral disks, are progressively smaller from the first to the fifth
(Embryo CXLIV, length 14 mm., Fig. 6). Otherwise they present no
characteristics of special note. In older embryos this difference becomes
less and less marked.

The neuro-costal processes present features of more specific interest.
In Embryo CXLIV, length 14 mm., centers of chondrofication may be
observed in the neural processes of the first two sacral vertebre. They
are not yet united to the bodies of the vertebre and are simple in form.
No cartilage has as yet appeared in the neural processes of the other
vertebre. In somewhat older embryos, CCXVI, length 17 mm., Fig. 38,
and CLXXXVIII, length 17 mm., Fig. 39, the neural processes of all
the sacral vertebre have become chondrofied and distinct. Separate
centers of chondrofication may be seen in each of the costal elements.
At a slightly later stage, Embryo XXII, length 20 mm., Big. 10, Plate
1V, the extremities of the costal elements of the first three sacral vertebree
kave fused with one another and have thus given rise to a cartilaginous
auricular surface, and each has fused with the neural arch of the vertebra
to which it belongs. The costal elements of the 4th and 5th sacral ver-
tebre are likewise fused to their corresponding neural arches and all of
the sacral neural arches are fused to their respective vertebre.

Cross-sections illustrate well the relations of the neural and costal pro-
cesses to the vertebra. Embryo CIX, length 11 mm., Fig. 29, Plate
VIII, shows an early blastemal stage in which the sacral vertebre resem-

272 Studies of the Development of the Human Skeleton

ble the thoracic. Figs. 30, 31 and 32 represent cross-sections through
the 1st, 2d and 3d sacral vertebree of Embryo X, length 20 mm. The
neural and costal processes in this embryo show cartilagenous centers,
but these are fused neither with one another nor with the vertebral body.
Embryo CCXXVI, length 25 mm., Figs. 33 and 34, shows fusion of
neural and costal processes with the body. The costal processes have by
this period given rise to a continuous lateral cartilagenous mass, a portion
of which is represented in Fig. 40. This shows an oblique section through
the 3d and 4th sacral vertebre of Embryo LXXXVI, length 30 mm.
This cartilagenous lateral mass may likewise be seen in the sacra of
Embryo CXLV, length 33 mm., Figs. 11 and 12, and LXXXIV, length
50 mm., Fig. 13.

From the primitive neural cartilages there develop pediclar, trans-
verse, articular and laminar processes. The pediclar and transverse
processes become intimately fused with the costal element as described
above. In none of the embryos examined was there seen a separation of
costal from transverse process marked by blood-vessels, such as one might
perhaps expect to find because of the occasional appearance in the adult
of transverse foramina in the lateral processes of the sacral vertebree
(Szawlowski, 02).

The articular processes in the older embryos retain a more primitive
condition than do those of the lumbar and thoracic regions. Figs. 33,
34, 35 and 36 show cross-sections through the articulations of the neural
processes.

The laminar processes of each side are still separated by a considerable
interval in embryos of 50 mm., although at this period the lumbar region
is nearly enclosed. Compare Figs. 28 and 36.

Changes in the cartilages preliminary to ossification occur both in the
bodies and in the neural processes of the sacral vertebre at a period
quickly following their appearance in the lumbar region. ‘Thus in
Embryo LX XIX, length 33 mm., changes of this nature may be followed
as far as the 5th sacral vertebra. It is well known, however, that actual
ossification in the more distal sacral vertebre takes place considerably
later than in the thoracico-lumbar region. The primary centers of ossi-
fication correspond with the centers of chondrofication except that there
is a single center in place of two centers for each body, and one center
instead of two for each neuro-costal processes of the two more distal
sacral vertebrae. Posth, 97, has recently contributed a valuable paper on
the subject of sacral ossification.

Charles R. Bardeen PA

COCCYGEAL VERTEBRA.

Since the valuable contributions of Fol, 85, who described 38 vertebrae
in the early human embryo, considerable attention has been devoted to
the development of the coccygeal vertebrx, especially from the point of
view of the number of vertebre in embryos. The literature on the subject
has been summed up by Harrison, or, and more recently by Unger and
Brugsch, 03. The later stages in the development of the coccyx have
been described by Steinbach, 89, who studied a large number of spinal
columns of foetuses, infants and adults. In a recent paper, 1904, I have
given a summary of the number of coccygeal vertebre found in various
embryos and of the number of hemal processes found in the embryos
belonging to the collection of Prof. Mall. 3

Without attempting here to enter into a detailed account of the condi-
tions found in these various embryos we shall pass at once to a consider-
ation of the more characteristic features of coecygeal development. The
differentiation of the coccygeal sclerotomes begins at about the end of
the fourth week, Embryo II, length 7 mm., Fig. 1, Plate I. As a rule
at least six or seven membranous vertebre are developed. The highest
relative differentiation of the coccygeal vertebre occurs in the fifth and
sixth weeks. At this time dorsal processes connected by interdorsal
membranes extend as far as the 4th and 5th vertebra, Figs. 3 and 4,
Plate III, Fig. 41, Plate X. No distinct costal processes are, as a rule,
developed on any except the first coccygeal vertebra, but at the height of
development most of these vertebra have distinct hemal processes, first
described by Harrison, or. These processes may also be seen on the more
distal sacral vertebre, Fig. 41, Plate X. They usually disappear before
the embryo has reached a length of 20 mm., but the coccyx described by
Szawlowski, 02, suggests that occasionally they are retained until adult
life. .

Usually the bodies of the first five coccygeal vertebree become chondro-
fied. The chondrofication of the more distal of these vertebre is, as
pointed out by Rosenberg, 76, often very irregular. There may be sepa-
rate bilateral areas of cartilage or the two areas may be connected merely
anterior to the chorda dorsalis. The bodies of successive vertebre may
be irregularly fused.

As a rule the neural processes of the first coccygeal vertebra alone
become chondrofied and fused to the vertebral body. The others, as well
as the connecting interdorsal membranes, disappear.

The cartilagenous coccygeal vertebre are thus relatively less developed
than the membranous. It is probable that the osseous are less developed
than the cartilagenous. Thus although Steinbach, 89, has made a strong

274 Studies of the Development of the Human Skeleton

plea for the presence of 5 vertebre in the adult coccyx, the more com-
monly accepted number of four seems to be a truer estimate of the defi-
nite number of bones usually found present.

The bend of the coceyx which takes place during the third month
is an interesting phenomenon. It seems to be associated with the
development of pelvic structures.

SUMMARY.

In the earlier stages of development the lumbar, sacral and coccygeal
vertebra resemble the thoracic. The blastemal vertebra arise each from
the contiguous halves of two original segments of the axial mesenchyme.
Each vertebra exhibits a body from which neural and costal processes
arise. The neural processes are connected by “interdorsal” mem-
branes.

As the blastemal vertebrae become converted into cartilage specific
differentiation becomes more and more manifest. The cartilagenous
vertebral bodies and the intervertebral disks are all formed in a similar
manner and except for size manifest comparatively shght differences in
form. The more distal coccygeal vertebrae are, however, irregular. But
the chief specific differentiation is seen in the costal and neural processes.

In the blastemal neural processes of the thoracic vertebre cartilagenous
plates arise from which spring pediclar, transverse, articular and laminar
processes.

In the lumbar vertebre similar processes arise from the neural carti-
lages. The pediclar processes resemble the thoracic but are thicker;
the transverse processes are shorter, much thicker at the base and remain
bound up with the costal processes; the superior articular processes
develop in such a way as to enfold the inferior ; the laminar processes are
broad, grow more directly backward than do the thoracic, and on meeting
their fellows in the mid-dorsal line fuse and give rise to the typical
lumbar spines. The mammillary and accessory processes are developed
in connection with the dorsal musculature.

In the sacral vertebre the neural cartilages give rise to very thick
pediclar processes; to articular processes the most anterior of which
develop like the lumbar, while the others long maintain embryonic char-
acteristics; to transverse processes which in development are bound up
with the costal processes; and to laminar processes which are very slow
to develop and of which the last fail to extend far beyond the articular
processes.

In the coceygeal vertebre the neural processes of the first, and rarely

Charles R. Bardeen 275

the second, alone give rise to cartilagenous plates. From these only
pediclar and incomplete articular and transverse processes’ arise. The
cornua of the adult coceyx represent fairly well the form of the early
neural semi-arches. The transverse processes develop in close connection
with the costal processes.

In the thoracic vertebre cartilagenous ribs develop from separate
centers in the blastemal costal processes.

In the lumbar vertebre separate cartilagenous centers sobanly always
arise in these processes, but they are developed later than those of the
thoracic vertebre and quickly become fused with the cartilage of the
transverse processes. The transverse processes of the adult lumbar ver-
tebre represent at the base a fusion of embryonic cartilagenous costal
and transverse processes, but in the blade an ossification of membranous
costal processes.

In the sacral vertebre separate cartilagenous costal centers are de-
veloped but they soon become fused at the base with the transverse
processes of the neural plates. Laterally by fusion of their extremities
the costal processes give rise to an auricular plate for articulation with
the ilium. :

In the coccygeal vertebre the costal processes of the first become fused
with the transverse processes and develop into the transverse processes
of the adult coceyx. I have been unable definitely to determine whether
a separate costal cartilage is developed in these processes or cartilage
extends into them from the neural processes. The costal processes of
the other coccygeal vertebre have merely a very transitory blastemal
existence.

For a brief period the more distal sacral and the coccygeal vertebrz
have membranous hzemal processes.

Centers of ossification correspond in general with centers of chondro-
fication, but, as in the case of the vertebral bodies and the more distal
sacral neuro-costal processes, a single center of ossification may represent
two centers of chondrofication.

B:

THE CURVES AND THE PROPORTIONATE REGIONAL LENGTHS OF
THE SPINAL COLUMN DURING THE FIRST THREE MONTHS
OF EMBRYONIC DEVELOPMENT.

In 1879 Aeby contributed an imoprtant paper dealing with the length
of the various regions of the spinal column at different ages, the height
of the constituent vertebre and the thickness of the intervertebral disks

276 Studies of the Development of the Human Skeleton

in man. He showed that in young embryos the cervical region is rela-
tively longer than the lumbar region but that as growth proceeds there
is a constant proportional increase in length of the latter over the former.
Taking the cervical region as 100, for instance, he found that in embryos
below 10 mm. in length the lumbar region equals 69, while in the adult
it is equivalent to 150. Thus, too, while from infancy to maturity the
spinal column increases three and one-half times in length and the thor-
acic region at about the same relative ratio, the lumbar region increases
four times in length and the cervical but three. Other investigators,
including Ballantyne, 92, and Moser, 89, have in general confirmed these
results. Of those who have studied the proportional length of the various
regions in the adult Ravenel, 77, and Tenchini, 94, have made note-
worthy contributions.

The post-natal lengthening of the lumbar region is associated with
those changes in the lumbo-sacral curve which accompany the assumption
of an erect posture during early childhood. Do similar alterations in
relative regional length accompany the straightening of the spinal column
which takes place during the first three months of embryonic develop-
ment?

In Fig. 44, Plate X, I have represented by curved lines the vertebral
columns of several embryos of this period and an adult column. The
cervical, lumbar and coceygeal regions are represented by heavy, the
thoracic and sacral regions by light lines. The 5th, 6th and 7th thoracic
vertebrae are made to coincide in each instance.

The anterior half of the spinal column is considerably curved in
Embryo II, length 7 mm. It gradually becomes straightened in succes-
sively older embryos until in Embryo CLXXXIV, length 50 mm., it is
nearly straight. The subsequent anterior convexity in the adult is asso-
ciated with the assumption of an upright position of the head.

It is, however, in the posterior half of the spinal column that the chief
alterations in spinal curvature are to be noted. In Embryo II, length
7 mm., the ventral surface of the sacral region faces the mid-thoracic
region ; in Embryo CIX, length 11 mm., the anterior end of the vertebral
column; in Embryo CLXXXIV, length 50 mm., almost directly ven-
trally ; and in the adult, in a posterior direction.

The relative lengths of the various regions of the spinal column during
the first three months of development may be gathered from the following
table, which is based in part upon data obtained from embryos belonging
to the Mall collection and in part upon those of the Born and His collec-
tions studied by Aeby.

=e
~

Charles R. Bardeen 2

TABLE A.

THE LENGTHS OF THE VARIOUS REGIONS OF THE SPINAL COLUMNS OF EMBRYOS
OF THE SECOND AND THIRD MONTHS, AND THE PROPORTIONAL LENGTH OF THE
THORACIC COMPARED WITH THE OTHER REGIONS.

Embryos. | Regions of the spinal column.
— —_ ———— —
| Cervical. ae Lumbar. Sacral. Coccygeal
§ lhc 5 iit emloonlas 2 qo) &
Designation. 3 epee =e eis Pe oe IS 25 De 335 = 8 = 25
& 5 ae Root) es Pal }~ eet) os ae ga aco
S | 8a See Sa 8, ge a,| go Se Sgo
s) atone Hel coil! | Allee, aA a] ae
II. Mall a 12 (60.6 |3.3 |1.2 |36.4 29 27.3
CCXXI. “ce 13 (1) |2 62-5 [Since Se AGe dele ou SO al 1.0 |31.2
1 (Aeby). His. 10 2.95 |76.6 18.85 |1.75 |45.5 |1.95 | 50.6(2)| .. ss
CIX. ‘Mall. 11 2.7 |67.5)/4 2 150.0|1.7 | 42.5 1.3 |82.5
CXLIV. “ 14 By NPB plo? |EGeOplsBee Boe ileal |ze)ote}
XLIII. ce 16 12.9 |61.9 |4.84]1.9 |48.7]1.55|] 35.7 |1.41 |82.5
2 (Aeby). | His. 10 3 165.214.6 |2.25 |48.9]1 88] 40.9(2)|
3 (Aeby). | 66 16 3 164.5 |4.65 |2.25 |48.2 |2.65 | 57.0(2)
XEXIT.< |Mall. 20 : Sree | ats) 4163 545/2.0 | 36.4 j|1 25
4 (Aeby). His. 21.5 |3.9 |61.9/6.3 [8.1 |49.2|8.4 | 53.92) me
CVIII. Mall. | 22 4.34/61.9|7.0 |3.5 |50.0/2.8 | 40 Woes” akaaal
CXL. [Pee | 88 |5.1 |59.518.55/4 |46.8/8.5 | 85.6 {1.75 |20
CCXXVII. ce 30 5.4 |60.0/9.0 |4 44.4 |3 33.3 1.38 |15.3
5 (Aeby). Born. na (OO 154, 5,10 SEQ A Rese re Wor
XCV. |Mall. 46 (8.25 59.0|14 = |7.25 |51.8|5.5 39.3 75 |\26.8
CLXXXIV. eect 50 8 61.1 138 6.35 |48.8 4.65) 35.8 (2.65 20.4
6 (Aeby). |Born. 10 (58.8/]17% [8.5 |50.0/9.5 | 55.8(?)
| | |

1. This is the measurement recorded before the embryo was sectioned. The embryo was
cut sagittally. The length of the sections in the median line is 7. mm. The general de-
velopment corresponds with that of an embryo of this length.

2. These figures represent the length of the pelvic portion of the spinal column.

This table discloses considerable individual variation. The length of
the cervical region is about 60% of that of the thoracic. In Embryos
CIX, 1 and 2, this ratio is much exceeded. The measurements for CIX
are calculated from obliquely transverse sections and hence are subject to
some error. The data concerning the measurements for 1 and 2 are not
given by Aeby.

The lumbar region, at first less than 40% in length of the thoracic,
in most embryos approximates 50%. ‘The length of the sacral region
varies from 33 to 42.5% of the thoracic. The coccygeal region, with a
nearly constantly diminishing comparative length, shows marked varia-
tions.

There is no good evidence that the straightening of the spinal column
is accompanied by a marked increase in relative length of the lumbar
region after the early stages in Embryos II and CIX.

278 Studies of the Development of the Human Skeleton

A comparison of the spinal columns of embryos of the second and third
months with those of older embryos and of children shows that it is dur-
ing the latter half of foetal life and early childhood that the chief relative
lengthening of the lumbar region takes place.

According to Aeby the average length of the cervical, thoracic and
lumbar regions in the new-born is respectively 45.1, 83.9 and 47.5 mm.
This makes the length of the cervical region 53.5% and that of the
lumbar region 56% of that of the thoracic. Corresponding figures from
Ballantyne, 92, for full-term fcetuses are: cervical, 33.6 mm. (42.8%) ;
thoracic, 78.4 mm. (100%); lumbar, 42.8 mm. (54.3%); and sacro-
coceygeal, 39.8 mm. (50.8%). Thus Ballantyne finds a greater propor-
tional reduction of the cervical region. —

The conditions in the adult, as given by various investigators, are as

follows:
TAB. B:

PROPORTIONAL LENGTHS OF THE VARIOUS REGIONS OF ADULT SPINAL COLUMNS.

Ratio of other
revions to,the
thoracic as 100.

Average leneth of
regions in mm.

Parade Sex and Size. 2 | 8 | 8 |e) 2 | 2°\ee
407 is S —& | 36 Ss = EES
(3) x=| | 5 | ne o 2 7p)
© i= | A 5 iS) | g
Ravenel, 1877. Male. £33) 1280 Wise 47.5 |65.0°
Female. 120 (260 {178 46.1 {68.5
Aeby, 1879. Male. 129.9 |273.4 |184.1 47.5 |67.3
Female. 122.9 |265.8 |190.3 | 46.2. \71
Tenchini. 3 | Short. 98. |222 [125 |151./44.1 (56.3 68
= | Medium. 100 |240 (187 - 1158 |41.% (57.1) |ebes
= | alll 104 (240 |184° |168 /48.3 [55.8 (61
Dwight, 1894. Male. 133. |287 19.9 | 46.3 |69.3
Female. 121 _ |265 18.7 | 45.7 |70.6

The chief point of interest in this table is the difference between the
results found by the German and American investigators and those of the
Italian. Apparently the Italians have proportionately shorter cervical
and lumbar regions than the Americans and Germans, but it is possible
that different ways of measuring were used. It is a subject worthy of
further investigation.

Both Tenchini and Ancel and Sencert, 02, have treated of variations
in measured length of individual vertebre associated with numerical
vertebral variation.

Charles R. Bardeen 279

C.
THE DEVELOPMENT OF THE SKELETON OF THE POSTERIOR LIMB.

One of the most studied subjects in morphology has been the develop-
ment of the vertebrate limbs. Fortunately critical summaries of its
immense literature have recently been given by several keen investigators,
among whom may be mentioned Wiedersheim, 92, Mollier, 93, 95, 97;
Gegenbaur, 98, o1, Klaatsch, 00, Rabl, o1, Fiirbringer, 02, Ruge, o2, and
Braus, 04. Therefore no attempt will here be made to review previous
work except in so far as it deals directly with the development of the
human limb.

During the third week of embryonic life the limb-buds become filled
with a vascular mesenchyme (Bardeen and Lewis, o1, p. 17, Figs. 18 and
19). The source of this tissue is uncertain. In part it may come from
the primitive body-segments, but it seems probable that in the main it
comes from the parietal layer of the unsegmented mesoderm.

During the fourth week the mesenchyme increases in amount and the
limb-bud begins to protrude further from the body-wall. Observed
structural differentiation does not, however, begin until the early part of
the fifth week, at the time when the lumbo-sacral spinal nerves are
beginning to form a plexus. At this period the tissue at the center of
the base of the limb becomes greatly condensed, Embryo CLXIII, length
9 mm., Fig. 45, Plate XI. The boundaries of the mass are not perfectly
definite, but a wax-plate reconstruction based upon ‘drawings made as
definite as possible gives rise to the structure shown in Fig. 2, Plate IT.
The relations of this tissue mass to other structures are shown in Plate
II, Fig. B, and Plate III, Fig. C, of the article by Bardeen and Lewis, or.
The condensation represents the acetabulum and the proximal end of the
femur. This is indicated by its relations to the nerve plexus.

Once begun skeletal differentiation proceeds rapidly. In Embryo CIX,
length 11 mm., Figs. 3 and 4, it may be seen that from the original center
of skeletal formation the condensation of tissue has extended both dis-
tally and proximally, but much more rapidly in the former than in the
latter direction. Distally the sclero-blastema shows femur, tibia, fibula,
and a foot-plate; proximally, an iliac, a public and an ischial process.
A series of sections through the skeletal mass (Figs. 46 to 52) shows that
in the femur, tibia and fibula chondrofication has begun. At the centers
of the blastema of the ilium, ischium and pubis a still earlier stage of
chondrofication has made its appearance. The leg of this embryo, there-
fore, represents a stage of transition from the blastemal to the chondro-
genous stage of development. Fig. 55, Plate XII, shows a longitudinal

280 Studies of the Development of the Human Skeleton

section through the leg of Embryo CLXXV, length 13 mm. The de-
velopment is slightly more advanced than in Embryo CIX.

The further general development of the skeleton of the limb may be
followed in Figs. 8 to 13. For the sake of convenience the development
of the several parts of the skeleton will be taken up in turn as follows:
(a) pelvis; (b) femur, hip-joint, tibia and fibula, and knee-joint; (c)
ankle and foot.

A. PELVIS.

Petersen, 93, has given a good account of the early development of the
human pelvis. His work was based upon embryos belonging to the His
collection. In embryos Ru, length 9.1 mm., Ko, length 10.2 mm., and
N, length 9.1 mm., the formation of the lumbo-sacral plexus has begun
and there is a condensation at the center of the limb-bud. The conditions
here resemble those in Embryo CLXIII, length 9 mm., Fig. 2, described
above. Petersen believed that the condensation of tissue in embryos Ru
and Ko represents the germinal area for the muscles and skeleton of the
lower extremity while that of N shows a further differentiation of the
diaphysis of the femur. The last ends anteriorly in a small undifferen-
tiated cell-mass but there is nothing further to indicate the future pelvis.
Yet, as I have mentioned above, the relations of this cell-mass to the
rerves arising from the plexus indicates that it is the fundament of the
future pelvis. The nerves pass about it as they do later about the ace-
tabulum. In Embryo §,, length 12.6 mm., Petersen found what he con-
siders the first traces of a definite pelvic fundament. This embryo is
evidently of about the same stage of development as CIX, length 11 mm.,
Figs. 3 and 4 and 46 to 53. But CIX is slightly more advanced and
shows early stages of chondrofication not seen in S,. The pelvis of S,
has a slightly more anterior position and the iliac blastema extends rather
toward the 24th than the 25th vertebra (Fig. 1, Plate I, of Petersen’s
article).

The pelvic scleroblastema of embryos of the stage found in CIX
undergoes a rapid development. Its iliac portion extends in a dorsal
direction toward the vertebre which are to give it support. The costal
processes of the latter at the same time become fused into an auricular
plate. With this the iliac scleroblastema comes into close approxima-
tion, Figs. 5 and 6, Plate III, although for some time separated by a
narrow band of tissue staining less densely than the blastemal, Fig. 37.

Anteriorly the iliac blastema extends toward the abdominal muscula-
ture, to which it finally serves to give attachment.

While the blastemal ilium is thus becoming differentiated the pubic

Charles R. Bardeen _ 281

and ischial processes of the pelvic blastema extend rapidly forward and
ventral to the obturator nerve they become joined by a condensation of
the tissue lying between them. Thus the obturator foramen of the
blastemal pelvis is completed, Figs. 5 and 6. Between the crest of the
ilium and the ventral extremity of the pubis dense tissue is formed to
give attachment to the oblique abdominal musculature. This represents
the embryonic Poupart’s ligament and completes a femoral canal, Figs.
5 and 6.

While the blastemal pelvis is being differentiated the formation of
cartilage in the ilium, ischium and pubis extends rapidly from the centers
indicated in Embryo CIX. In CXLIV, length 14 mm., Figs. 5, 6
and 54, the three cartilages are distinct.

The iliac cartilage is a somewhat flattened rod with anterior and pos-
terior surfaces, Fig. 38. The anterior surface of the iliac cartilage at
first faces slightly laterally as well as anteriorly. Lubsen, 03, in an
interesting paper has shown the importance of this from the standpoint
of mammalian phylogeny. He considers a flat plate with median and
lateral surfaces to be the probable primitive form of ilium from which
the triangular form, on which Flower, 70, laid stress, is derived by a
lateral projection serving to divide the lateral surface into an anterior
iliac and a posterior gluteal portion. In man and some other mammals
the anterior iliac surface, according to Lubsen, comes to be turned
medially by a great extension of the lateral projection and a secondary
union of the abdominal musculature to this. In man, however, the
primitive iliac cartilage is a rounded plate of which the long axis of the
cross-section lies nearly at right angles to the median plane of the
embryo. On the whole it suggests the prism described by Flower.

The pubic and ischial cartilages when first formed are mere rounded
masses of tissue lying in the center of their respective blastemal processes.
The acetabulum at this time is composed mainly of blastemal tissue but
the iliac and ischial cartilages form a part of its floor, Embryo CXLIV,
length 14 mm., Figs. 5 and 6, Plate III. The pelvis of CR, length
18.6 mm., described by Petersen and pictured in Fig. 2, Plate I, of his
article, is of a stage of development similar to that of CXLIV. The
cartilages are slightly more advanced in development. The iliac carti-
lage is broader in an antero-posterior direction and extends to the
sacrum. He observed no dense tissue completing the obturator foramen
but this tissue is quite plain in the corresponding embryos of the Mall
collection, and it is described by Petersen for embryos Wi, length
15.5 mm., and Ob, length 15 mm., which are slightly more advanced

than CR.
22

282 Studies of the Development of the Human Skeleton

While the human embryo is growing from 15 to 20 mm. in length,
there occurs a rapid development of the pelvic cartilages. About the
head of the femur each gives rise to a plate-like process. The fusion of
these processes produces a shallow acetabulum, Figs. 9 and 10, Plate IV.
Those from the ilium and ischium are larger than that of the pubis and
fuse with one another before the pubic cartilage fuses with them. The
proportional areas of the acetabulum to which each pelvic cartilage con-
tributes seem to be essentially the same as those later furnished by the
corresponding pelvic bones, 24 + ischium, 34 —ilium, YS pubis. While
growing about the hip-joint so as to complete the acetabulum each of the
pelvie cartilages has a centrifugal growth within the blastemal pelvis.
It is convenient to consider each cartilage in turn.

The iliac cartilage of the stage represented in Embryo CXLIV, Figs.
5 and 6, represents essentially that portion of the ilium which borders
the entrance to the true pelvis and which has hence been called the pelvic
portion of the ilium. From this cartilage extends dorsally into the sacral
area of the blastemal pelvis, Figs. 9 and 10, and there gives rise to the
sacral portion of the cartilagenous ilium. At the same time cartilage
extends into the blastema which passes anteriorly to give attachment
to the abdominal musculature and from this arises the abdominal por-
tion of the cartilagenous ilium. A slight extension of cartilage into the
anlage of Ponpart’s ligament forms the anterior superior spine, but the
blastemal covering of the femoral canal is not converted into cartilage as
is the similar covering of the obturator.

While the cartilagenous ilium is being developed the ischial and pubic
cartilages extend ventrally into their corresponding blastema. The
ischial cartilage rapidly increases in thickness and at the same time gives
rise to two processes. One of these, the ischial spine, extends toward the
sacrum. This seems to indicate a commencing enclosure by cartilage of
the ilio-sciatie notch. The other projects toward the developing ham-
string muscles and gives rise to the ischial tuberosity.

The pubis broadens as it extends forwards and beyond the obturator
foramen sends down a process which fuses with a longer one extending
up from the ischium.

The various stages in the formation of a cartilagenous innominate
bone have been followed in detail by Petersen in a series of six embryos
from 17.5 to 22 mm. in length. In general what he describes coincides
well with the appearances presented in a somewhat more extensive corre-
sponding set of embryos belonging to the Mall collection. The distal
position of the ilium, represented in Fig. 12, Plate VII, of Petersen’s
article, is to be looked upon as an individual variation and not as a

Charles R. Bardeen - 283

regular step in the process of attachment of ilium to spinal column.
This I have previously pointed out (1904). Hagen, 00, has given an
account of the pelvis of an embryo of 17.5 mm., which corresponds with
the description given above.

During the further development of the cartilagenous pelvis the ventral
extremities of its two halves, at first widely separated, in embryos of
20 mm. in close proximity, are finally united by a symphysis when the
embryo reaches a length of 25 mm. In an embryo of this length,
CCXXVI, the blastemal tissue of each half is fused in the median line
but the cartilages are separated by 44 mm., although the width of the
pelvis at the rim is only 3 mm. In this embryo the obturator foramen
is completely enclosed by cartilage. In embryos of 30 mm. the pubic
cartilages are closely approximated in front.

Petersen reconstructed the pelvis of an embryo of 29 mm., Lo,, and
has given an extensive description of it. In essential features it corre-
sponds with the pelvis of Embryo CXLV, length 33 mm., Figs. 11 and
12. The sacrum of this latter embryo is, however, composed of the 25th
to 29th vertebrae, while Petersen found the 30th vertebra of Lo, belong-
ing to the sacrum. This variation is common in the adult.

Of the characteristic features common to Lo, and CXLV may be
mentioned the relatively great development of the sacral portion of the
ilium, with a large posterior-superior spine, the relatively slight develop-
ment of the abdominal portion, and the comparatively large part of the
pelvic entrance which is bounded by the sacrum. In the adult, according
to Engel, the sacrum bounds 26.2% of this. For the new-born female
the following figures are given by Fehling: sacrum, 28.9% ; ilea, 29.2% ;
. pubes, 42.8% ; for the new-born male, 30.4%, 26.9% and 42.2%. For
Lo,, the percentages are: 37.0%, 31.7% and 31.3%; for CXLV, 34%,
33% and 33%.

In Lo, the rim of the acetabulum is deepened by dense blastemal tissue.
In CXLV this has been in part converted into cartilage by extension
of processes from the ilium, ischium and pubis. The processes of the
ischium and pubis are fused with that of the ilium but not with one
another, so that the cotyloid foramen is well marked.

In both embryos the iliac blades bend more sharply than in the new-
born and resemble in this respect the adult. The ischial spines are
relatively more developed and project more into the pelvis than do those
of the adult.

The pelvis of Lo, is that of a female; the pelvis of CXLV that of a
male. It is of some interest to determine whether or not sexual differ-
entiation is apparent. Fehling, 76, showed that during fcetal life differ-

284 Studies of the Development of the Human Skeleton

ences of this nature, though by no means marked, are none the less to be
made out. His conclusions have been confirmed by Veit, 89, Romiti, 92,
KKonikow, 94, Thomson, 99, and Merkel, 02. Petersen has made most
careful comparisons between the measurements of the pelvis he recon-
structed and the structural data furnished by Fehling: The variations
due to sex are, however, so slight that they are likely to be obscured in
wax reconstructions of early embryos. Allowance must be made for
errors of technique and for the difficulty of determining corresponding
points between which measurements are to be taken. ‘Thus, for instance,
the proportional widths of the entrance to the true pelvis, the pelvic
cavity and the pelvic exit I find to be in CXLV as 100: 75:54, while
Petersen makes them for Lo, as 100: 74.7:46.1. According to theory
the width of the exit should be proportionately less in the male than in
the female pelvis. For foetuses of 30 to 34 em. Fehling gives for females
100: 88:70; for males 100: 87:60; for new-born females 100: 84: 76;
for new-born males 100: 82:65; for adults 100: 92:81. Without the
possibility of a direct comparison of the two reconstructed pelves it is
therefore scarcely possible to determine accurately to what extent they
may show sexual differences. A characteristic on which Merkel, 03,
lays especial stress, the more posterior position of the greatest width at
the pelvic entrance in the male, does, I think, exist in CXLV in com-
parison with Lo,.

These two embryos show that at the beginning of the third month of
development the cartilagenous pelvis is well formed. At this time, also
ossification begins in the ilium. Preliminary changes in the cartilage
may be seen in embryos of 25 mm. in an area corresponding to that in
which chondrofication commenced. These changes are further advanced
in CXLV, but in this embryo neither deposit of calcium salts nor true
ossification has commenced in the ilium although ossification is under
way in the clavicle, inferior and superior maxille, occipital, humerus,
radius, ulna, femur, tibia and fibula. Another embryo of the same
length, 33 mm., LUXXIX, does, however, show a well-marked area of
ossification, Fig. 58. In the endochondral region calcium salts are de-
posited while on each side of this perichondral ossification takes place.
In an older embryo, LXXXIV, length 50 mm., this latter process shows
well, Fig. 36. It is known that ossification of the ischium and pubis
takes place considerably later, that of the ischium beginning in the 4th
month and that of the pubis in the 6th to 7th (Bade, 00).

Aside from the ossification of the ilium nothing especially noteworthy
seems to take place in the development of the pelvis during the period

Charles R. Bardeen 285

when the embryo is growing from 30 to 50 mm. in length. Fig. 13 shows
the form of the pelvis in an embryo of the latter size.

The relations of the pelvis to the sacrum during the second and third
months of life deserve some attention. I have endeavored to illustrate
them in Fig. 44. The curves of the spinal columns of several embryos
and an adult are there shown. A point “a” represents the place where
a line joining the centers of the two acetabula would cut the median
plane of the embryo. From this point dotted lines are drawn to each
extremity of the sacral region and one is projected perpendicularly to a
line joining these two extremities. A fourth line from point “a” indi-
cates the direction of the long axis of the femur.

In Embryo II, length 7 mm., the leg skeleton is not differentiated.
“a” represents there the approximate position where the pelvic blastema
will first become marked, as in Embryo CLXIII, length 9 mm. The
perpendicular falls on the body of the Ist sacral vertebra and points
toward the mid-thoracic region.

In Embryo CIX, length 11 mm., the perpendicular falls on the 1st
sacral vertebra; in CXLIV, length 14 mm., about at the junction of the
2d and 3d; in CVIII, length 22 mm., on the 3d; in CXLV, length
88 mm., at the junction of the 2d and 3d; in CLXXXIV, length 50 mm.,
on the anterior portion of the 3d.

At birth, judging from the figure of Fehling, the perpendicular would
strike at about the junction of the 2d and 3d. In the adult the area
where it strikes shows much individual variation, but in most of the
specimens which I have examined it strikes on the 2d sacral vertebra
not far from the junction of this with the 3d. In some specimens it
strikes the 3d. The material at my disposal has been chiefly dried
specimens from the dissecting room and has been subjected to some
warping.

Fig. 44 shows that when first differentiated the pelvis occupies a posi-
tion anterior to that which it takes when it becomes attached to the
vertebral column, but that after this attachment the position of the
central area of the acetabulum is altered but slightly with respect to the
sacral region of the vertebral column. At the beginning and at the end
of the period under consideration it probably occupies a position slightly
more anterior than that which it takes during the latter part of the
second and first part of the third month of development. The chief
alteration of the position of the pelvis with respect to the long axis of
the body is due to change in the position of the sacrum in relation to
the rest of the spinal column.

Merkel, 02, has contributed an important paper on the growth of the

“

286 Studies of the Development of the Human Skeleton

pelvis and refers to the previous hterature on that subject. He shows
that the sacrum and the innominate bones exhibit a certain independence
in rate of growth.

B. FEMUR AND HIP-JOINT, TIBIA, FIBULA, AND KNEE-JOINT.

The rapid development of the blastemal skeleton of the lower limb has
been briefly described above. Soon after the fundament of the femur
makes its appearance condensation of tissue marks out the anlage of the
tibia and fibula and the skeleton of the foot. This last seems to be at
first a somewhat irregular continuous sheet of tissue. It is not clear
whether or not the anlage of the tibia and fibula also begins as a con-
tinuous tissue sheet which becomes divided, by ingrowth of blood-vessels,
into tibial and fibular portions. The incomplete development of the
interosseous fissure in Embryo CIX, length 11 mm., Figs. 3 and 4, sug-
gests this. The blastemal anlages of the tibia and fibula are here very
incompletely separated.

In Embryo CIX the femoral blastema is continuous at one end with
that of the pelvis, at the other with that of the tibia and fibula and that
of the last two with the foot-plate.

Within the blastema of the femur, tibia and fibula chondrofication
begins as soon as the outlines of the blastemal skeleton are fairly com-
plete (Figs. 3 and 4). The embryonic cartilage appears slightly knee-
wards from the center of the shaft of each bone and then extends toward
the ends. In CIX, Figs. 3 and 4, and 46 to 53, the cartilage of the
femur consists of a bar largest at the knee whence it tapers off toward
the hip. The cartilages of the lower leg lie nearly in a common plane.
That of the tibia is larger than that of the fibula and toward the knee
it broadens out considerably. At this stage the joints consist of a solid
mass of mesenchyme, Fig. 55. The tissue uniting the femur and tibia
has something the appearance of precartilage.

The further development of the thigh and leg may be conveniently
studied by taking up at first the development of the femur and hip-joint,
and then that of the tibia, fibula and knee-joint.

The cartilagenous femur expands rapidly at the expense of the sur-
rounding blastemal perichondrium and at the same time acquires adult
characteristics.

In an embryo of 74 mm., CXLIV, Figs.’5 and 6, the shaft of the femur
extends almost directly into the hip-joint. Here there is a simple rounded
head, distal and dorsal to which a slight projection marks the beginning
of the great trochanter. There is nothing corresponding to a true

Charles R. Bardeen 287

“neck.” Similar conditions have been pictured by Hagen, oo, for the
His embryo So, length 17 mim.

In Embryo XXII, length 20 mm., Figs. 9 and 10, the head of the
femur is proportionately larger and between it and the great trochanter
the cartilage has developed in such a way as to give rise to a short neck.
Blastemal extensions serve to give attachment to the musculature of the
hip and indicate the lesser trochanter and the intertrochanteric ridge.
In Embryo CXLV, length 33 mm., Figs. 11 and 12, the cartilage has
extended into these projections and the main characteristics which dis-
tinguish the proximal end of the femur have become established. Even
at this stage, however, the neck is proportionately very short and thick.
In an embryo of 50 mm., LXXXIV, Fig. 13, the neck is relatively more
slender and the head of the femur has become more rounded.

The hip-joint is represented at first by a dense mass of scleroblastema,
Fig. 55. The development of the acetabulum by ingrowth and fusion
of processes from the iliac, ischial and pubic cartilages has already been
described. The cartilagenous joint-cavity is at first quite shallow, Fig.
56. But extension of cartilage into the blastemal tissue which passes
from the pelvis over the head of the femur serves greatly to deepen it
on all sides except in the region of the cotyloid notch.

The joint-cavity is at first completely filled with a dense blastemal
tissue, Fig. 56. While the embryo is growing from 20 to 30 mm. in
length cavity formation begins in the tissue lying between the cartilage-
nous fioor of the acetabulum and the head of the femur. The first stage in
the process is marked by a condensation of the capsular tissue immedi-
ately bordering upon the joint and of the perichondral tissue which at
this stage covers the cartilages on their articular surfaces as well as
elsewhere. In the region of the ligamentum teres a fibrous band is like-
wise differentiated from the blastema of the joint. ‘The rest of the tissue
becomes looser in texture and ultimately is absorbed, Fig. 57. Henke
and Reyher, 74, gave a good account of the development of the hip-joint.
Moser has discussed the ligamentum teres.

The shaft of the femur at the stages of Embryos CIX and CXLIV,
Figs. 3, 4, 5 and 6 is proportionately very short and thick. For a time
it then grows so rapidly that it may become distorted and bent from the
resistance offered at each end. But soon adjustment takes place between
the skeletal and the neighboring parts and the femur becomes straight
and slender. Yet in Embryo CXLV, length 33 mm., Figs. 11 and 12,
it is relatively thicker than in the adult.

The linea aspera is marked during the early development of the femur
by a thickening of the perichondrium in the region where the various

288 Studies of the Development of the Human Skeleton
muscle tendons and fascia are inserted. But in embryos up to 50 mm.
in length there is no extension of cartilage into this area. Since by this
period the shaft is ossified it is evident that no cartilagenous linea aspera .
is formed.

Ossification begins at an early period kneewards from the center of the
shaft. Endochondral calcification begins here in embryos about, or
slightly less than 20 mm. in length. Perichondral ossification usually
begins in embryos about 25 mm. long, although in Embryos LXXXVI
and LXXV, length 30 mm., the clavicle alone shows actual bone forma-
tion. Ossification of the femur takes place at about the same time as
that of the humerus, radius and ulna, and very slightly, if at all, precedes
that of the tibia. Ossification of the clavicle and the superior and in-
ferior maxillary bones seems always to begin a little earlier, that of the
scapula, ilium, occipital, and ribs, slightly later.

The distal extremity of the femur is large at an early period of differ-
entiation, Embryo CIX, Figs. 3 and 4. In Embryo CXLIV, length
14 mm., Figs. 5 and 6, it has expanded laterally and each lateral process
has extended dorsally so that fairly well-marked condyles are apparent.
These are better formed in Embryo XXII, length 20 mm., Figs. 9 and
10. In CXLYV, length 33 mm., Figs. 11 and 12, the form of the distal
extremity of the femur resembles the adult.

Fhe tibia and fibula at first lie nearly in the same plane, Embryo CIX,
length 11 mm., Figs. 3 and 4. As the head of the tibia enlarges toward
the knee-joint it comes to lie dorsal to the proximal extremity of the
fibula. This may be seen in Embryo CXLIV, Figs. 5 and 6, and more
marked in Embryo XVII, length 18 mm., Fig. 59; XXII, length
20 mm., Figs. 9 and 10; and CXLV, length 33 mm., Figs. 11 and 12.
In the last embryo the relations of the head of the fibula to that of the
tibia are nearly like the adult.

In Embryo CIX, Figs. 3 and 4, the fibula points toward the lateral
condyle of the femur and the tibia toward the median, but the long axis
of the femur much more nearly meets that of the tibia than that of the
fibula. As the head of the tibia enlarges the anterior extremity of the
long axis of the bone is carried toward the center of the distal end of the
femur while the head of the fibula is pushed toward the side, Figs. 5, 6,
59, 9, 10, so that the long axis of the fibula comes to point lateral to
the extremity of the femur. The head of the fibula is held in place by
ligaments developed from the skeletal blastema.

The development of the knee-jowt in man has been studied by a
number of competent observers. Bernays, in 1878, gave a good review
of the previous work of von Baer, Bruch, Henke and Reyher, and an

Charles R. Bardeen 289

accurate description of the processes which take place. Of the more
recent articles that of Kazzander, 94, deserves special mention.

Until the embryo reaches a length of about 17 mm. the knee-joint is
marked by a dense mass of tissue, Fig. 59. The medullary tissue at the
knee, like that at the hip and other joints, is less dense than the surround-
ing cortical substance, so that when the cartilages of the femur, tibia and
fibula are first differentiated they seem to be connected by a tissue which,
in some respects, resembles the prochondrium of which they are com-
posed, Fig. 55. But as the cartilages become more definite the apparent
continuity disappears. As the musculature becomes differentiated a
dense tendon for the quadriceps is formed in front of the knee-joint.
This is shown well in Fig. 56. In it the patella becomes differentiated.

In embryos of about 20 mm. the tissue immediately surrounding the
cartilages becomes greatly condensed into a definite perichondrium. The
peripheral blastemal tissue at the joints becomes transformed into a
capsular ligament strengthened in front by the tendon of the quadriceps.
Within the joint most of the tissue begins to show signs of becoming less
dense, Fig. 56, but the semi-lunar disks and the crucial ligaments, hke
the ligaments of the capsule are differentiated directly from the blastema,
Figs. 61 to 65. A kmnee-joint cavity first appears in embryos about
30 mm. long.

The shafts of the tibia and fibula are incompletely separated in the
blastemal stage. The cartilages which arise in the scleroblastema are,
on the other hand, separated by a distinct interval, Fig. 50, and as the
blastemal elements give way to cartilage the interosseous space becomes
larger. This is shown in Figs. 3, 4, 5, 6,9, 10, 11,12 and 59. At first
short and thick the shafts gradually become more slender in proportion
to their length. The fibula at all times smaller, becomes increasingly
more slender in comparison with the tibia. In embryos of 30 mm.,
Figs. 11 and 12, both bones, and especially the fibula, are relatively thick
compared with the adult bones.

During a period of rapid development, in embryos of 15 to 20 mm.,
the tibia and fibula, like the femur, may extend so rapidly in length as to
become temporarily distorted by resistance at the ends. This is often
especially marked in hardened specimens. Holl, 91, Schomburg, oo,
and others have called attention to this distortion.

Ossification begins in the tibia at about the same time that it does in
the femur and a little earlier than it does in the fibula. It is usually
under way in embryos 25 mm. long. In older embryos it is generally
well marked, Figs 11 and 12, Plate V. It begins in both bones knee-
wards from the center of the shaft and from here spreads toward the

290 Studies of the Development of the Human Skeleton

ends of the bones (Fig. 13, Plate V). The development of the distal
extremities of the tibia and fibula may best be taken up in connection
with the development of the foot.

C. ANKLE AND FOOT.

Of the papers dealing with the early development of the skeleton of
the human foot the more important are those of Henke and Reyher, 74,
Leboucgq, 82, v. Bardeleben, 83, 85, Lazarus, 96, and Schomburg, oo.

Since the work of Schomburg is the most recent of these and is based
on a considerable number of well-prepared embryos, I shall discuss his
results somewhat at length in connection with the results which I have
obtained. He recognizes four periods in the development of skeletal
structures, a mesenchymal, a prochondral, a cartilagenous and an osseous.
For the sake of ready comparison I shall take up each of these periods in
turn. The fourth period falls within the scope of this paper only in
so far as it overlaps the third.

Mesenchymal (blastemal) period ——This commences during the fifth
week of embryonic development. The free extremity of the limb-bud
becomes flattened and differentiated into the anlage of the foot and its
axial blastema becomes’ differentiated into a foot-plate, from which later
the bones of the foot are derived. Schomburg states that the axial blas-
tema becomes distinct at the end of the fourth week. In Embryos
CCXLI, length 6 mm.; II, length 7 mm.; CLXIII, length 9 mm.; and
CCXXI, length 13 mm.,’ I find no distinct signs of a foot-plate. In
each of the following embryos I find a foot-plate which has not distinctly
undergone further differentiation: CIX, length 11 mm.; CLXXV,
length 13 mm.; and CVI, length 17 mm. The last is a somewhat path-
ological specimen. In Fig. 3 a reconstruction of the foot-plate of CIX
is shown, in Figs. 51 and 52 transverse sections through this are repre-
sented, in Fig. 66 is pictured a longitudinal section through the foot-
plate of CLXXV.

Toward the end of the fifth week, in embryos usually 14 to 16 mm.
long, the first differentiation of definite bones is manifested by a conden-
sation of tissue in specific areas. Within these areas of condensed tissue
precartilage soon makes its appearance. Schomburg says that the first
metatarsal is differentiated distinctly later than the other metatarsals.
This I find to be the case in none of Prof. Mall’s embryos. I do, how-
ever, agree with Schomburg that the metatarsal bones become well
differentiated before the tarsals. When the metatarsals and phalanges

1See note 1, Table A, p. 277.

Charles R. Bardeen 291

become differentiated the portions of the foot-plate between them serve
for a short time to form a thick web, Fig. 67.

Prochondrium period.—Schomburg gives a detailed account of the
early differentiation of the anlages of the bones of the foot and illustrates
his belief as to their nature by several diagrams. Unfortunately he does
not picture the wax-plate reconstructions which he reports having made
of a number of early embryos. In Prof. Mall’s embryos I find no evi-
dence of the archipterygium-like conditions which Schomburg describes.
While it may be true that the somewhat slow development of cartilage in
the tarsus is owing to the great alterations from primitive conditions
which the human foot has undergone during its phylogeny, and to a cer-
tain extent has to repeat during its ontogeny, still the development of
the bones of the foot is far more direct than Schomburg’s diagrams
indicate. In the embryos studied I also fail to find the rudimentary
tarsal bones described by vy. Bardeleben, 83, 85. I have examined six
embryos between 15 and 20 mm. long without finding a trace of either
the os intermedium tarsi or the triangularis tarsi. In only one instance
have I found the I cuneiform distinctly portioned out into dorsal and
plantar divisions by a lateral fissure. Study of adult variation statistic-
ally, as so admirably carried out by Pfitzner, 96, for the foot, coupled
with comparative anatomy, in this, as in so many other fields, throws
more light on a possible phylogeny than is gained from ontological
investigation.

Embryo CXLIV, length 14 mm., is, of those I have studied, the
youngest showing definitely tarsal and metatarsal elements. The
general form of the skeleton is shown in Figs. 5 and 6. The differentia-
tion of the tarsal elements is difficult to make out, that of the metatarsals
is clear. Webs between the latter still persist, Fig. 67. Webbed digits
are sometimes found in the adult (Robertson).

It is to be noted that the elementary condition of the foot of CXLIV
corresponds with none of the diagrams given by Schomburg. On the
whole the cartilagenous anlages have a position much more nearly resem-
bling the adult. Embryo XLII, length 16 mm., exhibits pedal char-
acteristics almost identical with those of CXLIV.

It may here be mentioned that in none of the embryos I have studied
is the fibula so long as the tibia. Schomburg states that at first it is
longer. ' |

The metatarsals when first formed are spread wide apart and gradually

become approximated. The diagrams of Schomburg indicate a different
condition.

292 Studies of the Development of the Human Skeleton

Cartilagenous period.—This Schomburg distinguishes from the pre-
ceding by the fact that cartilage cells at the centers of the areas of
chondrofication show definite cell boundaries and become larger than the
surrounding prochondral cells. These changes take place in the various
skeletal anlages in the order in which the anlages were originally formed.
With the active production of cartilage cells the broad surrounding zone
of mesenchyme gives way to a narrower, denser perichondrium. At the
same time the form of the skeleton becomes more definite, so that, as
Schomburg says, the cartilages of the foot of an embryo at the middle of
the third month give a good picture of the adult bones of the foot. The
articular surfaces acquire more or less their definite form. I quite agree
with Schomburg, in opposition to Henke and Reyher, that the joints of
the foot, like the other joints of the body, are laid down at the start in
their definite form and are not moulded into shape by use.

The skeleton of the foot at the time when the cartilage cells at the
centers in most of the bones are beginning to be distinctly outlined, has
the form shown in Figs. 7, 8 and 59. The tibia is much larger than the
fibula and extends further distally. The astragalus has somewhat the
form of a rhomboid plate which runs dorsally from the fibular side
toward the tibial side on the plantar surface. The calcaneus is rather
small and is in direct line with the long axis of the fibula but in a plane
lying further plantarwards. The navicular is in a direct line with the
astragalus. Its tibial edge lies near the lower end of the tibia. The
three cuneiform bones are proportionately broader and thicker than in
the adult skeleton. The cuboid is in direct line with the calcaneus.
The metatarsals lie less spread apart than at an earlier stage, Figs. 5
and 6. The first phalanx has developed in all of the toes, and in the
second toe, the second phalanx as well. At the region of the phalangeal
joints there is a swelling of the blastemal tissue.

If now these figures be compared with Figs. 9 and 10, which show the
foot of an embryo of 20 mm., the most noticeable change will be seen in
the astragalus. This has become considerably thicker. It extends fur-
ther than the calcaneus. Between the tibia and the navicular it has so
increased in size that the foot is bent toward the fibular side. A much
greater interval than in Embryo XVII, Figs. 7, 8 and 59, exists between
the two bones.

The calcaneus has extended considerably in length both in a proximal
and in a distal direction. The cuneiform bones are becoming crowded
together. The cuboid is larger than in XVII. The phalanges are at a
similar stage of development. The joints between the metatarsals and
phalanges are surrounded by a mass of dense tissue, while the tissue
of the joints themselves is of a light texture and resembles prochondrium.

Charles R. Bardeen ; | 293

In Embryo CXLV, length 33 mm., Figs. 11 and 12, the process of
cartilage formation has given rise to structures which resemble adult
bones. The tibia has greatly expanded at its distal extremity and now
articulates.directly with the fibula. These two bones in turn articulate
with the well-developed superior articular process of the astragalus. The
malleolar process of the tibia is larger and extends further distal than that
of the fibula. In an embryo of a corresponding age, however, Schomburg
shows that the fibula extends further distal than the tibia. Individual
variation may exist.

The astragalus exhibits perhaps more marked alterations in form than
any other bone of the foot during the period when the embryo is growing
from 20 to 30 mm. in length. Toward the tibia and fibula it develops
a well-marked articular process. While this resembles closely the similar
process in the adult it is less developed on its fibular side than it is in
the adult. As Schomburg has shown the definite adult form is not
reached before the fourth month. ‘Toward the calcaneus the bone is
well developed and against it exhibits the two characteristic articular
surfaces. The posterior of these, compared with the adult, is relatively
undeveloped. Distally the bone sends forth a rounded process to articu-
late with the navicular. In the material at my disposal the whole com-
plex astragalus seems to arise from a single primary center.

The calcaneus, like the astragalus, undergoes marked changes in form
during the latter part of the second and the first part of the third month
of development. ‘Toward the heel a well-marked tuberosity has made its
appearance in Embryo CXLV, Figs. 11 and 12. Distally the bone ex-
tends to form a joint with the cuboid. Tibially it has developed a
sustentaculum tali for articulation with the astragalus. It is still, how-
ever, short in proportion to its width as compared to the adult.

The nayicular exhibits no marked changes. On its plantar side and
tibial edge it shows a distinct tuberosity.

The cuneiform bones are crowded together and have their character-
istic wedge shape. The internal cuneiform is the largest and extends
farthest distal. The middle is the smallest.

The cuboid shows a tuberosity. The phalanges, all of which are de-
veloped, present no points of special interest.

The joint-cavities begin to develop while the embryo is growing from
25 to 30 mm. in length. As in other cases, so here the blastemal tissue
in which the cartilages are developed becomes condensed at their articu-
lating ends and about the joint, while in the region of the joint the tissue
becomes less dense and finally disappears leaving a joint-cavity. In

294 Studies of the Development of the Human Skeleton

embryos of about 30 mm. the joint-cavities of the foot are filled with a
loose fibrous tissue, in embryos of 50 mm. definite cavities are to be made
out. The sesamoid bones develop later than the period to which this
investigation extends.

During the progress of form differentiation above described the shape
of the foot is markedly altered. At the beginning of the development of
the foot the tarsal and metatarsal bones lie nearly, though not quite, in
the same plane as the bones of the leg, Figs. 7, 8 and 59. They are so
arranged, however, that the foot is convex on its dorsal surface and con-
cave on the plantar, and the projections of the calcaneus and astragalus
serve to deepen the plantar fossa. The metacarpals spread widely apart.
As differentiation proceeds the metacarpals come to lie more nearly
parallel to one another and the tarsal elements become compacted in such
a way as to give rise to the tarsal arch. The foot at the same time is
flexed at the ankle and turned slightly outwards. The toes are flexed.
Fig. 68 shows the extent of the tarsal arch in an embryo of 23 mm.

In the further development of the skeleton of the foot the various
constituent structures are elaborated and the foot gradually becomes
more flexed and turned toward the fibular side. Yet even in the infant
the head of the astragalus is directed more inwards than in the adult.
Leboucq, 82, pointed out that the first metatarsal is relatively short in
the foetus and points more toward the tibial side than later.

Ossification—This begins in the metatarsals and phalanges during
the third month and is perichondral in nature. The tarsals begin to be
ossified considerably later. The center for the calcaneus appears in the
sixth month, that for the astragalus in the seventh month of foetal life.
The ossification of the other bones begins during the first five years of
life. Authorities differ as to the exact time at which the process begins
in the various bones. In Quain’s Anatomy the following dates are given:
cuboid, at birth, external cuneiform, 1st year; internal cuneiform,
3d year; middle cuneiform, 4th year; navicular, 5th year.

I have studied the ossification in the third and fourth months of embry-
onic life. In an embryo about 4 em. long, cleared according to the
Schultze method, I have found centers of ossification in the 2d, 3d and
4th metatarsals, and in the terminal phalanx of the big toe of each foot.
In Embryo XCVI, length 44 mm., there is a very thin layer of bone
being laid down about the center of the shaft of the 2d, 3d and 4th meta-
tarsals. I have been unable definitely to determine whether or not bone
has been deposited in the terminal phalanx of the big toe. In Embryo
XCV, length 46 mm., ossification has begun in the 2d, 3d and 4th meta-
tarsals and in the terminal phalanges of the Ist and 2d toes; in Embryos

or

Charles R. Bardeen 29

LXXXIV and CLXXXIV, length 50 mm., it is apparent in the 2d, 3d
and 4th metatarsals and in the terminal phalanges of the first three toes.
In a cleared embryo, 6 cm. long, there are centers of ossification in all of
the metatarsals and terminal phalanges; in one, 8 cm. long, in the first
two basal phalanges as well; while in one, 10 em. !ong, ossification has
begun in all of the metatarsals and the basal and terminal phalanges.
We may therefore conclude that ossification in the foot begins in the
three central metatarsals and in the terminal phalanx of the first toe
toward the end of the third month, and that it is thence extended to the
other metatarsals and terminal phalanges before beginning in the basal
phalanges.

For a consideration of the development of the individual bones of the
foot reference may be made to the excellent paper of Schomburg, oo.
The chief points in which my observations conflict with what he describes
have been pointed out above. Hasselwander, 03, has recently published
a good account of the ossification of the bones of the foot; and Spitzy, 03,
of the structure and development of the infant foot.

SUMMARY.

In general the development of the skeleton of the limb in man corre-
sponds closely with that which is known to take place in other digitates
and which has been recently admirably summarized by Braus, 04. Three
stages may be recognized, a blastemal, a chondrogenous and an osseoge-
nous. During the chondrogenous stage the chief features of form are
reached which characterize the adult structure. The centers for chondro-
fication correspond closely with those for ossification. The development
throughout is fairly direct. No distinct evidences of phylogenetic
structures discarded during ontogeny were found in the embryos studied.

LITERATURE QUOTED.

Arpy.—Die Alterverschiedenheiten der menschlichen Wirbelsatile. Archiv. f.
Anatomie und Physiologie, Anat. Abth., 1879, p. 77.

ANCEL ET SENCERT.—Variation numerique de la colonne vertebrale. Comptes

rend. Assoc. des Anat. Lyon, 1901, p. 158-165.

Les variations des segments vertebro-costaux. Bibliographie Anatom.,

X, pp. 214-239, 1902.

Des quelques variations dans le nombre des vertebres chez l’homme.

Journal de l’Anatomie et de la Physiologie, XXXVIII, 217-258, 1902.

Babe, P.—Entwickelung des menschlichen Fuss-skelets von der neunte Embry-

onalwoche bis zum 18 Jahre nach Rontgenbildern. Verh. d. gesellsch.

deutschen Naturf. u. Aerzte 1899-1900, 463-466.

Die Entwickelung des menschlichen Skelets bis zum Geburt. Archiv.

f. mikr. Anatomie, LV, 245-290, 1900.

296 Studies of the Development of the Human Skeleton

BALLANTYNE.—Spinal column in infants. Edinburgh Medical Journal, 1892.

BARDEEN.—Costo-vertebral variation in man. Anatomischer Anzeiger XVIII,
p. 3877, 1900.

Vertebral variation in the human adult and embryo. Anatomischer
Anzeiger, XXV, p. 497, 1904.

Development of the thoracic vertebre in man. American Journal of
Anatomy, IV, p. 163, 1905.

BARDEEN AND LEWwISsS.—Development of the limbs, body-wall and back in man.

American Journal of Anatomy, Vol. I, p. 1, 1901.

vy. BARDELEBEN.—Das Intermedium tarsi beim Menschen. Sitzungsber. d.
Jenaische Gesellschaft f. Med. und Naturw., f. d., Jahr 1883.

Zur Entwickelung der Fusswurzel. Sitzungsb. d. Jenaische Gesell-
schaft f. Med. und Naturw., f. d., Jahr 1885. Suppl. Bd., XIX.

Hand und Fuss, Verhandl. d. anat. Gesellsch. auf der 8 Versammlung,
1894.

BERNAYS, A.—Die Entwickelungsgeschichte des Kniegelenkes des Menschen
mit Bemerkungen Uber die Gelenke im Allgemeinen. Morphol.
Jahrbuch IV, 403, 1878.

BoLtk.—Ueber eine Wirbelsatile mit nur 6 Halswirbeln. Morph. Jahrb.,
XXIX, 84-93, 1901.

Braus.—Tatsachliches aus der EHEntwickelung des Hxtremitaten Skelets bei

den niedersten Formen. Festschrift f. Haekel, 1904.

Die Entwickelung der Form der Extremitaéten und des Extremitaten
skelets. Hertwigs Handbuch der Entwickelungsgeschichte der
Wirbelthiere., 1904.

CUNNINGHAM.—Proportion of bone and cartilage in the lumbar section of the
vertebral column of apes and several races of man. Journal of
Anatomy and Physiology, 1889, p. 117.

Dwicut, TH.—Methods of estimating the height from parts of the skeleton.

Medical Record, 1894.

Description of human spines. Memoirs Boston Society of Natural
History, V, 287-312, 1901.

A transverse foramen in the last lumbar vertebra. Anatomischer
Anzeiger XX, 571-572, 1902.

FEHLING, H.—Die Form des Beckens beim Foetus und Neugeborenen und ihre

Beziehung zu der beim Erwachsenen. Archiv. f. Gynekologie,
X, 1876.

FLOWER.—Osteology of the Mammalia.

Fou.—Sur la queue de ’embryon humain. Comptes Rendus de l’academi, C,
p. 1469, 1885.

FURBRINGER.—Morphologische Streitfragen. Morph. Jahrbuch XXX, 1902.

GEGENBAUR.—Vergleichende Anatomie der Wirbelthiere mit Beriicksichtigung
der Wirbellosen, 1, 1898.

GRUBER.—Ueber die Halsrippen des Menschen mit vergleich. anat. Bemer-
kungen. Memories de l’Acad. des Sciences de St. Petersbourg, XIII,
1869, No. 2.

HAGEN.—Die Bildung des Knorpelskelets beim menschlichen Embryo. Archiv.
f. Anatomie und Physiologie, Anat., Abth., 1900, p. 1.

Charles R. Bardeen _ 297

Harrison, R. G.—On the occurrence of tails in man with a description of the
case reported by Dr. Watson. The Johns Hopkins Bulletin, XII, 121-
129, 1901. :

HASSELWANDER.—Untersuchungen iiber die ossification des menschlichen
Fuss-skelets. Zeitschrift f. Morphologie und Anthropologie, V, 438-
508, 1903.

HENKE UND Reyuer.—Studien tiber die Entwickelung der Extremitaten des
Menschen insbes. der Gelenkflaschen. Sitzungsb. d. K. Akad. d.
Wiss. Math.-naturw. Klasse Wien, LXX, 3d Pt., p. 217, 1874.

Horii, M.—Ueber die richtige Deutung der Querfortsaitze der Lendenwirbel

und die Entwickelung der Wirbelsaiile des Menschen. Sitzungsb. d.
K. Akad. d. Wiss. Math.-naturw. Klasse. Wien. LXXXV, pp. 181-
232, 1882.

Ueber die Entwickelung der Stellung der Gliedemassen des Menschen.
Sitzungsb. d. K. Acad. d. Wiss. Math.-naturw. Klasse. Wien., C, 3d
12h 105 al alkene

KAZZANDER, G.—Sullo svilluppo dell’ articolazione del ginocchio. Monitore
Zoologico Italiano, V, p. 220, 1894.

KLaatscu, H.—Die wichtigsten Variationen am Skelet der freien unteren Ex-
tremitat des Menschen und ihre Bedeutung f. das Abstammungs-
problem. Ergebnisse d. Anatomie und Entwickl., X, pp. 599-719, 1900.

Konixow, M.—Zur Lehre von der Entwickelung des Beckens und seiner
geschlechtlichen Differenzierung. Arch. f. Gynekologie, XLV, p. 19,
1894.

LazArus.—Zur Morphologie des Fuss-skelets. Morph. Jahrbuch, XXIV, 1896.

Lresoucg.—Le développement du premier metatarsien et de son articulation
tarsienne chez homme. Archives de Biologie III, 335, 1882.

Recherches sur les variations anatomiques de la premiere céte chez
Vhomme. Archives de Biologie, XV, p. 125, 1898.
Low.—Description of a specimen in which there is a rudimentary first rib,

with thirteen pairs of ribs and twenty-five presacral vertebre. Jour-
nal of Anatomy and Physiology, XXXIV, 451-457, 1901.
LUBSEN, J.—Zur morphologie des Ilium bei Satigern. Overdruk nit Petrus
Camper, Dl. II, Afl. 3.

MEHNERT, E.—Untersuchungen tiber die Entwickelung des Beckengiirtels bei
einigen Saiigethieren. Morphol. Jahrbuch, XVI, pp. 97-112, 1889.

MeERKEL.—Beckenwachstum: Anat. Hefte I, 121-150, 1902.

Mo.iurer.—Die paarigen Extremitaten der Wirbelthiere. Anatomische Hefte,
SOS el oo Hs Loot.

Moser.—Das Wachstum der menschlichen Wirbelsatile, Dissertation Strass-

burg, 1889.

Ueber das Ligamentum teres des Hitiftgelenkes. Schwalbes Arbeiten,
Tips o0:

PAPILLAULT, G.—Variations numeriques des vertébres lumbaires chez homme.

Bulletins de la Soc. d. Anthropologie de Paris, IX, 198-222, 1900.

Parerson, A. M.—The human sacrum. Scientific transactions of the Royal
Dublin Society, V, p. 123, 1893.

PETERSEN.—Untersuchungen zur Entwickelung des menschlichen Beckens.
Arch. f. Anatomie und Physiologie, Anat., Abtheilung, 1893, pp.
67-96. <

23

298 Studies of the Development of the Human Skeleton

PFITZNER, W.—Die Variationen im Aufbau des Fuss-skelets. Morphol. Ar-
beiten, VI, 245, 1896.

Posty, M.—Le Sacrum. Thesis, Paris, 1897.

Rasit.—Gedanken und Studien tiber den Ursprung der Extremitaten.

Zeitschr. f. wiss. Zool. LXX, 474-558, 1901.

Ueber einige Probleme der Morphologie. Anat. Anz. Erganz. Hefte,

MEX, 1903:

RAMBAUD AND RENAULT.—Origine et développement des Os. Paris, 1864.

RAVENAL.—Die Massenverhaltnisse der Wirbelsatile und des Riickenmarks
beim Menschen. Zeitschrift f. Anatomie und Entwickelungsg., II,
343, 1877.

RETTERER.—Ebauche squelettogéne des membres et développement des articu-
lations. Journal de |’Anatomie et Physiologie, XXXVIII, 473-509,
580-6238, 1902.

ROBERTSON, W. G.—A case of supernumerary and webbed fingers. Edinburgh
Medical Journal, XIV, 535-536.

RomitT1, G.—Sui caratteri sessuali nel bacino del neonato. Atti della societa
Toscana di Science naturali, VIII, 1892, Pisa.

ROSENBERG, H.—Ueber die Entwickelung der Wirbelsaiile und das centrale

carpi des Menschen. Morphol. Jahrbuch, I, p. 83, 1876.

Ueber eine primitive Form der Wirbelsaiile des Menschen. Morphol.

Jahrb., XXVII, 1-118, 1899.

RucKeErtT.—Ossif, des menschl. Fussskelts. Sitzungsb. d. Koenig. Bay. Akad.
Miinchen, Mat. Nat. Kl. 1901, 65-72.

RuGe.—Die Entwickelung des Skelets der vordern Extr. von Spinax niger.
"Morphol. Jahrb., XXX, 1-27, 1892.

ScHomBurG, H.—Untersuchungen der Entwickelung der Muskeln und
Knocken des menschlichen Fusses. Dissertation, Gottingen, 1900.

Spitzy, H.—Ueber Bau und Entwickelung des kindlichen Fusses. Jahrb. f.
Kinderheil, 1908.

STEINBACH.—Die Zahl der Caudalwirbel beim Menschen. Dissertation. Ber-
lin, 1889.

SZAWLOWSKI.—Ueber einige Seltene Variationen an der Wirbelsatile beim
Menschen. Anatomischer Anzeiger, XX, 305, 1901.

-TENCHINI.—Die una nuova maniera di compenso nelle anomalie numeriche
vertebrali dell’ uomo. Archivio per l’Anthropologia, XXIV, Firenze,
1894.

THILENIUS.—Untersuchungen tiber die morphologische Bedeutung accesso-
rischer Elemente am menschlichen Carpus (und Tarsus). Morphol.
Arbeiten, V, 1895.

Tuomson, A.—The sexual differences of the fcetal pelvis. Journal of Ana-
tomy and Physiology, XXXIII, pp. 3 and 359-380, 1899.

TopInaRD, P.—Anomalies de nombre de la colonne vertébrale chez l’homme.
Revue d’ Anthropologie, VI, 577, 1877.

UNGER AND Brucscu.—Zur Kenntniss der Fovea und Fistula sacro-coccygea s.
caudalis, etc. Archiv. f. mikr. Anatomie, LXI, 151-219, 1903.

VrEIT.—Die Entstehung der Form des Beckens. Zeitschr. f. Geburtsh. und
Gynekologie, IX, 347, 1889.

WIEDERSHEIM.—Das Gliedmassenskelet der Wirbelthiere, Jena, 1892.

Charles R. Bardeen 299

LIST OF ABBREVIATIONS USED IN LETTERING THE FIGURES.

A. A. Pr.—Anterior articular process.

Cel.—Celom.
Chd.—Chorda dorsalis.
Co.—Coccygeal.

C. Pr.—Costal process.
Cu.—Cuboid.

C. V.—Cardinal vein.
Der.—Dermis.

Disk.—Intervertebral disk.
D.L.— Dorsal ligament.

D.M.—Dorsal musculature.

F. D. M.—Fascia of dorsal musculature.

Fi.— Fibula.
F.N.—Femoral nerve.
F.. Pl.— Foot-plate.
F. v. E.—Fissure of v. Ebner.
H. Pr.—Hemal process.

Ids. M.— Interdiscal membrane.
Idr. M.—Interdorsal membrane.

Il. Bl.—lliac blastema.
Is.—Ischium.

Is. A—Intersegmental artery.

L.—Lumbar.
Ls. Pl.—Lumbo-sacral plexis.
L. T.—Ligamentum teres.
Myo.—Mvyotome.

M.D. R.—Membrana reuniens dorsalis.

N. Pr.—Neural process.
O. F.—Obturator foramen.
O. N.—Obturator nerve.
Pd.—Pedicle.
Pch. S.—Perichordal sheath.
P.L.—Poupart’s ligament.

P. A. Pr.—Posterior articular process.

P.—Pubis.
Rib.—Rib.
S.—Sacral vertebra.
S. Bl.—Scleroblastema.
S.N.—Sciatic nerve
Sptm.—Perichordal septum.
Sp. C.—Spinal chord.
Sp. G.— Spinal ganglion.
Sp. N.—Spinal nerve.
Sp. Pr.—Spinous process.
T. R.—Tendon of r. abd. muscle.
T.—Thoracic.

Ti.— Tibia.
Trap.—Trapezius muscle.
Tr. Pr.—Transverse process.
V. L.—Ventral ligament.

V. B.—Vertebral body.

DESCRIPTION OF PLATES.

PLATES I-V.

Fics. 1-12. A series of figures drawn from models made by the Born
wax-plate method. Fig. 13 from an embryo cleared by the Schultze alkaline,

glycerine method.

PLaTE I.

Fie. 1. Skeleton of Embryo II, length 7 mm. About 20 diam. In part
the reconstruction was made free-hand from drawings.

PEATE Ll,

Fie. 2. Right half of the distal portion of the skeleton of Embryo CLXIII,

length 9 mm. 25 diam.

Fics. 3 and 4. Right half of the distal portion of Embryo CIX, length 11
mm. 25 diam. (4) lateral, (5) median, view. The prochondrium of the
pubis, ilium, ischium, femur, tibia and fibula are represented by stippling.

PLATE III.

Fics 5 and 6. Distal portion of the right half-skeleton of Embryo
CXLIV, length 14mm. 25 diam. The prochondrium of the neural arches, the

300 Studies of the Development of the Human Skeleton

vertebral bodies, the ilium, ischium, femur, tibia, fibula and of the bones of
the foot is represented by stippling. The last are but slightly differentiated
at this period.

PLATE IV.

Fies. 7 and 8. Dorsal and plantar views of the cartilages of the left leg
and foot of Embryo XVII, length 18 mm. 20 diam.

Fies. 9 and 10. Lateral and median views of the distal portion of the right
half of the cartilagenous skeleton of Embryo XXII, length 20 mm. 20 diam.

PLATE V.

Fies. 11 and 12. Median and lateral views of the distal portion of the right
half of the cartilagenous skeleton of Embryo CXLV, length 33 mm. 10 diam.
The centres of ossification of the femur, tibia and fibula are shown.

Fic. 18. Lateral view of the left leg of an embryo 5 cm. long. 5 diam.
For the sake of facilitating comparison, a mirror picture has been drawn
and a technique has been used similar to that employed for illustrating the
models. The centers of ossification of the ilium, femur, tibia, fibula, the
three middle metatarsals and the terminal digits are shown. The position
of the various structures has probably been somewhat distorted during the
preparation of the specimen.

PLATES VI and VII.

Fies. 14-28. Transverse sections through the twelfth thoracic and first
two lumbar vertebre of a series of embryos. 14 diameters. Figs. 14-16,
Embryo CLXXV, length 13 mm.; Figs. 17-19, Embryo CCXVI, length 17 mm.;
Figs. 20-22 Embryo XXII, length 20 mm.; Figs. 23-25, Embryo XLV, length
28 mm.; Figs. 26-28, Embryo LXXXIV, length 50 mm.

PLATE VIII.

Fics. 29-36. Transverse sections somewhat oblique through several ver-
tebre of various embryos. 14 diam. Fig. 29, 4th sacral vertebra of Embryo
CIX, length 11 mm.; Figs. 30-32, 1st, 2d, and 3d sacral vertebre of Embryo
X, length 20 mm.; Figs. 33 and 34, Ist and 2d sacral vertebre of Embryo
CCXXVI, length 25 mm.; Fig. 35, 2d sacral vertebra of Embryo XLV, length
28 mm.; Fig. 36, 2d sacral vertebra of Embryo LXXXIV, length 50 mm.

PLATE IX.

Fics. 37-40. Obliquely cut frontal sections through the sacral region of
several embryos. 14 diam. Fig. 37, Embryo CLXXV, length 13 mm.; Fig.
38, Embryo CCXVI, length 17 mm.; Fig. 39, Embryo CLXXXVIII, length
17 mm.; Fig. 40, Embryo LXXXVI, length 30 mm.

PLATE X.

Fies. 41-48. Sections through the coccygeal region of several embryos.
14 diam. Fig. 41, frontal section of Embryo CLXXV, length 13 mm.; Fig.
42, sagittal section of Embryo CXLV, length 33 mm.; Fig. 43, frontal section
of Embryo LXXXIV, length 50 mm.

Charles R. Bardeen 301

Fic. 44. Diagram to show the curvature of the spinal column, the pro-
portional lengths of the various regions, the relation of the acetabula to the
sacral region and the direction of the long axis of the femur in a series of
embryos 7 to 50 mm. in length, and in an adult. Each curved line repre-
sents the chorda dorsalis of an individual. The cervical, lumbar and coccy-
geal regions of this are represented by the heavy, the thoracic and sacral by
the light portions of the line. The approximate position where a line join-
ing the centers of the two acetabula would cut the median plane is repre-
sented at “a.” For Embryo II, in which the skeleton of the leg is not yet
differentiated the position of the future acetabula is deduced from Embryo
CLXIII, length 9 mm. (See Fig. 2.)

The line passing in each instance from ‘“‘a” and terminating in an arrow
point represents the long axis of the femur. For Embryo II, this line is
pointed toward the centre of the tip of the limb-bud. From ‘‘a” in each
instance a perpendicular is dropped to a line connecting the two extremities
of the sacral region. The numbers refer to the following embryos:

ise USS SCORE RED ODI Olio olsc lo Length 7? mm.
TOK), AGHIGS 3 Sas Or PoROIe aid td tn OMI Old Sido otic n Ely 3s
MAAS CUNT Wey ais, 5 cneve cause aie Memeoecrete etsiaene s i
HO) SER CWB Tore rerrevevies arta vice toner Poe ens atalel eee he ae 20
BUA Bram @ENCIUIVE siev's: sae: a ave reteus, eeu aeet tiene roveieore le aaa ters a Gysy
US A CT NOXOC TING |. 5! hcp staan hohtets) ofeseysteeters Y Ro
Ad. Adult.

PLATE XI.

Fie. 45. Longitudinal section through the center of the limb-bud of
Embryo CLXIII. 14 diam. Compare with Fig. 2.

Fias. 46-52. A series of cross-sections through the right leg of Embryo
CIX, length 11 mm. -

Fie. 58. Outline of the blastemal skeleton with the regions marked
through which the sections 46-52 pass. 14 diam. Compare with Figs. 3
and 4.

PLATE XLT:

Fie. 54. Section from Embryo CXLIV, length 14 mm., showing the
pubic, iliac and ischial cartilages. 14 diam.

Fic. 55. Section passing longitudinally through the femur and tibia of
Embryo CLXXV, length 13 mm. A portion of the foot-plate is shown cut
obliquely. 14 diam.

Fic. 56. Longitudinal section through the ilium, femur, and tibia of
Embryo XXII, length 20 mm. 14 diam.

Fic. 57. Section through the pubis, ilium, ischium and head of the femur of
Embryo CCXXVII, length 30 mm. The hip-joint cavity shows well. It does
not extend into the region of the ligamentum teres. 14 diam.

Fic. 58. Section through the ilium, ischium and head of the femur of
Embryo LXXIX, length 33 mm. Calcification is beginning in the ilium.

Pram XL:

Fig. 59. Section through the leg and foot of Embryo XVII, length 18 mm.
The section does not pass through the cartilage of the 1st metatarsal.

302 Studies of the Development of the Human Skeleton

Fic. 60. Section through the pubis, ischium, femur, fibula, calcaneus, cuboid
and the 4th metatarsal cartilages of Embryo LXXIV, length 16 mm. 14
diameters.

Figs. 61-65. Sections through the knee-joints of several embryos. 14
diam; 61, CCXXIX, length about 20 mm.; 62, LXXXVI, length 30 mm.; 63,
LXXV, length 30 mm.; 64 and 65, CXLV, length 33 mm.

Fic. 66. Longitudinal section through the knee-joint, tibia and foot-plate
of Embryo CLXXYV, length 13 mm.

Fia. 67. Section through the foot of Embryo CXLIV, length 14 mm.

Fia. 68. Section through the foot of Embryo LVII, length 23 mm.

The models from which the illustrations in this article were drawn
have been reproduced by Dr. B. EB. Dahlgren at the American Museum
of Natural History, New York, N. Y., and arrangements may be made
for securing copies by applying to the Director of the Museum.

PLATE

DEVELOPMENT CF THE HUMAN SKELETON

Cc. R. BARDEEN

LENGTH 7 MM

Il,

AMERICAN JOURNAL OF ANATOMY--VOL. IV

: 4 . = a) al x ‘> Se sR et ee
: eae tg sag Ohana ae

ad

et &~

7 = + :

F . 7 : 2

= * a /-

~ , # e ‘ |

DEVELOPMENT OF THE HUMAN SKELETON ; PEAWHEN sil
Cc. R. BARDEEN

Fig.3
CIX, LENGTH 11 MM.

Fig At,

CLXIIl, LENGTH 9 MM. CIX, LENGTH 11 MM.

AMERICAN JOURNAL OF ANATOMY=-VOL. IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE III
Cc. R. BARDEEN mi

“Sy

CXLIV, LENGTH 14 MM.

Fig.

CXLIV, LENGTH 14 MM.

AMERICAN JOURNAL OF ANATOMY--VOL. IV

wT. a? See

ope ee-is
A

DEVELOPMENT OF THE HUMAN SKELETON
Cc. R. BARDEEN

Fig.4

XXIl, LENGTH 20 MM.

Fig.10

XXIl, LENGTH 20 MM

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE V
Cc. R. BARDEEN

CXLV, LENGTH 33 MM.

Fig.12

CXLV, LENGTH 33 MM.

Fig.13

LXXXIV, LENGTH 50 MM.

AMERICAN JOURNAL OF ANATOMY--VOL. IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE VI
Cc. R. BARDEEN

\,

PA.PrT.12 PA.PrTAY
AAPr.LA AAPITI2 BA

~—-.

S \aaert2 ”

Pe cer.

AMERICAN JOURNAL OF ANATOMY--VOL. IV

2
-

. >
“i ‘ y -_ ad .. P Y
: ? =a" ae (2 oT rary
7 - sae > hy rc
‘ ? .* , a
~ *¢ Pu f

’

_ DEVELOPMENT OF THE HUMAN SKELETON
Cc. R. BARDEEN :

PLATE VII

PA.Pr.T.Al

AA.PrT.12

AA.PrLl

CPR

AMERICAN JOURNAL OF ANATOMY--VOL. IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE VIII
C. R. BARDEEN :

P.A.Pr.l5 £m

P.A.Pr.S4. f

g

A.A.Pr.S2

Fig.32

PA.Prs1

AA.Pr.S2F

AMERICAN JOURNAL OF ANATOMY--VOL, IV

yy

4

wattal®

DEVELOPMENT OF THE HUMAN SKELETON PLATE |X
Cc. R. BARDEEN

AMERICAN JOURNAL OF ANATOMY--VOL. IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE X
Cc. R. BARDEEN

Fig.44

AMERICAN JOURNAL OF ANATOMY--VOL. IV

DEVELOPMENT OF THE HUMAN SKELETON
Cc. R. BARDEEN

PLATE XI

AMERICAN JOURNAL OF ANATOMY--VOL. IV
24

DEVELOPMENT OF THE HUMAN SKELETON PLATE XIl
C. R. BARDEEN

AMERICAN JOURNAL OF ANATOMY=-VOL, IV

DEVELOPMENT OF THE HUMAN SKELETON PLATE XIII
C. R. BARDEEN

AMERICAN JOURNAL OF ANATOMY--VOL. IV

A COMPOSITE STUDY OF THE SUBCLAVIAN ARTERY IN
MAN.

BY
ROBERT BENNETT BEAN, M. D.
Assistant in Anatomy, Johns Hopkins Medical School, Baltimore, Md.

WITH 7 FIGURES AND 18 TABLES.

Several years ago Hitzrot* made a study of the axillary artery based
upon records made in the Anatomical Laboratory of the Johns Hop-
kins University. ‘To supplement this, the following study of the sub-
clavian artery was made at the suggestion of Dr. Harrison. The clini-
cal features relating to the artery are given in another article.”

That there is need for further data concerning the ramifications of
this artery is apparent when the accompanying figures, taken from a
number of universally recognized authorities, are compared. From them
it is seen, that, while certain branches such as the vertebral and in-
ternal mammary are represented in the same manner by all, there is the
widest divergence with regard to the other branches.

The records which underlie the present study were made by myself,
from the dissections by students of anatomy, upon Bardeen’s charts.’
Dissections from 129 subjects are recorded, 60 from the left side of the
body and 69 from the right side. Some of these records are complete to
the minutest detail; nearly all give the origin of the main branches;
while a few are incomplete, giving only the subclavian artery and some of
its branches, or only a few branches without the subclavian artery. The
distribution of the vertebral artery inside the skull was not worked out,
because many of the cadavers were not obtained until after the brain had
been removed. The distribution of the internal mammary artery was
worked out completely in but 28 cases because of the removal of the ster-
num at the autopsy in the others.

1Hitzrot, Johns Hopkins Hospital Bulletin, Vol. XII, 1901.

?Bean, Johns Hopkins Hospital Bulletin, Vol. XV, 1904.

’ Bardeen, Outline Record Charts, Johns Hopkins Press, Baltimore, 1900.
AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

Fie. 1. Branches of the subclavian artery according to different authors. A,
according to Quain, Testut and Gray; B, according to Henle; C, according to Tiede-
mann; D, according to Spalteholz and Toldt (B. N. A); #, according to Gegenbaur;
F’, according to Sappey.

The lettering on all the figures is alike and as follows: I, II and III, the three
parts of the subclavian artery; A. V., arteria vertebralis; A. M. J, arteria mam-
maria interna; 7. 7. C., truncus thyreo-cervicalis; A. 7. /., arteria thyroidea inferior;
A. T. S., arteria transversa scapula; A. 7. C., arteria transversa colli; A. A: T. C.,
ramus ascendens transversa colli; Rk. D. T. C., ramus descendens transversa colli;
A. ©. §., arteria cervicalis superficialis; A. C. A., arteria cervicalis ascendens; 7.
C. C., truneus costo-cervicalis; A. 7. S., arteria intercostalis suprema; A. C. P.,
arteria cervicalis profunda; @. 7., common trunk.

Robert Bennett Bean 305

While numerous variations in the origin and distribution of the
branches of the artery are observed in my study, it is nevertheless pos-
sible to classify the cases, for they are found to fall naturally into a num-
ber of distinct types. In section A of this work it is proposed to describe
these types. This will be followed in section B by a description of the
origin and distribution of the individual branches, while in section C the
results of the present study will be discussed in their relation to the
previous work upon the subject, and illustrative tables will be appended.

Fig. 2. Type I, occurring in 30% of the specimens, 22% on the right side, and
8% on the left side of the body. For index to lettering see Fig. 1.

The three divisions of the subclavian artery referred to throughout this
work are: Part I, that portion medial to the scalenus anticus muscle;
Part II, posterior to it; and Part III, lateral to this muscle.

The records given are from 74 male Negroes, 16 female Negroes, 21
male Caucasians and 3 female Caucasians. ‘The race and sex are not
determined in 15 subjects. The Negroes are the American variety, and
possibly all of them have a trace of the Caucasian mixed with the Negro,
the proportion in any case being uncertain.

306 A Composite Study of the Subclavian Artery in Man

Section A.—Typres oF RAMIFICATION.

The mode of ramification of the subclavian artery is found to be
divided into five types, depending upon the origin of the large branches.
The distribution of these branches is practically the same in all cases.

Type I (Fig. 2) occurs in 30%. of the cases classified, 22% being on
the right side of the body and 8% on the left side. In this type the
vertebral and internal mammary arteries rise from Part I; the inferior
thyroid and suprascapular arteries rise from a common trunk which
comes from Part I, and between these two arteries rises the superficial

Fie. 38. Type II, occurring in 27% of the specimens, 22% on the left side, and
5% on the right side of the body. For index to lettering see Fig. 1.
cervical artery; the ascending cervical artery rises’ from the inferior
thyroid; and the transverse cervical artery and the costo-cervical trunk
rise from Part II. Each of the branches often has a separate origin.
There are in this type 19 male negro subjects, 2 female negro subjects,
5 male white subjects, 2 female white subjects, and 4 subjects in which
the sex and race are not determined.

Type II (Fig. 3) is found in 27% of the cases classified, 22% being
on the left side of the body and 5% on the right side. The vertebral

Robert Bennett Bean 307

and internal mammary arteries rise from Part I; the inferior thyroid,
suprascapular and transverse cervical arteries rise from a common trunk
which comes from Part I; and is known as the thyroid axis; the super-
ficial cervical artery is absent, its place being taken by small branches
from the transverse cervical artery; the ascending cervical artery rises
from the inferior thyroid artery, as it does in practically all the cases of
all the types; and the costo-cervical trunk rises from Part II. The
internal mammary artery rises from the thyroid axis five times in this
type—four times in infants—showing a bunching of the branches.

Fig. 1. 4. Type III, occurring in 22% of the specimens. For index to lettering see
There are in this type 18 male negro subjects, 2 female negro
subjects, 3 male white subjects, 1 female white subject, and 1 subject in
which the sex and race are not determined. Types I and II are the rep-
resentative types for the right and left sides of the body respectively.
Cf. Fig. 7, A and B, pp. 314 and 315.

Type III (Fig. 4) with slight variations occurs in 22% of the cases
classified, 25 times in all, 13 on the right side of the body and 12 on the
left side. The vertebral and internal mammary arteries, and the costo-
cervical trunk rise from Part I in this type and in the two remaining

308 A Composite Study of the Subclavian Artery in Man

types. The inferior thyroid artery rises from Part I, the transverse
cervical artery rises from Part II, and the suprascapular artery rises
from Part III, or from the axillary artery (9 times). This type may
be considered a subtype of the first, Type I, showing the extreme separa-
tion of the origin of the branches and no bunching. There are in this
type 11 male negro subjects, 4 female negro subjects, 7 male white sub-
jects, and 3 subjects in which the sex and race are not determined.
Type IV (Fig. 5) is found in 12% of the cases classified, 14 times in
all, present in equal number on each side of the body. The inferior thy-

Fig. 5. Type IV, occurring in 12% of the specimens. For index to lettering see
Hig 1

roid artery rises with a common trunk from Part I. From the common
trunk rise the suprascapular and transverse cervical arteries. This type
may be considered as a subtype of the second, Type IJ, in which the
branches are often bunched. There are in this type 4 male negro sub-
jects, 3 female negro subjects, 3 male white subjects, and 4 subjects in
which the sex and race are not determined.

Type V (Fig. 6) occurs in 10% of the cases classified. The inferior
thyroid and superficial cervical arteries rise by a common trunk from

Robert Bennett Bean 309

Part I; the suprascapular artery rises from the internal mammary
artery. The type is of interest from this fact and because of its fre-
quent occurrence in the cases studied.

Section B.—DEscrIPTION OF THE INDIVIDUAL BRANCHES.

In its origin the vertebral artery is the most constant of all the
branches of the subclavian artery. It arises in every case, with three
exceptions, from the posterior and superior aspect of Part I, and is the
first and largest branch. It is associated with other arteries in its origin

ges 6. Type V, occurring in 10% of the specimens. For index to lettering see
Los ol

from a common trunk but four times, with the inferior thyroid three
times, and the thyroid axis once. It comes from the arch of the aorta
between the origin of the left common carotid and the left subclavian
arteries three times. In one case on the right side the vertebral artery
is double, two small arteries arising from Part I and entering the 6th
vertebral foramen together. The artery enters the 4th foramen once;
the 5th, 4 times; the 6th, 88 times; and the 7th, 4 times. Tiedemann
states that this artery enters any one of the vertebral foramina from the

310 A Composite Study of the Subclavian Artery in Man

Ist to the 7th, most frequently the 6th, and most infrequently the 7th.
He also mentions a double vertebral artery, one arising from the inferior
thyroid artery, the other from the subclavian, uniting at the 4th cervical
vertebra. Quain gives a chart of a similar double vertebral artery.

The internal mammary artery arises alone from Part I in 80% of the
cases, and is associated with other arteries by origin in a common trunk in
20% of the cases, arising with the thyroid axis in 10% of the latter, with
the suprascapular in 10% of them, and with the transverse cervical and
suprascapular once. The distribution of the artery is worked out in
minute detail only 28 times. A lateral thoracic artery is found five
times in the 28. It is as large as the internal mammary, is derived from
the latter close to its origin from the subclavian, and passes between the
parietal pleura and the ribs along the anterior axillary line, sending
branches into the intercostal spaces from the 1st to the 9th, and losing
itself in one of these spaces or in the diaphragm. A lateral thoracic
artery is mentioned by Quain, Tiedemann, Henle, and other anatomists,
but is given as a very rare anomaly. Intercostal branches come from
the internal mammary as single arteries posterior to the intercostal
spaces, sending one branch to the superior part and another to the in-
ferior part of the spaces; or they arise posterior to the costal cartilages,
sending a branch above and one below the adjoining rib; or there are two
intercostal branches to each space, one below the rib above it, the other
above the rib below it. Any two or all three of these arrangements may
be found on one side of a subject. In 54% of the subjects there are
two branches to each intercostal space, in 46% only one.

The thyroid axis‘ is found as shown in Type I, Fig. 2, in 30% of the
cases, 22% of these being on the right side of the body, and 8% on the
left side. It is present as shown in Type II, Fig. 3, in 279% of the cases,
22% of these being on the left side of the body, and 5% on the right side.
Quain and Gray give Type II as normal. ‘Tiedemann, Henle, Gegenbaur,
Sappey, Testut and other French and German anatomists give Type I as
the most frequent.

The inferior thyroid artery* arises as shown in Type I in 35% of the
subjects; it arises from Part I as a single branch in 33% of the subjects,
and as shown in Type II, in 32% of the subjects.

The suprascapular artery* arises as shown in Type I in 36% of the
subjects; in Type II in 34%, and from the subclavian alone as a single
branch in 30% of them.

This artery is absent 4 times, double 3 times, and very small 4 times.
The long thoracic artery arises from it once.

‘Table 2, p. 318. SiMablerowDe cles STablesd.spiroly:

Robert Bennett Bean oul

The transverse cervical artery’ arises from Part II in 39% of the sub-
jects, from Part I in 36% of the subjects (alone or with the thyroid
axis), and from Part III, or from the axillary artery in 25% of them.
Quain gives the most frequent origin of the transverse cervical artery
from the thyroid axis, dividing into the posterior scapular and superfi-
cial cervical; the next in frequency being the posterior scapular from
Part III and the superficial cervical from the thyroid axis; the least
frequent mode of origin being from Part IIT, and dividing into posterior
scapular and superficial cervical arteries. We found the following ap-
proximately :

The transverse cervical artery arises on the right side from Part II,
dividing into ascending [superficial cervical (7) ] and descending (pos-
terior scapular) rami, and on the left side from the thyroid axis, dividing
into ascending and descending rami, having previously given off the
superficial cervical artery. The ascending ramus of the transverse cer-
vical artery arises lateral to the levator scapule muscle, and, dividing al-
most immediately, sends one branch parallel to the superior lateral bor-
der of the trapezius and beneath it to the occiput. The other branch
passes parallel to the inferior lateral border of the same muscle and
beneath it to the level of the seventh thoracic vertebra, sending a large
branch to the rhomboid muscles. The descending ramus follws the pre-
scribed course of the posterior scapular artery as given in English and
American text-books. The relation of the two sides of the body with
reference to the origin of the transverse cervical artery shows the two
sides alike in 29 subjects, unlike in 13. In 10 of the latter the artery
arises from Part II on the right side, and from the thyroid axis on the
left side. -

The superficial cervical artery * is considered to be a branch that passes
from the transverse cervical artery in 60% of the subjects, from the in-
ferior thyroid artery in 22% of the subjects, and from the suprascapular
artery in 18% of the subjects, terminating just beneath the lateral bor-
der of the trapezius muscle. The artery is more commonly a number
of small branches arising along the transverse cervical artery as it ,
traverses the neck. The ascending ramus of the transverse cervical artery
is described by some anatomists as the superficial cervical artery (Quain).

The costo-cervical trunk® a small short artery, arises from Part I in
90% of the subjects; from Part II in 9% of the subjects, and from

™Tables 6, 7, and 8, p. 319.
§ Table 9, p. 320.
° Tables 10 and 11, p. 320.

312 A Composite Study of the Subclavian Artery in Man

Part III in 1% of them, dividing almost immediately into the superior
intercostal and the deep cervical arteries. The origin from Parts II and
III is on the right side of the body in all cases.

The superior intercostal artery” arises on the right side of the body
from the costo-cervical trunk in 41% of the subjects; from Part II in
10% of the subjects, and from Part I in 3% of them. It arises on the
left side of the body from the costo-cervical trunk in 38% of the sub-
jects, and from Part I in 8% of them.

The deep cervical artery” arises from the costo-cervical trunk in 83%
of the subjects; from the subclavian artery in 13% of them, and from
the inferior thyroid in 4% of them. It passes above the first rib in
82%. of the subjects, and below it in 18% of them. The distribution
of the deep cervical artery varies in inverse proportion to that of the
ascending cervical and the superior intercostal arteries.

SECTION C.—DIScUSSION.

First—We have demonstrated that the branches are arranged in a
different manner on the two sides of the body. Fig. 7 shows this.

This figure represents the most usual arrangement of the branches
of the subclavian artery as found on each side of the body, the difference
between the two sides of the body being chiefly in the origin of the trans-
verse cervical artery.” The type shown on the right side of the body
occurred in 51% of all the cases classified on that side. The type shown
on the left side of the body occurred in 55% of all the cases classified
on that side. The distribution of the anterior branches is put on the
right side of the body, and that of the posterior branches on the left side
of the body in this figure.

Second.—The number of branches arising from Part II on the right
side is more than double those from the same part on the left side, count-
ing all cases. This is due to the origin of the transverse cervical artery
and occasionally (11 times) the superior intercostal artery from Part
II on the right side.”

Third.—The relation of the branches to age discloses the apparent
abnormality of infantile subclavian arteries. There are 25 infant sub-
jects worked out, 17 male negro, 4 female negro, and 2 male white. No
two subjects show the same arrangement, all being irregular.

” Table 10, p. 320.

iTable la pho20:

a Tables) 125 135 14> >and 16. ppio20, cel soca manduocee
13 Tables 5, 10, 13, and 14, pp. 319, 320, and 321.

Robert Bennett Bean 313

Two striking features are noticed. In the first place there seems
to be a tendency for the branches to be bunched from Part I. The inter-
nal mammary artery arises 4 times with the thyroid axis, and the supra-
scapular artery arises twice from the internal mammary artery. In the
second place the suprascapular artery is small in five cases, and does not
extend beyond the suprascapular notch in these cases, its place being
taken by the dorsal scapular artery. There are 2 other cases with inoscu-
lation around the neck of the scapular between these two arteries, the
suprascapular being lke a continuation of the dorsal scapular artery.
Knowledge of the previous condition of the subject as to age, habits, and
family history, and dissection of subjects that had died at the ages of 1,
5, 10, 15, 20, and 25 years, ete., would be of value in studies similar to
this one.

Fourth.—There are 74 male negro and 16 female negro subjects, and
21 male white and 3 female white subjects from which records were made.
In 15 subjects the race and sex are not determined. In view of the
small number of female and white subjects, the relation of sex and race
will hardly admit of discussion. Many of the subjects are mulattoes,
or mixed bloods. The number of anomalies, variations, and queer types
is uncommonly large. ;

May we not explain the occurrence of this large number of abnormali-
ties by the well-known biological law that hybrids tend toward variation ?
The question is an open and an interesting one.

Fifth.—Free anastomoses by a definite arterial trunk were found in
connection with the suprascapular, deep cervical and superior inter-
costal arteries. The suprascapular artery inosculates with the dorsal
scapular artery posterior to the neck of the scapula, 17 times, or in 16%
of all cases. The trunk was from 2 to 5 mm. in diameter. The superior
intercostal artery anastomoses with the superior aortic intercostal artery
31 times, in every case where it is looked for. The anastomosis is found
to take place:

Aine sesame aatereestal smace. wn. es. ceo ee ode d oe 16 times
muster board: upercostal SPACes . 22244 woes ae te acetone 10 times
At the fourth intercostal space....... Ae ee ee 2 ass) DCbUNES

The trunk is very small, only a minute tube in some cases. The
deep cervical, “‘ profunda cervicis,” anastomoses with the “ princeps cervi-
cis” from the occipital in 11 cases, 10% of all. The trunk is of good
size, frequently about 5 mm. in diameter.

Sizth.—Anomalies when present are found as a rule on each side of

the same subject. The most frequent anomaly met with in the dissec-
So
eae)

Bape” Ses

ACA

See Fig. 1

The left subclavian artery of the same body as Fig. 7-A.

Tele

for index to lettering.

FiG.

316 A Composite Study of the Subclavian Artery in Man

tions is the suprascapular artery arising from the internal mammary
artery (see Fig. 6). This occurs 12 times in 104 cases (practically
12%). Quain” found the same anomaly 4+ times in 264 dissections
of the subclavian artery (about 1%). Arthur Thomson” found it 9
times in 544 cases (less than 2%). Anotuer anomaly is found in con-
nection with the internal mammary artery. The latter divides into two
branches a few cm. from its origin, one of which takes the normal course
of the internal mammary artery, while the other follows the anterior
axillary line between the ribs and pleura, terminating at the diaphragm
in two cases, at the fourth rib in three other cases. 'This branch is as
large as the crdinary internal mammaty artery, and sends branches into
the intercostal spaces just as that artery does. This “lateral thoracic
artery ” is present 5 times in 28 cases (18%) that are carefully worked
out. Anatomists mention this anomaly, but consider its occurrence
much rarer than our findings indicate. Another important anomaly is
observed twice. The anomaly is in the trunks arising from the arch of
the aorta. The first trunk divides immediately at the aorta into the two
common carotid arteries. The second trunk is the left subclavian artery.
The third trunk is the right subclavian artery. This arises from the
distal part of the aortic arch on a level with the fourth thoracic vertebra,
and passes posteriorly between the cesophagus and the vertebral column
to its usual place on the right side. The right recurrent laryngeal nerve
passes directly to its distribution, without looping around the subclavian
artery. The pneumogastric and phrenic nerves occupy their usual places
and relations. This anomaly has been reviewed by Gotthold Holzapfel,”
who collected 200 cases from the literature, including 4 of his own (1 in
an animal). He concludes that this anomaly occurs 6 times in every
1000 cases. Quain and other anatomists fixed the ratio at 4: 1000.
Tiedemann gives the ratio 8:1000, nearly 1%. The middle thyroid
artery, “ Thyroidea Ima,” another anomaly, is found three times. It
arises from the innominate artery, and, passing to the median line,
supplies the lower lobes of the thyroid gland and the isthmus. Wenzel
Gruber “ records 125 anomalies of this kind, and concludes that the artery
rises most frequently from the innominate artery, but also not infre-
quently comes from the aorta and the common carotid artery. He found
it 16 times in 100 consecutive dissections.

*% Quain, Commentaries on the Arteries, London, 1844.

* Thomson, Second Report of the Collective Investigation of the Ana-
tomical Society of Great Britain and Ireland. Journal of Anatomy and
Physiology, London, Vol. XXVI, p. 78.

** Holzapfel, Anatomische Hefte, XII, I part, p. 373 (1897).

™ Gruber, Virchow’s Archiv, Vol. 54, p. 445.

Robert Bennett Bean ay

SUMMARY.

I. The branches of the subclavian artery differ in their origin on the
two sides of the body, the most frequent arrangement being similar to
Type I on the right side, and Type IT on the left side.

(a). The thyroid axis, dividing into the suprascapular, transverse
cervical, and inferior thyroid arteries, is not normal, except on the left
side.

(b). The transverse cervical artery and the costo-cervical trunk arise
from the second part of the subclavian artery more frequently on the
right side than on the Jeft side. .

(c) The superficial cervical artery is of infrequent occurrence, and is
found more often on the right side. See Type I.

(d). The transverse cervical artery terminates by dividing into as-
cending and descending rami, the latter being commonly called the pos-
terior scapular artery. The former divides underneath the trapezius
muscle and supplies the upper and middle part of the back.

(e). There is a tendency in the branches of the subclavian artery to
bunch themselves in their origin on the left side, whereas on the right
side there is a tendency in each branch to arise directly from the sub-
clayian artery.

II. There are five important, and not infrequent, anomalies to which
the attention is directed:

(a) The origin of the right subclavian artery from the descending
part of the arch of the aorta. This occurs 4-6-8 times in 1000 cases
(0.5% to 1% of all persons).

(b). Variableness in the origin of the transverse cervical artery,
especially on the right side.

(c). The presence of a middle thyroid artery (‘Thyroidea Ima).

(d). The suprascapular artery arising from the internal mammary
artery.

(e). The lateral thoracic artery arising from the internal mammary
artery.

III. Eighty per cent of the dissections were made in negro subjects,
a large number of whom may have been mulattoes or mixed bloods.
That hybrids tend toward variation is a recognized biological law. This
may explain the unusually large number of abnormalities encountered.

IV. Twenty-three infants were dissected and many of these show ir-
regularities, particularly in the distribution of the suprascapular artery,
which is frequently deficient, its place being taken by the dorsal scapular
artery.

318 A Composite Study of the Subclavian Artery in Man

V. The branches of the subclavian artery may be more numerous in
adults than in infants. The branches rise from all parts of the artery
in adults, whereas in infants the branches frequently rise in a bunch
from Part I.

TABLE 1.1

THE ORIGIN OF THE ARTERIA MAMMARIA INTERNA (INTERNAL MAMMARY ARTERY).

Origin. Quain. Bean.
FLOM HPA Ga is rota Siesedare voce uote tayisnehonariet siyoo ecteuctote dobiey eprowtst aha) ake got ranetek exerelone eheremeneteue 935 .806
OPQ ar eee ceneteuepeevads: Ge vekers Oi snchoie Orne eoreeeie bouche iO tele mak enoke gotteeee 0038 -000
SONU Gell sway cr irate ate fous teNonersts veteretessceve re cevarowers ieee omedsMeervon ienoben ea cNsconemells -020 000
HEGn the WLUN CUS HUM YyLeCO-GeLVIGAliS«., ceive sleretheneelencienel el iatoicnseteieneRenate 050 O97
From a common trunk with the A. trans. scap. and the A. trans. colli. .015 007
From a common trunk with the A. transversa scapule................ .012 .090

TABLE 2.

SHOWING THE ARTERIES THAT COMPOSE THE TRUNCUS THYREO-CERVICALIS (THYROID
AXIS) IN THE DIFFERENT TYPES.

Type I TypeII TypelIII TypelV Not

Truncus thyreo-cervicalis composed of Fig. 2. Fig. 3. Fig. 4 Fig.5. Classified.
Ae thiyroidea interior cece scence oe 1 1 il 1 0
PRICE ALS VEL S QmCOLliimersrstey such eualercbetetevenetane 0 1 0 0
AS ULaANSWeLSa: SCApPUlei.tiemererensiecienel esis al 1 0 hy 0
T. t.c. gives off A. mammaria interna.. 0 0 0 0 ul
Frequency of each type—Thomson..... 417 413 060 .064 044
Frequency of each type—Bean........ 273 257 273 120 O77
TABLE 3.

THE ORIGIN AND ANOMALIES OF THE A. THYROIDEA INFERIOR (INFERIOR THYROID
ARTERY).

Origin. Thomson. Quain. Bean.
EAS! AST O Mya AMOS OO: Tesesis, Saati troene amet colioveusizavielus tus fo ae ceuche eon checomere ke 432 “on .296
AS (SHOW) im sy pe HLM crete st uaicils secs hecs) erie a re) cps susiai@us teleeayey abet onquetons .430 .900 SACL
AS ESO Wal el ype envi civ evcksieba Gretel ewescne ole tanetode Pouce rake Barre aera .036 ° Bate -129
Prom. bart l(alone)s asa sine les pranchimersecivereciiecr creer O95 101 287
IM oa THAGY UNS CohMOLMS) Conmowwh gaonacooecoguodoadadopodade O01 0038 000
Hrom, che vA thy roides dma = creczcretscreeclela sincera eieioiere crevekewene 001 O11 009
With whe As transverse Collis 7S sc. c-cievecre chepei eleva etenavo henceectene -038 .005 .065
Wathethe yay mammariainterm ayes aercistacrsliatercueienelcretensners ine he Pe ANC .046
Waithisthevar sviertebLralis: ss.ciens secs cro istene a cxelere.coret suererenarehe Grete tere 005 O01 .027
With ithe A... ocCipitalis mys Act suse siesioe eet choi oe le iolenneereteuere nie eae 018 009
Withs theeay ‘cervicalis sup erhClalisti. cite tnrecccicle s ciele aieiers 053 .025 OT4
Withwthie AS cenvicalis sproriumncaleersssrcietcicneresciekekoreteneialercieereneiene oe steve .027

Anomalies.

DASDSOMIE ile consieite voy suave ai tate Yousreae aps ISTE Oe owes tuCreetetetetean aller acucdoheierete OG .022 018
Very large—larger than the A. subclavia.................. Rat, O11 027
Very small—not supplying the glands. ... .. .4.s1s2. sees ate .014 .065
One branch), to, the slam. soir) ee rcteuerete ve sctenel oc otoneieten stetoucn exerci Siete Sree .046
Two branches tonthen clandestine cnet eter ene reliant eeenenenentee ae Mad 954
Forming both the A. occipitalis and the A. vertebralis...... Rare aor .009

‘The figures in the tables are all given in percentages or may be considered as the
number of times per thousand by removing the decimal point.

tobert Bennett Bean 319
TABLE 4.
THR ORIGIN OF THE A. CERVICALIS ASCENDENS (ASCENDING CERVICAL ARTERY).
Origin. Thomson. Bean.
HICGiienle Ate ii UTOlded oMLEDLOL ic. jcr «ica wre tn cones anc eeaveieregen ceteneneiederelsie cer etelcireyrs .902 .656
Vvomunescruncus . Cy TEO-COLVICBITS) cars, cree evesens ceermtersacie ceneislsy ielisiiei sila te tet els 030 099
LON Teen) UK Na ith spd vest tiareloy Ube thes Gecmcc Glo ici c oot ocon Ming omaiedo oe .022 O91
rome iNew Acs COLViCce lis) SUDCLA CIALIS, =. <7. 072.0 0) a ochre ete: seu cuehty eh eushs lel wee eh viel ste .038 OTT
ERO TINS ISG lle atic: 5 once hieviesel aries lasers retin, oifS) ci-eeerere eisl-a bette Naoia ced theatVairen ce wen eee airoyell Ss heimbelr tte 004 054
HILOMIE LHe w AGT Eran VELSA e SCAG. creas an) so ae) ar'e' el aleh stele reiohenel sitetakotelotel sleraiiatciis .002 .015
TABLE 5
THE ORIGIN OF THE A. TRANSVERSA SCAPUL® (SUPRASCAPULAR ARTERY).
Origin. Thomson. Quain. Bean.
Hrom-Part I, with other arteries or alone... 2... ..0.80c6050 776 .885 .813
From Part-II, with other arteries or alone................ 024 015 042
Hrom. Part LLL, with other arteries or alone... .5..........- .OT4 O88 101
FORO UDR At ULLAL Ss sate scence ree ceereiene’ fo tere vel-s, oteleyetalege,ehebeteenenate 015 .015 042
From the A. subclavia alone (as a single trunk)........... 05T LG .219
WiththemAcatransversan Collis cic «cs ice cleilelne ce cierto .080 .O80 as
VR ae /Ay THERA EEE TO ey emisaplomcicocccd coG0 Oba aDUL 016 .020 101
WAH MEN GMA a SUD SGA ULAMI Sir teie7e lic) oie1c1s\iche o: o/eiie)o) teenonotalel ste reneiors ate 005 O17
IDRIS eS ode oot ob eo EO CIC Coie Orin ODE ro cod corns Coy ta DOC O15 miets .025
TABLE 6.
THE ORIGIN OF THE A. TRANSVERSA COLLI (TRANSVERSE CERVICAL ARTERY).
Origin. Thomson. Quain. Bean.
Hromeearpienwithsouners arteries! OF AlOME (er cjersahere clara) sicker ake O08 429 .365
Brom sear UW with) other arteries, or alone’ =. <1: < crete lest sre .014 191 .390
‘From Part III, with other arteries! Or alone = aac <teccire stekeioners ATT .380 .245
WOE OMA e LEAN S VETS SCP ULE «io: oi: c silerscvclole) ore cuecavel helevintolieneda O79 .048 .145
Wibhmihemwae tiayroiGed Anterior «2:0 <1<,- 1611s) ele miei solelelsfelstekel tei .0386 Hane 072
Wiha nherActnoraGalis nl teOba@lis: cc scores) cisieienaiel olcleneretenaraienehelentel Beye wae .012
VOYEV ACHES ORs Gat aig bce ERROR IC Tro mo cInIO Cites Oo cd oO Ob .005 waite .012
TABLE 7
THE ORIGIN OF THE RAMUS DESCENDENS (POSTERIOR SCAPULAR ARTERY) .1
Origin. Deaver.! Thomson. Quain. Bean.
HAY rai Ta ae Ml pdees 3 ores o scesreravelarlel avers ele) suevcisucyelsfaicieWelisiens Wits Beets 003
Nidopre TENE YALL ES Bae, CronG Che ee CCICrCnRCrORRO Tr Dio en 6 Gas or 005 008 -150 arene
UICEFITM ES AUT Lee bek = orc layateMe lens oye te sbeus soieters ioveuatonerateleheteys .023 508 43515 .046
MEOH DnemAe LEAMS Versa (COLL, «reyerateletelavete eletnistse 968 484 480 918
iean Were Gh ETRE Spee eioaeacac coon nots oo oc 001 Sree sth .009
TABLE 8.
THE ORIGIN OF RAMUS ASCENDENS (NOT GIVEN BY THOMSON OR QUAIN).
Origin. Bean.
UEC END A ESB Lo rss do. as 0 vee a) we) a '0ive porec® si alin Konno) whahe Ghapel abe betay’ol eve tay ne eel atectelece revelemeleraysienereires one .030
MT OTM Tee E55 ic. aslo oye, oes fe. ol'o, aperDeave orn soi sleet tone oxclae etemerorave oreuaieicie othe cutee eae siebes .010
URGE WEG LD Ms 3's oye. o oye eteuciarqcelane o-easllege ep ctenerducdecthcling® cpepel shal srota el ote Cuees aeeoees ene) atc herons War cey 030
MTOM bea. <irANSVeCLSar ‘GOLA soc ic evencveiecaietavousncte iota cnsrslevensiene Welele (cnetalionst sh ttenanchel eyenece -730
Hron toe iuncus. thyreo-cervicalis (or LES DANCES io) owe cloucle ol ete cro ual'e avec ciepaisicisis ie .200

1 Deaver, Anomalies of the Posterior Scapular Artery, University Medical Magazine,
Phila., Vol. II, p. 151, 1889.

320 A Composite Study of the Subelavian Artery in Man

TABLE 9.
THD ORIGIN OF THE A. CERVICALIS SUPERFICIALIS (SUPERFICIAL CERVICAL ARTERY).
Origin. Thomson. Quain. Bean.
rom the Ae transversa: :collit siccce ssa cterencl everett esovonchenede mesic O89 .9638 590
HrOM. the FAL th yroidea, Interior ics .ccis siels cNstcnerehet Tous (heed delerats -o12 004 220
rom whe AC Eransversa mSCaplllse) acre crchers eicicieielele eiakene ciara .092 026 .190
Hirom athe: (Av SUDCI AVI Ay tac mystic ls orcronce arse atouetel ce erence yaar .002 008
TABLE 10.
THE ORIGIN OF THE A. INTERCOSTALIS SUPREMA (SUPERIOR INTERCOSTAL ARTERY).
Aa Quain. Bean.
Origin. Right. Left. Right. Left.
Brome Par Ws atascats ots of ela.d Grote) aravtaneuersh ce evel stouere seeker erts .O8S6 .220 .030 O75
OOM. warty. WM sors, sccustaceiesss/ellas. ove) aneire bahohevoretenenelicte torsion atotehesera.s .482 194 105 wae
MrOmMmthe nw tsuneismCOSto-ceLvicaliss cies aeisiecieeienione cere Sgege aus .400 370
MROMmtherAre Verte bralis! ti. < sc aise levels ste terjenene oslexerevey ens fetes 007 .020
Hrom thevaAe thy toidea inferior 21... -ckeleteiare cents teu avere 004
TABLE 11.
THE ORIGIN OF THE A. CERVICALIS PROFUNDA (DEEP CERVICAL ARTERY).
Origin. Quain. Bean,
HOM athe tLUnCuUSsmCOSTO-CERVAGAlISmaaisttanckclonenchctele cle ciieleiel onsieielcucicretereicts sinc .820
KromatheAq intercostalismsuprenva: - ence oriare he ecchchele eveseucl a eleven onetenerene ous ‘932 ete
HromthereA.- Su Pelee cre syevsce vue eitstas Gy sole to) clues Shalerel store ae arena egewoue te telsifeushcllercs 049 130
HOE, TA ANS (A PROIKO, Ib OOYS og oocano 2000 RIodadodoonducaoGopaonS soc .040
Krom<the TramuspdesCen Gens) Fycue) oi seyscavas jote ter otione Bah cltarelle sie ove tee i cle euscetebel one 018 .010
The artery passes:
ADOVeG SEDC MAE Sts TaD erences s tener suene artelever onisweney eas \enslal ov'e names; Sie. hie loners vas cvereporeitepalteds .936 .820
Below, *the shes Perl Diora ote eslcce tee eee onee ad cxevey ccaoabe ads layiattc ait eco ne romeabey otlarctereneWe fates .064 .180

TABLE 12.

A COMPARISON OF THE MOST FREQUENT ORIGIN OF THE BRANCHES ON THE TWO SIDES

OF THE Bopy.

RIGHT SIDE. LEFT SIDE.
Artery. Origin. % Origin. %
From: From:

Wertebrailish: ©... sestere meets | SPOR cree oh steeteusvede ete OF Part lees are teneorene 88
Mammaria interna ....... Par taslviiawckedeta, ots ior erete QS PRart ly Sesencvacel. ober enetiene 82
Truncus thyreo-cervicalis ... Part ele (omciinies))n trier 13) Part (25) times): sae 45
Thyroidea) inferior) (s.r Particle kc acukaisetercisaes 74 Truncus thyreo-cerviealis 45
Transversa scapule ....... AL Gepl y-te tec edets tenors) svotetencin 54 Truncus thyreo-cervicalis 45
MransVversauCOliil ysjereiesers tere | Parti ies rors cistenaiicne 54 Truncus thyreo-cervicalis 49
JRE HUES NS athe ONT Go eal lehyte IDE sen weaasaqsood 66 Truncus thyreo-cervicalis 80
Ramus dese; ear etre Colliiecr|||PAc mtr aiCOllivmre sleireneriee rs 97| A. transversa colli..... 90
Cervicalis) Superiicialisy sci Aen CL COllitmereene ceiernerene 45| A. transversa colli..... 76
Cervicalis ascendens ...... lean thyroidea inferior... 70) A. thyroidea inferior... 70
MGWUNCUS! COSTO-CELVICAIISN: «=| PEZaT ta lermeectenoneomedete ren oteiene TSicPart lye de ace ae 100
Intercostalis suprema ..... Tr. costo-cervicalis. ... 86 Tr. costo-cervicalis 80
Cervicalis profunda .......| Tr. costo-cervicalis .... 82) Tr. costo-cervicalis LT

Robert Bennett Bean 321

TABLE 13.

COMPARISON OF THE NUMBER OF BRANCHES FROM EACH Parr OF THE SUBCLAVIAN
ARTERY ON THE Two SIDES OF THE Bopy.

Right. Left.
AT Ge 1m in venav ol wisn ayanerareievete lasena olersip ten ereL eebetist eh tenes 80% 83%
Parts Ty she cia os hoeue ee ienets tah oietavaier yale reneenehe 11% 5%
VEEN RADU Gl DARA SCP ER RE crore othe 8% 9%
PAS SMUIDEDPARE © ocrd en ouevarecon exe tebe ove come teseekataiaterel otras 2% 3%

TABLE 14.
THE PERCENTAGE OF BRANCHES FROM Parts II AND III OF THE SUBCLAVIAN ARTERY.

Part IT. Sbart Pu
Quain. Bean. Quain. Bean.
INGTON getcic lei anaieuharacce schevs 19% 67% 54% 66%
ONE iy Soya siessthaystay se orehaians 67% } 24% 46% 1 24%
IEA AG) Se eH Sa ee Pacis 12% 8% 3% 4%
SENPEY Wj Aoveusdete sate ators 9% T% 8% 8%

1Quain counted a muscular twig which was not counted in this work.

Man

y in

an Arter

1

-
of
-

.

ite Study of the Subclav
“‘yuesqy

4108 JO WoIy

>

>

“JNO poyIOM SITIOLIB “ON
“yundjy UOMO:

“BIAB[OGUS “Y JO osegq
*SI[BOLAIO0-04JS00 SNOUNAY,
*TT[00 BSIDASUBI] “WV
‘aB[ndevos BSIOASUBI) “VY
“IOLIOJUL BIPIOIAT -W
*SI[BOIA199-O91AY1 SNOUNAT,
“BIAB[OGNS “VW [II Wed
“BIAB[OQNS "YW [I 318g
“BIAB[OGNS “YW [ leg

A Compos

trees eres TOD,

"s+ Bpunjoid sI[BolAdeo *y

BUIBIdUS SI[BJSOD1e} UT “WV

*** SI[BOIA190-0]S00 SnoundAy,

ehejenels SUIPUDISR SI[BIJAdeo “y
‘** SI[Vroyaedns s{[BofAdeo “Vy
‘SULI] “Y SUspUsdSep snUeRyy
‘*1[]]00 ‘suB41jy “yY SUepusose snuEy

SOO OO ion) ASHE YS LOM Gl B47

‘++ gpndvos Bs.wASUR.y “VY

SOPDOOGS Spence uo) Cony pl Ar

** SI[BOLAL90-00144} SHOUNAL

Soma ereesese srgaqeweA “y

“AGO GHL JO aAGlg LH4a'T AHL NO AUOLUY NVIAVIDdNS GHL 40 SAHONVUG DHL AO NIDINO

323

Robert Bennett Bean

‘I[[09 BS1oASUBI} “Y ‘1OlJeJUL BeplordAyy “y—Soyouvsq v1}

PLO L 6F <i ScL

¢ I
6¢ = oe mG =
8¢ G I
Lg T +. a 8
6S i 4 I Z
€¢ 2 9 i &
SF iar qT 1G
SF a AS FF
OS i I! : cs
¢ GP , :
9 ; c i 9
69 ¥ A T G
Z oa oa =) ©
iS : ot S g
2 (e) n =] 5
r= o £ o 3
or I 5 =
Bs nr st ® Sc
3 3 2 z
© — > q
n a g (=|
= D a =
fe)
iy teas
$ eI
ro¥ >
im ie)
ES ©
ct =

DR

*‘SI[BOIAI99-0}809 snoUnIT, 7

Tol

‘T][]OO BSIDASUFAY “VY

*B[NABOS BSIIABI] “WY

I

“LOLLOJUT BOPIOIAY “Wh

0JUl 8PIAIp 0} podeprsuoo

oO
N
Ke
i

ne)

lar! rg
4 3
Be
seme!
an =
=

a >
Le n
2 S
9 =
o o
ae ro)
4 <
Sf =
re »
po

=

2)

‘BIAB[OQGNS “WV [I 41ed

*BIAB[OGNS “WY [ ed

‘e[ndvos BsdaAsuvs) “y puB
S[ SI[BOlA1e0-O01L4} SNOUNA} So[qe} aAoqR aq} uy]

5 O10 8 (oeb Goan 6.01 0 tired Ol choo wach: [eq OL,
Oat (Ouch CnC teat) Cetktied tech Cuero.) ete Bpunyoad SI[BOLAde0 ‘Vv
Siiyiveyia, '@)? sis] =)is) sq) elle] (5 ves eee eee valei1dns ST[BJSOD.107 UL ‘Vv
ee oe eG) tee RLS A aa ae We SI[BVOTA190-0}800 SHOULAL
LOR HOILCho, 0.000 tresses sgapmaost sI[tolAded “y

GF O1e) CLG (610 62a) €eue eRe ate etm te SI[Bloy.aedns SI[COTAII0 ‘Vv
BS es SS TT[OD! “SUBI ~“y SUspueosop’ SNuBy

a OR ee ier om I[[00 “SURI “Y SUspuUdosSe SHUeY
0, OU Core otro Orcid Doan OID neo co * [JOO BSaeasuRay “y
CECEONCEONONE CRDDODOCTOCSO GODS Nets! rZHeVNSUM Ay
9119) 14) ‘e)feuielial itera} (e se yore) ieternsh a) ie SOD ROO Gry eeni vaproasyy ay
oe age tak, tape ts ee a als SI[BOIA120-00.1A4 9) SnOUnAL
eevee tes eras) elis\rmdel lelieieneyieverisirs ** BUdeJUL BIaBWRU oy
sete encenes Shoo Hoo BOD aD OoDIS EATS Ay

“X00G GHL JO WdIg THOIY AHL NO AUALUY NVIAVTIOUNS GHL JO SHHONVUG AHL AO NIOIWO

‘OT ATAVL

TABLE 17%.
THE DISTRIBUTION OF THE BRANCHES OF THE SUBCLAVIAN ARTERY.

324

A Composite Study of the Subclavian Artery in

» ee Meee ae sis pine Lo evs WS WR eo ee

ee Ais y pag | : 3 “ A ; 5 : ye s ; :
ene pug | : : TAS Hees P

‘aovds ]BISO01E9UL IST | : : io Re Oe ae ee

_ ” ” Gable: g VEU | f See eS ee

ean 73 an Wols : CRED Cen ee eas 2 ae
a ee Wigla : bared Ovid “peel ona Rel 12 S/T

“USTIBIOS [BIGOIAAA YF] 4: ey ee ee meee bo, Se

‘SIqid¥o snsuol‘Ww| :: Ev evaben oe, ea ne ee Se

“WNIT, | = ae ee > J Lae

ij “UOTMLOIDY | ; ae ee en : - poe

"RI NOTARTO | g ol De a ee eee

“Baprloi dy) B[NpURTyS | : aS Proce ane Se

“quiof aeprnoygs | : jeunes 5 is 2

“STUITIOPYGR SN{DAI “I | = eras 3 ; 2 Se ae

“BUIsBIYNAdBIP “JX | x Site : fc) = Se

“BILR[NdBosqns “fq | : cS — 2 SS aes

*ISIOp SNULISSI}R] “JA | Sree = I eS Sea

‘TITOO snsuo “WW | Se : Ss Se eee

*10110}S0d ‘sntpoeul ‘10T.19]uB ; TUO[BOS “JT = : wy) Go te © 5 SRR Re eS ie!
“‘snoploqidéy OUIEYS pu SNaplodyousoyS “J | PE ES Paseo ee ; eae
“SneploAyoulo "FW J|BY IaMOT | eC ae a a ee cue ae
“SNoplOJSBUI-OPT9]9-OUIIIS "JT | SOR a a ee ee See

| *SIOTAIO9 SITBISOOOT]L PUB SIOTAIa0 STTRUIdS "Jq |: pie ee ee ae ey OS Sie oe
*plOJSBVUI-O[O OBI PUB SNXVBTAUIOD "JW | 7 IP eee eel oe >) Soba
*SIOTAIOO SNUIISSISUO|T PUB ST[BUTUSIMES " CF Se ie Oe oS oS
“SIOTAIOD puB ‘sIyIdBo ‘sniueyds “pw | Hae SOR ee ee ee ee.
“SNIBUICSBAJUL “J ee et by ee ere

“sn euldseadns ‘pw | ee. Sees eas Oe es ios

“1OTI9JUB SNIBIIIS “JL 4 RES CTE ei ray toe eee oi ee

“B[NdBos 10IVAVT “fA | ee aE mir Aca es eae

“1OL193SOd SN{IBIIOS PUB TOPIOgMIOY.A "JT | Hy Genre c Ama es ig ees
“‘SnIZodB.1} “JY SEAUS Gate on a esaires a oe Se

“rmsAqeyd *yq | aL | Base ese ass

NEN EY RUC 0 FN Ti ces cae cara eet Cae PePC ESE ae RI cer SHES RPE ae ENN ES [url feel Ba fll ae [fpr Pe

A7\ Lat yujeke yack ay ydeChooovond annenonabscan

AMTICO: Oreo rAV CHIN ae oacnnnoonanmunheoduasouaaancodoos

A. thyroidea inferior......

A ELATIS VE TBAT SCIP Ul esis rarcrayeiereierole clare crete. sai ne ieTornrstere a oye aiaroratnsrevatovsinvs

PA ADEA TIS VET SA COLT: gers seater vvenciacert lesan ccrerste oineteeee e a F

HeyASCENd CNSsUraAN SAC Ollilsnmias-yaccve nase cticariera eee meen

lade, Oke Verena leyaysh mths} Collie dauconndoncsovoudoobonoenus anno nooseor

A. cervicalis superficialis

A.

COLVICALSPASCONGCNSS ana aeei circ tenon DEC a One

PEUITICUSICOSLO-CET NACH IB et astircen earn eee in ee eee

A MLELCOBLAIISINUN Den Astin aneeiieeer reenter

MA SCORVACATIS DLO TUG Aine a caer oa caeten eee hat We emt ore lorena

7A AE At Bermon catebanccn Mea Aaa ne Obanetade suche HOG ane ad oea!

AG CIVCUMMOXASCADIL Gs. -leiseraniate relia -valelerie

Robert Bennett Bean BAe

TABLE 18.
MUSCLES SUPPLIED BY THE SUBCLAVIAN ARTERY, WITH THE BRANCHES SUPPLYING THEM.

M. platysma:

Almost constant.....:..--. A. cervicalis superficialis.
| A. cervicalis ascendens.

A. thyroidea inferior.
WP OCUIEN Gene eters ene aherel=)) one! « ax0 A. transversa scapule.
A. transversa colli.

erasionalwee cael ee aac es A. subclavia, Part I. Twigs.
| Truncus thyreo-cervicalis. Twigs.

M. trapezius:

[R. ascendens transversa colli. Two or more large branches
passing between trapezius and rhomboids.
Constant. ...-.- sees see ees R. descendens transversa colli. Several small branches.
Some twigs after passing rhomboids.
[ A. cervicalis superficialis. Several small branches.

A. transversa scapule. One large branch just at bend to
TGR C QUEM G5 «) ccevcceyslaren le) eelen) © 3 ieee
drop over scapula.
CYeOTR AD so boocoGomaUT

A. cerviealis ascendens. Small branches.
A. subelavia. Small branches.

M. rhomboids, and serratus posterior:

R. descendens transversa colli. One large branch under
Constant. ........--++seeee- rhomboids (anterior), and several smaller ones into
the muscles.

JOUR PUK Ssat Rete eRe ioc cro GRE R. ascendens transversa colli from inferior branch be-
tween rhomboids and trapezius.
(a. cervicalis superficialis. Small branch or branches.
Mecasionailey iter sei) -ies= 1) axe cervicalis profunda. Small branch or branches.
|. subelavia. Small branch or branches.

M. levator scapule:

_ R. ascendens transversa colli. Small branches.
Constant.....-++.++--+seees R. descendens transversa colli. Small branches.
A. cerviealis superficialis. Small branches.

MEMOS oo oe oo ca opan coor: A. cervicalis ascendens. One large, several small branches.
WeECaslondl sat ate csercys cies sponse’ A. transversa scapule. Small branch.
j SF eS cs eect Gao CMOS ACEO OCR ROL A. cervicalis profunda. Small branch.

M. serratus anterior:

MOONS EA cette ois eile) <Veitstie 8 «1 onsets ele R. descendens transversa colli. Small branches.
IN Ree (NTE oo. pO COI OIol 6 at _R. ascendens transversa colli. Small branches.
WGEASIONAM ew ciers cle elelele) =! oes ohare A. transyersa scapule. Small branches.

M. supraspinatus:

RIO TIREET Ce eh ee as ae Sa. transversa scapule. One large branch.
Lr. descendens transversa colli. Several small branches.

OCCASLOTIDl cy. dicho: oO nve te ta) tea. o honoree A. circumflexa scapule. One large branch.

i A “7 iy y 4 \ . 716 . rr 7 .
326 A Composite Study of the Subclavian Artery in Man
M. infraspinatus:
Gaunknnt fA. transversa scapule. One large branch.
“ Sit tahayt othe cat a'e ahel@iaie rs) « 2
‘LR. descendens transversa colli. Several medium branches.
NGO EN Geel «. crertuerennercretete retort . A. circumflexa scapule. One large branch.
M. latissimus dorsi:
= ; R. descendens transversa colli.
@onsSt@neec cake ain we esiens el vier. ,
A. subscapularis.
M. subscapularis:
CONStANIGE ceo clecetsterelsoherenehee R. descendens transversa colli.
FUSE QUCH state ouctarveces cry enenel levee sc A. transversa scapule.
M. sternocleido-mastoideus:
fa. transversa scapule. Small and medium branches.
Constant :
ONSLANE. ee eee eee ee ee A. cervicalis superficialis. Small and medium branches.
A. cervicalis ascendens. Small and medium branches.
Weaduent Truncus thyreo-cervicalis. Small and medium branches.
En sac ye or *\ A. thyroidea inferior. Small and medium branches.
RETO Aereeeke oie wie aye we tedee ena) tess R. descendens transversa colli.
M. omohyoideus:
(Ae cervicalis superficialis, or branches from.
(CyEM NES S GEG ao cao minIae C :
‘A. transversa colli.
JOM (OK IANS: Go Sinirce.o olalaolotcc dead © A. transversa scapule.
‘ { A. thyroidea inferior.
Occasgionalys 5. shes colo 's eis iee ee
aA cervicalis ascendens.
FRAT Grrr censter et nsvetciskes eners erate. oi benste A. subelavia. Three times.
M. sterno-hyoideus, and sterno-thyroideus:
Constant or frequent........+4 A. thyroidea inferior.
Occasional. aes.aen seers eee crerie A. transversa scapule.
VAG eyens rel teusWevenctioter eles at aurutoheyovenayte A. subclavia.
M. scalenus anterior, medius, and posterior:
“A. cervicalis ascendens. Many branches.
Gols EaNt A ore ens A jun ens costo-cervicalis, or its branches.
ie intercostalis suprema (R. dorsalis).
A. cervicalis profunda.
(a. cervicalis superficialis, or branches from,
PIPequentas cic cereus sponse aie

ale aa) A. transversa colli.

OCCASTOM A] Erect seca oketene susie ..R. ascendens and descendens transversa colli.
RGR O eS aot ecoue rebel crea stiahey otto ieneneray et A. subclavia.

VW. longus colli:

A. cervicalis ascendens.
Truneus costo-cerviealis, or its branches.

GOTMSUAIM Te remenel ecto epskataielenarenstae
A. intercostalis suprema.
A. cervicalis profunda.
BVeqQuent:. /.,s% cla outo reaver enaeteet A. thyroidea inferior.

WAVE). syienss ancviadeteted te is do ciao ahr omCnE A. subclavia.

Robert Bennett Bean oer

M. longus capitis:
APOVN TAM Gccte re ees shale cohses mieusyevel ere A. cervicalis ascendens.

M. splenius capitis, and cervicis:

A

AOS tert ce ches oledete i sviel talel ssc ;

: Ae

IBCCUICH Geraererere cislcyetstelele Musics

Oceasionals oe cisiiche sacle sss 1
PRADO enevereterst cisictevace voucns. cl sheveycs

cerviealis profunda. Through complexus M.
princeps cervicis. Many small branches.

ascendens transversa colli. Several medium branches
between M. splenii and trapezius.

cerviealis superficialis. Several medium branches be-
tween M. splenii and trapezius.

cervicalis ascendens. Branch through above fifth ver-
tebra takes the place of A. cervicalis profunda.

M. semispinalis capitis and longissimus capitis (complerus, and trachelo mastoid) :

(4.

Cine in he tooo doeka cols aide

A. princeps cervicis.

. ascendens transversa colli.

. cervicalis ascendens.

cerviealis profunda. Many medium branches beneath
the muscles.

Many medium branches beneath the

muscles.
vertebralis. Many medium branches beneath the
muscles.

Around edge of splenii, or
through them.

cervicalis superficialis. Around edge of splenii, or

through them.
Branch through above fifth ver-
tebra takes the place of A. cervicalis profunda.

M. semispinalis cervicis, and longissimus cervicis:

cervicalis profunda. Small branches.
intercostalis suprema (R. dorsalis).

ascendens transversa colli. Small branches.

descendens transversa colli. Small branches.
cerviealis superficialis. Small branches.
cervicalis ascendens. Small branches.

WM. spinalis cervicis, and iliocostalis cervicis:

intereostalis suprema (R. dorsalis).

descendens transversa colli. Small branches.
cervicalis superficialis. Small branches.
eervicalis ascendens. Small branches.

MV. multifidus spine, interspine, and intertransverse@:

ERIS epee erst chu cfosers fect eticcvats b at
A
INTIS EAM bet sceestere chayaers! cise cokels { :
A.
BP EMQUICT try evelshcts eles ce me ile ese lite
: f R.
ACMA LOU cctcie von ovens oeRene © miate Ne
Ue
(CHOPS TN Roe no Sicha aici Dea'o Ceo? A.
; J R
(Giaenicikey OF NS ese eupioic ceo coe iA
Ike
GOTSEAOE SS He: cic een ote i
A.

A. mammaria interna.

A. mammaria interna.

vertebralis.
cerviecalis profunda.

VW. diaphragma:

VM. rectus abdominis:

328 A Composite Study of the Subclavian Artery in Man

Shoulder joint:

Conciente { A. transversa scapule.
sere > (ie ie ioc i ae lA. cireumflexa scapule.

Frequent fA. transversa colli.
MUCNT. . 2 cee eee eee ee eree | 4. cervicalis superficialis.
Clavicula:

A. transversa scapule.

Processus acromialis:

A. transversa scapule.

A. cervicalis superficialis, or transversa colli.

Sternum:

A. mammaria interna.

A. transyersa scapule.

Glandula thyroidea:
A. thyroidea inferior.

Glandula mammaria:

A. mammaria interna.

ON THE OCCURRENCE OF SHEATH CELLS AND THE
NATURE OF THE AXONE SHEATHS IN THE CENTRAL
NERVOUS SYSTEM.

BY
IRVING HARDESTY.
(From the Hearst Anatomical Laboratory of the University of California.)

In a previous paper dealing with the development of the neuroglia
tissue, a brief note was made of the occasional observance of certain half-
moon or seal-ring cells encircling the medullated nerve fibers of the devel-
oping spinal cord of the feetal pig. _ It was stated that these cells appeared
more numerous in the spinal cords of pigs between 16 and 25 centimeters,

_ the period of most active medullation, and that in transverse sections they

_ appear as seal-rings or crescents encircling the medullating axones and
that they closely resemble the “ nerve corpuscles ” or “ Schwann’s corpus-
cles” which have been described clasping the medullated fibers of the
peripheral nervous system. It was also suggested that, while in the light
of recent investigations these cells have probably little or nothing to do
with the formation of the myelin of the medullary sheath, they may
have to do with the development of the supporting framework of that
sheath.

Adamkiewicz, who first described the cells upon the fibers of the devel-
eping peripheral nerves, referred to them as “ nerve-corpuscles,” and -by
way of distinction I have referred to the similar cells observed upon the
fibers of the central system as “ seal-ring cells.” This name only apples
to their appearance in transverse section and is non-commital as to their
particular function.

The purpose of this paper is to give a further description of these seal-
ring cells based upon a more extended study of their occurrence and varia-
tions, and to offer a suggestion as to their relation to the medullary sheath.
Some attention has necessarily been given to the nature of the support-
ing framework of the medullary sheath.

MATERIAL AND METHODS.

Pig material has been used almost exclusively in the observations in
that the different ages could be easily obtained. All the study has been

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.
26

»”»

330 Sheath Cells and Axone Sheaths in the Central Nervous System

made upon the spinal cord alone, and usually upon pieces taken from the
cervical region. Preparations from the adult hog were compared with
similar preparations from the adult human. After some study of younger
stages, it appeared that only pigs of about 16 centimeters in length and
above were necessarily concerned in the study of the structures involved.
Prior to this all the fibers of the spinal cord are non-medullated or in the
very early stages of medullation and none of the nuclei and their sur-
rounding protoplasm show evidences of the differentiation in mind. The
study, therefore, has chiefly involved foetal pigs. of 16, 19, 21 and 28
centimeters, suckling pigs of about two weeks, and the adult.

Both transverse and longitudinal sections were used, supplemented
by teased preparations. Some of the sections from each specimen were
prepared by the Benda neurogla method as employed by Huber. Others
were made from pieces fixed either in Zenker’s fluid or Van Gehuchten’s
mixture and stained hghtly with hematoxylin and counterstained with
congo red. The latter method was employed in that, with other tissues,
congo red is efficient for bringing out cell outlines. Also sections from
certain of the stages after medullation has begun and sections from the
adult were stained by Mallory’s method for white fibrous tissue. Other
sections of the adult spinal cord were subjected to the action of pancreatin
by the method of “ digestion on the slide ” described by Flint.

The pieces of spinal cord of different ages from which the teased prepa-
rations were made were first split with a sharp razor into thin longitudinal
strips about one centimeter long and these were fixed in a mixture of
equal parts of saturated aqueous corrosive sublimate and 1 per cent osmic
acid. In order to fix the strips straight and extended, they were at first
merely immersed in the fluid and then placed in adherence to the walls
of the vials containing the fluid with the lower end of each strip alone in
the fluid. Corking the vials and allowing the fluid to act for 10 or 15
minutes was found sufficient to stiffen the strips so that they would remain
straight and extended when shaken down into the fluid. After doing
this they were allowed to remain in the fluid for 12 to 24 hours. The
strips were then washed for several hours in water frequently changed.

Especially in the younger stages the closely bundled nerve fibers of
the spinal cord are so friable that it was found impossible to satisfactorily
dissociate them with even the finest teasing needles. Even after dehy-
dration and while in clearing oil (in which condition nerve fibers usually
tease more easily on the slide than when in water and may be mounted
in balsam immediately without disturbing their separated positions)
but few pieces of isolated fibers could be obtained of. sufficient length for
satisfactory study. Teasing in glycerine with needles gave no better

Irving Hardesty 331

results and glycerine mounts are very inconvenient for work with high
power objectives.

After considerable trial, teasing with a fine stream of water was found
to give much the best results with the material. The following simple
arrangement proved very efficient: A strip of the fixed tissue was pinned
by one end to the middle of a small piece of thin, carefully smoothed
wooden board 4 or 5 inches long by 1 inch wide. With the free end of
the specimen downward, the board was held in the left hand with its
lower end resting at a slight angle upon the bottom of a stender dish of
suitable dimensions, while with the right hand a stream of water the
size of a fine needle was directed upon the specimen. The bits of dis-
sociated tissue wash down into the stender dish, and a wooden board
seems to give less spattering and rebounding of the water than a strip of
glass, especially a glass slide with a cell or groove in it. A piece of
8-millimeter glass tubing drawn out to the required capillary dimensions
and broken off squarely and the large end thickened and bound securely
into a piece of rubber tubing was used in obtaining the sufficiently fine
stream of water. This was securely attached either to the tap direct,
and the tap water used, or attached to a Woulfe bottle containing dis-
tilled water under pressure, the pressure being obtained by attaching the
reverse side of the Woulfe apparatus to an air force pump.

With sufficient pressure, a strip of tissue is washed down to a shred
in a few minutes. The stream is best directed against it with a shght
up and down, brushing motion. After two or three strips have been
teased, or until the stender dish is nearly full of water, the dish is set
aside and another taken if required. ‘The nerve fibers being stained
black by the osmic acid, the stender dish is placed upon a white surface.
The material soon settles to the bottom of the stender and the water may
be practically all withdrawn. Then the material from several stenders
may be all transferred to one smaller stender and again allowed to settle.
Then more water may be withdrawn and the material counterstained if
desired.

For counterstaining I used 1 per cent aqueous acid fuchsin, which is
known to stain osmic material readily and can be easily controlled. It
was only desired in the teased preparations to demonstrate the shape
and position of the sheath cells. After being in the fuchsin the required
time, the material was washed first in 50 per cent and then in 70 per
cent alcohol. It settles in alcohol more quickly than in water and the
alcohol may be drawn off more completely.

Dehydration was completed with 95 per cent alcohol and finally with
two changes of absolute. Then the absolute was replaced with a clearing

332 Sheath Cells and Axone Sheaths in the Central Nervous System

oil composed of equal parts of carbolic acid crystals, xylol and oil of
bergamot. Xylol alone will clear the material, but it acts more slowly
and dries too quickly on the slide where it is necessary to spread out the
material before mounting.

To mount, the greater part of the clearing oil was removed from the
material settled on the bottom of the dish, and then with the points of
fine forceps or with a teasing needle, a sufficient mass of the dissociated
fibers were lifted out and placed upon the slide and the surplus oil then
drained off or taken up by holding a bit of blotting paper near the mass.
A drop of balsam was then added and the fibers gently spread through it.
The cover glass placed squarely down upon the mount tends to further
spread the fibers out. Thus a preparation is obtained which is per-
manent and upon which the oil immersion may be used with convenience.

Examination of the preparations shows that teasing with the stream
of water results in three advantages not obtained by teasing spinal cord
by the ordinary methods: (1) Bits of nerve fiber are obtained of con-
siderably greater length than could be obtained with needles and the
fibers are nearly all isolated instead of in compact, broken clumps as
often results from needles. (2) The preparation is cleaner. The fibers
are washed out, leaving most of the plexuses of blood-vessels and the
coarser masses of connective tissue behind in the shred remaining pinned
to the wooden board. (3) The fibers themselves are clean. The neu-_
roglia fibers and nuclei and the general protoplasmic syncytium other-
wise surrounding and adhering to the fibers is washed off, especially from
those having acquired a medullary sheath. This adds greatly to the value
of the method for the purpose here, in that almost every nucleus to be
seen is adhering closely to a nerve fiber and usually can be considered
to represent one of the sheath cells in question.

Of the stains employed upon the sections, the Benda method proved
best for the seal-ring cells. The toluidin blue of this method seems to
especially differentiate these cells, staining their granular cytoplasm a
deeper blue than the general syncytial protoplasm and rendering the cells
more easily found than is the case after the more ordinary staining
methods. Most of the syncytium is stained light brownish-red by the
alizarin of the method, thus giving a background of good contrast. The
nuclei and the adult neuroglia fibers when present are, of course, stained
a deep blue.

In all the figures the camera lucida was used in outlining the drawings
and the magnification in each was that obtained with ocular 4 and objec-
tive rz (oil immersion, Zeiss).

Irving Hardesty 333

Tur APPEARANCE OF THE MEDULLARY SHEATH IN THE SPINAL CorD
AND THE OCCURRENCE OF THE SEAL-RING CELLS.

Most of the observations upon the origin, development and structure
of the sheaths of the nerve fibers have been made upon the fibers of the
peripheral nerves. This is perhaps due to the greater ease with which
peripheral fibers may be studied. Satisfactory fixation is easily obtained
with bits of peripheral nerves, their fibers are more easily isolated, being
supported and separated by abundant connective tissue, and they have
thicker and stronger supporting sheaths than the fibers of the central
system. Indeed, the fibers of the central nervous system are described
as having no primitive sheaths, or sheaths of Schwann, at all.

Notwithstanding various conflicting theories, the general concensus
with regard to the origin of the nerve axones and their medullary sheaths
may be summed up in the following:

1. All axones arise as outgrowths or processes of “nerve cells.” *

2. Developing axones become invested by special cells which give rise
to the sheath of Schwann but probably have nothing to do with the forma-
tion of the myelin of the medullary sheaths.

3. In development, the axone precedes, secondarily the sheath of
Schwann appears upon it, and lastly, or simultaneously with the begin-
ning of the sheath of Schwann, the myelin sheath begins to appear.

4. No sheath of Schwann and therefore no sheath cells are described
for the nerve fibers of the central nervous system. In the peripheral
system the sheath cells and therefore the sheaths of Schwann are of
mesodermal origin.

5. In the order in which they have been advanced, the theories of the
origin of the myelin sheath are: (a) that the myelin is formed through
the agency of the sheath cells by a process something like that by which
fat is formed by the fat cells (Ranvier) ; (b) that the myelin is formed
at the expense of the outermost portion of the nerve axone (Kolliker) ;
(c) that the myelin is of exogenous origin, formed in the blood and dis-
tributed from the blood-vessels to the axones (Wlassak); (d) that the
myelin arises as the result of influences exerted by the axone upon the
surrounding stroma (Bardeen). By “stroma” Bardeen refers to an
apparent fluid substance enclosed about the axone by the already formed
sheath of Schwann. The theory of Ranvier (supported by Vignal and
others) is invalidated by the statement that there are no sheaths of
Schwann in the central nervous system, while there are medullary

1See Kolliker: Ueber die Entwickelung der Nervenfasern. Anat. Anz.,

iad

July, 1904, page 7.

334 Sheath Cells and Axone Sheaths in the Central Nervous System

sheaths present, and further, by the conclusion of Gurwitch that for the
peripheral nerve fibers the sheath of Schwann has nothing in common
with the medullary sheath. It has also been shown by Kolster and
Bardeen that the myelin may begin to appear about the axone before the
sheath of Schwann is evident, though, as a rule, the axone of the per-
ipheral fiber is enclosed by the sheath of Schwann before the formation
of the myelin is apparent.

6. The medullary sheath is composed of at least two parts, first, the
myelin (lecithin, etc.), and second, its supporting framework. As early
as 1862 Mauthner roughly depicted this framework for certain giant
fibers of the trout and the coarser portions of it were observed in 1876 by
Ewald and Kiihne, who gave it the name “ neuwrokeratin,” suggesting it
to resemble horn in that they found it to resist the action of certain
digestive ferments. Since then it has been studied in greater anatomi-
eal detail by various investigators, more recently by Wynn and Hata,
whose observations were also made upon peripheral nerve fibers.

The seal-ring cells which I have observed in the spinal cord of the
pig are somewhat puzzling both as to their origin and their function. In
the first place, they appear to have a period of maximum abundance. In
the spinal cord of pigs of about 21 centimeters they are more easily found
than at any other stage I have examined. In transverse sections stained by
the Benda neuroglia method, I have seen as many as three evident in a
single field of the oil immersion. While often a field contains none at all,
it usually requires but little search to find one, though the nucleus, situ-
ated in the thicker side of the ring, may not always be contained in the
section. In the older fcetus they seem less abundant, and in the suckling
pig it becomes difficult to find them in sections, while in sections of the
adult spinal cord satisfactory examples of them are even more seldom
found. Their protoplasm seems to have been used up and a free nucleus
clasping the side of a nerve fiber may possibly be a neuroglia nucleus
instead.

There are no direct indications of seal-ring cells prior to medul-
lation. None have been observed upon fibers in the earliest stages of
medullation. The conditions before the accumulation of myelin has
begun are represented in Fig. 1, which is taken from the frayed end of a
longitudinal section of the future white substance of the cord of a pig
of 11 centimeters. The axones (a) appear as well-defined threads much
smaller than in the older stages and separated by or imbedded in the
general protoplasmic syneytium (s) which is the early form of the
neuroglia tissue, no masses of which show differential blue staining by the
Benda method, nor any definite outlines indicating individual cells. The

Irving Hardesty 335

nuclei at this stage all appear nearly similar and are simply imbedded in
the common protoplasm of the syncytium, the amount about a nucleus
depending upon its position.

This study has not involved the spinal cord of the very young stages
.and, perhaps for that reason, by none of the stains I have employed does
the growing axone, before medullation or at any stage in my prepara-
tions, appear as a “ group of fibrils,” or fibril bundles, as described by
Bardeen for the peripheral nerves and by others cited by him. The
axone of the young spinal cord, at least from 6 centimeters up, appears
as a nearly homogeneous strand of even caliber which increases appre-
ciably in size with the growth of the specimen. At best its structure
shows nothing more than a fine, elongated reticulum, the threads of
which become heavier and more evident as the fiber approaches maturity.
Though Bardeen’s observations were confined to the developing peripheral
nerves, my failure to note the fibril-group form of the young axone must

Fic. 1. From longitudinal section of white substance of spinal cord of
pig of 11 centimeters. Benda method. a—nerve axones; — syncytium.
< 550. ¢

be due to unsuitable methods, or to the fact that I have not examined
the very young stages, for it is hardly probable that the axone in the
central system is essentially different from the peripheral axone. Also
Bardeen states that as early as 2 centimeters, most of the peripheral
(intercostal) nerve fibers of the pig are covered with embryonic myelin.
This indicates that myelination occurs very much earlier in the per-
ipheral than in the central nervous system of pigs, for not till about
16 centimeters have I observed any appearances at all suggestive of the
illustrations he gives as representing embryonic myelin.

Fig. 2 represents the frayed end of a clump of axones from the white
substance of the spinal cord of a pig of 16 centimeters teased by the water
method. This material was fixed in the corrosive sublimate and osmic
acid mixture and the teased fragments stained with acid fuchsin. The
appearance of the interaxone substance was verified from sections of
the same stage stained by other methods. The teased preparation is

336 Sheath Cells and Axone Sheaths in the Central Nervous System

preferable in that the axones, uncut, may be followed a considerable dis-
tance and being often washed clean of the interaxonic syncytium, they
may be studied more closely. It is seen that even in the pig of 16 centi-
meters, a stage when the medullation of the peripheral nerves is well
advanced, most of the axones of the spinal cord show no signs of medul- -
lation, being but slender threads (a) of more or less even contour, with
the substance of the syncytium (s) adhering to them.

The fibers indicated by b in Fig. 2 show the appearance of the first stages
of the accumulation of myelin, or at least the first stages to be observed
after the technic here employed. The myelin begins as small globules of
various shapes and sizes adhering to the axone, giving it a beaded appear-
ance. The globules are but very slightly blackened by the osmic acid
at this stage and then upon their surface only, making them appear as
small blisters which resist the action of the water in teasing. When
washed clean of other adherent substance they may be observed minutely
and there is no sign whatever of the presence of any other sheath. Be-
tween adjacent blisters and connecting them there is usually discerned
a thin film on the axone but not always. Usually the globule appears
adhering to one side of the axone rather than evenly surrounding it.
This form of the first appearance of the myelin upon the axone is similar
to that described by Vignal, Westphal, Wlassak, Kolster, Bardeen and
others.

Those of the observers who take into consideration the sheath of
Schwann of the peripheral fibers, give varying accounts of the time
of the appearance of the myelin. After examining the great amount of
literature upon the subject, it seems that the sheath of Schwann usually
appears upon the peripheral axone before the myelin begins to be depos-
ited, but often simultaneously with its appearance, and sometimes after
the appearance of the myelin. The latter sequence indicates that the
sheath of Schwann is not concerned in the origin of the myelin.

The fiber indicated by ¢ in Fig. 2 is an example of the most advanced
stages of myelination to be found in the spinal cord of pigs of 16 centi-
meters. It is the only fiber found after considerable search through the
preparations of this stage which apparently possesses a sheath cell, though
the protoplasm of this cell is not distinctly differentiated. There is posi-
tively nothing indicating such cells upon fibers of earlier stages of medul-
lation. In sections of specimens of this age stained by the Benda method
I have found no cells distinctly clasping the medullating axone, and
showing the definite outline and differential staining of those found in
the later stages, and especially in pigs of 21 centimeters. Occasionally
a nucleus may be seen upon the side of an axone with protoplasm sur-

Irving Hardesty 337

rounding it which is apparently more compact and which stains a deeper
blue than the protoplasm of the general syncytium, but instead of having
a definite outline the protoplasm seems to grade off into that of the
syncytium. ‘This condition is apparent in fiber c of Fig. 2. In general,
the nuclei of this stage are merely imbedded in the syncytial protoplasm
and show the various types of the neuroglia nuclei.

The appearance of definitely formed seal-ring cells is shown in Fig. 3.
In this figure are represented two small areas from transverse sec-
tions of the spinal cord of a pig of 21 centimeters stained by the Benda
neuroglia method. ‘The cells (c) here show the form suggesting the name
given them. Their finely granular protoplasm stains a decidedly deeper
blue than that of the now more sparse protoplasm of the general syncyt-

Fie. 2. From the spinal cord of a pig of 16 centimeters. Osmic acid and
fuchsin. Teased by water. a@—axones before medullation; 6—axones
showing beginning medullation; c— fiber in more advanced stage of medul-
lation and with probable sheath cell; s=syncytial protoplasm. > 550.

lum and their boundaries are definite. The fields were chosen because
of each having two cells near together, three of the cells containing nuclei
in the section. With the nucleus in fhe thicker side, the cells usually in
this stage completely enclasp the fiber as a ring, but sometimes the pro-
toplasm on the side away from the nucleus is either absent or so thin
as to give the appearance of a crescent. Frequently a cell is found of
the shape presented in d, where the protoplasm seems mostly extended
from one side. In sections, the cells seldom seem to produce a depres-
sion in the medullary sheath for the fiber usually appears circular. Very
probably none of the fibers possessing these cells are full grown, for the
cells are found more abundant at about this age and the average diam-
eter of the medullated axones here is much less than that of the adult. At

338 Sheath Cells and Axone Sheaths in the Central Nervous System

21 centimeters there are still many fibers in the spinal cord which have
not acquired a myelin sheath (a). The syncytial protoplasm is less
abundant probably because it is being transformed into neuroglia fibers
(n) which begin to appear at this stage.

After birth cells of the seal-ring type are less numerous than in the
pig fetus. Fig. 4, c, shows one as found in the suckling pig of two
weeks. It is upon a larger fiber than those in Fig. 3 and the protoplasm
of the cell is relatively less in amount and merely forms a crescent about
the medullary sheath. The nucleus represented by e of this figure is
probably a nucleus of a seal-ring cell which has no blue staining proto-
plasm about it. This is only inferred from its position, resting upon
and here slightly indenting the medullary sheath. It may possibly be

Bi Rosery: ct

Fie. 3: Fie. 4.

Fic. 3. Areas from transverse sections of spinal cord of pig of 21 centi-
meters. Benda method. c and d=—seal-ring cells; @a—axones before medulla-
tion; n=beginning neuroglia fibers; nn —neuroglia nuclei. > 550.

Fig. 4. From spinal cord of suckling pig of two weeks. Showing stage
more advanced than Fig. 2. Otherwise same as Fig. 2. c—=seal-ring cell; e=
probably nucleus of former seal-ring cell (taken from a different field); n —
neuroglia fiber; nc—neuroglia cell. X 550.

one of the larger “ free” neuroglia nuclei which has acquired this posi-
tion. At this stage neurogla nuclei are sometimes observed which have
an area of more compact protoplasm about them and which stains darker
than the, now scarce, granular protoplasm of the syncytium. One of these
“ neuroglia cells ” is shown in the figure (mc) and such are described in
many of the papers dealing with the neuroglia. Both this and the con-
dition represented in the nucleus e are found in the adult material (see
Pic ey. See

In Fig. 5 are arranged some types of fibers selected from teased prepa-
rations of the spinal cord of the 21-centimeter pig fixed in the osmic
acid mixture and counterstained with fuchsin. Three of these pieces of

Irving Hardesty 339

fiber possess sheath cells and it will be noted that all three are fibers upon
which the processes of medullation are well advanced. In the group
indicated by a are three fibers not separated by the teasing and there is
still present about them the syncytial protoplasm and some of the nuclei
belonging to it (s). One of the axones of this group as yet shows no
evidence of myelin, a condition which is quite frequent in pigs of this
age. The fibers b and ¢ were selected as showing the next stages in the
acquirement of myelin. The medullated fiber with group a@ was included
as illustrating the corrugated or ruffled outline of the growing sheath, an
appearance frequently found and which suggests that it is an earlier
stage than either of the fibers indicated by e, being a further develop-

Fic. 5. Types of nerve fibers selected from teased preparations of spinal
cord of a pig of 21 centimeters, showing stages of medullation and nature of
seal-ring cells. Osmic acid and corrosive sublimate. Fuchsin. s—syncyt-
ium; fr—framework of sheath washed out in teasing; sc—seal-ring cells.
X 550.

ment of the smaller blistered or beaded form (b and c). The fiber d is
perhaps in about the same stage, but it is doubtful whether either of the
nuclei adhering to it represents sheath cells. The fiber f, showing an
even contour of its sheath, is considered as illustrating the type of the
most advanced stage in the growth of the medullary sheath found in
pigs of 21 centimeters. This type is fairly numerous and often sheath
cells are found upon it. The myelin of this type stains more darkly than
that of the others, especially that of b and c, where it is much lighter and
corresponds with certain of the descriptions of so-called “ embryonic
myelin.”

In the late foetus and in the new-born (suckling pig of two weeks)
the conditions more nearly resemble those of the adult. Medullation

340 Sheath Cells and Axone Sheaths in the Central Nervous System

has proceeded till there are much fewer fibers of types b and c¢, Fig. 5.
At birth there may be found in the white substance, but very rarely
indeed, fibers totally void of myelin such as one of those shown in group a.
Sheath cells of the form of the seal-ring cells of the 21-centimeter pig
are also more difficult to find in the later stages. This is apparently due
to their being relatively less numerous and to the fact that when found
they show relatively less protoplasm about their nuclei and about the
fibers they clasp.

In the adult especially are unquestionable examples of these cells diffi-
cult to find. Usually the protoplasm has apparently been used up or so

Fic. 6. Types of fibers selected from teased preparations of spinal cord
of adult hog. Fixation, etc., same as in Fig. 4. e and g=types of small,
thinly medullated fibers from white substance; fr—framework of sheath
from which myelin has been washed in teasing; sc—sheath cells (sheath
nuclei); p= peripheral sheath. X 550.

dispersed that little more than the nucleus remains and the only sugges-
tion that such nuclei do not belong to the general class of the neurogha
nuclei present throughout the inter-axone spaces, is their flattened shape
and their position upon the nerve fibers, and, when observed in the teased
preparations, the fact that they are not washed off in the process of teas-
ing. It is possible, of course, that during growth neurogha nuclei proper
may be flattened against the medullary sheaths.

In Fig. 6 are presented some of the types of fibers to be found in the

Irving Hardesty 341

spinal cord of the adult hog. They are all selected from teased prepa-
rations of the same specimen. Three of these pieces of fiber (a, 6 and c)
possess what may be reasonably considered sheath cells. They were
found after considerable search. Sometimes cells could be found adher-
ing to the fibers in the manner similar to that shown upon the fiber d,
but these were not considered as examples of the type sought. Cells
situated in indentations of the sheath and with a more or less even outer
contour were sought as more probable examples of the cell in mind.

There are many fibers in the adult cord relatively larger than the
type d, but so far I have found no cells upon their sheaths, which latter
are always deeply blackened by the osmic acid. Of the smallest medul-
lated fibers in the adult cord, the types indicated by e and g are inter-
esting. In the peculiar bulbous enlargements of their sheaths and in
their relative size, they are identical with certain of the fibers described
by Ranvier in the spinal cord of the dog. Ranvier pictured a nucleus
with protoplasm about it adhering to what appears to be one of the larger
of this type of fibers. So far I have not found examples of sheath cells
upon any of them in the adult hog, but this may be due to the fact that
such fibers are much less numerous in the cord than fibers of the larger
type and that therefore a much less number of them was examined.
Their peculiar appearance can hardly be considered wholly artifact, for
the type is quite constant and often one of them may be followed un-
broken for several millimeters and throughout shows the same form. It
would be difficult to determine whether or not they are younger fibers in
the process of medullation. They resemble certain of the Remak fibers.
The few undoubted collateral branches I have seen in the preparations
were of this general type of fiber.

The relation the sheath cells bear to the medullary sheaths of the
central nervous system is as much of a question as it is in the peripheral
nerves. When, in the study of this question, one examines into the
nature of the medullary sheath in the spinal cord, it is immediately evi-
dent that it differs from that of the peripheral nerve fibers in several par-
ticulars.

While the fiber of the spinal cord is, of course, devoid of any structure
similar to the capsule or sheath of Henle possessed by certain peripheral
medullated fibers, it also lacks a distinct sheath of Schwann. Many deny
that the nerve fiber of the central system possesses anything similar to the
sheath of Schwann found on the peripheral fiber. Under certain con-
ditions when the myelin is crushed or shrunken away, there may be seen
occasionally evidences of a very delicate sheath about the periphery of
the medullary sheath (p, Fig. 6). More usually, however, such must

342 Sheath Cells and Axone Sheaths in the Central Nervous System

adhere so closely to the myelin as to be invisible and to break with
the breaking of the myelin sheath. This thin membrane-like appear-
ance was first noted by Ranvier in the cord of the dog and he referred to
it as a membrane. It has since been discussed by Schiefferdecker and
others and its existence has been repeatedly denied. |

The nerve corpuscles or Schwann’s corpuscles of the peripheral nerve
fibers are described as lying under the sheath of Schwann—hbetween it and
the myelin sheath. The cells here observed upon the fibers of the spinal
cord seemingly he upon the medullary sheath, being attached or in some
way in close relation to it, without being enclosed upon it by a percep-
tible membrane. Sometimes in the teased preparations the protoplasm
of the cells seems to blend into the blackened myelin (e, Fig. 5), but in
the stained sections the protoplasm appears distinctly outlined from the
myelin. The latter is perhaps the true condition, for whatever the func-
tion of the cells, the granular, more deeply staining portion probably
only represents*the untransformed endoplasm of the cell. In Fig. 7,
sc, is shown one of the seal-ring cells as found in transverse sections of
the spinal cord of the adult hog when stained by the Benda method. It
is merely a nucleus practically free of endoplasm and its shape and posi-
tion are the only features which suggest its being one of the cells in
question.

I am as yet unable to reach a definite solution of the exact significance
of these cells. If one follows them through the preparations with the
idea that they are a distinct type of cell, having probably to do with the
development of some part of the medullary sheath, the following observa-
tions may be made in support of this idea:

1. They do not appear differentiated till after the acquisition of mye-
lin has begun.

2. When first indicated they do not appear as definitely outlined and
differentiated cells, but rather their more deeply staining protoplasm
seems to grade off into that of the general syncytium and to be continuous
with it. This, and their being attached to a medullating fiber suggests
that they are derived from nuclei and protoplasm formerly a part of the
syncytium, and that their differentiation may be due to influences ex-
erted upon it by the developing myelin.

3. During the period in which the process of myelination is at its
height, they appear as distinctly differentiated cells with considerable
protoplasm (or probably endoplasm) about their nuclei, often sufficient to
completely encircle the growing sheath at the level of the nucleus.

4. With the further growth of the medullary sheath, the protoplasm or
endoplasm of the cells is apparently used up gradually, till as the sheath

Irving Hardesty 343

nears completion only the nucleus appears adhering to the periphery of
the sheath. The fact that even such nuclei are rare in the adult suggests
that they also may disappear.

5. In the relative abundance of their protoplasm at the different
periods, these cells resemble the nerve corpuscles described for the per-
ipheral nerve fibers and which are interpreted as having to do with the
structures of the sheath.

On THE FRAMEWORK OF THE MEDULLARY SHEATHS OF THE SPINAL
Corp.

In none of my preparations of the spinal cord of the hog is there evi-
dent any arrangement in the medullary sheaths producing the appearance
of the Schmidt-Lantermann clefts and segments described in the sheaths
of the peripheral fibers. Especially is there no evidence of the heavy,
separate interfitting cones described by Wynn. Hatai, who used a
method much superior for the purpose to that used by Wynn, describes
the structure of the peripheral medullary sheath as consisting of a net-
work which contains the myelin. This “neurokeratin” network con-
sists, he says, first, of two thin layers, one beneath the sheath of Schwann,
and the other around the axone, the two being continuous at the nodes
of Ranvier; and second, of a chain of cone-like formations, the bases of
the cones being attached to the peripheral layer and the apices to the
axone layer. Both the cones and the layers are highly reticular, exhib-
iting meshes of various sizes and shapes. He thinks the Schmidt-Lan-
termann clefts are produced artificially.

Hatai found formalin the best fixing agent in his study of this frame-
work. Formalin is the fixing agent required in the Benda method and the
alizarin used in the staining procedure of this method apparently brings
out the framework of the medullary sheath in greater delicacy of detail
than the stain used by Hatai. The toluidin blue of the Benda method
stains the axone a dense blue as it does the developed neuroglia fibers,
while the white fibrous connective tissue and the framework of the medul-
lary sheaths is stained a light brownish-purple even to the finest fibrille.
After formalin, as after many other fixing agents, the axone shrinks to
considerable density and decrease in its normal diameter, but on the other
hand, formalin seems to produce a slight swelling of the medullary sheath.
In this process the framework of the sheath remains attached to the
axone and is thus drawn or distended into a more open condition which
renders the study of its detailed structure less difficult.

On comparing sections of peripheral nerve fibers with sections of the

344 Sheath Cells and Axone Sheaths in the Central Nervous System

white substance of the spinal cord, from both of which the myelin has
been removed, it is evident at once that, whatever the detailed structure
of the sheaths of the two, the framework of the peripheral medullary
sheaths is somewhat stronger and heavier than that of the sheaths of
the central fiber. This is true for human material and is especially true
for the hog. The comparison may usually be made very readily with
sections of the spinal cord, for nearly always portions of the ventral and
dorsal roots have remained attached to the cord and, involved in the
sections, are subjected to the identical technique in staining, etc., as the
cord itself. In the peripheral nerve proper the framework of the medul-
lary sheath is slightly heavier than in the nerve roots close to the pia
mater. The nerve roots also do not possess the heavy connective tissue
investments of the nerve outside the dura mater. Furthermore, in the
hog the framework of the medullary sheath, both in the peripheral and
central system, is apparently heavier than in man and the other verte-
brates more usually studied. This is coincident with the well-known fact
that in the hog the fibrous tissue framework of the organs, especially
white fibrous and reticular tissue, is peculiarly abundant.

Under higher magnification the framework of the medullary sheaths
of the hog shows a structure and arrangement capable of an interpreta-
tion somewhat different from that usually given. The structure cer-
tainly differs considerably from that pictured by Wynn. Wynn, however,
used the Weigert staining method for medullated fibers and many of the
appearances to be obtained by this method suggest that it is somewhat pre-
cipitative in its action upon the medullary sheath or that it may be classed
among the impregnation methods. Its tendency certainly is to clog the
finer structures rather than merely to dye them. Wlassak in his study
of the origin of the myelin claims that the Weigert method stains only
one substance of the medullary sheath. This substance he calls “ cere-
brin” and states that it is one of the constituents of the myelin. In
this case, as Hatai points out, Wynn probably did not study the real
framework of the sheath, the neurokeratin network of Hatai, but rather
obtained pictures indicating the distribution of the cerebrin.

As stained by the Benda neuroglia method, the framework of the me-
dullary sheath of the hog spinal cord appears as represented in Fig. 7.
This figure contains fibers from both transverse and longitudinal sections
of the cord. Each group was taken from an area near the periphery of
the section or in the neighborhood of the pia, for the reason that near
the periphery the framework appears heavier than toward the center and
is always less collapsed and shrunken, due perhaps to better or earlier
fixation near the surface of the specimen. It is seen that in transverse

Irving Hardesty 345

sections stained by this method the framework supporting the myelin
appears arranged in the form of concentric lamelle. The different
lamelle, however, cannot be followed as distinct and individual mem-
branes, for they apparently anastomose with each other and are further
connected by still finer threads or branches. The structure is better de-
scribed as a lamellated reticulum in the meshes of which the myelin is
contained. Were it fibrillar in structure, lamellation could not appear
in both transverse and longitudinal section. In longitudinal section the
meshes of the reticulum appear considerably elongated in the direction
of the long axis of the nerve fiber.

At the periphery of the fiber there 1s a slight condensation of this
lamellated reticulum, giving under certain conditions the appearance of
a membrane (p, Figs. 6 and 7), but close examination of the uncollapsed
sections shows this membrane continuous with the more open network
further in. This explains the difficulty with which the membrane is
seen and the disputes concerning its existence, for consisting of but a
condensed peripheral portion of the reticulated framework, the meshes
of which are intimately occupied by the myelin throughout, the mem-
brane is really continuous with the framework and therefore necessarily
seems to adhere closely to the myelin. The breaking of the medullary
sheaths in the fixed preparations consists, of course, of a breaking of the
framework, and in the usual osmic ‘acid and Weigert preparations espe-
cially, one could hardly expect to see the apparent membrane except in
fortunate cases where the myelin is crushed or shrunken away from the
periphery in such a way. as to expose it. Quite frequently in the teased
corrosive-osmic preparations the broken end of a fiber showed frayed por-
tions of the framework of the sheath as indicated in Figs. 5 and 6, fr.
These appearances are due to the myelin having been washed out of the
meshes of a short extent of the framework by the action of the water in
teasing and are probably not to be seen except in preparations teased by
water.

The lamellated reticulum also usually shows a slight condensation
about and upon the axone of the medullated fiber. This corresponds to
the second thin layer of neurokeratin as described by Hatai. It probably
corresponds to the axolemma frequently mentioned in the books. Being
of the same nature and formed in the same way as the peripheral mem-
brane, the usual difficulty with which it is seen is perhaps due to the
same reason as that given to explain the difficulty with which the per-
ipheral membrane is seen.

In the longitudinal sections of the fibers of the spinal cord it is seen
that even the heavier lamelle of the framework do not run uniformly

o7

al

346 Sheath Cells and Axone Sheaths in the Central Nervous System

parallel with the contour of the fiber. At least after the manipulation
in making the preparations, lines of adjacent lamellae often appear col-
lapsed upon each other, giving a resemblance of heavier lines running
obliquely in the sheath. Often this collapse may be so great as to give
openings in the framework, and sometimes, especially nearer the center
of my sections of the cord, where fixation is perhaps less perfect, almost
the whole framework may appear clotted against the axone or massed,
axone and all, at one side of the section of the fiber. A partial collapse
of this kind often occurs near the periphery also and gives rise to the

Fic. 7. Small areas from the outer portion of a transverse and of a
longitudinal section of the spinal cord of an adult hog. Benda neuroglia
method. a—axone; al—=axolemma; p=—peripheral membrane; = neuro-
glia nucleus; nf=neuroglia fibers; sc—seal-ring cell as seen in adult; d=
sheath with collapsed framework. Detail of figure may be better observed
with hand lens. X 550.

appearance indicated in fiber d, Fig. 7. Again, in the less collapsed
condition, one of the lines of collapse may appear in the longitudinal
section to run slantingly from the axone to the periphery of the fiber
(c, Fig. 7) suggesting one of the Schmidt-Lantermann clefts of the
usual osmie preparations of peripheral fibers.

The framework of the medullary sheaths of the peripheral nerve fibers
appears not only somewhat heavier than that of the fibers in the spinal
cord, but also its arrangement is more varied and generally more com-
plex. The form of the reticulated framework may be said to consist of

Irving Hardesty 347

two general types with, however, all gradations between the two. In
Fig. 8, A and B, there are represented examples of the extremes of the
two types. They are chosen from among the dorsal root fibers of the
adult specimen and show the nature of the framework as brought out by
the Benda stain. The majority of the fibers in both the nerve roots show
an arrangement of the framework conforming in various degrees to type
A. The framework of type B, though somewhat heavier in structure,
conforms quite closely with the general type found in the spinal cord.
To illustrate type B, a piece of fiber involving a node of Ranvier (nF)
was chosen to show the interesting fact that the lamellated reticulum of
the sheath framework is not interrupted at the node. Not only are both

Fig. 8. Transverse and longitudinal sections of dorsal root (peripheral)

lad

nerve fibers from same preparations as Fig. 7. Showing extremes (A and B)
of the two general types of framework of medullary sheaths. sch —sheath
of Schwann; n—nerve corpuscle; nR—=node of Ranvier; ¢—cone arrange-
ments of framework; other letters—=same as in Fig. 7. Use hand lens to
observe details cited. » 550.

the outer and inner “membranes” continuous through the node, as
pointed out by Hatai, but also some of the intermediate lamellx pass
from one internode to the other. From the fact that but shght condensa-
tion is apparent at the nodal constriction, it is probable that the frame-
work is less abundant at that point. Close examination shows that the
reticulum suddenly narrows down by both a diminution in the number
of its lamelle and a diminution in the size of its meshes. The same
general behavior is also apparent at the nodes in fibers of type A.
Sheath frameworks of the type B seem less frequent in the peripheral
nerves proper than in the nerve roots within the dura mater. So far,
however, I have examined sections of only one piece of peripheral nerve

348 Sheath Cells and Axone Sheaths in the Central Nervous System

stained by the Benda method. The root fibers were chosen for the illus-
trations because, being on the same slides as the sections of the spinal
cord, they were subjected to the identical fixation, treatment and decol-
orization as the fibers in the cord, and were therefore deemed better for
comparison with the conditions found in the cord.

The form of sheath framework shown in A, Fig. 8, is no doubt the
form to which attention has been usually given in the literature. In
this the reticulated framework is more or less condensed into interfitting
conical partitions between masses of less intimately supported myelin.
In this condensation the outer and inner “ membranes” (p and al) are
maintained and rendered even more evident. The partitions when in
the form of cones are so arranged that the bases of the cones are con-
tinuous with the peripheral membrane and the apices with the axolemma.
This is necessarily the case. Otherwise they would not be conical. When
a sheath shows the decided conical arrangement for any considerable
distance, which it very seldom does, the cones are not necessarily arranged
parallel and interfitting throughout the distance. For a short extent
they may be parallel, then irregular cones may be interposed or cones
with their apices pointing in opposite directions. The most perfectly
formed cones themselves contain openings of the same optical properties
as the spaces between the cones. As stained here the cones appear to
consist of a fibrillated reticulum with the meshes greatly elongated in
the direction occupied by the cone between the axone and the periphery.
In A, a fiber was found through which the knife passed obliquely at a
region of more or less perfectly formed cones. This allows a certain
amount of perspective. Luckily the region also possessed a sheath nu-
cleus. It is here especially evident that the cones do not fit cleanly
upon the axone but are continuous with the axolemma both above and
below by a more irregularly dispersed portion of the reticulum in such a
way as to give an appearance resembling an open umbrella, the rib-braces
of which often extend to the apex of the adjacent cone (c, Fig. 8). The
appearance shown in the transverse section indicated by A is frequently
seen in the sections of peripheral fibers and is interpreted as a section
involving the base of one cone and the apex of another. ‘The heavier
fibrille show a pecuhar whirled arrangement which is probably due to
the lamelle of both cones being cut at levels where they are curving upon
the axone in the one case and the periphery in the other.

Though frequently cones are to be found in the hog material somewhat
longer than the example chosen in A, I have as yet seen none as long as
certain of the cones pictured by Hatai and Wynn for the peripheral nerves
of the animals they studied. The finely fibrillated reticular nature of

Irving Hardesty 349

the framework suggests that the heavy cones pictured by Wynn may after
all refer to the arrangement pictured here. His pictures probably rep-
resent the conical arrangement of the reticulated framework, the fine
meshes of which had been clogged by the precipitate of the Weigert
method.

The Schmidt-Lantermann clefts seen in the usual osmic prepara-
tions of the peripheral nerves are interpreted as representing the cones
in longitudinal section. The cones, being condensed portions of the
framework, which for that reason contain less myelin than other portions
of the sheath, are therefore less blackened by the osmic acid. Light pass-
ing through them more readily results in the familiar appearances.

As is well known, in the ordinarily fixed material, every peripheral
fiber does not show the clefts, and when they are shown they seldom
appear in straight, even course slanting from the axone to the periphery.
Further, the clefts never appear parallel along a considerable extent of a
fiber, indicating that the cones do not all le in the same direction. Again,
it is invariably claimed that the clefts are not present in the fresh con-
dition of the medullary sheath; that after death the sheath soon under-
goes changes resulting in their appearance.

All the statements concerning cones and clefts in the medullary
sheaths have been made with reference to the peripheral nerve fibers
alone. I have found neither in the fibers of the spinal cord. In the
peripheral fibers it seems to me that a partial explanation of the cones is
suggested in the many varieties of arrangement of the framework of
sheath to be found in the fixed and stained material. Considering types
A and B of Fig. 8 as the two extremes, a study and arrangement in series
of the intermediate forms may be made which will suggest that the inter-
mediate forms are gradations from type B into type A; in other words,
that the form of framework shown in type B, though less frequent in the
preparations, may represent more nearly the normal arrangement and
that type A is probably derived from type B through a procession of
post-mortem changes. In the intermediate forms may be noted: (1)
those in which the lamellated reticulum of type B, arranged more or less
parallel with the axone, contains occasional small, oval or globular spaces
interrupting the parallel arrangement; (2) those in which the spaces
are more numerous, some of them much larger than others and so shaped
and situated with reference to each other as to suggest that, the larger
spaces arise from a coalescence of the smaller; (3) those sheaths in
which the larger spaces predominate, giving the framework a marked
blistered or honeycombed appearance with the smaller spaces in the
partitions between the larger; and finally, those which conform in

350 Sheath Cells and Axone Sheaths in the Central Nervous System

general to type A of Fig. 8. Here a continued coalescence of the
spaces has resulted in some so large that they more or less completely
encircle the axone and are bounded by necessarily condensed portions
of the reticulated framework which, at the periphery of the sheath
and about the axone, amount to httle more than the outer and inner
layers of Hatai. The partitions between adjacent large spaces them-
selves often contain numerous smaller spaces. These larger spaces
encircling the axone may be often so shaped and arranged that in
longitudinal section, either optically or by the microtome, the result-
ing partitions of the framework between them may easily appear in’
the more or less conical form. In the left-hand end of the bit of fiber
indicated by 4A, the cones are less apparent than further to the right.
Many of the fibers with numerous larger spaces do not show the conical
arrangement of the framework, even as distinctly as is indicated in 4.

The spaces are interpreted as occupied by globules of myelin, the
larger resulting from a coalescence or fusion of the smaller globules. In
the normal fresh condition the myelin is probably not in the form of
globules at all but as a more finely divided emulsion evenly distributed
throughout the framework supporting it. Globulation, beginning as very
small globules which later coalesce into the larger, apparently results in a
distortion of the natural arrangement of the framework of the myelin
sheath. The beginnings of such may be noted even in B of Fig. 8 and in
all fibers of its type. In accordance with this view the conical arrange-
ment of the framework and all the intermediate forms may be looked upon
as artifacts. The material of my preparations was taken at the slaughter
house shortly after death and immediately placed into the fixing fluid.
That the peripheral nerves were more exposed to the atmosphere and to
handling probably explains why globulation is so marked in their fibers
and not in those of the spinal cord. However, the material of my Brenda
preparations of the human cord was taken 48 hours after death and in this
also the sheaths of the central fibers do not show the conical arrangement
of the framework. But it also was placed into the fixing fluid immedi-
ately upon removal from the body.

That the medullated sheaths of the peripheral nerve fibers after fixation
in weak solutions of osmic acid or after poor osmic fixation usually
present the appearance of a coarse network, is a matter of frequent obser-
vation. Lantermann himself noted such an appearance. Ewald and
Kiihne named it newrokeratin. The appearance is considered due to
the myelin being in the form of imperfectly blackened globules, the
interstices between the globules giving a merely optical impression of
a network. Pertik, however, interpreted it as indicating the presence

Irving Hardesty 351

of a substance between the globules which colored differently by the
osmic acid. Boveri and Kupffer thought the appearance of the network
to result from a stage of the decomposition of the myelin, while on the
other hand, Gedoelst affirmed the preformation and preéxistence of the
network, agreeing with Pertik that it indicated the presence of two sub-
stances in the myelin sheath. More recently (1904) Chio, experimenting
with different solutions of osmic acid upon the peripheral fibers of the
frog and guinea-pig, reached the conclusion that the globules are a con-
stant form of the myelin. He finds globules present not only after the
action of the osmic acid but after various other reagents as well, both
isotonic and anisotonic. His illustrations, all of them after the action
of osmic acid, present the appearances usually found in imperfectly
blackened medullated fibers. By removing the blackened portions, the
globules of myelin, from his pictures, all of them may be homologized
with the form here indicated by A in Fig. 8 and with the gradation
forms between A and type B.

Without doubt myelin exists normally in a finely divided form—an
emulsion. For this reason the nerves possessing it appear white by
reflected light. In the fresh condition, however, the individual droplets
are much smaller than after post-mortem exposure and treatment with
reagents. The larger globules, arising after death by a continual coal-
escence of the smaller, perhaps bear a similar relation to the condition
in the fresh nerve as do the globules of cream bear to those of fresh milk.
In the finely divided condition the myelin is distributed evenly through-
out the sheath framework supporting it; in the coarser globular form the
normal arrangement of the framework is distorted in the characteristic
manner by the continued coalescence of the smaller into the larger glob-
ules, the conical appearances resulting from the necessary shape of the
large globules within the confines of the cylindrical sheath.

In the central system there is no sheath of Schwann conforming to
the distinctly formed membrane investing the fibers of the peripheral
system. In my sections the sheath of Schwann of the peripheral fibers is
well differentiated by the Benda stain. It seems everywhere to be com-
pletely separate and distinct from the framework of the sheath, as is
shown in Fig. 8, sch. The sheath nuclei, usually pictured as ad-
hering to the sheath of Schwann, by no means necessarily do so.
Usually surrounded by a small amount of protoplasm, they may as often
be found adhering to the surface of the medullary sheath as to the inner
surface of the sheath of Schwann. The sheath of Schwann closely invest-
ing the medullary sheath, the nuclei are usually in contact with both.

The sheath of Schwann in both structure and staining properties re-
sembles the ordinary basement membranes of the epithelia of the body.

‘
352 Sheath Cells and Axone Sheaths in the Central Nervous System

Basement membranes are of connective tissue origin and are not cellular.
The sheath nuclei may represent certain of the cells having to do with
the development of the sheath of Schwann which were enclosed within
the sheath and therefore separated from the similar nuclei distributed in
the endoneurium outside the sheath. It is well known that, while an
internode of the medullated peripheral fiber usually has a sheath nucleus,
one is not present in every case, and further, a single internode may
sometimes show two or more sheath nuclei.

On the other hand the sheath nuclei may have to do with the develop-
ment of the framework of the myelin sheath as well as with the sheath
of Schwann. The framework, both of the central and peripheral fibers,
stains like the sheath of Schwann. It is suggested that the seal-ring
cells of the foetus and the sheath nuclei of the adult spinal cord repre-
sent “cells” derived from the syncytium and whose activities result in
the reticulated framework of the medullary sheath; that, wherever found,
the protoplasm about the nucleus represents only the endoplasm which is
being transformed into exoplasm, which in its turn is transformed into
the lamellated reticulum by a process similar to that described by Mall
in the development of the fibrous connective tissues; and finally, that the
origin of the framework and the origin of the sheath of Schwann of the
peripheral fibers may be similar.

The framework of the medullary sheath in the spinal cord resists the
digestive action of pancreatin as first noted by Ewald and Kiihne for
that of the peripheral fiber. However, in the digested sections it does
not appear so abundant as after the Benda neurogha stain. After fixa-
tion in other fluids it does not appear as abundant as it does after fixa-
tion with formalin and also it stains very hghtly or not at all by the
ordinary staining methods. Mallory’s method for white fibrous con-
nective tissue stains it but lightly. The digested sections of the spina
cord stained by this method or with strong fuchsin solutions show more
or less collapsed circles representing the transversely cut nerve fibers.
These circles contain remnants of the framework usually so collapsed and
washed together that little semblance of the original arrangement can be
made out. Occasionally there is a small inner ring showing the opening
from which the axone has been digested and representing the inner layer
of the framework or axolemma with certain other portions of the frame-
work collapsed upon it. The outer ring, representing the periphery of
the fiber, is better maintained, due probably to the presence of the inter-
stitial framework of the white fibrous connective tissue of the spinal cord.

Irving Hardesty 353

SUMMARY.

1. There are present upon the medullated fibers of the spinal cord cells
similar to the nerve corpuscles or sheath cells of the peripheral nerve
fibers. :

2. These cells are more abundant and possess more protoplasm during
the period of the most active formation of the myelin sheath than during
the later stages.

3. They do not appear upon the fibers till after the fibers have begun
to acquire myelin.

4. They are apparently derived from the nuclei and protoplasm of the
syncytium of the developing spinal cord and are perhaps differentiated
through some influence exerted upon the syncytium by the developing
myelin upon the axones.

5. These cells occur much more rarely upon the adult fibers and when
found possess little or no protoplasm.

6. The framework of the medullary sheaths of the spinal cord occurs
in the form of a lamellated reticulum in the meshes of which the myelin
is supported.

7. This framework differs from that of the medullary sheaths of the
peripheral nerve fibers in that it is not quite so heavy and always shows
an arrangement parallel with the axone of the fiber.

8. The more or less parallel arrangement of the reticulum fs probably
the normal condition of the framework in the peripheral nerves also, the
post-mortem appearance of the usually described coarse “ neurokeratin
network ” being but a distortion of the normal arrangement produced by
a continued coalescence of the much finer globules of the original myelin
emulsion, the occasional conical arrangement of the framework repre-
senting the final result of the further coalescence of the globules.

9. The framework of the medullary sheaths of the spinal cord resists
digestion as does that of the medullary sheaths of the peripheral nerves.

10. While there is a supporting contingent of white fibrous tissue
among the nerve fibers of the spinal cord, the statement is confirmed that
there is no distinct, separate membrane investing the fibers of the central
system corresponding to the sheath of Schwann investing the medullated
fibers of the peripheral nerves.

11. The sheath cells of the spinal cord are probably concerned in the
development of the framework of the medullary sheath and probably in
a manner similar to that in which the other fibrous supporting tissues of
the body are developed.

354 Sheath Cells and Axone Sheaths in the Central Nervous System

AUTHORS CITED.

ADAMKIEWICZ, ALBERT.—Die Nervenkorperchen. Ein neuer, bisher unbekannt-
er morphologischer Bestandtheil der peripherischen Nerven. Sitzb.
der Kais. Akad. der Wiss. (Wien), Bd. XCI, Abth. III, 1835.

BARDEEN, C. R.—The Growth and Histogenesis of the Cerebro-spinal nerves
in mammals. Am. Jour. Anat., Vol. II, No. 2, 1903.

BoverRrI und Kuprrer.—Ueber die Bau der Nervenfasern. Sitzungsbr. der Ge-
sellsch. ftir Morphol. und Physiol. (Munich), Bd. I, 1885.

Cu10, M.—Sur quelques particularités de structure de la fibre nerveuse
myelinique soumise a l’action de l’acide osmique. Arch. Ital. de Biol.,
TT Xu, Fase: 2; 1904:

Ewatp und Kit'une.—Ueber einen neuen Bestandtheil des Nervensystem.
Verhand. des naturh. med. Vereins zu Heidelberg, Vol. I, Heft. 5,
1876.

Fuint, J. M—A New Method for the Demonstration of the Framework of
Organs. The Johns Hopkins Bulletin, Vol. XII, 1902.

GreporLtst.—Etude sur le constitution cellulaire de la fibre nerveuse. La
Cellule, T. III, Fase. 1, 1886.

Gurwitscu, A.—Die Histogenese der Schwannischen Scheide. Arch. ftir Anat.
und Physiol., Anat. Abthl., 1900.

Harpesty, I.—On the Development and Nature of the Neuroglia. Am. Jour.
Anat., Vol. III, No. 3, 1904.

Hatar, S.—The Neurokeratin in the Medullary Sheaths of the Peripheral
Nerves of Mammals. Jour. Comp. Neurology, Vol. XIII, No. 2, 1903.

Huser, G. C.—Studies on the Neuroglia. Am. Jour. Anat., Vol. I, No. 1, 1901.

Kouster, R.—Beitrige zur Kenntniss der Histogenese der peripherischen
Nerven, ete. Beitrage zur path. Anat., X XVI, 1899.

* KOLLIKER, A. von.—Ueber die Entwickelung der Nervenfasern. Anat. Anz.
Bd. XXV, 1904.

LANTERMANN, A. J.—Ueber den feineren Bau der Markhaltigen Nervenfasern.
Arch. fiir Mikros. Anat., Bd. XIII, 1877.
Matt, F. P.—On the Development of the Connective Tissues from the Connec-
tive Tissue Syncytium. Am. Jour. Anat., Vol. I, No. 3, 1902.
MAUTHNER, L.—Beitrage zur nadheren Kenntniss der morphologischen Ele-
mente des Nervensystems. Kais.-Kénig]. Hof- und Staats-druckerei,
1862.

PertIK, O.—Untersuchungen iiber Nervenfasern. Arch. ftir Mikros. Anat.
Bd. XIX, 1881.

Ranvier, L.—Traité Technique D’Histologie, 1889, p. 814.

SCHIEFFERDECKER, P.—Beitraige zur Kenntniss der Nervenfasern. Arch. fur
Mikros. Anat., Bd. XXX, 1887.

WeEsTPHAL, A.—Ueber die Markscheidenbildung der Gehirnnerven des Men-
schen. Arch. fiir Psychiatrie., Bd. XXIX, 1898.

W.Lassak, R.—Die Herkunft des Myelins. Arch. fiir Entwickl.-mech., Bd. VI,
1898.

Wynn, W. H.—The Minute Structure of the Medullary Sheath of Nerve Fibers.
Jour. Anat. and Physiol., Vol. XIV (N.S.), 1900.

VIGNAL, W.—Mémoire sur le developpement des tubes nerveux chex les
embryons des Mammiféres. Archives de Physiologie, XV, 1883.

THE DEVELOPMENT OF THE LYMPHATIC NODES IN THE
PIG AND THEIR RELATION TO THE LYMPH HEARTS.

BY
FLORENCE R. SABIN.
From the Anatomical Laboratory, Johns Hopkins University.

WitH 17 TEXT FIGURES.

Notwithstanding the numerous investigations on the lymphatic nodes,
there are many points in regard to their structure and development in
which our knowledge is not yet clear." For example, such a fundamental
question as the relation of the nodes to the lymphatic system as a whole
and to the vascular system, in other words, the problem of general
morphology: or, more special questions in regard to the nodes them-
selves; primarily the existence of a structural unit; and secondarily the
relations of the endothelium of the channels to the reticulum within the
node, whether the channels are open or closed, and the origin of the
lymphocytes.

In regard to the problem of general morphology, we have previously
shown * that the lymphatics are modified veins. They develop as blood-
vessels do, by the budding of endothelial cells, and the direction of their
growth is determined by the arteries and veins. In the lymphatic sys-
tem there develop four lymph hearts which pulsate in the amphibia; but
in the mammalian forms, at least in pig and human embryos, have no
muscle in their walls. These lymph hearts drain the body, that is to
say all the lymph passes through them before entering the veins. It
will be shown in the present paper that the first lymph nodes in the body
develop from the lymph hearts. That is, the organ which was a pulsat-

1 During the past eleven years, there have been four extensive researches
on the development of lymph nodes by Gulland, Saxer, Retterer, and Kling.
Each of these authors has reviewed the literature. Gulland, Journal of
Pathology and Bacteriology, Vol. 2, 1894; Saxer, Anatomische Hefte, Bd. VI,
1896 Retterer, Journal de L’ Anatomie et de La Physiologie, 1901 Kling,
Upsala Lakareférenings Foérhandlinger, 1903, and Archiv f. mik. Anat. u.
Entwicklungsgeschichte, Bd. 63, 1904.

2Sabin: American Journal of Anatomy, Vol. I, 1902.

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

356 Development of the Lymphatic Nodes in the Pig

ing heart in the amphibia becomes transformed into a node in the higher
forms. The lymph heart is the point from which ducts radiate in de-
velopment to drain a large area, generally speaking a quarter of the
surface of the body, and here the primary lymph node develops. Subse-
quently there are secondary points, from which ducts growing from the
lymph heart radiate to drain lesser areas, and here other nodes are
formed. These are subcenters through which lymph passes before reach-
ing the primary node. ;

As to the nodes themselves, the study of their development brings out
that they are made of two fundamental structural elements. Tirst, a
lymphoid or adenoid tissue, consisting of lymphocytes in a reticulum
around the blood-vessels making the lymph cords and germ centers; and
secondly a lymphatic tissue, or sinus, made of large numbers of lymph
duets closely packed together. In brief, an ordinary lymph node is a
blood vascular organ, made in part of structures derived from the blood-
vessels and in part of structures derived from the lymphatics. The vas-
cular unit of the node is the terminal artery and its capillary plexus, the
artery being bordered by the cord and the capillary tuft surrounded by
the germ center. These two elements, the vascular and the lymphatic,
are found in varying proportions in the ordinary node. However, both
in the embryo and in the adult, either element may be found alone. In
the embryo and probably in the adult pig, there are small lvmph follicles
consisting of a tuft of blood capillaries surrounded by lymphocytes and
entirely without a sinus. In the hemolymph node and in the spleen the
same type of lymphoid tissue is found, but here the sinuses are made,
not of lymph ducts, but of veins. Thus in the blood vascular organs
the lymphoid element consisting of lymphocytes in the adventitia of an
artery is constant, while the sinus element varies, being absent, or made
of veins, or of the modified veins which are called lymphatics.

Throughout the paper certain terms have been adopted. Lymph node
is used as a general name to cover all lymphatic glands; the term follicle
is used to represent a simple node consisting of the structures that go
with a single artery. The simplest node consists of one follicle; other
nodes are groups of follicles. The follicle is the anatomical or structural,
unit; it is also the vascular unit. The follicle may be without a sinus,
or surrounded by a lymphatic sinus or surrounded by a venous sinus.

The term lymph heart has been retained notwithstanding that in the
pig there is no muscle in its wall at any stage. It is, however, a sac
from which all the ducts for the skin radiate and in that sense is
homologous with the lymph heart of the amphibia.

Florence R. Sabin Bor

Material and methods.—The material for the present study has been
embryo pigs of all stages. In studying the development of the nodes,
just as in studying the development of the ducts, it is essential to employ
injections, both lymphatic and arterial, and these injections have been
made in every stage. The lymphatics have been injected by means of a
hypodermic syringe, with either saturated aqueous Prussian blue or
with India ink. The material has been preserved for the most part by
the injection of a saturated solution of bichloride of mercury, either into
the aorta or into the umbilical vein. The blood-vessels are usually first
washed out with warm salt solution, then the bichloride introduced and
continued until the embryo is hard and white. The injection is made
slowly with a pressure of about 100 mm. of mereury and the bichioride
allowed to stay in the vessels from one-half to two hours. It is then
washed out thoroughly by injecting 70 per cent alcohol through the same
canula. The embryo is then placed in 80 per cent alcohol over night and
the next day transferred to 95 per cent alcohol. This method involves
the least possible shrinkage, indeed it may be made to produce a slight
distension of the tissues, which is an especial aid in studying lymphatic
nodes.”

In studying the developing nodes in fresh tissue, it is readily noticed
that they are sometimes found distended with fluid and sometimes col-
lapsed. It is just as easy to tell with the umaided eye when a node is
thus distended with lymph as to distinguish between the mesenteric
lymph nodes distended with chyle or collapsed and empty. This method
of injection produces the same distension of the spaces that occurs nor-
mally when the node is in active function; that is to say, it makes the
lymphatic ducts rounded rather than collapsed. This explains the espe-
cial value of the method as applied to lymphatic tissue.

A yaluable aid in the localization of the nodes, and especially in
studying the relations of the lymph hearts to the developing nodes, is
found by making injected embryos transparent.’ 'The lymphatics are
first injected with India ink and then the entire embryo is placed in
95 per cent alcohol. They are left in the alcohol until they are shrivelled.
This takes at least two weeks. The embryos are then cleared in a dilute
solution of potash from 1 to 2 per cent, taking from 1 to 4 hours. The
specimens are preserved in glycerine, at first 20 per cent and later in
pure glycerine.

’McFarland: Jour. of App. Microscopy, Vol. II, No. 10, and Myers, Ibid.,
Volaviee INO, 12, and! Je He Bullet Oo:
*Mall: American Journal of Anatomy, Vol. IV, 1905, p. 6.

358 Development of the Lymphatic Nodes in the Pig

The lymph heart—The present paper is a continuation of two papers
previously presented in this journal, the first in Vol. I, 1901, and the
second in Vol. III, 1904. It has been shown’® that the lymphatics bud
off from the veins at the root of the neck, grow along the internal jugular
vein on either side, and expand into a sac in the neck. This sac or lymph
heart is shown in Fig. 1 as it appears in the neck of a pig 2.7 cm. long.

Fic. 1. Embryo pig, 2.7 em. long, showing the anterior lymph heart in-
jected. x about 4%. #, extravasation at the point of injection; /d, lymph
ducts leading from the extravasation to the lymph heart.

The lymphatics were injected with India ink by a hypodermic needle
introduced into the ducts over the shoulder at the point marked by the
extravasation in the figure. The relation of this sac to the veins is
shown in the accompanying diagram, Fig. 2. One duct of the sac lies
along the course of the internal jugular vein. Just below the ear, under
cover of the sterno-cleido-mastoid muscle, this duct widens into a sac
which makes an arch in the neck. The sac curves outward and down-

*Ibid., Vol. I.

Florence R. Sabin 359

ward with the apex, marked a@ in the diagram, near the surface, adjacent
to a superficial vein éver the shoulder. This apex is to be found in the
triangle between the sterno-cleido-mastoid muscle and the trapezius.
From the apex of the sac, a duct follows along the vein of the shoulder
and empties into the duct along the internal jugular vein. This duct
becomes the main duct of the sac.

Figure 3 is a section of the base of the sac at about the level marked 6}
in Fig. 2. The section is cut transversely through the neck of a pig
2 cm. long and passes through the larynx. It shows the relation of the
lymph heart both to the internal jugular vein and to the sterno-cleido-
mastoid muscle. The lining of the sac consists of a single layer of
endothelium.

Fig. 1, which shows the lymph heart and
its ducts in a pig 2.7 em. long, is to be com-
pared with Figs. 1 and 2 in this journal,
Vol. III, 1904, pp. 184 and 185, which show
the superficial lymphatics in pigs 2.5 em.
and 3 cm. long.

From these three figures it can be seen
that the lymph ducts or capillaries grow

from the apex of the sac first to the skin paldtiocion aie (eens Wea Ge
of the shoulder and back of the head. Then heart'is in solid black. 4, apex
of the lymph heart; b, base of
ducts grow forward from the apex of the sac the lymph heart; e¢jv, external
over the surface of the sterno-cleido-mastoid ijv- internal jugular vein.
muscle, and form a long plexus around the
external jugular vein parallel to the anterior border of the muscle. This
plexus is shown in a little later stage in Fig. 6, and in section in Fig. 5.
From this long plexus ducts first grow to the face as is shown in Fig. 2,
Vol. III, p. 185. Later on it will be shown that at this stage, namely,
when the pig is 3 em. long, the apex of the sae begins to be transformed
into a lymph node.

The condition of the lymphatic system in the neck of a pig 3 em. long
is as follows: There is first the lymph heart with its efferent ducts con-
nected with the veins; then ducts have grown from the apex of the sac
first to the skin of the shoulder and back of the head, and secondly to the
face. The sac shows also the first rudiments of a lymph gland.

The maximum size of the lymph heart is attained when the pig is
3.6 cm. long, and in Fig. 4 is given a flat reconstruction of the sac at
this stage. It was made from a set of serial sections and gives the size

more accurately than the potash specimens. Fig. 5 is a section from the

360 Development of the Lymphatic Nodes in the Pig

same series taken at the level marked b on Fig. 4. It corresponds with
Fig. 3, and shows a similar relation to the internal jugular vein and the
sterno-cleido-mastoid muscle. It shows that the sae comes nearest the
surface between that muscle and the trapezius. It also shows the ex-
ternal jugular vein at the anterior border of the sterno-cleido-mastoid
muscle and the neighboring plexus of lympa ducts. A section about half
way between the letters a and b on Fig. 4 shows the heart near the surface
and the duct adjacent to the vein as two distinct cavities.

Fic. 3. Transverse section through the neck of an embryo pig, 2 cm. long,
Showing the anterior lymph hearts. x15. Jjv, internal jugular vein; lJ,
larynx; p. pharynx; scm, sterno-cleido-mastoid muscle; sn, sympathetic nerve;
un. vagus nerve.

Fig. 6 is another specimen made transparent in potash. It is from a
pig 6 em. long and is to be compared with Fig. 5, Vol. III, p. 188, which
is from a pig 5.5 em. long. The former shows the lymphatics in the
depth and the latter those of the surface at about the same stage. ‘The
injection for this specimen was made in two places: one just back of the
fore leg as marked by the extravasation; from this injection the ducts
over the shoulder, the lymph sac, and two large ducts running to the
long plexus were filled. The second injection was made between the eye

Florence R. Sabin 361

and ear, by which the capillaries over the face and the long plexus in the
neck were filled.

In Fig. 6 the lymph heart is still plain, showing as a triangle in the
depth. It has modified somewhat in shape, inasmuch as lymph nodes
are forming at the apex and base, making the angles of the triangle
appear as knobs. The apex of the sac is now labeled primary lymph
node, and the base is lettered 6. The actual size of the triangle as a
whole is not greater than when the embryo was 3.6 cm. long. That is to
say, the distance from the apex to the base is about 4 mm. in either case.

The ducts over the shoulder from the apex of the sac are well injected,
these being the first set to develop. The set of ducts which grows for-
ward from the apex of the sac over the surface of the sterno-cleido-
mastoid muscle to make the long plexus in the neck shows somewhat, but
is not as well injected in this specimen as
in Hig. 7. They are present at this
stage but the injection from the region
near the eye was not pushed quite far
enough to bring them out well. This set
of ducts develops into the long and abund-
ant plexus which follows the course of the
external jugular vein as it lies parallel to
the sterno-cleido-mastoid muscle. From
this long plexus, the entire face, front of
the neck, fore leg, and thorax are supplied
with lymphatics, and these: different sets 1 wo4 Diagram of the anterior
can be seen in Fig. 6. All of these sets CEOS ee ee Dien ee
of ducts anastomose in the skin, as can be of Fig: B Te tomatoe jevel
seen in Fig. 5, Vol. III, p. 188.

In brief, the ducts for the shoulder and back of the head grow directly
from the lymph heart; those for the face, neck, and fore leg grow from
the lymph sac, but form a long plexus along the external jugular vein
before reaching the skin. As has just been said, both sets of ducts, dis-
tinct in the depth, anastomose in the surface. The line of growth of
the lymphatics has been tested by a large number of injections in every
stage from the time the lymphatics first appear up to the time of birth.
Every injection made into the ducts of the skin of the anterior part of
the body will run to the lymph sac or the gland derived from it, if pushed
far enough. The different systems of ducts of the neck can be brought
out by injecting in four different places. When the needle is entered
over the shoulder the injection mass invariably runs to the apex of the

28 ;

Duct

362 Development of the Lymphatic Nodes in the Pig

lymph sac; occasionally it enters the surface ducts that anastomose with
the long plexus. When the needle is introduced into the layer of the
lymphatics between the eye and ear, or over the lower jaw and front of
the neck, or into the pads of the fore feet, the injection mass runs into

Fic. 5. Transverse section through the neck of an embryo pig, 3.6 cm. long.
< about 11. The shape of the entire heart of which this figure shows a
section is given in Fig. 4, in which the line b is the level of Fig. 5. Ca,
carotid artery; ejv, external jugular vein; ijv, internal jugular vein; 7, larynx;
lh, lymph heart; Jp, lymph plexus along the external jugular vein; nv, vagus
nerve; oe, wsophagus; scm, sterno-cleido-mastoid muscle; sn, sympathetic
nerve.

the long plexus and across the sterno-cleido-mastoid muscle to the apex
of the lymph sac. This general relation is not only true in the stages
already pictured, but in the later stages when the apex of the lymph
heart is a lymph node and the long plexus has been replaced by a chain
of lymph nodes.

Florence R. Sabin 363

As is seen in Fig. 6, the ducts which connect the lymph sac and the
long plexus, join the plexus half way between the ear and the fore leg.
Once or twice, out of many injections in which the needle was introduced
between the eye and ear, the injection mass has reached the veins in
two ways: one the usual course through the lymph heart, and secondly,
through ducts that follow the course of the external jugular vein to its
junction with the internal jugular, showing that the ducts along the
two veins anastomose.

Fic. 6. Lymphatics in the neck of an embryo pig, 6 cm. long, showing the
modified lymph heart in the depth and the plexus of lymphatics along the
external jugular vein. X about 3. B, lymph node developing in the base of
the lymph heart; e, extravasation at the point of injection; lp, long plexus of
lymphatics along the course of the external jugular vein; pln, primary lymph
node developing in the apex of the lymph heart.

Fig. 7 is from a pig 11 cm. long and shows an injection of the
lymphatics made from two points: one between the eye and the ear, and
the other into the foot pad. The injection mass, both from the ducts of
the face and from the fore leg, has entered the long plexus and then

364 Development of the Lymphatic Nodes in the Pig

passed through ducts that lie over the sterno-chleido-mastoid muscle into
the node representing the lymph heart (pln). At this stage there are
two lymph nodes at the angle of the jaw, nf, one deeper, receiving the
ducts around the eye and cheek, the other more superficial, receiving
the ducts just in front of the ear. The rest of the long plexus is also
being modified into lymph nodes, one of which is in the middle of the
plexus where the ducts join with the lymph sac, the other is at the
posterior end of the plexus and drains the fore leg (nfl).

Fic. 7. Lymphatics in the neck of an embryo pig, 11 cm. long. Xx 1%.
Lp, lymph plexus which lies parallel to the external jugular vein; nf, nodes
developing in the long plexus draining the face; nfl, node developing in the
long plexus and draining the fore leg; pln, primary lympa node between the
trapezius and sterno-cleido-mastoid muscles.

Since the spread of the superficial lymphatic capillaries in the skin
of the pig is practically complete when the embryo is 6.5 em. long, it
may be well to sum up the superficial lymphatic system at that stage.
In the neck there is, in the depth, the lymph heart now considerably
modified by the formation of lymph nodes. It has one large efferent
duct along the internal jugular vein and one along the main superficial
vein of the shoulder. Secondly, there is a plexus of ducts along the
external jugular vein; this plexus connects freely with the lymph heart.
Lymphatic nodes in the neck are to be found developing first from the
lymph sae, secondly in the long plexus on the course of the external jug-

Florence R. Sabin 365

ular vein, and thirdly in the depth along the internal jugular vein.
In the surface the capillaries have grown from the apex of the lymph
heart to the shoulder and back of the head and from the long plexus to
the face, neck, thorax, and fore leg. The capillaries of all these sets of
ducts anastomose freely in the skin and there are no valves to check the
spreading of an injection mass. The lymphatics of the axilla belong to
the deep set which grow along the arteries rather than the veins.

The spreading of the lymphatics for the lower part of the body can
be constructed from Fig. 5, Vol. III, p. 188. The position of the pos-
terior lymph heart is just caudal to the kidney, and at this point a
lymph node develops. The superficial lymphatics for the lower part
of the body grow in two directions, one set following the vein to a point
over the crest of the ileum, where a node is formed which drains the
skin of the back and hip; the other set coming to the surface in the
inguinal region where a long node is formed which drains the abdominal
wall and the hind legs. These three nodes with the abundant chain of
nodes along the aorta represent the distribution of the glands which
drain the skin of the lower part of the body. In following the histo-
genesis of the lymph nodes all of these different nodes have been studied,
but most of the figures presented are from different stages of one node,
namely, the first one to develop in the body.

Histogenesis of the primary lymph node-—We turn now to ie histo-
genesis of the lymph node, which will involve determining the structural
unit, and tracing the two elements, the vascular element with the adenoid
tissue, and the lymphatic element or sinus.

The first lymph node in the body develops from the apex of the lymph
heart and will be referred to as the primary lymph node. This node
will be traced in its development until its condition is practically adult.

The first evidence of the formation of lymphatic nodes occurs when
the embryo is 3 cm. long. At this stage the lymph heart, which has
been a smooth walled sac, as shown in Figs. 3 and 5, lined with a single
layer of endothelial cells, begins to show a slight modification at the apex
in that bands of connective tissue begin to push into the lumen without
destroying the lining. The apex of the lymph sac is pictured in Fig. 8
from a pig 3.6 cm. long. The section is taken from the same series as
Fig. 5, by which the transverse plane of the section can be noted. The
figure shows the character of the surrounding tissue consisting of a syn-
eytium of protoplasm with nuclei in the nodes. The wall of the sac
consists of a single layer of endothelial cells, and in the left hand side
there is no perceptible modification of the connective tissue. On the right

366 Development of the Lymphatic Nodes in the Pig

side, however, bands of connective tissue project into the sac without
destroying its endothelial lining at any point. The connective tissue in
these bands and on the right border of the sac appears different from
the surrounding tissue.

Studied with the oil immersion lens, the surrounding connective tissue
appears as deseribed by Mall” to be a network of granular protoplasm in

NS Ua M
$3

Avia &
at Soh
oe
ONE

MANNY

ies x ei ‘
D4 NP ANG ‘ hay
ig, 2 iu :
a ’ Sa my
SAAN RD :
LY NY
iy

A | . a | f i
fea’ Oy NY td ‘
3 Ny Udi
Pes i

pe
Nai
ai
A RANG Cd CBER H
SM AN ‘ Be Gite Ae
Fic. 8. Primary lymph node from a transverse section of the neck of an
embryo pig, 3.6cem. long. x about 60. The left side of the figure is the mesial
side of the heart, the right side is toward the skin and shows the afferent
ducts. The top of the figure is the position of the hilum of the node. Ld,

lymph duct (afferent); 7h, lymph heart; v, vein.

which are distinct anastomosing fibrils. The nuclei he in the nodes of
the network and each one has around it a drop of clearer protoplasm
which he calls endoplasm, distinct from the rest or exoplasm. Near the
lymph sac, on the right hand side in the figure, are numerous blood capil-
laries and around each one are clumps of from 8 to 20 nuclei. These

*Mall: American Journal of Anatomy, Vol. I, 1902.

Florence R. Sabin 367

nuclei lie definitely within the syncytium and belong to the connective
tissue. They are only to be distinguished by the fact that they are in
clumps and that some show karyokinetic figures while others are smaller
and take the deep stain of a newly divided nucleus. In short, cell pro-
liferation takes place around the capillaries.

Passing now to the bands or bridges of connective tissue, the first point
to be noted is that there are numerous blood capillaries filled with red
blood cells, many of them nucleated. The same sort of protoplasmic
network is present as in the surrounding tissue, but the network is denser
and the meshes finer. The increase seems to be in the granular proto-
plasm rather than in the fibrils. In this dense network of protoplasm
are crowded many connective tissue nuclei; the mature ones are oval in
shape and take the stain faintly. Many of the nuclei are dividing, and
there are numerous small, round, deeply staining, young nuclei. These
round nuclei belong, however, to the connective tissue and there are no
true wandering cells outside of the blood-vessels. Thus the modifica-
tion of the tissue around the sac consists merely of an increase in the
blood capillaries, and in the connective tissue protoplasm and nuclei.
The cell increase does not take place independently of the blood capil-
laries. There is no muscle in the wall of the sac at any time. By the
time the embryo is 3.6 cm. long a second node is just beginning at the
other end of the sac. This second node from the lymph sac develops in
the same manner as the first, but slightly later.

We pass now to the primary lymph node when the embryo is 4.9 em.
long, as shown in Fig. 9. The section is taken in the same plane as Fig. 8,
that is, it is from a set of transverse sections. The efferent ducts are
on the right and the hilum at the top of the section. There are no strik-
ing changes between this and the preceding stage. The node as a whole
has increased considerably in size. The lymph heart is about the same
actual size as in Fig. 8, but the lymphatic plexus is greater. From Fig. 1
it can be seen that when the ducts first start out from the sac they grow
directly to the skin, but in Fig. 9 there has been an anastomosis or plexus
formation of the ducts on the border of the sac. This greatly extends
the area of the node. On the left side of the sac there are a few blood
capillaries with clumps of dividing nuclei around them. The bands of
connective tissues show the same abundance of blood capillaries and in-
crease in the protoplasm and nuclei. Young and dividing nuclei are
abundant, but no true wandering cells are present.

A more important stage is met with when the embryo is 7 cm. long.
From this stage on, the development of the primary lymph node is shown

368 Development of the Lymphatic Nodes in the Pig

in a series of five diagrams. Each diagram is made from a single section
traced with the aid of the camera lucida. The blood-vessels are put in
freehand from the study of the complete set of serial sections from which
each diagram was made. All the figures are of the same magnification,

tn / f
‘e
vale
i)
ait

Fic. 9. Primary lymph node from a transverse section of the neck of an
,embryo pig 4.9 cm. long. xX about 44. The section is placed similar to Fig. 8.
Ld, afferent lymph duct; 7h, lymph heart; v, vein.

about 33 diameters. The lymphatic vessels are shown in solid black as
if injected, while the connective tissue of the lymph cords and follicles is
dotted. In the later stages the increased number of the dots represents
the increase in lymphocytes and the lines show the beginning capsule
and trabecule.

Florence R. Sabin 369

In the first diagram, Fig. 10, the step in advance beyond the stage of
Fig. 9 is in the proliferation of the lymphatic capillaries. The sac has
been completely cut up into ducts. The entire node consists of a plexus
of lymphatics which differs in no way from the plexus in the skin
pictured in my first paper. There are the same swollen bulbs, the same
blind sprouts and slender channels. The connective tissue bridges are
similar to those of the preceding stages. They contain many dividing
cells but no true wandering cells. In the bridges is an abundant plexus
of blood capillaries which are not shown in the
diagram. This diagram might also represent any
lymph node which develops in a plexus.

To sum up, the figure marks the culmination of
the first stage of the development of the lymphatic
nodes in early embryos, namely, the stage in which
the node consists of a plexus of lymphatic capil-
laries separated by bands of connective tissue
which is denser than the surrounding tissue. ‘This
stage is shown in Kling’s‘ models, Fig. 1. The
connective tissue is embryonic in type, consisting
of a net work of granular protoplasm with a few
fibrils and with many nuclei. The bands or
bridges have blood capillaries and the increase in
connective tissue does not take place independently
of them. There are no true wandering cells out-
side of the blood capillaries. It is the stage of
lymphatic ducts and pure connective tissue bridges.
All of the nodes of the early embryos, the primary
nodes in the sense of Gulland pass through this 4, 49, Dinas tee
stage. That is to say, the nodes which develop in’ the Primary lymph node

in an embryo pig 7 cm.
the long plexus in the neck from which ducts one, abouts. Tne
radiate to the face, neck, fore legs and thorax Piactand the connecte
(Fig. 6); or the node which comes in the in-
guinal region at the point where the ducts radiate over the abdominal
wall and hind legs; or in the node over the crest of the ileum where they
radiate over the back (see Fig. 5, Vol. III, p. 188). All of these nodes
come in places, where plexuses are formed because ducts radiate over a
wide area, which is shown well in the figure just quoted. They are
primary nodes in the sense of Gulland because they develop early and
drain large capillary areas. It will be shown subsequently that lymphatic

*Tbid

370 Development of the Lymphatic Nodes in the Pig

nodes which develop later in the hfe of the embryo, after lymphocytes
occur, hurry through the primary process, and show a considerable modi-
fication of it.

Up to this time the node has had none of the structures characteristic
of the adult node; there are no lymph cords, nor germ centers, no
lymphoeytes, and no sinuses.

The next stage, pictured in Fig. 11, shows the beginning of some of
these structures. The diagram is made from a section of the primary
lymph node in a pig 8 em. long. In the center of the node the blood
capillaries have proliferated, giving a tuft of capillaries surrounded by

Fic. 11. Diagram of the primary lymph node in an embryo pig, 8 cm. long.
x about 33. This represents the primordial follicle. The hilum is marked
by the artery. A, artery; ald, afferent lymph duct; eld, efferent lymph ducts;
f. follicle.

connective tissue. The artery is shown leading up to the node but
reduced to capillaries on entering it. The vein is not shown in the
diagram, but the artery and vein lie parallel, up to the point where the
node or follicle is entered, where they separate. This is an important
and characteristic point in the relation of the blood-vessels. At this
stage there are only capillaries within the node.

Here for the first time we can speak of the lymph follicle, which is the
vascular unit and consists of the structures that go with a single artery.
At this stage the entire node is one follicle. Here also for the first time
two elements are differentiated, a lymphoid element connected with the
artery and a lymphatic element made of lymph ducts.

Florence R. Sabin oie

The point of entry of the artery determines the hilum of the node. The
position of the hilum is determined from the beginning of the formation
of the node by the lines of growth. Blood-vessels and lymphatics grow
from the center of the body to the periphery, so that the proximal surface
of the gland has from the start the entering blood-vessels and the efferent
lymphatic ducts, while the peripheral surface of the node is the place
from which the efferent lymphatic ducts radiate to the area they are to
drain. In the central core of connective tissue the lymphatic capillaries
are reduced in number and size; they are never quite absent but do not
appear except in well-injected specimens. The presence of these ducts
within the connective tissue core may have some bearing on the pathology
of lymph nodes. The disappearance of the lymph capillaries in the
center of the node involves the retrogression and absorption which is
characteristic of developing tissues. Throughout the evolution of the
lymph node there is continual building up and tearing down. This will
be evident in later stages where there is a continual change in the pro-
portion of the lymphoid structures or cords and the lymphatic struc-
tures or sinuses.

Beside being the stage which marks the beginning of the adult struc-
tures of the node, that is to say, of the follicle, this stage also shows
fundamental changes in cell differentiation. It marks the beginning of
the wandering cell in lymph nodes. Up to this time the connective
tissue part of the node has consisted of a network of granular proto-
plasm with many nuclei, young, dividing, and old. At this time three
types of wandering cells appear, the lymphocyte, the polymorphonuclear
form, and the eosinophile.

Lymphocytes are present in the thymus at a much earlier stage, they
are abundant there in the sections from the embryo 3.6 cm. long. In
the sections of the lymph node at 8 em., there are a few lymphocytes in
the connective tissue core of the node, and in little clumps in the con-
nective tissue just without the node. These little clumps of cells are
found near the capillaries. The differences between the connective tissue
cell and the lymphocyte are as follows: The nucleus of the former is
large, faintly staining, and oval in shape, and the protoplasm belongs
definitely to the network, while the latter has a small, round, deeply
staining nucleus, with a more distinct nuclear membrane. The nuclear
network and the chromatin granules are coarser, and there are one or
more nucleoli. Moreover, the protoplasm makes a narrow but definite rim
around the nucleus. Between the connective tissue cell, especially the
young forms, and the lymphocyte one can see every possible transition.

372 Development of the Lymphatic Nodes in the Pig

Often the connective tissue nuclei appear as if being extended from the
protoplasmic network of exoplasm, the irregular endoplasm still clinging
to the nucleus.

This form of observation cannot be considered as proof of the origin
of the lymphocyte from connective tissue—it is obvious that the position
of a wandering cell cannot give evidence of its origin. With the same
type of tissue to examine, Gulland, noting the occurrence of the lympho-
cytes in clumps around the capillaries, concluded that they were filtered
from the blood stream. The evidence does not suffice to prove either
that the lymphocyte develops from the connective tissue in the lymph
node, nor that it reaches the node through the blood stream. We must
await some new method of attacking this problem. One point is, how-
ever, definite in my specimens—that cell division in the connective tissue
takes place in little clumps around the blood capillaries, and, as will be
shown later, the division of the lymphocytes takes place also around tufts
of capillaries.

Besides the lymphocytes there are a few polymorphonuclear cells at
this stage, perhaps not more than twenty or thirty in each section. They
are quite typical, having irregular nuclei and finely granular protoplasin.
They oceur within the follicle. Eosinophiles appear also for the first
time. Within the follicle there are numerous red blood cells outside of
the capillaries, showing signs of degeneration, that is, a vacuolization
and a breaking up of the protoplasm into granules. These granules are
all of the same size and cannot be distinguished from the granules of the
eosinophilic cell. This is the same evidence that has led Weidenreich to
the conclusion that the eosinophilic granuie comes from the red blood
cell. It is suggestive, but not conclusive.

To sum up the stage represented by Fig. 11, it marks the beginning of
the differentiation of the node into its two elements, lymphoid and
lymphatic. It shows the beginning of the follicle and of the wandering
cell. There is a marked proliferation of the blood capillaries and a con-
sequent increase in the connective tissue in the center of the node. This
involves a retrogression or destruction of some of the lymph ducts. At
the same time wandering cells appear, lymphocytes in greatest numbers
and also polymorphonuclear leucocytes and eosinophiles. There is also
evidence of degeneration of the red blood cells.

The next stage is shown in Fig. 12. It was made from the primary
lymph node of a pig 13 em. long. The first point to be noted is the de-
velopment of the artery. Without the limits of the node, the artery has
divided into two branches. These two branches enter the node and

Florence R. Sabin aie

divide into five main branches and two much smaller ones. Consequently
there are five definite primordial follicles, and two small ones. Both of
the small ones and two of the large ones show in the section. At this
stage there is no definite capsule, the limits of the node being determined
by the lymphatic vessels. The nodes increase in size by invading the sur-
rounding tissue, for example, the artery which here branches without the
node is subsequently included in the gland. This stage marks several
important changes. The first has already been noted as being the
division of the artery and the corresponding multiplication of the fol-

Fic. 12. Diagram of the primary lymph node in an embryo pig, 13 cm. long.
<x about 33. The section shows two large follicles and two small ones. A,
artery; ald, afferent lymph ducts; bf, beginning follicle; eld, efferent lymph
duct; f, follicle; gc, germ center.

licles. The vein, which is not shown in the diagram, runs parallel to
each branch of the artery up to the point where the artery enters the
follicle. On entering the follicle artery and vein separate and both break
up into plexuses. At this stage the artery without the follicle can be
distinguished by a thickening of the connective tissue around, there being
no media, and by its smaller caliber. The vein has only a lining of
endothelium and there is as yet no thickening of the adventitia. Within
the follicle, the vessels are all capillaries, but the plexus which con-
nects directly with the artery is made of smaller vessels than the plexus
which connects with the vein.

374 Development of the Lymphatic Nodes in the Pig

The second point in advance is the formation of the germinal center.
Within the follicle, as will be seen in the diagram, there are small clumps
of cells, definitely lymphocytes, heaped around a capillary tuft. In the
entire node at this stage there are eight of these germinal centers. The
lymphoeytes are closely packed in them, and there are more lymphocytes
near these centers than elsewhere in the node. In regard to the wan-
dering cells, the follicles contain in general four types: First, the lympho-
eytes which are found in the germinal centers almost to the exclusion of
any other free cell. A few of them are to be seen near the germ centers.
Second, polymorphonuclear forms, which are scattered throughout the
follicle except in the germ centers. Third, eosinophilic cells. Fourth,
mononuclear forms which have larger nuclei and more protoplasm than
the lymphocytes. All of these cells are found in the follicle. Red blood
cells are also present, many of them being free in the connective tissue
meshes and showing a breaking up of the protoplasm into granules. This
appearance of the red blood cell is seldom seen in the corpuscles within
the capillaries. The bridges of connective tissue between the lymph
ducts have fewer wandering cells than the follicle.

Beside the two changes already noted, namely, the division of the
artery and consequent multiplication of follicles, and secondly, the differ-
entiation of the follicle into germ center and lymph cord, there is a third
important change, namely, the beginning of the formation of lymphat-
ic sinuses out of the lymphatic ducts. It will be noted in the dia-
gram that around the border of the follicle the lymph ducts are arranged
in rows closely packed together and that the connective tissue bridges
between them are slender. This is still more definite in the next diagrams
given in Figs. 13 and 14. In section, this point is to be seen in Fig. 15,
where the surface of the node adjacent to the capsule is made of a plexus
of lymph ducts, while in the depth of the node the ducts are closely packed
together in certain areas making sinuses. The section shows all grada-
tions in the width of the connective tissue bridges, some being just wide
enough to contain a single nucleus while others are broad bands.

Kling was, I think, the first to note this method of the formation of
the sinus, though he clouded the clearness of his picture by saying that
subsequently the bridging of the sinus is made by the stretching of endo-
thelial cells across the lumen of the ducts, so that there are bridges made
of endothelium alone. This appearance is undoubtedly found in sections
just as in sections of the lung the epithelial lining of an air sac sometimes
shows as a membrane.

By following the evolution of the sinus it is possible to understand
clearly the relation of the reticulum of the adult sinus to the endothelial

Florence R. Sabin 375

cells. The reticulum fibers develop subsequently in the connective tissue
bridges (not, as Kling says, from the endothelium). The lymph ducts
from which the sinus is made have a complete endothelial lining. More-
over, the increase in the ducts which form the sinus takes place by the
same process of the budding of endothelial cells which characterizes the
development of lymphatics or blood-vessels elsewhere. Thus the spaces
of the sinus are lined throughout by endothelium. The sinus can be pic-
tured in three dimensions by imagining the follicle surrounded by a plexus
of ducts so dense that the bridges between them are reduced to the thick-
ness of a single network of fibers. Such a structure in cross section would
give the appearance of fibers with endothelial cells around them cutting
the lumen of the sinus. As a matter of fact the fibers are between the
endothelial cells and without the lymph channels. They connect with
the rest of the connective tissue framework of the node as is readily seen
in Fig. 12.

The next stage is from a section of the primary lymph node of an
embryo 15 cm. long (Fig. 13). There is now a great development of
the artery. The wall of the artery has developed considerably and shows
one row of smooth muscle cells beside the adventitia. The vein which lies
beside it has only an endothelium and a thickened connective tissue sheath.
The artery divides into three main branches on the edge of the node, and
within the node each branch subdivides several times. As in the early
stage, the artery and vein run parallel until the follicle is entered and
there they separate. The amount of lymphoid tissue has increased par-
allel with the development of the artery. The sinuses are growing down
into the node between the arteries, thereby surrounding and limiting the
lymphoid masses around the blood-vessels. By this process the lymph cord
along the blood-vessel becomes defined, as will be clear in the next dia-
gram.

The especial advance in this stage lies in the connective tissue, for here
the reticulum fibers within the node first begin. Up to this stage the
connective tissue of the node has been a protoplasmic network with delicate
anastomosing fibrils which, however, do not stain sharply by Mallory’s
method. Now for the first time there are a few fibrils which stand out
clearly in the Mallory stain. They occur in the germ centers where they
are laid down in concentric circles. With the oil immersion lens it can
be seen that the fibers of the germ centers make a definite mosaic of poly-
gonal spaces in concentric rows. All of these polygonal spaces thus out-
lined are filled with cells. This appearance of the mosaic can be seen
in thin sections of the adult node and can be brought out by silver nitrate.

376 Development of the Lymphatic Nodes in the Pig

The diagram shows the first beginning of the trabecule in the con-
nective tissue that pushes down between the peripheral sinuses as these
surround the follicles (see dt on the figure). Neither the capsule nor
the trabecule have fibers different from the surrounding connective tissue
at this stage.

There are certain interesting points in regard to the cells of the node
at this stage. In the germ centers there is a marked division of the

Fic. 13. Diagram ot the primary lymph node in an embryo pig, 15 cm.
long. X about 33. The node shows several follicles. A, artery; ald, afferent
lymph ducts; dt, developing trabecula; gc, germ center; ps, peripheral
sinus.

lymphocytes. In any one section each center contains from two ‘to
fifteen or more dividing lymphocytes. They are easily distinguished
from the dividing connective tissue nuclei, which are always larger and
have much more protoplasm. There are many eosinophiles in the
cords, but at this stage none are to be found in the germ centers. There
are numerous degenerating red blood cells.

The next diagram (Fig. 14) is from the primary lymph node in a
pig 23 em. long. Only a portion of the section was drawn, in order to

nw
4d

Os

Florence R. Sabin

Keep the diagram at the same magnification as the others. An outline
of the entire section is shown in the margin. It gives the artery and a
few efferent ducts at the hilum and also a large trabecula (#), carrying
afferent ducts. This trabecula connects with the cortex in another
section of the series. The left hand part of the section shows a consid-

we

= lc.cs.

Fic. 14. Diagram of the primary lymph node in an embryo pig, 23 cm.
long. X about 33. The diagram was made from a segment of a section, the
outline of the entire section being given in the margin. A, artery; c, capsule;
cs, central sinus; dts, developing trabecule of the sinuses; ed, efferent duct:
ge, germ center; Jc, lymph cord; lp, lymph plexus; ps, peripheral sinus; f¢,
trabecula with afferent ducts.

erable advance over the preceding stage; it represents practically adult
conditions where the entire node has been transformed into lymph cords
and sinuses. This change has come about in the following manner:
The lymph duct plexus or sinus, which was only on the edge of the
preceding section, making the peripheral sinus, has now grown into the

29

,

i]

378 Development of the Lymphatic Nodes in the Pig

depth between the arteries, thereby extending the sinuses and definitely
limiting the lymph cords. Indeed, the true lymph cord of the adult
condition now appears for the first time. The diagram shows well the
nature of the lymph sinuses consisting of rows of closely packed lymph
ducts. The bridges between them are still protoplasmic and _ slightly
wider than they appear in the adult nodes.

The diagram shows the reiation of the trabecule to the capsule and
the sinuses. Here for the first time there is a definite capsule and it is
interesting to note that it is not complete. It extends along the margin
of the peripheral sinus but ends abruptly on the right hand side of the
section, where the node consists of a plexus of lymph ducts. It will be
noticed that there is a definite indentation in the margin of the node
where the capsule ceases. At this place the node can increase in size by
invading the surrounding tissue.

The structure of the capsule itself is best studied in good Mallory
specimens with the oil immersion lens. For this study it is necessary
to note the condition of the surrounding connective tissue. This has
been described by Professor Mall... There is in the first place in the
surrounding tissue a delicate network of fine but definite anastomosing
fibrils. The nuclei of the network are in the nodes, and most of them
form part of the characteristic spindle cells. This is the prefibrous
tissue of Mall; the fine fiber network is still slightly granular, repre-
senting the protoplasmic syncytium of the earlier stages. Most of the
definite protoplasm is around the nuclei. Beside this fine fiber network
there are large bundles of definite fibers. The fibers of these bundles
are several times the width of the fibers of the network, and the bundles
themselves are often eight or ten times the diameter of a red blood cell
in width. Many of these bundles have several spindle cells clinging to
them. The large bundles are found near the border of some gland or
muscle, while in the less differentiated areas every transition between
the fine network and the coarse bundles can be made out. As Mall has
shown, these are the white fiber bundles developing from the prefibrous
tissue. The fibers of the bundles are close together and are straight.
The capsule of the lvmph node is different both from the fine network
and from the white fibrous bundles. It consists of a dense network of
anastomosing fibers. These fibers are larger and sharper than the fine
fiber network, but they are much finer than the fiber bundles. They
are wavy and closely packed. In other words, they differ from the fine
fiber network by being larger, sharper, and more closely packed, and

* Ibid.

Florence R. Sabin 379

from the fiber bundles by running as separate, wavy fibers, having in
general the same direction but forming abundant anastomoses. These
are obviously the beginning reticulum fibers.

There is one small, fat organ within the capsule. The trabecul are
formed, as can be seen in the figure, by the folding in of the capsule,
thereby bringing the trabecule in the center of the sinuses (dts). This
shows especially well in the border of the large central trabecula (t¢) of
the figure and explains why the trabecule of the adult node are bordered
by sinuses. The trabecule often carry blood-vessels from the capsule.
The tissue around the blood-vessel at the hilum thickens to form trabe-
cule; these trabecule follow the veins farther than the arteries for the
arteries soon enter the lymph cords. The trabecule around the blood-
vessels are less developed at this stage than the capsule.

The connective tissue framework within the node itself is a definite
protoplasmic network with a few fine fibrils which do not stain sharply
in the Mallory stain except possibly in the germ centers. The bridges
in the sinuses are all protoplasmic at this stage. In other words, the
connective tissue framework of the node is less advanced than the sur-
rounding tissue.and the reticulum fiber is a later development.

The diagram shows the relation of the lymph cords and the germ
centers. As has been said, the cords are first definitely outlined and
restricted to the borders of the artery by the development of the lymph
sinuses. Some of the cords are narrow and have the single central
artery, but the majority have an abundant plexus of blood-vessels. The
artery and vein never run side by side in the lymph cord, and the veins
leave the cords to enter the trabecule that grow in from the hilus. At
this stage of development only the large veins are in the trabecule. The
germ centers are around the capillary tufts where the lymphocytes
actively divide. The ]ymphocytes then wander down the borders of the
artery, filling up the lymph cords. In the adult node, as is well known,
the germ centers may present two different appearances. In the one
case, the center is uniformly filled with lymphocytes which are actively
dividing; in the other case the lymphocytes are crowded to the edge of
the germ center, giving the appearance of a dark rim in sections stained
in hematoxylin. In the embryo where cell division is active, the germ
centers are always uniformly crowded with lymphocytes.

The right-hand part of the section is far less developed. It consists
of a plexus of lymph ducts with connective tissue bridges which contain
a network of blood capillaries; it is essentially in the stage of Fig. 7,
though the lymphatie plexus is more abundant. Most of the nodes in

30

380 Development of the Lymphatic Nodes in the Pig

the pig at this stage have the center made of definite cords and sinuses
and the entire periphery made of a plexus of lymph ducts; and this is
the condition of the primary lymph node in a pig 2 weeks old. That is
to say, there are three kinds of tissue in the nodes: first, lymph cords
and germ centers; second, lymph sinuses; and thirdly, plexuses of lymph
ducts not yet transformed into sinuses. It is clear that the lymph
plexus is a stage in the development of the lymph sinus, and hence is a
less highly developed structure.

These points are clearly shown in Fig. 15, which is taken from the
inguinal node in a pig 24.5 cm. long. A portion of the capsule is shown
which is not as yet a limiting membrane. The outer part of the node
consists of a plexus of lymph ducts with connective tissue bridges. The
nuclei in these bridges are large and oval. The inner portion of the
node, away from the capsule, consists of lymph cords and sinuses.
Within the cords the predominating cell is the lymphocyte, with which
the germ centers are closely packed. The sinuses are groups of lymph
ducts and transitions between the lymph plexus of the edge of the node
and of the developed sinus are to be seen.

In the pig the lymph plexus persists up to adult life. In specimens
where the lymph ducts-are all collapsed, these areas look like masses of
connective tissue, hence Delamere® pictures them and calls them homo-
geneous areas. Ranvier” found the same tissue in the mesenteric nodes
of the pig, but since his specimens had the lymph ducts distended he
called the areas cavernous or erectile tissue. As has been said, in study-
ing large numbers of lymph nodes in fresh specimens one often finds the
lymph ducts or sinuses so distended with fluid that all the spaces are
rounded. The presence of this lymph plexus not developed into a sinus
in the nodes in the pig is, I believe, an important point not only in the
study of development but in the understanding of the structure of
allied organs. In all of the hemolymph nodes I have seen, there have
been three types of tissue, the lymphoid areas, true sinuses, and zones
filled with blood but not definitely sinuses. These zones appear like
veins not crowded enough to be sinuses.

We have thus followed the development of the primary acl node
which comes from the lymph sac. The stages are, in brief: first, a
preliminary stage in which the entire node consists of a plexus of
lymphatic capillaries with connective tissue bridges containing blood

°Poirer, Cunéo, and Delamere: The Lymphatics, p. 100, translated from
Poirier’s Anatomie.
Ranvier: C. R. Acad. Sciences, 1895, p. 800, or C. R. Soc. Biol., 1895, p. 774.

Florence R. Sabin 381

20n6. red
Spon oO ra dOas
Se ons Se.

20

a)

ade ls
ro)

Fig. 15. Portion of the inguinal lymph node in an embryo pig, 24.5 cm.
long. about 200. Drawn with a camera lucida. The drawing shows the
transition between the lymphatic plexus near the capsule and the sinuses

farther within the node.

382 Development of the Lymphatic Nodes in the Pig

capillaries; then a stage in which the development of the blood-vessels
determines the primordial follicle. Following the evolution of the artery
the adenoid tissue is determined while the lymphatic part of the node,
or the sinuses, develops from the proliferation of the lymph ducts. In
the pig the adult node still contains some of the lymph plexus which in
higher animals is completely transformed into the sinuses.

Development of nodes in the primary plexuses——The other early
lymph nodes which drain the skin develop in certain definite areas. If
we limit the group to those which receive ducts directly from the skin
without passing through other nodes, these glands develop in the long
plexus around the external jugular vein in the neck and in the plexus
of the inguinal region, and over the crest of the ileum. The develop-
ment of the nodes in all of these areas has been studied. These nodes
all begin in a plexus of lymph ducts rather than in a lymph sac; the
connective tissue bridges at first show no thickening but soon the nuclei
become more crowded and the nodes pass through the various stages
shown by the primary lymph node. The node over the crest of the
ileum is the simplest, for like the primary lymph node it develops as a
single node. The long plexus has the most comphcated development
and it drains a varied area. In the series at 7 cm. long the entire plexus
shows a thickening of the connective tissue bridges. Subsequently, as
shown in Fig. 7, the plexus represents a chain of nodes. In this figure
four macroscopic nodes are shown, two at the angle of the jaw, one about
the center of the plexus, and the fourth at the lower end of the plexus.
Of these nodes the one nearest the angle of the jaw receives ducts directly
from the skin of the face around the eye; the other node receives ducts
both from the skin over the head and from the first node. The node in
the middle of the plexus receives ducts from many sources, in fact from
all of the other glands in the chain and from the front of the neck as
well, while the fourth node receives ducts from the fore leg. Thus it
will be seen that some of the nodes are secondary or intermediate in the
sense of receiving ducts from other nodes. Some nodes, for example
those along the internal carotid artery, receive only ducts that have
passed through other nodes, that is, they develop along the efferent ducts
of a node, while some receive ducts both directly from the skin and
from other nodes. In short, a gland may be secondary for some ducts
and primary for others.

The histogenesis of these various nodes follows the same general lines
as the primary lymph nodes, beginning with the stage represented in
Fig. 10, but there is the widest possible variation in the relative pro-

Os

Florence R. Sabin 38:

portions of the lymphoid tissue and the lymphatic tissue. Sometimes
the lymphoid tissue, that is, the connective tissue portion with lympho-
cytes around the artery, nearly fills the node while in other cases the
lymphatic ducts predominate.

Development of the follicle independently of the lymph ducts.—
One important fundamental point comes out in the study of these

KET a GRO

<a

Ties 5

ee

ry
;

Fic. 16. Group of microscopic lymph follicles in the neighborhood of the
inguinal lymph node in an embryo pig, 24.5 cm. long. x about 72. The
condition of the main inguinal node at this stage, is shown in Fig. 14. A,
follicles consisting of lymphocytes around a tuft of blood capillaries; b,
follicles with a single peripheral sinus; c, follicles which are made mainly of
a plexus of lymph ducts; /d, large lymph duct; Jp, lymph plexus in the border
of a large lymph node.

various nodes. After a certain stage in development, the large nodes,
especially those that develop in the long plexus and in the inguinal
region, show many microscopic nodes in their neighborhood. These
microscopic nodes are abundant in embryos from 22 ecm. long on. Fig. 16
is a section in the border of the inguinal node in a pig 24.5 em. long,

384 Development of the Lymphatic Nodes in the Pig

showing seven microscopic nodes. These different small glands repre-
sent all stages in development.

In the figure there is an abundant plexus of lymphatic ducts and
around this plexus is a group of small follicles. Two of these nodes,
marked a, consist of a collection of lymphocytes around a tuft of blood
capillaries. Studied through serial sections, these small follicles (a)
have no lymphatics whatever. Two of the small nodes, marked b, have
a single peripheral sinus around the follicle while others, marked c, have
an abundant plexus of lymph ducts not transformed into sinuses. The
node in the lower edge of the section is shown only in part. It is the
margin of a larger node and shows the peripheral zone of lymph ducts.
This entire mass of follicles will be subsequently fused with the large
inguinal node, which is one method of increasing the size of the nodes.
The inguinal node at birth contains fewer of the smaller follicles in the
border than at earlier stages. The study of these microscopic nodes
proves an important point, namely, that the lymphoid part of the node
or the lymphocytes in the reticulum occurs primarily with the arteries
and blood capillaries rather than with the lymphatics. This was sug-
gested by the angiogenesis of the primary node, but is proved by the fact
that small nodes are found around the capillaries before they are reached
by the lymphatics. That is to say, the follicle is primarily and essen-
tially associated with the artery. That the follicle occurs in the adult
without lymphatics is shown in the Malpighian corpuscles of the spleen.
In Fig. 17 is a small lymph node found in the lung of an adult pig. It
hes in the edge of a lymph eapillary but is without a true sinus forming
an integral part of the node itself.

To return to Fig. 16, the section shows that nodes which develop at
later stages after the lymphocytes occur in the glands, hurry through the
long preliminary process of the first nodes; that is to say, they begin
at once with a heaping up of lymphocytes around a blood-vessel.

Hemolymph nodes—The study of the development of these nodes is
incomplete. It has not yet been extended to all the areas in which the
hemolymph nodes occur, but confined to the neck region and along the
course of the thoracic aorta. The hemolymph node does not occur in
the neck of the pig until the embryo is about 23 cm long, showing that
it is a considerably later development than the lymphatic node. From
the time the embryo is 23 em. long there are one or two small nodes to
be found near the node which forms in the center of the long plexus
(see Fig. 7). In the specimen at 23 em. long the hemolymph node is
in the simplest possible stage consisting of a single follicle with a

Florence R. Sabin 385

peripheral sinus filled with blood. It looks exactly like the follicles
marked b in Fig. 16, except that the sinus is filled with blood. The
next stage looks like the nodes marked ¢ in the same figure, while a third
stage shows an increase in the lymphoid tissue.

The hemolymph node thus parallels the stages in development of the
other nodes, except that its sinuses from the beginning belong to the
blood-vessels rather than to the lymph vessels. In the lymphatic nodes
the sinuses are made of the modified veins called lymphatics, while in
the hemolymph node they are made of the veins themselves. The espe-

08.0 Bea £3
ears 5)

*) i
Ons
us soar 7Oe y

Z Y ¢ 5 we Eme
Garces
Bie eee OE \

o Fk

id alc) ES
Vv Nail o/ i} \s ass
/ 4 gC) L 8) [ Is ‘
Na) Wh, ay
/

Fie. 17. Lymph follicle without a peripheral sinus, found in the edge of a
lymph duct in the lung of an adult pig. x 66.

cial point in the development of the nodes which is not yet clear is the
relation of the veins which make the sinuses to the rest of the blood-
vessels of the node.

In all of the specimens of the hemolymph nodes as far as they have
been studied, that is in pigs up to 3 weeks old, the venous or sinus
portion of the node predominates over the lymphoid element. The
nodes show both the true sinus and the plexus formation observed in
lymphatic nodes, but the developed sinus is much more limited in
amount. ‘This is also true in all of the hemolymph nodes I have seen
from the adult pig. -They consist of three elements, the lymphoid masses
surrounded by true sinuses, and a considerable amount of a venous

386 Development of the Lymphatic Nodes in the Pig

plexus not transformed into the sinus. Thus the hemolymph node is a
later and less developed organ occurring along the blood-vessels.

The hemolymph node found in the neck of the pig is from the begin-
ning a distinct organ, different in type from the lymphatic node. In the
adult pig these hemolymph nodes in the neck are sometimes fused into
the same capsule with an ordinary lymph node, but the two remain as
distinct structures with trabeculee between.

SUMMARY AND GENERAL DISCUSSION.

Follicle——From the study of the development of lymph nodes we find
that there are in general three types of follicles. The simplest follicle
consists of a collection of lymphocytes in a reticulum around an artery
and its capillaries. This type occurs in the embryo and probably in the
adult. The second type consists of the follicle of the first type sur-
rounded by the lymphatic sinus. This type occurs either singly or in
groups in the ordinary lymphatic gland. In the third type the lymphoid
follicle is surrounded by a blood sinus or a less developed plexus of blood-
vessels. This occurs in the hemolymph node and spleen.

Lymphocytes.—In connection with the nature of the lymphoid tissue
it has already been brought out that the lymphocyte occurs in a fine
reticulum around the artery and its capillaries. For the ultimate origin
of the lymphocyte we have as yet no proof, but the lymphocytes divide,
as Flemming showed, in the germ centers. The germ centers are definite
organs around capillary tufts or glomeruli. In the embryo the division
of lymphocytes is so constant that the germ centers are always filled
with them. The lymphocyte develops independently from the lymph
ducts.

General structure and growth—In regard to the general structure
a lymphoid

of the node, it has been shown to consist of two elements
or vascular and a sinus element, which is either venous or lymphatic.
Through the study of the development of the nodes, it becomes clear
that the sinus comes from a plexus of ducts or vessels by increasing the
number of ducts and reducing the size of the connective tissue bridges.
In the pig the plexus element of the node is not completely transformed
into the sinus in the adult, so that the node has three structures, the
lymphoid element, the plexus, and the sinus. There are wide variations
in the proportions of these elements.

The question of growth and increase in size of the lymph node is an
interesting one, closely bound up with the process of absorption in devel-
opment. All of the diagrams show that the nodes in the early stages
increase in size by the invasion of surrounding tissue by lymph ducts.

Florence R. Sabin ~ 387

This point is brought out by comparing Figs. 5 and 6, from the primary
node at 3.6 and 4.9 em. long. The sections are cut in the same plane
and can be related by the sac. It will be seen that the increase in size of
the second is due to the development of the lymph plexus. This invasion
of ducts can take place as long as there is no definite capsule. The cap-
sule is a late development, as shown in the last diagram, where it is still
incomplete. After the capsule is formed the nodes increase in size by
the fusion of many small nodes in the neighborhood. The capsules of
the smaller nodes become the trabecule of the larger ones. The marked
tendency of the fusion of nodes in the pig has been noted by other ob-
servers. In the pig not only do the ordinary lymph nodes fuse, but a
hemolymph node may fuse with a pure lymph node and the two remain
distinct but enclosed in a common capsule. The primary lymph node
from the lymph sae, and the node over the crest of the ileum show but
few of the small nodes in the neighborhood and hence few evidences of
fusion. On the other hand, the inguinal node which represents a group
of nodes in higher forms, appears hike a conglomerate of small nodes in
the new-born pig.

The lymph nodes give much evidence of a repeated tearing down and
rebuilding in the process of growth. For example, in passing from the
stage of the first to the second diagram, there must be a destruction of
many lymph ducts in the formation of the primordial follicle. It will
be noted that this destruction is not complete in Fig. 8, for there are a
few lymph ducts scattered through the follicle. In following through the
diagrams it will be noted that there is a constant change in the propor-
tions of the lymphoid or vascular portion and the lymphatic portion of
the node. The small nodes of Fig. 16 make this point clearer. The
youngest ones have a predominance of the lymphoid portion, the very
youngest ones having no lymph ducts at all. Others have merely a peri-
pheral sinus, while others slightly larger are made almost entirely of
a plexus of lymph ducts. As these nodes develop farther the lymphoid
tissue will increase until the balance of lymphoid tissue, sinus, and
lymph plexus characteristic of the adult node is reached. This point
illustrates the extreme variation met with in the development of the
lymphatic apparatus.

Sinus.—The lymph sinus develops out of the lymph ducts by a mul-
tiplication of the ducts along certain lines. The areas in which the
lymph ducts multiply until the connective tissue bridges are reduced to
the thickness of fibers, are determined by the blood-vessels. That is, the
sinuses grow in between the arteries thereby bounding the lymph cords.

388 Development of the Lymphatic Nodes in the Pig

The sinus develops by a proliferation of the endothelium of the ducts,
and so, as Kling pointed out, each space in the sinus has its complete ring
of endothelium. The fact that the sinus is made of a great number of
small lymph ducts packed closely together explains why one cannot get
the silver picture of a membrane of endothelium far beyond the periphery
of the node. None of the membranes are large enough. In the em-
bryonic stages the connective tissue bridges are largely protoplasmic and
show connective tissue nuclei. In the adult the bridges are a network
of reticulum fibrils so that the endothelial cells appear on the fibers.

The study of the contents of the sinuses will prove, I think, an
important point in the physiology of the lymph nodes. There is a
marked difference between the embryonic nodes and the adult in this
respect. In the embryonic nodes there are few free, wandering cells
in the sinuses as compared with the adult. In the early stages there are
almost no free cells in the ducts. After the lymphocytes appear, they
occur occasionally in the ducts and sinuses, as well as a few large mono-
nuclear forms. As a rule the sinuses are nearly empty of the free cells,
so that the bridges stand out with great beauty and clearness. In the
adult node, on the other hand, the content of the sinus in free cells is
most varied, both in the number and in the kind of cells. Sometimes
the sinuses are so packed with cells that the reticulum and endothelium
are almost covered up. It often happens that the sinuses of one por-
tion of a section are densely packed with cells, while in another portion
they are almost empty. These free cells may be any type of white blood
cell or may be large phagocytic cells. The sinuses may be filled with
phagocytic cells that are crowded either with red blood corpuscles or, in
the mesenteric nodes, with fat globules.

Reticulum.—The complete relation of the connective tissue frame-
work or reticulum can only be clear after noting the nature of the sinus.
The reticulum fibers first appear in the germ centers where they make
a mosaic pattern around the capillary tuft. The trabecule develop from
the capsule in connection first with the sinuses and secondly with the
blood-vessels, especially the veins. The reticulum fibers are laid down
in a very close protoplasmic syncytium, and this syncytium remains pro-
toplasmic long after the surrounding connective tissue has become pre-
dominately fibrillar. The reticulum fibers do not show until the embryo
is 15 or 16 cm. long. They appear first in the germ centers and in the
capsule. In an embryo 23 cm. long they are limited to the capsule and
a few trabecule, while up to the time of birth the connective tissue frame-
work is still largely protoplasmic.

Florence R. Sabin — 389

The reticulum framework in the adult makes a complete anastomosis
throughout the node. The framework can be readily traced in the last
diagram. Starting from the capsule, fibers enter the trabecule, pass
between the ducts of the sinus and enter the lymph cord. The large
trabecule carry veins. In a specimen of reticulum from which the cells
have been digested out, the entire node can be reconstructed for the
sinuses occur in the looser reticulum that borders the trabecule while
the cords and follicles are between trabecule and show a much finer and
denser network than the sinuses. The germ centers often appear as holes,
since the fibers are delicate there. :

Lymph capillaries.—The subject of open or closed lymphatics, within
the lymph node as elsewhere, has given rise to most definite but oppo-
site opinions. The study of development touches but one aspect of
the question, unless it is combined with many injection experiments.
The lymphatics develop as blind sacs from the veins and have a complete
endothelial lining. The sinuses result from a multiplication of the lymph
ducts and the only difference between the sinus and the preliminary
lymph plexus is in the width of the connective tissue bridges between the
ducts. In the sinus these bridges are reduced to the thickness of the
reticulum fibers of the adult. Thus from the purely embryological argu-
ment the spaces of the sinuses have a complete endothelial lining. Nu-
merous injections of lymph nodes in every stage have been made with
Prussian blue and with India ink. The ink always runs farther than
the Prussian blue. In injecting the nodes through the afferent ducts it
is always easy to avoid undue pressure within the node, for the lymphatic
area of the node is so much greater than the caliber of the ducts leading
to it. It is possible in all stages up to two weeks after birth to obtain
injections without extravasations. But many injections do show extrava-
sations and these occur in the walls of the sinus rather than the less de-
veloped plexus of ducts which always forms a part of the node. This
shows that the wall of the sinus is weaker than the walls of the plexus.

To conclude, the present study shows the close relation between the
lymphatic system and the vascular system. In the formation of lymph
nodes, hemolymph nodes, and spleen, there is one fixed element in com-
mon, namely, the lymphoid tissue associated with the artery; the fluc-
tuating element or the sinus belongs either to the venous system or to
modified veins called lymphatics. The sinus may be absent, or venous
or lymphatic.

ON THE ANGLE OF THE ELBOW.

BY
FRANKLIN P. MALL.
From the Anatomical Laboratory of the Johns Hopkins University.

WitTH 1 FIGURE AND 8 TABLES.

Artists consider a woman’s arm beautiful when, in its extended
position it is straight, or nearly so, and sufficiently plump to give it
delicate curved lines. With the elbow flexed the upper arm is considered
the more beautiful the nearer its section approaches a circle. The same
is claimed for the forearm near the elbow. As the section approaches the
wrist the circle becomes an ellipse, and the farther it is from the elbow
the more marked is its eccentricity. With the forearm flexed and semi-
prone the long axis of the ellipse is directed diagonally downward and
outward. It is self-evident that artists do not make either cylindrical
or conical arms; they prefer as a model an arm whose section is nearly
circular. ‘The forearm is always a little flat, more so in supination than
in pronation. In antique statues the upper arm is found to be more
nearly circular, while in those of the renaissance a lateral flattening is
shown. It appears then that the ideal arm of artists changes from time
to time, possibly because the models before them changed correspond-
ingly. At any rate the shape of the upper arm of the renaissance
approaches the modern anatomical arm more than does that of antiquity.

The form of the ideal woman’s arm is caused in great part by the layer
of subcutaneous fat drawn over the structure below. In the ideal man’s
arm the structures below, especially the muscles, protrude and form
marked lines indicating strength. And it is considered beautiful by
many to show some of these lines in a delicate way in woman’s arm. A
slight outline of the deltoid, biceps and triceps does not make the arm
appear masculine provided it is built upon a delicate skeleton.

The beautiful arm, then, is one that is plump, round, tapering and
relatively straight. Differences are, of course, to be expected, due to
race, sex andage. The amount of fat upon the arm differs much between
the ages of 15, 25 and 45. The same seems to be true regarding the
angle of the elbow. The arms of young girls are said to be straighter and

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.
2]
vo

©
we)
©

On the Angle of the Elbow

the motion of the elbow seems to be greater than are those of young
women. At any rate, from an artistic standpoint, a slight amount of
hyperextension is permitted in a child’s arm, but it makes a bad impres-
sion when it is present in the arm of a muscular man. It is evident
that the standard of the beautiful must change for different periods of
hfe, and the question is whether the beautiful and anatomical normal
correspond. It was mentioned above that the arms of antique statues
were unlike in form those of the renaissance, the latter being more
realistic.

The Greeks constructed the canon of the human body with its length
eight times the length of the head, which was taken as the modulus.
This gave a rather short body, too much so, for Michael Angelo found it
necessary to add one-third of a modulus, making the body 84 heads long.
This third of a modulus was added to the legs above, which were also ex-
tended one-sixth of a modulus below. Thus Michael Angelo’s canon
is half a head longer than the Greek canon, all of the difference being
added to the legs. This may account for the plump arms of the Greeks
and the thinner arms of the renaissance. Since that time many systems
of measurement have been invented, differing mostly in the modulus,
and contributing little to the proportion of the body. It appears that
the first scientific step was taken by Quetelet, who drew averages from
the measurements of some 30 soldiers. So much variation was encoun-
tered, however, that his results proved to be of little value. Others made
many measurements, as is shown by Sargent’s admirable work on the
average figure of American students of both sexes. The next
step in advance was made by measuring from the principal joints to
obtain the proportions of the trunk, and the most satisfactory system is
that of Fritsch, whose modulus is the length of the spinal column. His
canon, to a certain extent, outlines the human body, giving at a glance
many ratios. When this is compared with numerous recent outlines it
is remarkable how well they coincide. It may be added that the best
recent outlines of the body were constructed by anatomists and that there
is now a tendency for artists to accept a canon which is anatomically
correct. Furthermore, this canon is much more like that of antiquity
than like that of the renaissance, being half a head shorter than the
Greek canon.

The difficulty is not to be solved by inventing a new modulus but by
establishing its length. This in turn will establish the length of all
other important measurements, bringing ultimately the artistic ideal
and the anatomical normal together. Furthermore, measurements from
the centers of the main joints are most desirable, for then exact measure-

Franklin P. Mall 393

ments can be made in a statue, since a femur or a humerus measures the
same in all positions. Fritsch’s canon fulfills this requirement. That
the average measurement is the most beautiful is further proved by
observing composite photographs of many average faces, few of which
are considered beautiful, while together they are decidedly so. Nature
has here been and must continue to remain our best standard.

To what extent artists may idealize variations is not for me to consider
now. They must, however, remain within bounds, and when they empha-
size a variation of one part their convention must make the rest of the
body the anatomical normal in order to bring out well the difference.
So if straight arms are the artistic ideal, the rest of the canon must
correspond with the anatomical normal.

In a model an arm is not considered beautiful if it is too long, for
this is said to be indicative of a lower race. Neither is hyperextension
nor a lateral angle considered desirable. It is especially inartistie to
have the two combined. Hyperextension of the elbow may be overlooked
in children and in delicate girls, for it helps to indicate the flexibility
of the body, which is a characteristic of youth.

It appears that those who write upon the angle of the elbow from the
standpoint of the artist are not altogether familiar with its anatomy, for
the straight arm is considered the more usual. According to Briicke, the
lateral deflection of the forearm is more common in women than in
men, while according to Stratz the opposite is the case.

The latter author makes the normal arm so straight that the wrist in
turning rotates in a circle with the radius as its center, thus in prona-
tion the lower end of the ulna moves lateral as much, or more, than the
radius does medial (see Stratz’s Fig. 67).

It has been known for a long time that the trochlea is not set at right
angles to the shaft of the humerus but obliquely to it, making an acute
angle directed outwards. It is this lateral angle which, according to
Langer, causes the forearm to bend out when the arm is extended, and
assuming that the articular surfaces of the ulna and radius are at right
angles to the forearm the hand must fall upon the chest when the elbow
is flexed. The amount of lateral deflection when the arm is extended
equals the extent the hand moves in when the arm is flexed. The ob-
servations of Langer are accepted by Briicke, but apparently he does
not consider the normal arm beautiful, and suggests various methods
by which it can be corrected in a picture or a statue by always presenting
the arms either partly pronated, or partly flexed, or both.

The subject was carefully reworked by Braune and Kyrklund in their
study of the elbow joint, and they show that not only is the axis of

394 On the Angle of the Elbow

the elbow joint set obliquely to the humerus but also to that of the
forearm. As a rule the angles formed by the axis with the humerus
and with the forearm are nearly equal, each measuring about 83°, both
acute angles pointing outward. The styloid process of the ulna which
marks the long axis of the forearm in both pronation and supination,
deflects fully 14 degrees from the sagittal axis of humerus when the
elbow is extended, and gradually approaches it as the elbow is flexed, for
the angles of the humerus and of the forearm neutralize each other in the
flexed position.

This study was made to test these results, and to determine the extent
of the motion of the elbow in the European and in the negro, male and
female, for during a number of years past I have felt conscious that a
sexual difference exists.

In order to make satisfactory measurements fixed points had to be
established and after extending Braune and Kyrklund’s reliable method
a modified method for measurement was hit upon which when applied
repeatedly to the same arm gave an error of less than one degree. Un-
fortunately there are more variations than I had anticipated, but I
venture to give my data with the hope that they may be of more general
use, and that I may be able to add to them in the course of some years.
To collect and measure 100 specimens is not altogether a small task, but
this kind of work must be multiphed many fold before the foundations
of anatomy—descriptive, regional and artistic—become anthropological.

The arms studied were taken from the dissecting rooms and carefully
cleaned, leaving all of the ligaments of the elbow intact. Then the axis
of the elbow was determined by fixing the ulna and radius, and moving
the humerus to and fro. By doing this it is quite easy to find a point
over each epicondyle which does not move. A line extending through
these two points passes through the middle of the trochlea and marks
the axis of the elbow joint. Frequently there is not an immovable point
over one of the epicondyles, but instead a line is determined and in most
cases the middle of the line is taken as the axis. Next the humerus was
fixed with this axis and’a point in the middle of the upper third of the
shaft in a horizontal plane 15 cm. above the plane of the table. The
perpendicular plane was then passed through the middle of the upper
third of the humerus and through the coronoid process, for it has been
shown by Braune and Kyrklund that this process keeps within a milli-
meter of this plane, passing first to one side of it and then to the other.
The screw motion which is said to be present by Meissner, Henke and
Langer does not exist. Five centimeters from this plane a glass plate
was fixed from which the measurements were taken. The horizontal

Franklin P. Mall 395

plane becomes a coronal plane when the arm hangs down along the side
of the body, and therefore I still call it the coronal plane. In the same
position the perpendicular plane is parallel with the sagittal plane of
the body and may therefore be called the sagittal plane of the arm. The
intersection of the two planes forms the axis of the humerus passing
through the center of the upper third of the shaft and through the axis
of the elbow below the coronoid process.

Diagram of the bones of the arm with the planes from which measurements
were taken indicated. D, circle of maximum deflection of the ulna; d chord
of the arc of deflection when the elbow is extended; d’, the same at about 110
degrees; e, chord of the are of extension when reduced to degrees to be sub-
tracted from 180 degrees; f, chord by which the extent of the flexion of the
elbow is measured.

The angle of the axis of the joint with the long axis of the humerus
was first determined by direct measurement, the right arm being clamped
with the humerus to my left, and the left arm in the opposite position.
In all cases the degree of flexion, extension and deflection was deter-
mined by measuring from the styloid process of the ulna. The degree

w
Lo)
ron

On the Angle of the Elbow

of flexion was determined by the chord of the are which would be de-
scribed by moving the styloid process from maximum flexion to the
long axis of the humerus with the axis of the elbow as the center. The
degree of extension was determined by the chord of the are described
by the styloid process between maximum extension and the projected
axis of the humerus. Accordingly when the elbow joint did not extend
to a straight line or when it hyperextended this amount was respectively
subtracted from or added to 180 degrees. The amount of deflection was
determined for three positions measuring from the styloid process with
the elbow flexed, extended and at 90 degrees. The degree is determined
by the chord of the arc described by the styloid process intersecting the
sagittal plane at right angles in these three positions named above. In
case the deflection is out it is marked plus and in case it is in it is
marked minus.

In the appended tables the first column gives the number of the cadaver.
The second column gives the length of the ulna from the styloid process
to the axis of the elbow joint. Next the angles of the humerus and the
ulna with the axis of the arm are given. These, together with the degree
of deflection of the forearm with the elbow joint extended to its maxi-
mum, always equal 180 degrees. Then follow the columns with the
degree of motion, from maximum flexion to maximum extension, 180
degrees being a straight line. The lateral angle is next given in three
positions and when it is marked minus it indicates that the arm turns in.
This takes place frequently when the elbow is flexed to its maximum,
occasionally when at right angles, and not at all when it is extended. In
other words, when the arm is extended and supinated the whole wrist
including the styloid process lies to the outward of a line drawn through
the middle of the upper third of the shaft of the humerus and the
coronoid process of the ulna. All arms are deflected laterally.

Braune and Kyrklund have shown conclusively that the elbow joint
is a pretty perfect hinge joint and that there is practically no screw
motion in it. As it flexes the ulna shifts a little, first outward then
inward, which motion causes the shaft of the humerus to rotate outward
nearly 6 degrees in case the forearm is flexed. For all practical purposes
the joint is a hinge joint, the axis being set obliquely at nearly 84 de-
grees to both humerus and ulna. In all cases the styloid process of the
ulna deflects about 12 degrees when the arm is extended and when flexed
becatse the angles of the humerus and ulna are about equal, the ulna
lies in the sagittal plane of the humerus, 7. e., the ulna comes to lie
directly upon the humerus and not upon the chest as is claimed by
Langer. In ease the angle of the axis of the ulna is less than that of

Franklin P. Mall 397
the humerus, the styloid process still deflects when the elbow is flexed
and in case it is greater it is turned in. Braune and Kyrklund’s few
cases (nine in number) seem to bear out these statements, but they are
by no means always borne out by my records. The sigmoid cavity does
not hug the trochlea closely and the slight rotation of the coracoid and
olecranon processes may be sufficient to account for my figures. Fur-
thermore, the inequalities in diameter and form of the two conical sur-
faces of the trochlea may cause sufficient shifting to counteract a slight
difference between the angles of the humerus and ulna. This is already
indicated when the points below the epichondyles through which the axis
passes are determined. One of them is usually extended into a line
several millimeters long showing that the section of the cone of the
trochlea is not circular on that side. Even my averages do not confirm
Braune and Kyrklund’s notion. In my 89 specimens, the average of
the angle of the humerus is 82.5 degrees, and of the ulna 86.5 degrees, and
yet the styloid process still deflects .5 degree when the arm is flexed to the
maximum. This, of course, is when the elbow is flexed to 39 degrees,
and could it be flexed to 0 the styloid process should turn in about 2
degrees. In general the irregularities of the surfaces of the elbow joint
fully neutralize the fine difference between the angles of the humerus and
ulna and only in a general way is the assertion of Braune and Kyrklund
correct. In about three-fourths of my measurements the angle of the
ulna ‘is greater than that of the humerus, while in Braune and Kyrklund’s
measurements (but one-eighth as many) they were just the opposite.

The extent of motion of the elbow from flexion to extension gives some
interesting results. It is well known that the extent of movement in the
joint of children is much greater than that of adults, and artists often
try to express this in the arms of children and young, delicate girls, I
have often observed this difference in examining arms of infants in the
dissecting rooms. In fact, in numerous specimens which I have exam-
ined not a single infant’s arm was found in which the elbow could not
be hyperextended. The measurements from the adult arm which I
have made give equally interesting results, for they point towards a
sexual difference. The following table gives the degree of extension of

Degrees of extension....... 156° 1602 1652 iOS o> St SsOo a tsha loos loos

Number of males... . 0... 1 6 19 14 14 6 3 1

Number of females........ ae re 2 6 q¢ 5 Pd 1 2
SETS TAISE ae, co asi steht ey sinralls 1 6 21 20 Zi 11 5 2 2

89 measurements. The straight arm is 180°. It is evident that
the female arm is straighter and more frequently hyperextended than the

398 On the Angle of the Elbow

male. The degree of flexion gives a similar table, which becomes more

Degree of flexion........ 202 25° 30° 35° 40° 45° 50° 55°

Number of males......... i oes 3 ny, 23 ay, 3 te

Number of females....... 1 i] 6 8 6 3 1
ANTHEMS 4c O60 COUDOOC 1 1 9 25 29 20 3 1

marked when it is expressed in differences, that is the degree of motion,
from maximum flexion to maximum extension.

Merreeron motion. - 110% 1S ped 20° wel 25 eo SOS 135 oe 140 O14 el 5 Oe ae

Number of males... 2 5 9 6 15 ala 12 2 2 nie

Number of females.. .. eae oe 1 2 4 8 ° 2 3
Totals ccs kiaiedven nes 2 5 9 7 ul? 15 20 7 4 3

The two lines now move away from each other more than before,
the greatest number of cases for each sex being 10° instead of 5°
apart. In constructing this table the degree given is each time the
middle figure; for instance, 130° includes 128° to 132°. Further-
more, there seems to be a slight racial difference which tends to make
the sexual difference rather less marked than it really is. The greatest
number of European males occurred in maximum extension under 170°,
in maximum flexion under 35°, and in degree of motion under 135°.
In other words, the motion of the elbow of the European male is more
nearly like the female than the male negro. So when the joint of the
negro alone is considered the sexual difference is more pronounced than
when it is considered with that of the European. The following table

Weeree of motion... =. os 6. T1098 L152 120° 125° 18088 185°" 1402 145° o0emiaoe

Ne eal Number of males.... 1 5 7 (aly iG 8 2 1 ae

ae ( Number of females... zis 43 ae se 1 4 8 4 1 3
MOCA sivscvale scars avert ovate ts if 5 7 6 13 ipl 16 6 2 3

includes only the arms taken from negroes. On account of an insufficient
number of records further tabulations give no results which are definite.

As far as the records go, they indicate that the elbow joint of the female
is more flexible than that of the male—is more of the infantile type—and
that of the European male holds an intermediate position between the
negro male and negro female. Practically all of the subjects considered
came from the laboring class, so a difference on account of muscular de-
velopment cannot be entertained.

The amount of deflection of the forearm is shown in the data which
follow. In all cases the styloid process deflects when the arm is extended

Franklin P. Mall 399

and every specimen verifies the statement of Langer and Briicke, that the
whole wrist falls to the outside of the sagittal plane of the humerus when
the forearm is extended and pronated. The assertion of Stratz that in
_ this position the line falls in the middle of the wrist is absolutely incor-
rect. Furthermore, his diagram (Fig. 67) which is apparently based
upon Merkel’s normal figure, is also incorrect, for Stratz’s own copies of
Merkel’s figures (Figs. 31 and 32), as well as the originals, coincide with
Briicke’s as regards this point.

With the elbow extended the average deflection of the styloid process
of the ulna from the sagittal plane of the humerus is 11° from my
measurements. The average length of the ulna from the axis of the
elbow to the styloid process is 258 mm. With these two measurements
11° equals a chord about 5 centimeters long, so the styloid process deflects
normally 5 cm. or about the width of the wrist. Therefore, with the
arm extended the wrist should fall outside of the sagittal plane of the
humerus in both supination and pronation. In both positions the
styloid process falls about 5 em. to the outside of the sagittal plane of
the humerus and in pronation it passes through the styloid process of
the radius. In the extended arm all of the wrist, or at least its greater
part, falls lateral to the sagittal plane of the humerus in both pronation
and supination. This marked deflection, more so in my records than is
stated by any author, is no doubt due in part at least to a racial differ-
ence, for 70 of the arms are from negroes and but 19 from Europeans.
A glance over the tables shows that some difference does exist, which I
shall now consider. When the differences in deflection are grouped for
every 5°, as was done when discussing the motion of the elbow, nothing
definite is noted, and when they are grouped under single degrees the
figures scatter so much that it is again difficult to see any marked result.
The negro male, however, shows some 3° greater lateral deflection in the
movement from flexion to extension than does the European male. The
difference between the European male and female is much greater, but
the number of cases are so few that this also cannot be considered. It
would indeed be remarkable if more records showed that the European
female had the greatest lateral deflection and that the European male
the least, that of the negro lying between. If it should prove to be so,
then artists have secured their ideal straight arm of females from the
males and infants’ where the lateral deflection is the least.

The racial difference becomes more marked when the total amount of

1Braune and Kyrklund state that the angle of the humerus in infants is
much less than in adults.

400 On the Angle of the Elbow

deflection (that is, the difference between that at flexion and that at
extension) is divided by the number of degrees of motion of the elbow.
The deflection being greatest in the negro male, the result becomes
still greater because the motion of his elbow is the smallest. This quotient
becomes the degree of lateral deflection for one degree of elbow motion.
If this in turn is multiplied by 180, the amount of deflection is obtained,
in case the elbow joint could be moved from zero to 180 degrees. This
quotient I shall speak of as the total deflection, it being the amount of
deflection in case the elbow joint had a motion of 180 degrees. For ex-
ample, in the right arm of subject No. 925 the deflection is between
— 4.5 degrees and 1 degree or 5.5 degrees. This divided by the number
of degrees of motion (169—50) 119 makes .0462 or the number of
degrees of lateral deflection of the forearm for each degree of flexion or
extension. In turn this multiplied by 180 gives the total deflection
could the forearm move through an entire semicircle. In this case it
amounts to (.0462 & 180) 8.3 degrees. Now it is found that the deflec-
tion per degree varies for different positions of the forearm, as is shown
in the following table. In the first column the deflection per degree is

Degree of deflection for each degree of motion from

Maximum Flexion 90° to Maximum Maximum Flexion

Race Sex Arm to 40° Extension tote
RASC ee ces cehere ere O8 .10 .09
Male THOLGS Woe See canes kota! eis OT .O8 OTS
| 18a) Ssssocogasod= O75 10 -08
Negro RishGul.y scare steer 07 .09 -08
Female Wiehe See ew reves sere OTS 08 .08
| BOtHINe aciak ee eteas 07 -085 08
| Both Pinte eieiotolesorh areata OTS .O9 .085
{ QING oss G omaeae .04 OT .06
Male [et (acer oe 025 055 .046
(i Both aet oe eens 03 06 05
European Rights ses, ci. otters eck ste silat 10 -10
HRomale. Welt) .a-5.006ee oo. .10 .05 OT
| Bothtcseat sors 105 075 085
Bothit sctr rine ce aes oer orecoetelers 05 O75 .065
AViCL ALO Me oreueiusaterc loko omuensescels oe cistei sis eile OT O85 -08

given between maximum flexion and 90 degrees, in the second from 90
degrees to maximum extension, and in the third the average degree of
deflection for the whole motion of the forearm. Of course to determine
each figure it was necessary to start with an average. It is seen from
the table that the lateral deflection per degree of motion is generally less
when tne elbow is flexed less than a right angle than when it is extended
beyond it. The deflections seems to increase as the maximum extension
is approached. This is to be accounted for in part by the irregularity of

Franklin P. Mall 401

the surface of the elbow joint. When the averages are considered it is seen
that there is a marked difference between the deflection in the negro and
in the European which becomes more pronounced when the males only
are considered. The race of fully 70 per cent of the cadavers from
which these arms were obtained can be determined by these measure-
ments. The average deflection here is .08° and .05° for each degree
of motion, and this difference is pretty constant, as can be seen from
the table. The deflection between maximum flexion and maximum
extension for negro males and European males is 10.5° and 6.50°,
which, considering the differences in the lengths of the ulnas equals
5 em. and 3 cm. respectively. If the total deflection is considered, that
is, if the motion of the elbow were 180°, the deflection would be 14.5°
and 9° for the negro and European respectively, which when the average
lengths of the forearms are considered equals 6.5 and 4 centimeters. In
a measure this difference is obscured for the flexed arm of the European
deflects more than does that of the negro.

The conclusion of this study is that the degree of motion of the elbow
is greater in the female than in the male and that the lateral deflection
of the hand, from flexion of the elbow to extension is much greater in the
negro than in the European. The lateral deflection of the hand in the
extended arm is much greater than the artistic ideal.

MEASUREMENTS.

Nearly all of the measurements are from arms taken from individuals
belonging to the laboring classes. The American negro is more or less
intermixed with European blood; those in Baltimore are, however, usu-
ally over three-fourths black.

402
No. of Uina.
GOD Rts pis nsichaventes 255
Qe eras costes iauctae 270
Ob Ge rice. wcadeioe tote 260
MOG Bele o se cners cheters 290
MO Gecrarclncdais Qicteis 268
UB G ro oe ote (ena nists 290
HAD Gt Ss icieis ctoseranste 270
a ba I: ae ee eR eeNE 275
PVG ate.c eve 8 ee seis 265
TWA ie Se ehsus diaetete 240
NGO ra cusve creer sl 260
MAG Sirrck-Rersianers,orete 260
PAG Drrrerensonerehstsie4ers 280
TNS Cie aaetetisrcherey shares 240
UNG Os aires wares 290
ANZSO\ shore ctavees sieve ete 245
P2GW sale she sevatel eevans 275
BN FNL e Faire nay ayaa ca/oal one 270
i PY eee Se ecaro ices 265
NZS acre cusre ts, af eed
ABU ere cuareraleceteress 270
myaielsis henetevetes 300
Brey ict Sitay wate pepe 285
Average (23) 269.5
QD Bie eersis exe ai o0dh 280
GOODS sacha sieve eos 255
G56 Bias oyels shea) s 270
NOG 2 eos 5 FS eclee es 285
MORO. Sos whe sss os 266
UDOG tarerccevsvsoays eves 290
MVD es acters lone 250
MDS steers stevens ss 270
ADD Gisiercncteeks, 38a 275
MIMD SNe orateieveye as 290
NS CR crolPoeteteicus 265
3B ener ARCRONATT 245
MED ete er cistvece siete 290
MAES OE cretaya.c one ate chs 234
AMS eee cae autor 240
POOR eres Srevehegereve 260
PSU OM perescys, cs ceekene 255
T2OSS Ma years 230
2S Ole hers orcte eee 232
U 26st d clecsine wre 278
5 [OAT fd Wee eeCR CRORE PaO 273
U2 TB rciayansne eerelent 270
W2SO Sst weete wuss 285
USED Gacreuevenskebencucueks 260
Sense Kawane Teena ete 295
SNE apeeo dtemeces s 255
Average (26).... 265

*Boy.

On the Angle of the Elbow

Angle of axis
of elbow joint.

Humerus.

86
82
85
80
67
76
87
83
75
84
83
86
86
79
85
87
85
83

2
81
87
84
83

82.%

83
85
82
82
75
73
75
84
84
83
78
75
87
80
90
85
79
83
83
85
85
83
86
88
80
80
82

NEGRO, MALE.

Right Arm.

Ulna. Flexion.
83 45
97 50
92 47
84 41.5

100 33
90 35
88 35
94 438
94 43
84 48
87 39.5
80 38
84 47.5
91 31
85 33
81 42
82.5 39
85 42

5 idl 40
94 39
81 40
86 40
85 44
87.5 40.5

Left Arm,
92.5 46
86 47
91 44
82 38
94 42
89 36
81 38.5
85 43
90 34.5
88 40
91 42
94 39
85 34
91 36.5
82 45
84 46
85 41
81 44.5
80 48
80 36
85 ot
86 38
91 36
90.5 34.5
84 38.5
95 39
87 40

Degree of movement
of ulna.

Extension.

162.5
169
162
173
165
176
175
168
168
180
181
162
163
167
166
161
170.5
168
167
162
163
168.5
155
167.5

192
164
168
174
179
173
161.5
177
163
170.5
72
180
175
184.5
158
167
184
173
164
164
167
173
166
163
173
181

172

Lateral angles of ulna in
different positions

Maximum Right
Flexion. Angle.

4.5 5:3
—4.5 —1
—Ss —)
Ld 8
—7.5 0
—1 4
3 3

—1l11 —6.5

—3.5 3.5

—3 3)
0 2,
5D 1
—1 1
—2 9
4 5

3.5 7.5
3 5
0 4
—8s —2
—4 0
5 6
—3 1
—_4 2

—1.3 2.6
3 6
6 6

—8 —3.5
0 5
—T 0
3.5 9
2.5 6
5 7
—3 it
6 4

2 4.5

—3 1.5
—2 —1
3 6
—1.5 0
i 4

7.5 11.5

5 8.5

3 8.5
6 8
—2 3
—1 5
0 1

3 1.5

0 Or
—16 —7T
a1) 4

Maximum
Extension.

11
1
3

16

13

14

10.5

Franklin P. Mall 403

NEGRO, FEMALE.
Right Arm.

Silisee cys ieee ets, 6 270 83 83 46 184 —l1 5 14
THIS s, coareinay Ot NceeiCne 250 Souo 79 36 175.5 4.5 6.5 16
MELEE evaivece) the, ‘ssi 0 265 80 87 30 171 —2 5.5 13
LANTOS Dee ORE 245 85 85 32 Ali (é Dae 2.5 10
MUS Sicevatsts sieves LOS 80 88.5 40 180 —1l1 3.5 11.5
WASOs OSoRI eae 240 86 81 31 164 5 9 13
sc Siets,sysiie ss 6 as 240 83 85 38 180 4 8 12
Teese Sveroie).c',0:.¢,e1'stsis 240 80 88 39 175 0 5 12
Sed tieieree: sisisss aveiel.eve 240 87 76 37.5 168 C0 7 alr
US OD ctor crsieysist so. 255 88 76 36.5 193 6 i (3) 16
Average (10).... 248 83.5 83 36.5 Wer 2.5 6 13.5

Left Arm.

SOM eR tetie sale tesa le 260 19 90.5 43 182 —4.5 0 10.5
hi 19) OOS Bic teiceercaene 245 83 Kates 19.5 173 7 ial ish;
Pere anon save hy tis 260 80 92 29 alral —3.5 2.5 8
AMIE Ges dete hes crea) «y'5).2 PSNI 85 87 PAG 170.5 4 4 8
MNS Si reralec dea dis euens. 235 80 92 3D.D 180 —3.5 »D 8
ME OVertatsiisy see) sie) 3s. 243 80 83 29 173 "i 13 ul
Meee leotens ers clsyecansheus 240 83 86 41 174 3 6.5 ual
Meer Snax cease shes les 245 75 94.5 36.5 175 —t 1 10.5
MDGS ir oes c! cya se 8° 255 80 91 30 179 —6.5 0 9
ia ercus) suse sv 3s 240 80 83 ith 170.5 3 7 17
STOR 6 Ap. CO cere 250 80 85 37 1938 8.5 11.5 15
Average (11)..... 246 80.5 87.4 33 176.5 1 5.2 12:1

EUROPEAN, MALE.
Right Arm.

QS Baits cae: cts) ors sos 240 83 88 41 172 —1.5 3 9
Nese fcners. species via Leo 83 81.5 45 165.5 6 8.5 15.5
ANS aoe sence de oe LOO 84 88 43.5 175 —9 —5 8
Le pete es ccvens sts 270 85 81 30 169 9 10 14
UA Gemedlssies aso) LOD 90 81 41 ETE TE 8 4 9
EN SS erat civei.e. sien ete. 255. 85 84 35 167 4.5 4.5 ial
MRNGH ot erereess co! sveieveus sr 200 80 85 33 181 0 5D 15
MS Ge aere cs) e's\eisiaye LOO 83 87 43 163 2 5 10
12 Sie eee eae 260 89 84 37 173.5 6 6.5 1

AEROS 255 84 89 43 185 0 25 T
Average (10) 249.5 84.5 85 39 iS 2:5 4.5 10.5
Left Arm.
AG Ate veie! s: <<! ove 00" OO 87 89 40.5 173 8 6.5 4
MEN GA aiiee8e ciel oe sas 255 75 87 57 166 6 9.5 i8
T1555 Sea ie Ri Exper 270 83 84 a ilyfal it 1 13
W2RGre ee sis s+) o. <1 OO 80 93.5 35 Wr ps —2 —1 6.5
ND aire) s). 01101 >: c0\%e, 5 2 265 85 84 31.5 169 5 8 11
Average (5)...... 255 82 87.5 39.5 170 5 6 10.5
EUROPEAN, FEMALE.
Right Arm.
MELE ee yey dios. crove Sie GO! ver kee 90 37 183 —2.5 20 13
ear ctelsis) a) ¢le} oraneep ae 82 88 44 175 —-4 af 10
Average (2). 225 79.5 89 40.5 179 —3.25 1.75 Hales
Left Arm.
17/55 Ae SA Coen 220 82 87 40 188 3 6 11
OTe aisle a «i Sistas ais 223 83 85.5 40 167 3) 16 Thi bis;
Average (2)...... 221 82.5 86.5 40 Lito Lr 6.5 11.3

404 On the Angle of the Elbow

BIBLIOGRAPHY.

BRAUNE AND KyRKLUND.—Archivy fiir Anatomie, 1879.

BrucKE.—Schonheit und Fehler der menschl. Gestalt. Wien, 1873.

DUrer.—Von der menschl. Proportionen. Niirnberg, 1527.

Fick.—Handbuch der Gelenke, 1904.

FritscH.—Verhandl. der Berlin. Anthropol. Gesellsch., 1895.

FritscH vu. HAarLAss.—Die Gestalt des Menschens, 1879.

Hay.—The Geometric Beauty of the Human Figure Defined, 1851.

HeNKE.—Anatomie der Gelenke, 1863.

Die Menschen des Michel Angelo im Vergleich mit der Antike. Rostock,
1892.

HULTKRANTZ.—Das Ellenbogengelenk. Jena, 1897.

KoLLMANN.—Plastische Anatomie. Leipzig, 1886.

LANGER.—Anatomie des Menschen, 1865.

Anatomie der a4usseren Formen des menschl. Kérpes. Wien, 1884.

MEISSNER.—Zeitsch. fiir ration. Medicin, 1857.

MeERKEL.—Handbuch der Topographischen Anatomie, 1896.

MEyYeER.—Statistik u. Mechanik des menschl. Knochengeriistes, 1873.

Parson.—Jour. of Anat. and Physiol., 1900.

Porter.—Jour. of Anat. and Physiol, 1895

QUETELET.—Des proportiones du corps humain. Bull. de l’acad. royal des

scien., Belg., xv.

Ricuer.—Anatomie artistique, 1890.

Canon des proportiones du corps humain, 1893.

SARGENT.—Scribner’s Magazine, XIV, p. 130, 1893.

Scuapow.—Polyklet, oder die Maszen des Menschen nach dem Geschlechte, ete.

Berlin, 1834.

Scumipt.—Proportionslehre des menschlichen Korpers. Tiibingen, 1882.

Scumip.—Arch. ftir Anthropologie, 1873.

Stratz.—Die Schoénheit des Weiblichen K6rpers. Stuttgart, 1900.

WELCKER.—Archiv fiir Physiologie, 1875.

ZAHN.—Bau und Mechanik d. Ellenbogengelenkes. Wiirzburg, 1862.

ZEISING.—Neue Lehre von den Propositionen des menschl. K6rpers, 1854.
Die Unterschiede in den Proportionen der Racentypen. Archiv ftir
Heilkunde, 1856.

Ueber die Metamorphosen in den Verhaltnisse der mensch]. Gestalt
von der Geburt bis zur Vollendung des Langenwachstums. Ver-
handl. d. K. Leopold. Carolin. Akad. d. Naturforscher, XXVI.

A STUDY OF THE LOCATION AND ARRANGEMENT OF THE
GIANT CELLS IN THE CORTEX OF THE RIGHT HEMI-
SPHERE OF THE BONNET MONKEY (MACACUS SINI-
GUS).

BY
E. LINDON MELLUS.
From the Anatomical Laboratory of the Johns Hopkins University.

WITH 3 FIGURES.

In view of the recent experimental results regarding the representation
of movements in the cerebral cortex, a more exact study of the distri-
bution of the large pyramidal cells as seen in microscopic sections has
seemed to me desirable. The following paper deals with the distribu-
tion in the bonnet monkey.

The right hemisphere of a healthy adult monkey (Macacus sinicus)
hardened in Miiller’s fluid, dehydrated and imbedded in celloidin, was
eut in horizontal sections 50 microns thick. The sections were num-
bered from below upward, and stained first by Pal’s modification of
Weigert’s hematoxylin method, counter-stained with ~carmine and
mounted in balsam. ‘This double stain has the advantage of accentuat-
ing the contrast between cells and fibers and facilitating the study of
the relation of the cells to the various fiber tracts.

A careful study of these serial sections reveals the following arrange-
ment of the Betz (giant) cells in the motor cortex and, if confirmed by
the study of their disposition in other specimens, may lead to some
modification of the present ideas in relation to the extent of the so-called
“motor areas of the cortex.

On the external surface of the hemisphere the lowest point at which
any giant cells are found corresponds to the lower extremity of the
fissure of Rolando. A few scattered cells are found here upon the an-
terior lip, but quite in the depth of the fissure. From this point upward
they gradually increase in number and at a point about 0.5 mm. higher
up there is a small group of giant cells between the corona radiata of
the ascending frontal convolution and the surface, but the majority of
these cells is still within the fissure contiguous to what may be called
the posterior aspect of the corona radiata of the ascending frontal con-

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

406 Giant Cells of the Bonnet Monkey

volution. Another small group of giant cells appears in a corresponding
position, after an interval of 10 sections (0.5 mm. higher on the sur-
face) contiguous to the external aspect of the corona radiata of the
ascending frontal convolution. This arrangement of the giant cells in
groups does not obtain within the fissure of Rolando, but they appear
here as a continuous layer in gradually increasing numbers from below
upward. Nor is there any further appearance of grouping of the giant
cells on the external aspect of the corona radiata of the ascending
frontal convolution, but from this point upward they extend farther
and farther forward until in the level of the anterior limb of the frontal
suleus they cover the entire antero-posterior extent of the external aspect
of the corona radiata of the ascending frontal convolution. At a slightly
higher level the giant cells entirely envelope this process of the corona
radiata; that is, they are present upon its posterior, external and
anterior aspects, and as we reach still higher levels extend a short dis-
tance forward in contiguity with the external surface of the corona
radiata of the frontal lobe (Fig. 3).

This arrangement is maintained throughout the remainder of the
upward extension of the corona radiata, which, in the monkey, cor-
responds to the cortex of the ascending frontal and the posterior portion
of the superior frontal convolution (Fig. 1).

This distribution of the giant cells upon the external surface of the
brain is by no means uniform. They are most numerous within the
suleus of Rolando and in that portion of the cortex covering the ascend-
ing frontal convolution, while in the cortex of the superior frontal they
are more scattered. Within the sulcus of Rolando they extend to the
base of the sulcus, but are confined entirely to the anterior lip; that is,
they nowhere pass beneath the base of the fissure to the parietal lip.

There are two small groups of large cells in the cortex of the ascend-
ing parietal convolution: one just above the lower extremity of the
intra-parietal fissure appearing in fourteen consecutive sections, the
cells diminishing in number and size from below upward. These cells
extend into the intra-parietal fissure but not into the fissure of Rolando.
There is another small group of large cells in the cortex of the upper
extremity of the ascending parietal convolution, only present in four
sections. With the exception of a few very large cells within the intra-
parietal fissure, the cells described in the cortex of the ascending
parietal convolution are much smaller than the majority of the giant
cells anterior to the fissure of Rolando. Those anterior to Rolando
measure from 20 to 60 microns in length by from 10 to 40 microns in

E. Lindon Mellus 407

Meee

DISTRIBUTION OF GIANT CELLS (BETZ) IN CORTEX OF BRAIN (MELLUS).

32

408 Giant Cells of the Bonnet Monkey

breadth, while those in the cortex of the ascending parietal convolution,
with the exception noted, are from 15 to 22 microns long by 12 to 18
microns wide.

On the mesial surface of the brain the superior border of the cortical
area in which the giant cells are found corresponds exactly with the
superior border of that upon the external surface; that is, the anterior
and posterior borders of this area may be followed directly over from
the external convex surface of the brain upon the mesial surface. The
area occupied on the mesial surface is a somewhat irregular triangle
with its apex corresponding to the upper extremity of the fissure of
Rolando and its base directed toward the frontal pole. The giant cells
are relatively more numerous upon the mesial surface. Instead of
being arranged in a single layer, as upon the external surface, they are
more irregularly scattered about in groups of several superimposed
layers. Cells of the larger diameters are also relatively more numerous
on the mesial than on the external surface. The area extends downward
to the calloso-marginal sulcus about 8 mm. below the crest of the hemi-
sphere. The cells in the lower portion of this area are less numerous
and rather smaller than elsewhere on the mesial surface.

EXPLANATION OF FIGURES.
Note that, for the sake of comparison, Fig. 3 is placed above Fig. 1.

Fig. 1. External surface of right hemisphere; R. Fissure of Rolando; S.
Fissure of Sylvius; F. Frontal pole; O. Occipital pole; AB. Plane of section
of Fig. 3. Area of distribution of giant cells is striated.

Fic. 2. Mesial surface of right hemisphere. C. Calloso-Marginal sulcus;
F. Frontal pole; O. Occipital pole; AB. Plane of section of Fig. 3. Area of
distribution of giant cells is striated.

Fig. 3. Horizontal section at line AB, Figs. 1 and 2, showing-arrangement
of giant cells at that level. R. Fissure of Rolando; F. Frontal pole; 0. Oc-
cipital pole.

A THREE WEEKS’ HUMAN EMBRYO, WITH ESPECIAL
REFERENCE TO THE BRAIN AND THE NEPHRIC
SYSTEM.

BY

SUSANNA PHELPS GAGE,
Embryologic Laboratory, Cornell University.

WITH 5 PLATES.

The specimen under consideration was loaned to me sometime ago by
Dr. Mall for the investigation of the evidence of segmentation in the
brain of the higher mammals.

While the study of this embryo is of necessity incomplete in the present
state of knowledge, certain points were shown so clearly by it, that they
are here presented.

These points are:—For the central nervous system (a) the approxi-
mate location of the morphologic front of the brain, and (b) the location
of folds and lobules of the fore- and hind-brain which may have segmental
significance; for the nephric system, (a) the presence of a pronephros
and (b) a generalized state of the mesonephros, illustrating well-known
conditions in the lower as well as in the higher forms.

Other facts which agree with or differ slightly from those generally
accepted are presented incidentally.

The specimen is No. 148 of the Johns Hopkins University Medical
School Collection gathered by Dr. Mall; sectioned under his direction,
and most generously opened for inspection to students. In the catalogue
of this collection * is found a list of nine papers which use this specimen
to illustrate special points. These will be referred to under the proper
headings.

The embryologic collections of Cornell University and of the Harvaré
and Johns Hopkins Medical Schools were also freely placed at my com-
mand for study and comparison.

1Mall, F. P., Johns Hopkins Hosp. Bull., XIV, 1903. Catalogue of the
collection of’ human embryos in the Anatomical Laboratory of the Johns
Hopkins University. Baltimore, 1904.

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

410 A Three Weeks’ Human Embryo

MopELs AND DRAWINGS.

A model of the entire specimen was made by the Born method, modi-
fied as described by Bardeen.*

As the sections were not equally perfect, the best of each group of
three was drawn at a magnification of 6624. The sections are 10» thick,
and as but one-third of them was drawn, each wax plate of which the
model was made, represents the thickness of three sections, or 30u. The
reconstruction was laid out on an outline enlarged from the photograph
made by Dr. Mall.“ The completed model has a length of nearly 300 mm.
A comparison of the model with the photograph taken before embedding
shows that in the process, there had been a shrinkage of 10 to 12%.

The model was used as a basis for the drawings, the contour line of
every fifteenth or thirtieth section being indicated. In this way the
different levels can be correlated throughout the set of drawings which
represent the model sectioned at different planes. The drawings in each
case, were corrected by repeated comparison with the specimen.

The form of the ccelom was obtained by building up the parts cut from
the entire model.

Models at a much greater magnification were made of various details,
as the fore-brain, the mesonephros and mesonephric tubules.

A part of the details are illustrated by reproductions of photographs
of sections of the embryo.

EXTERNAL Form.

This embryo was photographed by Dr. Mall at three times enlarge-
ment. The photograph, published in an article by him,* shows the speci-
men from the right side, lying upon the opened chorion and is described
as “* An embryo three weeks old.” He also mentions that the umbilicus
is at the right instead of the left, as is the more usual position. In
another paper, he shows an enlarged outline drawing.’

The general form is shown in Figs. 1 and 1a, the body being about once
and a half as long as the head and at an angle of about 65° with it. The
face is featureless except for the wide slit-like mouth (Fig. 2). At the
corner of the mouth (Figs. 2, 5, 6) is a depression with a thin gill-cleft-
like membrane. The four gill-clefts are irregularly spaced giving the

? Born, G., Morph. Jahrb., II, 1876; Arch. f. mikr. Anat., XXII, 1883.

° Bardeen, C. R., Johns Hopkins Hosp. Bull., XII, 1901.

‘Mall, F. P., Welch Anniversary Volume, 1900; also Johns Hopkins Hosp.
Reports, IX, 1900.

°Mall, F. P., Johns Hopkins Hosp. Bull., XII, 1901.

Susanna Phelps Gage 411

second arch great prominence while the third ‘and fourth are in a common
depression, the sinus precervicalis. On the right the gill-clefts are less
crowded and the sinus not quite so deep. The first gill-cleft is wider at
its dorsal end indicating probably a beginning of the external meatus.

The heart is prominent. The yolk sac is extensive turning to the right
instead of the left as is more usual, thus making the umbilicus more
apparent in a view from the right than from the left side (Figs. 5, 6).

The limb-buds are remarkably prominent in comparison with other
specimens supposed to be of the same age.

Epithelial thickenings——In Fig. 1 are mapped out the regions having
a thickened epithelium. Details of these thickenings as shown by indi-
vidual sections are found in many of the figures, as the arm (Fig. 10,
11), the leg (Fig. 5), anal plate (Fig. 1), gill-arches (Figs. 1, 11),
mouth (Fig. 5), olfactory region (Fig. 2), lens (Figs. 1, 2, 5, 8, 16),
and the neuropore (Figs. 1-8, 16). In Fig. 1 the thickness of the
epithelium corresponds with the density of the dots. The portions left
white represent one layer of cells which become flattened over the heart
and near the dorsimesal line. Over the entire oblongata this layer of
flattened epithelium coalesces with the wide, thin roof of the brain (Fig.
4). Mall® calls attention to the thickening of the neuropore in this
specimen. Especially noticeable is the H-shape of the thickening over
the olfactory and cerebral regions (Fig. 2). The continuity of leg-bud
thickening with that of the anal plate (Fig. 1) is comparable to the con-
_ dition in amphibian embryos.

AGE OF THE SPECIMEN.

The exact age of this specimen cannot be determined any more closely
than has been done by Dr. Mall, who considers that it is of about twenty-
one days development. Its relative age is somewhat important since it
presents certain features not hitherto fully described. It is necessary to
determine whether it may be a transitional stage or a pathologic or
arrested condition which is under observation.

Though it presents one feature (the small number of thoracic myo-
tomes) not universal, and others (the umbilicus turning to the right)
which are not common, and still others rare or not previously observed,
still on the whole, it so well fits into the series which has been described
by various authorities or examined by the author that the weight of evi-
dence seems to point in the direction that this is in general a normal
transitional condition, though possibly somewhat exaggerated in a few
particulars or retarded in others.

412 A Three Weeks’ Human Embryo

Comparing this specimen with the His models and Atlas,’ it is seen
that there are some resemblances to his specimen Lr estimated at twenty-
one days, but in external appearance, it seems to he between a and R
estimated by him to be from 21 to 25 days.

The specimens No. 164, 209, 148, and 80 in the Johns Hopkins collec-
tion show that in certain features of development they form a progressive
series (see Nephric System, below). The formula devised by Mall,’ for
determining the age (days = 1/100 x length of embryo), approximately
worked out gives the following:

TABLE I.
INN irasieenctereieiytie ronoeucie 209 164 148 80
Wenge theiny mim veyetaerenese i 3 + 3.5 4.3 5 or 4.5
WAN ROUOUN SI. cea gdogs sou0se 19 28 to 29
DAY Stott cian ers ota ees IG SF 19 — 21— 22 +

THe ALIMENTARY CANAL AND ITS APPENDAGES.

Mouth.—The mouth (Figs. 1-+) is simply a cleft between the fore-
brain region and the mandible, extending laterally to the just forming
maxilla and at its tips having a thin membranous portion (Fig. 5) in
section strongly resembling the membranous tips of the gill-pouches.

No remnant of the oral plate was found. The position of the dorsal
limit of the ectodermic portion of the oral plate is indicated by the hypo-
physis (Figs. 9, 3,4). The latter is a bi-lobed, widely opened pouch in
contact with the infundibulum of the brain and partially surrounding it.

Pharynx.—No signs of Sessel’s pocket, the most cephalic of the ento-
dermic structures of the pharynx could be found since the sections are not
favorable in this region.

The lateral or lingual folds, described by Kallius* as forming the first
rudiment of the tongue, are represented in the mesal view (Fig. 4), as a
ridge lying at the side of a mesal pit. The tuberculum impar has not yet
arisen in front of the median thyroid. The latter body with the pit from
which it arises is present, but the tubular connection between gland and
pit has nearly disappeared (Figs. 3, 4).

The floor of the pharynx is partially exposed in Fig. 11, showing lateral

®‘ His, W., Anatomie Menschlicher Embryonen. Text and Atlas. Leipzig,
1880-1885.

™Mall, F. P., Age of Human Embryos. Ref. Handb. Med. Sci., 2d Ed., III,
1901.

®Kallius, E., Anat. Gesell., Verhand, 1903.

Susanna Phelps Gage 413

extensions into the four gill pouches, each of which ends in a thin plate.
The location of the membranous tips of the gill pouches is indicated upon
the mesal view (Fig. 4). The gill pouches have also small ventrally
projecting blind pouches ending in a somewhat thickened epithelium,
beginnings or protons of the thymus and lateral thyroid bodies.

The larynx is represented by a slight depression, on the ventro-lateral
borders of which is a pair of minute pouches (Fig. 11) similar in appear-
ance to the ventral processes from the gill pouches. From their general
relations it seems probable that they represent the rudiments of a 5th
pair of gill pouches.

A tubular projection in the roof of the pharynx over the entrance to
the esophagus is apparent in several sections. Killian ° identified a mesal
pouch occupying a position just caudal of the pharyngeal tonsils as the
Bursa pharyngea of Meyer. He traced this back to the 11th week of
embryonic development. The specimens of the 7th to the 12th week in
Cornell University make it apparent that the pouch seen in Figs. 3, 4,
at the left of the abbreviation ch, is this same Bursa.

Trachea.—The trachea (half a mm. long) ends in a pair of widely-
spreading bronchi, each with a single slightly enlarged end, the lung-bud
(Figs. 3, 11), surrounded by an enlargement in the mesentery.

Alimentary Canal.—The esophagus is small and practically closed
through part of its length. It extends to section 107, where it merges
gradually into the stomach which shows a spindle-shaped enlargement
increasing at its caudal end and turning to the left (Figs. 3, 11).. The
lesser peritoneal cavity pushing the stomach to the left (Figs. 2, 17) is
shown at its opening into the ccelom (Fig. 11, crossed by line pointing
to mesentery ).

The stomach narrows again as it merges into the duodenum (Fig. 10).
A minute dorsal enlargement of the duodenum is the rudiment of the
pancreas (Fig. 3). On the ventral side is found the short bile duct
(Figs. 3, 6). As somewhat diagrammatically shown in Figs. 6 and 10,
the trabecule of the liver are in a great sinusoid along the path of the
vitelline vein. In Fig. 11 both lobes of the liver are shown from the
dorsal side, and in Fig. 6 there is a section through it at the level of the
bile duct. The duodenum is enlarged at this point of union with the bile
duct, and continues as a tube to its wide (260) union with the vitelline
sac (Fig. 5).

The caudal intestine within its free dorsal mesentery (Fig. 5) con-
tinues from the vitelline sac in a curve following the back. It enlarges

® Killian, G., Morph. Jahrbuch, XIV, 1888.

414 A Three Weeks’ Human Embryo

again as it leaves the free mesentery and curves around the end of the
cceelom, and unites with the allantois to form the wide cloaca which is
joined by the Wolffian duct (Figs. 5, 17).

The cloaca is closed by the anal membrane, a thickened plate with only
a slight indentation (Fig. 17).

Allantois—The allantois extends as a narrow tube along the abdomi-
nal stalk and bending over the caudal end of the ecelom (Fig. 5), enlarges
to form the bladder as it unites with the cloaca (Fig. 17).

From the standpoint of the development of the diaphragm Mall ”*
gives the following description of the organs of the thoracic and abdomi-
nal cavities :—“ Sections of the embryo 4.3 mm. long (No. 148) show
the liver sprouts growing in all directions through the septum trans-
versum, encircling and ramifying through the omphalo-mesenteric veins,
making a condition slightly in advance of that in His’s embryo Zr. The
sections of this embryo show clearly, that the heart, lungs, liver, and
lower peritoneal cavity arise in tissues surrounded by that portion of the
ceelom extending into the head in Embryo XII. ... The lungs arise
when the pericardial ccelom goes over into the pleural, 2. e., high up in
the region of the head. Immediately on the dorsal side of them is the
beginning of the lesser peritoneal cavity, and the intestinal tube struck
in this section is the stomach. All these structures he on the cephalic
side of the first cervical myotome. Projecting into the peritoneal ccelom,
encircling and penetrating the omphalo-mesenteric veins, are the projec-
tions of the liver. The two lobes reach from the tip of the lungs and
the foramen of Winslow to the point where the entodermal cells of the
liver arise from the alimentary canal, or in this case, the duodenum.
The lobes of the liver lie entirely within the canals of the ccelom on either
side of the head. The caudal ends of these celom canals have migrated
from opposite the second cervical myotome in Embryo XII, to opposite
the second thoracic myotome in Embryo 148. It has moved toward the
tail, while the cephalic end of the canal, the pericardial ccelom, has been
kinked over to correspond with the growth of the heart. ... We have
in the embryo the necessary stage to locate the organs which arise in the
neighborhood of the septum transversum, as well as to give the fate of
the ccelom in their immediate neighborhood.”

The points of the above description illustrated by the figures are:
the penetration of the septum transversum by liver substance and blood-
vessels (Fig. 10); the continuity of pericardial and abdominal ccelom
(Figs. 2, 11); the position of lung, stomach, liver, and it may also be

” Mall, F. P., Johns Hopkins Hosp. Bull., XII, 1901.

Susanna Phelps Gage 415

said, duodenum, in the ccelom canals which lie dorsad of the septum trans-
versum (Fig. 3); the caudal end of the ccelom canals opposite the 2d or
perhaps the 3d thoracic myotome (13th myotome).

It seems to me that the most cephalic portion of the pericardial ccelom
might be said to be that surrounding the bulbus arteriosus, in which case
it would be difficult to decide that it had retreated so far from its origi-
nal position opposite the ear as Mall has mentioned. ;

The specimen and the model show more clearly than any of the draw-
ings, the caudal traction of the organs which has taken place while their
original cephalic attachment can still be traced. The furrows of the
mesentery (Figs. 3, 6, 10, 11) show this more clearly than the varying
caliber which indicates the division of the entodermal tube into organs.
The furrow, for instance, separating duodenum and yolk sac extends
cephalad to the 8th myotome (6th cervical), and the attachments of all
the entodermal organs cephalad of it are spread out between this point
and the 1st occipital myotome, that is in the neck. From this, it hardly
seems correct to say that lung and stomach are really so far cephalad as
mentioned by Mall (7. e. in the head).

MESODERM.

General—The general appearance of the mesodermic tissue is shown
in Plate V, being entirely cellular. The condensations in the nephric
region are described below. Other condensation is seen in the limbs, but
without clear differentiation into separate masses. Condensation also
occurs at the side of the pharynx and markedly so in the floor. This
continues uninterruptedly into the thick mesentery surrounding lungs,
esophagus, stomach, and intestines (Figs. 2, 3, 6, 10, 11), and in the
region of the cloaca joins the condensation around the Wolffian duct and
continues into and fills the leg-bud.

A distinct spindle-shaped condensation of cells occurs ventrad of the
eye (Figs. 7, 8).

Myotomes.—Bardeen and Lewis” give the following description:
“Length, neck-breach, 4.3 mm.; age about three weeks. ... Though
more advanced in development than Lr (His), but twenty-seven myo-
tomes are present (20, 8c, 10t, 51, 2s). This has been determined by
careful counting of the myotomes in serial sections of the embryo. The
base of the arm-bud appears to lie opposite the seventh to the eleventh
myotomes. It is, therefore, probable that two occipital myotomes are
present. But nine myotomes lie in the area between the arm-bud and the

11 Bardeen, C. R., and Lewis, W. H., Amer. Jour. Anat., I, 1901.

416 A Three Weeks’ Human Embryo

leg-bud. The base of the latter les opposite the 21st to the 25th or 26th
myotomes. If two myotomes be considered occipital myotomes, the leg,
in this instance, lies two segments nearer the head than usual. It is
therefore probable that this embryo has an unusually short body-wall.”

Twenty-nine myotomes were counted and modelled on the left side,
and twenty-eight on the right, the discrepancy occurring in the caudal
region. This count in general, agrees with Bardeen and Lewis; 2 occi-
pital, 8 cervical, 10 thoracic, 5 lumbar, and 3 or 4 instead of 2 sacral,
as noted by them.

The arm-pads have even a longer cephalo-caudal enlargement than
noted by Bardeen and Lewis, and cover the 7th to the 13th myotomes,
thus leaving only 7 complete myotomes between the arm- and leg-buds.

Many of the myotomes do not model in the regular forms usually
shown. The first occipitals are small and imperfect (Figs. 14,15). The
3d, 4th, and 5th are dorsally composed of two distinct, hollow horns
(Fig. 14), which merge ventrally into a common cavity. More ventrally
they become solid and are marked across the middle by a band of cells
(Fig. 15, at left). The lumen is not large until the 11th (Fig. 11),
from which point until near the end of the series it is a marked feature.
The largest and most regular myotomes are opposite the legs, the most
irregular among the cervical.

In several myotomes (Figs. 5, 11), careful examination could detect
no limitations between a certain area at their ventral end and the con-
densed mesoderm of the limb-bud, in fact in these cases, the appearance
would indicate the origin of limb tissue from myotomes.

Evidences of segmentation in the mesoderm are also seen cephalad of
the two clearly recognizable, occipital myotomes 1 and 2, and immediately
in line with them. Cephalad of the Ist is a minute area with apparently
the identical structure of a myotome, including the familiar corrugation
of the epidermis (Fig. 14). Still more cephalad are two other lesser
condensations and corresponding epidermal corrugations (Fig. 15). That
is, there are indications of three more occipital myotomes than are dis-
tinctly figured.

Sclerotomes.—On their mesal aspect the mesodermic tissue is shrunken
away from the myotomes in loops (Fig. 15), and the tissue shows a con-
densation (sclerotome) corresponding with each loop down to the 18th
myotome. In the cephalic region a slight condensation occurs ventrad
of the notochord. That is, these last two points indicate a slight differ-
entiation and segmentation of sclerotogenous tissue. In some sections,
the continutiy of these sclerotomes with the myotomes can be seen.

Susanna Phelps Gage 417

Dermatomes.—This specimen by itself does not throw much light on
the question as to the fate of the outer wall of the myotome, 7. e., whether
it is in reality a dermatome or not.

VASCULAR SYSTEM.

Heart.—In Figs. 1, 3, and 5 the position of the heart is seen, and in
Fig 12, a ventral view showing a compact, somewhat square form, remind-
ing one of the His’ model from a 10 mm. (4 week) specimen rather than
those from younger specimens. The greater length of the right auricular
portion is similar to the His model of a 5 weeks’ embryo. A very young
embryo, modeled and presented by J. L. Bremer” at the American Asso-
ciation of Anatomists, in December, 1904, shows an exaggeratedly long
right auricle. In external form, the heart of this specimen resembles
older, rather than younger stages, but the internal relations accord with
the descriptions given for the 3 weeks’ stage. That is, the tubular heart
is already dividing into right and left (Figs. 2, 6); the entrance of the
sinus venosus is decidedly to the right (Fig. 2, left of Fig.), the con-
nection of the auricle with the ventricle is through a narrowed tube, the
auriculo-ventricular canal (Fig. 10), entering the ventricle at the left
(Fig. 6), while the bulbus arteriosus makes its exit at the right (Figs.
12, 6), turns sharply cephalad and extends along the ventral aspect of
the heart.

The walls of the heart contain only undifferentiated muscular tissue.
In parts the endothelial tube is closely applied to the walls; in parts,
notably along the path of the auriculo-ventricular canal (Fig. 10), the
ventral end of the right ventricle (Fig. 6) and of the entire bulbus (Figs.
10, 2), the endothelial tube is connected with the outer wall by only a
‘delicate mesh-work of tissue.

Arteries—The endothelial tube of the bulbus plunges into the floor of
the pharynx, expands into a wide sinus, giving off on each side near the
middle line, a small branch which divides into the 1st and 2d gill arches
and a mesh-work of capillaries supplying the mandible. The Ist and 2d
arches join with the dorsal aorta by a very slender capillary connection.
‘The 3d and 4th gill arches (Fig. 11) are given off from the side of the
sinus and unite dorsally to form the main portion of the dorsal aorta.
‘The 5th gill arch is given off caudally near the middle line, divides into
-capillary branches, supplying the larynx and trachea (5th and 6th
arches), reaching the dorsal aorta by very small branches. Each dorsal
-aorta sends forward a branch which can be traced in the roof of the

2 Bremer, J L., Amer. Jour. Anat., IV, No. 2, p. VIII.

418 A Three Weeks’ Human Embryo

pharynx nearly to the hypophysis. This portion of the vascular system
resembles closely others described of the same age, but reminds one, in
the extreme difference in size of the aortic arches, of the condition de-
scribed by Miss Lehmann,” where in lower mammals the six arches are
not all complete at any one time.

The two dorsal aorte unite (Fig. 3) near the cephalic border of the
liver and the caudal margin of the arms, giving off just before their union
the brachial arteries and soon after, the small: vitelline or omphalo-
mesenteric artery (Figs. 10,17). Branches could occasionally be traced
to the tubules of the mesonephros (Fig. 6). In contrast to the other
arteries the umbilicals are large and in the body stalk anastomose across
the middle (Fig. 5), but continue as a pair for some distance in the body
stalk.

Veins.—The veins are remarkable for the great variation in caliber.
The much branched jugular (precardinal) can be traced from a point
lying between the eyes and cerebrum (Fig. 7), keeping near the surface
(Fig. 8), sending branches through, and then passing mesad of the
Gasserian ganglion (Fig. 9), laterad of the ear vesicle and the ganglia
of the 7th and 8th nerves (Fig. 11), and by a breaking-up in the 9th and
10th ganglia, comes to lie mesad of the 10th ganglion as it passes over
into the vagus nerve, then unites with the cardinal (Fig. 11, at left) and
the umbilicals (Fig. 2, at left) to form the ducts of Cuvier. The course
of the jugular in the head seems to illustrate one phase of the change of
position of the bloodvessels with relation to the nerves as demonstrated
by Dr. Mall” in a recent article.

The ducts of Cuvier unite across the middle to form the sinus venosus ;
and this connects by a small opening with the great sinusoid “ of the liver
formed in the course of and by the union of the vitelline veins (omphalo-
mesenteric) (Fig. 10).

The umbilical veins coming from the body stalk are joined by veins
from the legs, become enlarged and break up in the body wall into a great
sinus (Figs. 5, 6, 10, 11), and finally as they enter the duct of Cuvier
become so small as to be scarcely traceable (Fig. 2, at left, and Fig. 10,
at right).

THE NEPHRIC SYSTEM.

General—The nephric system (Fig. 17), as is usual at this age, con-
sists of a pair of Wolffian ridges extending along the back from the arm-

13 Lehmann, Harriet, Anat. Anz., XXVI, 1905.
14 Mall, F. P., Amer. Jour. Anat., IV, 1904.
18 Minot, C., Boston Soc. Nat. Hist., Proc. X XIX, 1900.

Susanna Phelps Gage 419

pad nearly through the region of the leg-pad, and of Wolffian ducts
opening into the cloaca. Sections of the ridges and their contained
structures are shown at different levels, beginning cephalad, in Figs. 11,
18, 10, 19, 6, 20, and 5.

There is no definite thickening in the ccelomic epithelium or the meso-
derm which can be designated as a rudiment of the genital ridge or of the
Miillerian duct.

In the ridges, the Wolffian duct is traceable as indicated by dots in
Fig. 17 from the 1st mesonephric tubule along the lateral border of the
ridge and extending beyond the ceelom in a curve (Fig. 5), to open into
the lateral border of the cloaca (Fig. 17). Extending still farther
laterad along the Wolffian ridge is the cardinal vein, the diameter of which
varies greatly in different sections (Figs. 18, 19).

Adrenal.—Near the cephalic end of the Wolffian ridge (Figs. 17, 11,
suprarenal), on the mesenteric border are slight folds in the epithelium
which remind one of the structures in the cephalic part of the meso-
nephros found by Aichel * in the rabbit to be associated with the develop-
ment of the adrenal or suprarenal body.

Pronephros.—A wide open funnel (Figs. 17, 18), opening to the
celom and connecting with a small, blind tube extending cephalad for a
few sections, is here called a pronephric tubule on account of its position
in the cephalic portion of the Wolffian ridge. It has no connection
whatever with the Wolffian duct and is separated by a marked interval
from the beginning of that duct. This is apparently only a further
retrogression from the condition found by MacCallum” in which a
separated portion of duct extends opposite the 6th, 7th, and 8th myotomes
of an embryo 3.5 mm. long. MacCallum calls attention to the possibility
that this separated tube with its cephalic opening through a funnel to the
ecelom may be a pronephric remnant. In an older embryo (5 mm. long)
he finds such a disconnected remnant with a single tubule not opening
to the ccelom.

Tandler“ says that in human embryos from 5 to 20 mm. he found
pronephric tubules in communication with the ccelom eight times. These
were at the level of the 5th to 6th segment.

In Embryo 148, the isolated pronephrie tubule lies opposite the 11th
myotome. Whether the difference in position relative to the myotomes
from those reported by MacCallum and Tandler, indicates that this is not

1% Aichel, Otto, Arch. f. mikr. Anat., LVI, 1900.
7 MacCallum, J. B., Amer. Jour. Anat., I, 1902.
1% Tandler, J., Centralbl. f. Physiol., XVIII, 1904.

420 A Three Weeks’ Human Embryo

a true pronephric remnant, or whether a caudal shifting of the Wolffian
ridges has taken place, or whether there is an increase in the number of
myotomes in the occipito-cervical region of No. 148 cannot be determined.
But at present, it is unmistakable that in three embryos of 214 to 3 weeks
in development (Nos. 164, 148, and 80 of the Mall collection) there
exists a structure which has many resemblances to a pronephros.

In a 3 mm. human embryo, Janosik” found that there is a distinct
glomerulus protruding into the ccelomic cavity. The position of this
glomerulus is similar to the tubules found in the embryos of the Mall
collection.

Tandler “ found one glomerulus in eight cases. It seems, therefore,
that in spite of the apparent discrepancy of position, there should be no
hesitation in expressing the homology of the structure under considera-
tion with the pronephros of lower forms.

Mesonephros——A preliminary report on the generalized constitution
of the mesonephros of Embryo No. 148 was presented with models at the
December, 1903, meeting of the American Anatomists.” The main facts
noted’ were, that in the Wolffian ridges an open pronephric tubule occurs
on éach side followed by two groups of mesonephric tubules, the first
group of eight being of the usual embryonic type, S-shaped, with Bow-
man’s capsule, rudimentary glomerulus, and union with the Wolffian
duct; the second group of eleven or twelve, consisting of tubules, none
of them uniting with the Wolffian duct and varying from solid aggrega-
tions of cells to hollow vesicles and finally to vesicles open to the ccelom.
The two sides present the same general arrangement although the tubules
do not form symmetrical pairs in precisely the same stage of development
(ef, Higs 17-19).

MecMurrich”™ says, page 391, “ One of the characteristics of the mam-
malian mesonephros is that it possesses no nephrostomes.” Minot * in
his Embryology, page 237, says, “In all amniota the nephrostomes all
become completely separated from both the myotomes and peritoneum
throughout the region of the Wolffian body, except that possibly in a few
anterior segments, the connection with the peritoneum is retained as is
suggested by Sedgwick’s observations (in birds).”

In an embryo 4.25 mm. in length, Meyer ~ showed that in the cephalic

12 Janosik, J.) Arch: f. mikr, Anat., X&GxS ssi:

*» Gage, Susanna P., Amer. Jour. Anat., III, 1904, p. VI.

2 McMurrich, J. P., The Development of the Human Body. Phila., 1902.
Minot, C. S.. Human Embryology. New York, 1892.

7 Meyer, H., Arch. f. mikr. Anat., XXXVI, 1890.

Susanna Phelps Gage 421

half of the Wolffian ridge, certain of the tubules are directly continuous
with the epithelium covering the ridge, but there are no hollow tubules
opening into the ccelom. In the caudal half the unsegmented blastema
which he considers the fore-runner of the tubules, was not connected with
the epithelium. From the description, this is probably a less advanced
specimen than No. 148.4

Janosik’ found in an embryo, 3 mm. long when fresh, that in the
cephalic part of the Wolffian ridge there are a Wolffian duct and a number
of independent tubules; the concentrated blastema serially following the
tubules, is not segmented, but connects with the ecelomic epithelium at
intervals.

MacCallum,” the latest special investigator on the subject, shows in a
3.5 mm. embryo (164 of the Mall collection) that extending from the
10th to the 19th or last formed myotome there are thirteen enlargements
of the Wolffian duct, the 5th to the 8th being considerably elongated, but
showing no glomoruli or Bowman’s capsules. In No. 80 of the same
collection (4.5 to 5 mm. long), there are 17 to 18 tubules with the char-
acteristic S-shaped curve, enlarged Bowman’s capsule, and union with the
Wolffian duet. He showed in this case a close connection of the Wolffian
duct at intervals with the ccelomic epithelium. In none of these did the
tubules open to the ccelom.

It has been my privilege to look over the two specimens described by
MacCallum and also a more recent specimen of the same collection
(No. 209). These with No. 148 form a series which may throw light on
the true development. The following apparently consecutive history is
drawn from a study of these embryos ranging from the 17th to about the
23d day (see above, Age of Specimen), and from comparison with various
embryos of the Cornell University collection as cat, shark, lamprey, ete.

In the mesonephric region of the least developed stage (No. 164), in
respect to the nephric region, the Wolffian duct lying at the dorso-lateral
part of the Wolffian ridge has enlargements, some of them, the 5th to the
7th, quite pronouncedly thorn-like. In the next stage (No. 209) in which
the sections were thick and details difficult to determine there exists a
series of about fourteen rounded tubules with thickened epithelial walls,
and a lumen well-defined or in process of formation. In the ventral
most prominent part of the Wolffian ridge some of these, as the 5th, are
connected by a solid out-growth with the ccelomic epithelium of the most

*a Meyer’s specimen was 2 mm. long after being hardened, thus making the
discrepancy in length less apparent than would be indicated from the
measurement 4.25 mm., taken while fresh.

422 A Three Weeks’ Human Embryo

protruding part of the ridge. At the other extremity of the tubule it
joins the Wolffian duct.

In a cat from which a part of the mesonephros was modeled, the next
stage apparently was found. Some of the centrally located sub-cylindrical
tubules open from the ccelom at the crest of the Wolffian ridge, and at the
other end of the tubule into the Wolffian duct. The general appearance
of these tubules and their connections is perfectly comparable to the con-
dition found in the early shark, the tubules of which were also modeled
and found to differ from those of the cat in the fact that they were more
S-shaped.

No. 148 (Figs. 17-20) furnishes the next stage. It has apparently
in its series of twenty or more tubules a complete recapitulation of the
preceding stages co-existing with the commonly recognized type-form
of tubule.

There are :—Ist, The Wolffian duct in the caudal part, having no con-
nection with the tubules, but presenting a series of distinct enlargements ;
2d, Aggregations of cells with no apparent lumen, and no connection
with the ccelomic epithelium or the duct; 3d, Cavities or vesicles deep in
the Wolffian ridge surrounded by a several-layered epithelium, and having
no connection with either ccelomic epithelium or duct; 4th, Similar
vesicles lying near to the ccelomic epithelium of the crest of the Wolffian
body, connected with that epithelium by (a) a solid string, (b) a narrow
open channel, or (c) a wide open funnel (Fig. 20) ; 5th, In the cephalic
half, one tubule, the 7th, with a slight opening from the Bowman’s
capsule to the ccelomic epithelium; 6th, The remainder (8th, 6th, 5th,
4th, 3d) of the cephalic group have the typical Bowman’s capsule, the
S-shaped tube, and the connection with the Wolffian duct but no connec-
tion with the ceelom (Fig. 19) ; 7th, The 2d tubule having two Bowman’s
capsules, one dorsad of the other, that is, an apparent beginning of the
dorso-vyentral division of the tubules; 8th, The 1st small tubule, appar-
ently consisting only of a Bowman’s capsule, sessile on the beginning of
the Wolffian duct.

The artery and vein supplying the rudimentary glomerulus of these
Bowman’s capsules (Fig. 6) could in a few places be clearly seen. The
specimen not being favorable to the study of blood-vessels, the vascular
supply could not always be made out.

Such a tubule as was found in the cat with both ends open (see above)
did not occur in No. 148. The tubules which had attained connection
with the duct had in this specimen apparently lost connection with the
celomic epithelium. In No. 80 (MacCallum”™) the transitional forms

Susanna Phelps Gage | 423

of tubules described above are all lost. The 17 to 18 tubules are complete
and like Fig. 19.

In brief, there is, in human embryos, of the 17th to the 23d day, the
history of the origin of the tubules from a cellular blastema. This
blastema segments into rounded masses, these masses develop cavities that
first unite with the celomie epithelium, then losing this connection,
unite with the Wolffian duct which already had definite enlargements,
and then finally each tubule forms the thin-walled Bowman’s capsule with
its glomerulus.

The connection which probably existed between the myotomes and the
ccelom through the intermediate cell mass which gives rise to the blastema
forming the tubules, is of an earlier stage than these here considered.
Dr. Minot in the discussion of this paper when presented at the meeting
of the American Anatomists stated that in a very early rabbit in the
Harvard University Medical School collection, such connection can
actually be seen between myotome and nephrotome to the ccelom. I have
recently seen the same in the chick.

Metanephros.—If any trace of the metanephros or true kidney exists
in specimen No. 148, it consists merely of a slightly condensed portion
of the blastema, caudad of the mesonephric region where the Wolffian
duct extends beyond the ccelom (Fig. 17).

CENTRAL NERVOUS SYSTEM.

General.—Most of the figures illustrate some features of the central
nervous system. The general outline follows closely the profile shown in
Fig. 1. The lateral and mesal views are seen in Figs. 3, 4, and with
segments of the model and photographs of the sections together give
better than words, the idea of shape. The form more nearly approaches
that of embryo Lr (4.2 mm. long, with 32 myotomes) shown in His’s”
atlas (Plate IX, Fig. 13), and the model of the same than any other
specimen figured.

The stage of development of eye, olfactory pit, and ear vesicle also put
it in the same class, that is, the nervous system is like, in general features,
this well-known specimen of three weeks. Details which bear upon the
object of the present investigation are not shown by His nor, as far as I
know, by other writers on mammalian brains. Mall’ in an early presen-
tation of this same embryo mentions the neuropore, which in the present
article is fully illustrated and in a way made the starting point for
definite conclusions.

The evidence from human and mammalian as well as immammalian

oo
vv

424 A Three Weeks’ Human Embryo

material has been accumulating in the form of specimens, models, draw-
ings, and notes with preliminary papers,*” until it seems clear that the
statements presented with regard to the central nervous system of this
embryo are not artifacts due to shrinkage, nor abnormalities, but that the
individual characteristics of this specimen may be depended upon to
represent one phase of development. This phase is probably transitory
because nothing exactly lke the neuropore in this specimen has been
found in any other specimen examined, though stages both older and
younger are seen. On account of the clear demonstration of important
facts in transitional stages it seems worth while to record them fully.

The cephalic end of the brain tube and tts relation to a serial order of
parts.—von Baer * in 1828 represented as perfectly obvious the original
cephalic end of the body, including the neural plate, at the point where
the hypophysis arises. By more refined methods, Keibel * in 1889 showed
with apparent conclusiveness, that in the rabbit the neural plate, the
enteron, and the notochord end in an undifferentiated mass of tissue
which is the true cephalic end of the body, and corresponds with the point
indicated by von Baer.

From another view point, the place of final closure of the neuropore has
been considered by some a crucial test for determining the end of the
brain tube. His” found it at the optic recess and found also that just
before final closing, the neuropore is a slit, includes the recessus infundi-
buli, chiasma, recessus opticus and the olfactory lobes, and extends to the
dorsal end of the lamina terminalis. That is, in fact, His really agrees
with von Baer and Keibel as to the original condition of the cleft between
the neural plates.

Kupffer,” in sharks and some other fishes, found that the final closure
of the neural tube occurs between the olfactory lobes at a point which he
calls the lobus olfactorius impar. THerrick*™ in discussing these widely
divergent views concluded that the final closure had no necessary relation
to the morphologic front of the brain but that the recessus infundibuli is
the primitive cephalic end. Herrick, therefore, also goes back to von
Baer’s original location of this point.

Accepting then the von Baer-Keibel observations as correct, I

“von Baer, K. E., Ueber Entwicklungsgeschichte der Thiere. K6nigsberg,
1828.

2 Keibel, F., Arch. f. Anat. u. Phys. Abth., 1889. es

72 His, W., Arch. f. Anat. u. Phys., Anat. Abth., 1892.

7 Kupffer, C., Studien zur vergleichenden Entwicklungsgeschichte des Kop-
fes der Kranioten. Heft 1. Miinchen u. Leipzig, 1893.

2 Herrick, C. L., Jour. Comp. Neur., III, 1893, p. XVI.

Susanna Phelps Gage | 425

believe that at all stages of development, the hypophysis and its corre-
sponding fold of the brain-wall represent the morphologic cephalic end
of the body and of the brain. It logically follows that the cephalic end
is the point at which the dorsimesal and ventrimesal lines meet. The
ventrimesal line is present from the beginning as the middle line between
the neural plates. The dorsimesal line is that in which the original mar-
gins of the neural plates unite in closing the brain tube.

To redetermine the exact location of this point of union of the two
lines, young specimens of both immammalia and mammals, including man
as far as material was available, have been examined step by step from
the open neural plates to the closed neural tube and until adult land-
marks become unmistakable. From these observations reported in 1903 ”
and 1904,” it becomes certain to me that the original margin or dorsi-
meson extends as far as the hypophysial fold of the brain (Figs. 3, 4).
This is irrespective of the exact place where the final closure of the neural
tube takes place. In all the higher forms examined this final closure is
between the eye-stalks, its adult representative being the pre-optic recess.
In torpedo, as Kupffer found in sharks, this point of final closure lies be-
tween the olfactory lobes. But even here, in earlier stages the original
margin of the neural plate extended as in mammals to the hypophysial
region. The difficulties of determining the cephalic end of the tube in
lamprey and other forms having an original solid neural plate are, that
when the cavity does form, it has seemed somewhat uncertain as to the
location of the front of the tube. From my observations it seems that
even in the lamprey the front of the tube can be placed at the hypophysis.
In amphibia the large size of cells and the consequent thickening produce
an obscurity as to the exact formation, but here again, the weight of evi-
dence seems to show that the original cleft between the two sides of the
neural plate extends to the hypophysial region.

The human embryo here under special consideration is the most illumi-
nating of any specimen examined for the purpose of determining the
exact point at which final closure takes place, because that event is delayed
until the surrounding parts are so well developed that identification is
unmistakable. At the point recognized by Mall as the neuropore, lying
between the eye-stalks, a connection of brain and skin tissues exists and
extends through a number of sections (100% or more). In parts the
arrangement of cells indicates that the margins have only recently united
(Fig. 16). Figs. 1-8 show the relations of the neuropore, and the extent

* Gage, Susanna Phelps, Science, N. S., XVII, 1903.
*° Gage, Susanna Phelps, Amer. Jour. Anat., IV, 1904, No. 2, p. VIII.

426 A Three Weeks’ Human Embryo

of the epidermic thickening. The thickened epidermis, separated from
the brain, extends cephalad from the point of contact toward the olfac-
tory region and hence away from the hypophysis (Fig. 7). Following
still farther away from the hypophysis the olfactory region is found
(Figs. 3, 4, 5), separated by a sharp fold from the eye-stalk. From the
above, the only logical conclusion seems to me that this specimen gives
positive evidence that the olfactory region of the brain is not its mor-
phologie cephalic part, but that the eyes are relatively to the original
margin of the neural plate cephalad of the olfactory region, 1. e., nearer
the hypophysis. ‘This is in contradistinction to the arrangement that has
always been accepted namely, that in the vertebrate brain the olfactory
region is the most cephalic, forming the first of the series. Even His
after his acceptance of the demonstration of Keibel and his own state-
ments concerning the neuropore ignored the logical conclusion as to the
order of the parts in the series.

Following von Baer, Reichert, and Gotte, Studnicka®” has finally de-
monstrated that the olfactory lobes and the cerebrum are essentially
dorsal and paired outgrowths from the neural tube. The present investi-
gation confirms this and further places the eye-stalk and the retina in a
similar category as dorsal paired organs serially in front, 7. e., toward the
hypophysis from, the olfactory lobe. The natural corollary follows that
the optic chiasma crosses the original margin or dorsimesal line; that it
is in serial order with the pre-commissure, forni-commissure, callosum,
supra-commissure, and post-commissure, each binding together paired
dorsal organs, the chiasma being as truly dorsal as the post-commissure.

Nor is the conclusion above reached based merely upon logic. The
study of a model made of a mouse in which the neural plates are cleft to
the hypophysis show that in tracing the series of folds, which have rela-
tion to the margin, there are on each side: Ist, Hypophysial rudiments
both in the skin and the brain, consisting of folds which reach the margin ;
2d, A fold which ends in the outgrowing eye, and extends to the margin,
while along the outer surface of this fold the skin is in direct contact with
the brain; 3d, A flap or margin, still undifferentiated, lying between the
2d fold and the mesencephal. By comparison of this specimen with later
stages of various specimens, it is seen that the flap (3d) becomes differ-
entiated into olfactory, cerebral, and diencephalic rudiments.

A model made of the neural plates of the human embryo 12 of the
Johns Hopkins University collection, lends strong confirmatory evidence

31 Studnicka, F. K., Kgl. béhm. Ges. Wiss., Math.-natur., XIV, 1901; Zool.
Centralbl., VIII, 1901.

Susanna Phelps Gage - 427

to the above observations, while the comparative studies on lower forms
lend their quota to the result.

These conclusions were stated by me in the above-mentioned
abstracts.” Johnston,” the latest reviewer of the serial order of seg-
ments of the head, does not agree with my view. His work is largely
based on immammalian material which in my experience, as above stated,
does not show the facts clearly being obscured by either thickened walls
or secondary formation of a neural cavity. The crucial point on which
he rests the conclusion that the olfactory is in front of the eye is depend-
ent upon those observations which tend to show that; (1) The segmental
mesoderm extends past the eye to the olfactory region; (2) The hypo-
physial thickening of the skin is continuous with the nasal epithelium
in petromyzon. With regard to the first point attention is called to a
recent paper by Froriep “ in which he shows that the original mesoderm
of the head does not primarily pass cephalad of the hypophysis upon the
mesal line. As to the second point, Lubosch “ shows that the thickened
epithelium of hypophysis and olfactory plates is not continuous but is
separated by an interval of thin epithelium. Moreover, it is shown inci-
dentally in his figures, that it is in close connection with this thin mesal
plate that the incipient cavity of the eyes originates, that is between
hypophysis and olfactory.

In order to make his contention good, Johnston is driven to the con-
clusion that the eye is a dorsal organ lying between the olfactory region
and the diencephal and in the course of development, is dragged ventrad
to its final position, the cerebrum and the eye being portions of the same
neuromere. The observation that the eye vesicle is originally at the edge
of the neural plate between hypophysis and olfactory region seems to
make this device unnecessary.

The question of the cephalic end of the brain is not, to what point the
neural plate is cleft after the formation, and final growth of the meso-
derm into the head, but to what point it was cleft at the outset. As above
stated, in mammals and torpedo, the cleft originally extends to the hypo-
physis. The eye lies next to the hypophysis and distinctly intervenes
between hypophysis and olfactory region.

More recent investigations on invertebrate brains,” seem to have
established the fact that the lobe of the brain connected with the com-

% Johnston, J. B., Jour. Comp. Neurol. & Psychology, XV, 1905.

3% Froriep, A., Anat. Gesell., Verhandl. 16, 1902.

4Tubosch, W., Morph. Jahrb., XXIX, 1902.

35> Comstock, J. H. and Chujiro Kochi, Am. Nat., XXXVI, 1902. They
summarize the work, including Patten’s, from. 1775-1900.

428 A Three Weeks’ Human Embryo

pound eyes is really cephalad of that connected with the antenne, now
proved to function as organs of smell. This fact seems to fit into the
finding given above on the seria] order of parts.

Total folds—In Figs. 3, 4, it is seen that the neural tube is divided
more or less clearly into lobules and these again into folds. Those now
under consideration are not the total folds of the cerebrum considered by
some authors as transitory fissures and by others as artifacts. Indeed,
some of them antedate the distinct formation of a cerebrum, some being
formed in early human specimens with open neural plates (No. 12 of
Johns Hopkins University Collection, and No. 714 of the Harvard
University Embryological Collection). In the present study, the cere-
brum itself is the name applied to one of these total folds.

Nor are they newly recognized structures. Bischoff * in 1845 published
a minute figure of the brain of an embryo dog showing such folds in the
oblongata.

Total Folds in the Immammalia.—Orr," in 1887, found in the lizard
a series of such total folds. McClure® followed with similar results.
Locy’s * remarkable dissections of shark and also of chick and of Ambly-
stoma show the beginning of these folds as marginal structures before the
mesoderm had reached the parts, and therefore indicating the segmenta-
tion of the epidermis antecedent to that of the mesoderm. Locy’s results
have been questioned by Neal,” but in going over some of the same ground
it seems to me that Locy’s observations are well founded. The careful
confirmatory work done in Locy’s laboratory by Hill” seem to put the
essential points beyond controversy. He shows 3 neuromeres in the fore-
brain, 2 in the mid-brain, and 6 in the hind-brain.

Orr, McClure, and Locy call the folds, neuromeres. This term is here
avoided because it leads too far afield into a consideration of related
questions of theory and fact concerning nerve distribution and meso-
dermic segments on which the literature is extensive and well-known.

Total Folds in Mammals.—In mammals, since the time of Bischoff,”
figures of such folds appear occasionally in literature. Mihalkovies *
shows folds in the rabbit’s oblongata; Prenant“ in that of the pig.

%®von Bischoff, T. L. W., Entwicklungsgeschichte des Hunde-eies. Braun-
schweig, 1845.

37Orr, Henry, Jour. Morph., I, 1887.

88 McClure, Chas. F., Jour. Morph., IV, 1890.

3° Locy, W. A., Jour. Morph., XI, 1895.

49 Neal, H. V., Bull. Mus. Comp. Zool., Harvard Coll., XXXI, 1898.

“1 Hill, C., Zool. Jahrb., Abth. f. Anat. u. Ontog. d. Thiere, XIII, 1900.

* Mihalkovics, V. von, Entwicklungsgeschichte des Gehirns. Leipzig, 1877.

* Prenant, A., Soc. de la science de Nancy, Bull., Ser. 2, IX, 1889.

Susanna Phelps Gage 429

Zimmermann ™ in rabbit and chick shows 2 in the fore-brain, 3 in the
mid-brain, 8 in the hind-brain, and 4 in connection with the accessorius
nerve. Froriep“ found in the mole 3 in the diencephal, 3 in the mid-
brain, which disappear later, and 7 in the hind-brain, the last being
connected with the vagus nerve. Schultze“ figures folds in the pig.
Lewis“ and Minot * show four neuromeres in the oblongata of the pig.
Bradley ® is the latest investigator to study these folds in the pig, finding
7 in the hind-brain, the 1st belonging to the cerebellum and the 7th con-
nected with the Xth nerve.

Total Folds in Homo.—Kupfter ® merely mentions in a human embryo
of three weeks, five pairs of total folds in the oblongata. His” shows
that in the region of the oblongata, certain folds in disappearing, leave
behind cell-nests as in the olive. However, in his monumental work on
human embryology,’ he seems to have avoided giving any hint that such
structures exist, while in his models of early specimens, a glittering
smoothness occurs in regions which are really full of significant form.
In models of human embryos made in Dr. Mall’s laboratory, certain facts
shown in his specimens were avoided since they did not bear on the
subjects he was investigating. However, this seems to be a case where a
little positive evidence more than counterbalances a vast amount of
silence.

Granting the existence of total folds in the neural tube of mammals
at certain stages of development, the question has been put, are they
artifacts due to shrinkage of the mesoderm? In answer I would say :—
1st. The crowded cellular growth of the neural tissue and the scattered
cells of the mesoderm would seem to indicate that though the latter might
shrink away from the neural tube, it would not throw it into such sharp
foldings as occur. Direct observation seems to corroborate this argument.
2d. The folds are found at the margins of the neural plate in man before
the mesoderm has grown up to this margin. 3d. In lower mammals
corresponding in general to this human specimen, the mesoderm as far as

“Zimmermann, W., Anat. Gesell., Verhandl. 5, 1891.

“4 Froriep, A., Anat. Gesell., Verhandl. 6, 1892.

Schultze, Oscar, Grundriss der Entwicklungsgeschichte des Menschen u.
der Sdugethiere. Leipzig, 1897.

7 Lewis, F., Amer. Jour. Anat., II, 1903.

* Minot, C. S., Laboratory text-book of Embryology. Phila., 1903.

49 Bradley, O. C., Rev. Neurol. and Psychiatry, II, 1904.

*° Kupffer, C., Konig. baierische Akad. der Wiss., Math.-Phys. Cl., Sitz.
XV, 1885.

51 His, W., Konig. sachs. Ges. d. Wiss., Abhand. d. Math.-Phys. Klasse,
XVII, 1890.

430 A Three Weeks’ Human Embryo

it has grown around the sides of the neural tube shows little indication
of shrinkage and forms complete contact with the foldings of the tube,
indicating that the folds give form to the mesoderm rather than the
reverse. 4th. Many of the young mammalian specimens examined are
cut so accurately, after such perfect preservation, that no question of
asymmetry can be raised as might be the case with the human material
examined. It may be said that in all the models made by me, no matter
how twisted or imperfect the specimen, the evidence of essential symmetry
is clear. 5th. As is natural to suppose, the folds arise step by step with
the growth and development of the tube. In human specimens they are
in their most typical condition during the third week, after which their
external creases are bridged by the growth of white matter and later
their internal sharp lines are gradually obliterated. To realize the exist-
ence and probable significance one should study them when most typical
and when they all have approximately the same size as in this specimen
(148).

Serial arrangement of folds—It must not be understood that the table
given below represents a final conclusion as to the number of folds in the
brain tube. It is an attempt to bring into as nearly definite relations as
possible at this time, the early structures shown, with those of the adult.
The lobules and folds cannot be said to accurately represent the definitive
segments which Wilder” proposes for convenience in studying later
stages, nor do they more nearly coincide with the divisions settled upon
by the German committee on Anatomical terms. In fact could we start
without so many preconceptions from the complicated adult structure
and the names which have been applied to the parts our task would be
simpler. As it is, in the figures, as few names as possible have been used
and even these do not always agree with the customary usages; for
instance, the term Diencephal is here used for a part of the roof and
lateral wall, but does not, as usually understood, include the infundi-
bulum and the eye-stalk. Perhaps the old and indefinite term thalamus
would better fit the case.

The common characteristic of all these folds (except the albicantial,
see below), is that each pair takes its origin in a common pocket at the
dorsimeson (see above Cephalic End of Brain Tube), or at the edge of
the membranous roof or metatela and thence radiates a greater or less
distance along the lateral wall of the brain. Each fold, no matter how
obliquely, tends in general direction across the axis of the brain tube.

52 Wilder, B. G., Reference Handbook of the Medical Sciences, 2d Ed., Vol. II,
1901. See also Bibliography in Wilder Quarter Century Book.

Susanna Phelps Gage 431

TABLE II.
e .
o =I 72)
E : , s8
. to | © om
< oe Ze 5 af Fos Wi reats
2 es ut B§ Ie | ae
s Z2 Ia] 9) 4* gre | (ge
Albicantial a alb., albi, 8, 4, 6, 8, 9,16
1, Infundibular \ Hypophysial | 1 infund. 3, 4, 6, 8 Hyp
II, Visual | j eae f eye 8, 4, 5, 9, 16
Striatal 4 |str., stri
III, Cerebral | 4 Olfactory 5 jolf. Seed Ont
! Cerebral 6 |cer.
IV, Thalamic or | Dieuteplalic UN ey :
Diencephalic | 2| 8 \ \Dien. 3, 4,5, 7, 8
V, Mesencephalic | Mesencepualic ; os entice 10. t 4, 5-10
VI, Cerebellar i See ee Some af 4,11
= Ghlongats 1)18 jobl i
VII, Gasserian ( a 5
or Pontial eet aleailae 7 dieses Vth N.
Oblongata 416 |obl. 4 Vv III
VIII, Otic see aa = 3, 4, 10 nes Vv
_§ Oblongata 618 jobl. Cr Serdeuatel IX N.
eee! Vie, ee 719 | « 7 8,4, 13-15 XN
X, Accessorial F meer Ye. ah 3, 4, 18, 14 { me

NOTES ON THE ToTAL FOLDs.

Albicantial.—In early stages of the chick and amblystoma, the first
folds to form in the cephalic region before the neural plates close, is this
pair lying at either side of a middle piece which is molded over the
cephalic tip of the blind cephalic end of the enteron. This connection
is soon lost, a great mass of mesoderm filling the cephalic bend, and
intervening between the albicans and the pharynx (Fig. 3, 4). The
prominence of these folds in this specimen (Figs. 9, 16) is especially
marked. In mammals, they seem to decrease in relative prominence as
the gill region begins to transform.

In many studies which have been made of the brain, a fold from the
infundibular region is shown to extend entirely across the brain tube to

432 A Three Weeks’ Human Embryo

the roof of the diencephal. In this specimen, at first glance, it seemed
that the only one from this region which could possibly extend to the roof
of the diencephal, is the albicantial. The early history of the fold as
seen above, does not make the interpretation seem probable and, moreover,
a careful modeling of the region as shown in Figs. 6, 8, 9, seems to indi-
cate that the albicantial and the diencephalic folds, originating at widely
separated parts of the brain in the middle line, end near each other in the
lateral wall but are distinctly not continuous.

This albicantial fold is only tentatively put at the beginning of the
series, since from its original close approximation to entodermal rather
than ectodermal tissue, it does not seem to belong to a truly dorsal series
nor to be the ventral end of a diencephalic fold.

1. Hypophysial——These folds are strongly developed in later human
embryos. They are not sharply outlined in this specimen but the pair
can be distinguished lying opposite to the pair of widely open pouches
representing the hypophysis at this period (Fig. 9). There seems to be
no fact thus far found which might bar these folds from the series. In
a young mouse (see above), the folds as modeled show distinct relation
to the margin of the neural plate, while the associated organ, the hypo-
physis, is a really paired organ” from the ectoderm as distinguished from
the early entodermal relationship of the albicans. In this connection,
the position of the hypophysial as the first in the series, it is significant
that Boeke™ finds in Amphioxus and certain fishes a ciliated pit in the
region of the infundibular process, having according tb him, the physical
appearance of an organ of sense. Should there be confirmation of this it
would represent a lost sense organ, the first of a dorsal series.

2. Optic or Eye-stalk.—In its earlier stages, this region is represented
by a pair of wide furrows extending from the margin of the neural plate
to the pouches forming the optic vesicles. With the closure of the plates,
each furrow forms a wide vesicle connected in the present specimen, with
the epidermis through the neuropore (Figs. 4, 6, 8,16). As development
proceeds, the portion of this vesicle toward the hypophysis, is constricted
by a complicated folding to form the so-called optic nerves, the portion
toward the olfactory remaining single as the pre-optic recess.

3. Eye-vesicle or eye proper.—As shown in Figs. 6, 8, and 9, the eye
is distinctly constricted off from its stalk. In Figs. 6-8, it shows second-
ary folding, but a typical cupping does not occur.

The eye-vesicle seems to have relationship with the original margin,

8 Gaupp, E., Arch. f. mikr. Anat., XLII, 1898.
54 Boeke, J., Anat. Anz., XXI, 1902.

Susanna Phelps Gage 433

only through the stalk, the two together forming a lobule. Even at the
point where the eyes approach the skin, they are separated from it by a
thin layer of mesoderm (Fig. 16). The lens thickening has only a slight
development, showing no tendency to bend towards the eye. Its borders
are ill-defined.

4. Striatal.—A deep fold separates the visual region from the cerebral.
On the side of the fold toward the cerebrum is a smaller total fold cross-
ing the middle line (Fig. 4). From a careful comparative study this is
identified with the fold later bordering the striatum.

5,6. Olfactory and Cerebral.—tIn Figs. 4, 7, are shown the slight total
folds, the forerunners of the olfactory and cerebral regions proper. Each
pair of folds, begins in a mesal pocket but does not pass far across the
brain tube. Mall” considers these among the artificial fissures of the
cerebrum but in fact only one of them is cerebral and it represents the
whole of that organ.

Relatively to the dorsimesal line, the three folds included in the
cerebral lobule are seen to be caudad of the eye or as expressed by
Studnicka,” they are dorsal.

While the neural plate is still open it was found both in man and
mouse that the region of the cerebrum and also of the diencephal (see
above, Cephalic End of Brain Tube) is comprised in a narrow undiffer-
entiated flap beyond the eye and including a portion of the margin. The
flap becomes relatively wide before closure and shows some total folds
which need more careful identification.

The olfactory epithelium shown in Fig. 3 has an irregular H-shape
with a bar across the meson. This seems to agree with the idea that the
olfactory epithelium shifts from the margin to its final, lateral position.
Bedford” has in the pig, found a certain amount of lateral shifting of
the olfactory plate. van Wijhe” found that in shark, both olfactory
organs and nerve arise out of the neuropore, thus lending confirmation to
the fact shown in this specimen.

7, 8. Diencephalic.—Are the two folds seen in the roof of the dien-
cephal, each meeting its fellow of the opposite side, in the dorsimeson.
7, is in the region which ultimately forms the membranous roof and 8,
is apparently to form the epiphysial outgrowth from its mesal pocket
(cf. Minot”).

5% Mall, F. P., Amer. Jour. Anat., II, 1903.
Bedford, E. A., Jour. Comp. Neur. & Psych., XIV, 1904.
5%’ van Wijhe, J. W., Zool. Anz., IX, 1886.
58 Minot, C. S., Science, N. S., XIV, 1901.

434 A Three Weeks’ Human Embryo

9, 10. Mesencephalic—The roof of the mesencephal or mid-brain is
in this specimen so broken that details of form were impossible to work
out, but in general, it is possible to see that there are two pairs of total
folds, one, the 9th, beginning in a mesal pocket lying caudad of the deep
notch dividing the diencephal from the mesencephal, or in other words,
caudad of the future post-commissure and extending obliquely caudad
for half the length of the mesencephal, the other, the 10th, arising near
the cephalic border of the metatela and ending abruptly near the point
where the [Vth nerve will take its origin (4 of Fig. 10).

12. Cerebellar.—These are two total folds represented clearly only in
Fig. 11, rising at the cephalic border of the metatela and involving the
part of the lateral wall which at this stage of development represents the
cerebellum. Their history has not been traced.

14-18. Oblongata.—Are folds which arise at the edge of the mem-
branous roof of metatela, and extend across the brain wall near to the
ventrimeson. If one remembers that the neural plate on closing in this
region as well as in the fore-brain, at first, is a tube with as thick walls
dorsally as elsewhere, it is easily seen how the origin of these folds at
the edge of the membrane, may represent the dorsimesal pockets occur-
ring farther cephalad, especially since Locy’s “ work shows marginal folds
in the early stages.

18. Oblongata 1.—In this and several other specimens studied, this
fold seems fully separated from the following, obl. 2, but the sections are
so cut as to make it difficult to trace it with certainty to the dorsal edge.
In other human specimens studied I was not certain of its presence. In
this region some authors find a ventral representative of the cerebellum,
but in this specimen, at least, there is no such relation.

14. Oblongata 2.—This fold is one of the most strongly defined of the
series. Its invariable connection with the roots of the Vth nerve and the
Gasserian ganglion makes it a land-mark in the embryos of all vertebrates
studied. The Gasserian ganglion is large but loosely formed and pene-
trated by branches of the jugular vein.

15. Oblongata 3.—This is also a sharply defined fold and has been
widely recognized though as yet no structure has been definitely associated
with it.

16. Oblongata 4.—The roots of the VIIth and VIIIth nerve are as
invariably associated with this fold as the Vth with its fold. In the
present specimen, in its dorsal portion, it is divided into two folds, the
more cephalic being connected with the roots of the VIIth, the more
caudal with those of the VIIIth nerve. The roots are very short, soon
uniting with their corresponding ganglia. The ganglia lie close together

Susanna Phelps Gage 435

yet are for the most part distinguishable, the auditory following the
thickening of the auditory vesicle on the cephalic and lateral portions,
that of the VIIth extending without special differentiation into its nerve
which forms a union with thickened epithelium at the dorsal end of the -
Ist gill-cleft (Figs. 1, 11).

17. Oblongata 5.—In the chick and all mammals examined, this fold
lies opposite the otic vesicle but has no connection with it, unless possibly
at a very early stage while the ear is merely a thickened plate pushed
close to the neural plate. Dr. Johnston called my attention to the fact
that such a fold is wanting in Amblystoma and an examination of
material at hand, confirms his observation that no fold exists between
that to which the VIIIth and IXth nerves are attached, in early stages
of Amblystoma. However, in the shark Sewertzoff” shows such a fold
to exist and it seems probable that modelling of the region in Amblystoma
might reveal its rudiment.

18. Oblongata 6.—In all forms examined, this permanent fold has been
found connected with the roots of the IXth nerve.

19. Oblongata 7.—This is a large fold arising at the caudal end of the
metatela, extending obliquely cephalad and ending in the floor of the
oblongata in close relation to the previous fold. It is connected with the
roots of the Xth nerve. Apparently in other specimens, this fold cannot
be so clearly defined as the others in the oblongata since it has rarely been
recognized. Froriep “ in one human specimen, and Bradley in the pig,
have observed this fold.

20, 21. Oblongata 8, 9.—These two pairs of folds are really dorsal
pockets extending only through the dorsal half of the neural tube.
Opposite their ventral portion, roots of the XIth nerve arise (Fig. 14).
This nerve is interrupted in its course to join the Xth nerve by masses
of ganglionic cells.

The relations of the VIIth, [Xth, Xth, and XIth nerve of this speci-
men to their ganglia and the sensory epithelium have so recently been
fully discussed by Streeter” that they will not be treated. The sensory
epithelial thickenings to which he calls attention, are here figured
(Pig. 1).

Beyond the clearly formed folds, above discussed, there occur several
others each corresponding with an enlarged part of the ganglionic cord.
As this cord has no further indication of dorsal nerve roots, the exact
relations cannot be determined. Moreover, the following total folds in

5° Sewertzoff, A. N., Anat. Anz., XXI, 1902.
© Streeter, G. L., Amer. Jour. Anat., IV, 1904.

436 A Three Weeks’ Human Embryo

the myel are not strongly marked, and in other specimens it is only in
favorable sections that they can be seen at all.

THE SpecIAL Pornts APPEARING FROM A STUDY OF THIS EMBRYO
ARE:—1. Both external form and internal organs show with diagram-
matic clearness a normal development but with individual differences
from other specimens of about the same age, some of these differences
indicating greater, some less development. It seems probable that a care-
ful study of such embryonic peculiarities in man and higher mammals
may throw light on very important questions of heredity and variation.

2. Epithelial thickenings occur at the neuropore, olfactory region,
lens, gill-clefts, and about the mouth, at the summit of the limbs, the
thickening of the leg being continuous with that of the anal region.

3. There are 29 myotomes, 2 being occipital, and also remnants of 3
other occipital myotomes.

4. The nephric system is in a generalized condition presenting a re-
capitulation in one specimen of several distinct stages of development.
This is shown by :—An open pronephric tubule on each side, independent
of the Wolffian duct; each mesonephros having in its cephalic half, 8
rudimentary glomeruli opening by tubules into the duct; in its caudal
half, 11 or 12 tubules not opening into the duct, but part of them opening
to the ceelom. The mesonephric tubules vary in structure from solid
masses of cells to tubules with glomerulus and Bowman’s capsule.

5. The developmental stage of the central nervous system shows with
definiteness the position of the neuropore and its relation to the hypo-
physial region. In comparison with other specimens examined this makes
it possible to determine the front end of the brain tube and of the body.

6. I believe that the morphologic cephalic end of the body is as figured
by von Baer, in the region of the hypophysis and, furthermore, I believe
as a generalization, that in all stages of development the hypophysial
region is at the morphologic, cephalic end of the body, and consequently
that parts which in the exigencies of growth have gone beyond this point
are morphologicly caudad of it, as the eye and olfactory region.

7. The brain tube shows both at this stage and at earlier and later
stages total foldings which are directly correlated with definite nerves or
epithelial thickenings. Other foldings have not yet been correlated with
definite organs. These foldings are so uniformly present in mammals,
birds, and selachians that they cannot be conceived of as artifacts but are
believed to be true morphologic features.

Susanna Phelps Gage

NAMES AND ABBREVIATIONS USED ON THE FIGURES.

alb. or albi.—Albicans, albicantial fold.
allant. st.—Allantoic stalk.
amn.—Amnion.
anal pl.—Anal plate.
ao. ar., 1st-4th.—Aortic arches, 1st to 4th.
A.umb.—Arteria umbilicalis.
aur.—Auricle, left, right.
auric-vent. c.—Auriculo-ventricular canal.
A. vit.—Arteria vitellina.

vb. art. or b. arter.—Bulbus arteriosus.
B.c.—Bowman’s capsule.
bile d.—Bile duct.

cbl.—Cerebellum.
cbl. 1, 2—Cerebellar fold, 1, 2.
cer.—Cerebrum.
1st-Sth cer.—1st to 8th cervical myotcmes.
ch.—Chorda.

d. Cuvier.—Duct of Cuvier.
Dien.—Diencepha:.
duod.—Duodenum.

g. c. 1st-4th—Ectodermal gill clefts.
glom.—Glomerulus.
gn.—Ganglion.
gn. Fror.—Froriep’s ganglion.
gn. Gsn.—Gasserian ganglion.
g. p. 1st-4th.— Entodermal gill pouches.

Hypoph.—Hypophysis.

inf. or infund.—Infundibulum or infundibular fold.

intest.—Intestine.

lach.—Lachrymal furrow.
lens ep.—Lens epithelium.
1. perit. cav.—Lesser peritoneal cavity.

mand.—Mandible.
mar.—Maxilla.
Mesen.—Mesencephal.
mesent. —Mesentery.
mesoneph.—Mesonephros.
metat.—Metatela.
my. 1-32.—Myotomes.

438 A Three Weeks’ Human Embryo

N. [1I-XII.—Nerves III-XII.
nas. ep.—Nasal epithelium.
neur. or neurop.—Neuropore.
neph. or neph. t—Nephric tubule.

obl. 1-9.—Oblongata folds, 1 to 9.
w@s.—Esophagus.
olf.—Olfactory epithelium or olfactory fold.

pron. t—Pronephric tubule.

S. venosus.—Sinus venosus.
sec.—Section.
sept. trans.—Septum transversum.
stom.—Stomach.
str., stri.—Striatum.
supra-ren.—Supra-renal capsule, adrenal.

t. 1-21.—Mesonephric tubules, 1-21.

V. jug.—Vena jugularis.
V. postcard. or V. pe.—Vena cardinalis.
V. umb.—Vena umbilicalis.
V. vit.—Vena vitellina.
vent.—Ventricle.
vit. ves.—Vitelline vesicle.

W.d., Wolff. d., Wolffian d.—Wolffian duct.

EXPLANATION OF PLATES.
FIGuRES 1-13.

Drawings made from a model of Embryo 148 of the Mall collection, with
sections and dissections of the same (see above, Models and Drawings).
Magnification of the figures, X 3314.

PPAT Ee

Fics. 1 anp 1a. View of the left side of the model. Compare with figures
of this embryo in articles by Mall 1, 5.

It shows: The head comparatively small in diameter but great in length,
and forming at the neck-bend an angle of 65° with the body; the position of
the neuropore; the eye and ear scarcely apparent as external features; the
prominent heart, limb buds, and tail; the umbilicus turning to the right
(cf. Fig. 5); the wide undeveloped mouth and small maxillary process; the
crowding of the 2d, 3d, and 4th clefts into the precervical sinus; and 29
myotomes, the 3d being noted as the Ist cervical.

The density of the stippling on Fig. 1 indicates the relative thickness of
the epithelium (see above, External Form).

The topographic lines show the direction of the sections, the numbers

Susanna Phelps Gage 439

upon them indicate the corresponding sections of the series. The following
figures have either topographic lines or the section number at which they are
cut, and hence can be located with reference to Fig. 1.

\ IPrAmE Dis

Fig. 2. A face view of the head. As shown by the topographic lines, it is
tilted to give a clear view of the parts about the mouth which is merely a
wide slit between the hypophysial region of the head (cf. Figs. 3, 4) and the
mandibular process.

There are seen: The small maxillary process with the depression at the
corner of the mouth lying between maxilla and mandible; the H-shape of
the nasal epithelium extending also over the cerebral region; the large
neuroporic thickening; the lens epithelium with a tract extending along the
lachrymal furrow.

The cut surface (Sec. 125) shows: The division in the dorsal part of the
auricles (cf. Fig. 3); the entrance of the sinus venosus into the right part
(left of Fig.); the liver lying in the septum transversum; the folds about
the duct of Cuvier, pushing across the space to help form the diaphragm;
the connection of pericardial and abdominal ccelom; the opening of the lesser
peritoneal cavity into the abdominal celom.

Fie. 38. A view from the left side showing the central nervous system,
pharynx, heart, lung, and liver. The lateral wall has been removed. Pro-
jections of the ear vesicle and myotomes 1-15 are indicated by dotted out-
lines.

The approximately uniform tube formed by the central nervous system
and the strong cephalic flexure characteristic of this stage of development
are evident. The mesoderm in the flexure has been removed.

The brain shows: The series of total folds; the great prominence of the
albicantial; the relation of the neuropore to the epidermis and to the visual
lobe or eye; the small size of the striatal, olfactory, and cerebral folds; the
relation of the fold, oblongata 2, to the Vth nerve root (shown by a dotted
circle); of the fold, obl. 4, to the VIIth and VIIIth nerves (dotted circles) ;
of fold, obl. 6 to N. IX; the great size of fold, obl. 7 and its relation to
N. X; the continuity and segmented character of the ganglionic chain in
the neck region with the roots of the XIth nerve extending along its dorsal
side; and the folds in the myel.

There are seen: The inner tube of the bulbus arteriosus as it enters the
floor of the pharynx and divides into the aortic arches; the median thyroid
cephalad of this branching; lateral folds just cephalad of the thyroid, the
only rudiments of the tongue present; the bursa pharyngea, the dorsal pocket
at the division of trachea and esophagus (at left of abbreviation ch.); dotted
lines indicating the outline of the epithelial tubes forming lung and alimen-
tary canal; the wide communication of the pericardial and abdominal celom
dorsad of the septum transversum; the point of union of the aorte (aorta) ;
the hypophysis cut to the left of the middle line.

Fic. 4. A mesal view of the brain, myel, and pharynx. The brain shows
from the interior the same total folds as Fig. 3, but brings out somewhat
more clearly the grouping of folds into lobules (cf. Table II, in the text).
The elevations in Fig. 3 correspond to the depressions in Fig. 4.

34

440 A Three Weeks’ Human Embryo

Especially noticeable in this figure are: The short striatal folds; the
apparent continuity of the albicantial fold with the roof fold (cf. Figs. 8, 9);
the cerebellar folds 1 and 2 (cf. Fig. 11, cbl.); the cleft in the oblongata fold
4, on either side of which arise the roots of the VIIth and VIIIth nerves
(shown by dotted circles); the cut ends of the ring of mesoderm surround-
ing the neuropore; the intrusion of mesoderm into the cephalic flexure; the
fact that no mesoderm intervenes between hypophysis and infundibulum on
the middle line; the union of all layers in the roof of the oblongata; the dark
areas in the pharynx, indicating with their dotted extension the membranous
parts of the 1st to the 4th gill pouches; the notochord touching the caudal
wall of the hypophysis and coming in close contact with the roof of the
pharynx between the level of the 1st and 3d gill pouches.

Prats III.

Fic. 5. A ventral view of a segment of the model (Fig. 1) extending from
section 247 to 155.

The left side of the head is dissected away to show: The relation of the
visual lobe, the neuropore, the olfactory and cerebral regions, and the roof
of the diencephal.

The right side of the head shows: The thickened epithelium of the neuro-
pore; the olfactory region and its extension over the cerebrum; the future
lens; the mandible and gill-cleft-like pocket at the corner of the mouth.

The caudal portion of the figure (Sec. 247) shows the 23d myotome appar-
ently continuous with the mesoderm of the leg-bud; the division of the aorta
into the umbilical branches, the left one looping over the caudal end of the
cceelom; the left umbilical vein with a branch from the leg; the right plexi-
form umbilical vein passing along the body-wall toward the heart.

The middle part shows: The wide umbilicus turning to the right and
containing the thick-walled vitelline sac with its veins and arteries and its
union with the caudal intestine inclosed in mesentery; the reappearance of
the intestine in section near the union of the Wolffian duct and allantoic
stalk (cf. Fig. 17).

Fic. 6. A ventral view of a deeper segment of the model. It extends from
section 185 to 155. In the head region (cf. Figs. 2-5) it is cut through; the
neuropore; the eye vesicles, partially constricted off from the stalk; the
albicantial folds (cf. Fig. 9); and the cephalic end of the thick-walled mesen-
cephal as it dips into the albicantial region. At N. III is a strand of meso-
dermic tissue but apparently no true nerve fibers.

The caudal part of the figure (Sec. 185) shows: On one side, the appear-
ance of a myotome cut through the middle; on the other side the over-
lapping ends of two myotomes; the 8th mesonephric tubule opening into the
Wolffian duct and with its cap of thickened peritoneal epithelium (cf. Fig.
17) and its artery and vein.

In the middle portion are: The umbilical veins at either side, the left show-
ing the greatly divided sinuses; the vitelline veins as they approach on either
side of the alimentary canal; the vitelline artery; the liver near its caudal
part embedded in the transverse septum and near the level where the bile
duct unites with the duodenum; the umbilicus turning to the right of the
specimen.

Susanna Phelps Gage ~ 441

The heart (Sec. 168) cut near its middle shows: The undivided chamber of
the ventricle; a strong fold arising between the two sides and separating
the exit of the bulbus arteriosus (cf. Fig. 12) from the entrance of the
auriculo-ventricular canal (cf. Fig. 10).

Fic. 7. From a segment of the model cut at section 200, near the edge
of the neuropore, looking into the roof of the curved brain tube and showing
that the striatal, olfactory, and cerebral folds, and those of the roof of the
diencephal and of the mesencephal, meet in the mid-dorsal line, there being
no division by a middle partition.

Fie. 8. A segment of the model extending from Sec. 198 to 155. It cuts
the neuropore (cf. Fig. 16), showing a pit on its neural aspect, and looks
into the visual lobe and eye vesicles in the opposite direction from Fig. 7.
It shows the notch at the tip of the optic vesicle, apparently the beginning
of the optic cup.

The deep projection of the floor of the mesencephal into the albicantial
region is here shown and the independence of the albicantial folds from those
dorsad-of if (ck. Fig: 9).

T2awNaua} JEN

Fig. 9. A segment of the model from section 162 to section 200 (cf. Figs.
1-4), showing: A caudal view of the eye vesicle and visual lobe; the V-shaped
union of the albicantial folds and their independent dorsal ending; the
mesencephal with its sharpened beak-like ventral ending between the albi-
cantial folds; the strand of tissue, at the point where the III N. would later
appear; the hypophysis forming a bi-lobed, ectodermic organ surrounding the
end of the hypophysial fold; part of the neuroporic thickening; and the Gas-
serian ganglion.

Fic. 10. A segment of the model extending from Sec. 155, through the
cephalic flexure, to Sec. 96. With dissections at Secs. 130 and 140. It shows:
The base of the mesencephal and oblongata, with the large protuberance (4)
at the end of the second total fold of the mesencephal; the oblongata folds
1-5, and the relations to the Vth and VIIth Ns.; mouth; pharynx; the ending
of the gill-clefts 2-4 in the precervical sinus; the entrance of the vitelline
veins into the liver at the side of the duodenum and their union in the
dorsal part of the liver with the sinus venosus; the vitelline artery; and
the mesonephros.

Fic. 11. A view from the dorsal side of the same segment of the model
as is shown from the ventral side in Fig. 10, 7. e., it extends from Sec. 96 to
Sec. 155 (cf. Figs. 1, 3, 4). It cuts the arm buds, looks into the floor of the
pharynx and cephalad into the pons, mesencephal and ear vesicles.

There are seen: A portion of the cerebellum with its folds; the mesen-
cephal with its narrow opening cephalad and its floor protruding deeply into
the pons region; the interior view of the pons lobule with its three folds,
obl. 1, obl. 2, obl. 3; the otic lobe showing the ventral ends of folds, obl. 4, 5;
obl. 4 connected with the VIIth and VIIIth nerve; at the right the relation
of the ear vesicle to obl. 5; at the left the ganglion of the VIIIth lying next
the otic vesicle, that of the VIIth crossing dorsad of the first gill-pouch; at
the right the intimate union with the epidermis of the ganglion of the IXth
nerve; the ventral ends of folds obl. 6.

442 A Three Weeks’ Human Embryo

On the left, dissections down to Secs. 99, 103, 110, and 128 show: The
relations of the gill arches, and the four gill pouches to the pharynx, the
larynx, esophagus; the celomic cavities separated by the mesentery and only
partially divided into pericardial and abdominal regions by the lateral
infolding formed by the ducts of Cuvier; at the left the dissected cardinal
vein arching over the celom and uniting with the jugular vein to form the
duct of Cuvier and thence dipping ventrad to join the sinus venosus (cf.
Figs. 10, 2); the right and left aorte near their point of union; the right
arm-bud with its thickened epithelium and the branches of the terminal
blood-vessels; the 10th myotome merging into the mesoderm of the left arm-
bud near its dorsal portion; well developed motor nerve roots.

Fic. 12. A ventral view of the heart showing the somewhat greater
length of the right part of the common auricular chamber. The figure is
labeled as though the right and left sides were separate. By comparison
with Figs. 2-6, the relations are seen of the bulbus arteriosus and the sinus
venosus to the single tube forming the heart.

PLATE V.

Fig. 13. A segment of the model (Fig. 1) from Secs. 5 to 20, showing:
The total foldings obl. 7, 8, 9 of the neural tube dorsad of the roots of the
Xth, XIth, and Ist cervical nerves (cf. Figs. 3-4); the foldings of the skin
corresponding with those of the neural tube; the metatela rapidly widening
from the cephalic end of the cut surface (cf. Fig. 14); the close connection
of neural and epidermic epithelium.

Fic. 14. From a photograph (xX 4714) of Sec. 25 of the human embryo
148 (cf. Figs. 1, 2, 4). It shows: The neural tube just at the ventral border
of the folds, obl. 8, 9, represented in Fig. 13; the cephalic, Ist, root of N. XI,
attached to the base of fold, obl. 8; the 2d root of N. XI, attached to obl. 9;
the XIth N. as it passes through the Ist cervical and Froriep’s ganglia; the
intimate union of the 2d and 3d cervical ganglia.

The cilia lining the tube appear faintly and stop short of the dorsal margin
of the neural tube.

At the right, the first four myotomes are seen, the 4th showing especially
well the dorsal division into two separate horns. Noticeable is a continua-
tion cephalad of similar cell-groups and epidermal corrugations representing
remnants of still more cephalic, occipital myotomes.

Fig. 15. A photograph (xX 47%) of part of Sec. 44 (cf. Figs. 3, 4), showing
the neural tube with its cilia, metatela, and the relations of the Xth, XIth,
and XIIth nerves. :

At the left of Fig. 14, the fold, obl. 7, is seen. At the right, it has dis-
appeared, to reappear as a more marked depression at the lower level of
Fig. 15, where the attachment of the Xth N. is found.

At the left is seen the appearance of the 1st, 2d, and 3d myotomes at a
lower level than in Fig. 14, and at the right, at a still lower level, as they
recede farther from the skin and where the nuclei, related to the developing
muscle fibers, form a band across the myotome. At the left is a continua-
tion cephalad of the same segmented appearance of the epidermis as that
which lies over the myotome.

Susanna Phelps Gage 443

Fig. 16. A photograph (xX 47%) of Sec. 192, through the neuropore to
show: The point of most intimate union of the thickened epidermis with
the neural epithelium (for the extent of the neuroporic thickening, (cf. Figs.
1-8); the total folds entering into the formation of the visual lobe; the
sharp angle formed by the albicantial furrow at the right; the external
filling up of this angle at the left, indicating its independent ending (see
text); the mesoderm continuous over the eye-vesicle, but interrupted at the
neuropore (cf. Fig. 4). The cilia present in this section do not show clearly
here.

Fie. 17. Ventral view of a large model of the nephric region of Homo
148, extending from Sec. 94 to See. 295 (x 75). It shows the two Wolffian
ridges, each including the pronephric remnant; the mesonephros; the dorsal
portion of the mesentery; a part of the stomach and lesser peritoneal cavity;
the union of the allantoic stalk with the intestine; the cloaca and the
imperforate anal plate; the right Wolffian duct and its union with the cloaca.

The opening of the pronephric tubules (Fig. 18) on the ccelomic surface is
shown. Crosses indicate the position of the first eight mesonephric tubules
which do not open on the celomic surface (Fig. 19) but connect with the
Wolffian duct. The 9th to the 20th tubules have no connection with the
Wolffian duct. The 9th and the 13th to the 18th are open to the celom.
The 10th to the 12th are hollow but connect neither with duct nor celom.
The 19th to the 21st are thickenings touching the celomic epithelium but
showing no cavity. The general arrangement of the tubules of the other
side is similar but not identical.

The slight furrows at the cephalic end may represent a rudiment of the
supra-renal or adrenal body.

Fic. 18. From a photograph (x 160) of Sec. 104, Homo 148, showing the
pronephric tubule of the right side (Fig. 17) with its opening into the
celom and its duct traceable for a few sections farther cephalad. The
cardinal vein is also seen. :

Fic. 19. From a photograph (x 160) of Sec. 166 through the 7th meso-
nephric tubule of the right side (Fig. 17), showing the small rudiment of the
glomerulus; the typical S-shaped tubule connecting with the Wolffian duct
and thin-walled Bowman’s capsule which a few sections farther caudad unites
with the celomic epithelium and forms a small opening. The remainder of
the first eight tubules are of this same general type but are completely
separated from the celomic epithelium.

Fic. 20. From a photograph (x 160) of Sec. 195, through the 10th meso-
nephric tubule of the left side (Fig. 17), showing a wide opening to the celom
and its independence of the Wolffian duct.

The photographs reproduced on this plate were made by Henry Phelps
Gage and are part of a complete series of 96 taken from the sections.

A THREE WEEKS’ HUMAN EMBRYO

SUSANNA PHELPS GAGE

AMERICAN JOURNAL OF ANATOMY--VOL. iV

Ee
= a
eee

tr.

\ THREE WEEKS’ HUMAN EMBRYO
SUSANNA PHELPS GAGE

2 Soa
Des ney

IERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE

A THREE WEEKS’ HUMAN EMBRYO

SUSANNA PHELPS GAGE

2 IYIV4L)

s ;
| ra
A5 |

YID1Uu97S.

4074 nq P

r.
/ 7y ay
2)2)4n2

snbny foseo
bun

4 AJLD 9
? b 52

PUP UL
/

PM oe V

LXA
XA
4/7 ky
VIA ale
NN la?
99
‘ ke
/7
$79
“7
Whe

AMERICAN JOURNAL OF ANATOMY--VOL. IV

PLATE

SUSANNA PHELPS GAGE

A THREE WEEKS’ HUMAN EMBRYO

9 :
DJZAOD

P1v2psou 4

au,

~ DJ4LOKD

: P uoiffoy

matt

6a) j
SOALY U8 UOS ILE
PA
(¢44 ga |
fF oper ~ \
3 }
rN i
ay ad be ho Hil ER : \
hk es _: var
0 IS gun a Fes
A r Pes
N S, _® of :
¥/ #75 FIOZU DTI O- cm Z te
oe
AdDQUaSI Ud ‘ { nN
Wa x
>)

P40U04 HAL

MAN

ues WW

MERICAN JOURNAL OF ANATOMY VOL. IV

PLATE IV

A THREE WEEKS’ HUMAN EMBRYO

»

SUSANNA PHELPS GAGE

$$ 225>

P40 2-98 24 j
eae

42a 17

pice SAM

bury (p422)5 04 A

(yhs quny yoDuas
bun} -
(76-4) LAB 4 > : i : ¥ \ 431029 P ‘4 s4olosiett
: ’ | Aas
Se iby ne D9usn-21Kne

sisky lary SASOMLIZLD FZ
eae 2 y

wb
977° 2b
= s/
bnl: 7
“Dd
yRNod
S74
479?
2 217 ae

TAT“

UPS? WW

IV

AMERICAN JOURNAL OF ANATOMY--VOL.

A THREE WEEKS’ HUMAN EMBRYO >) ya SV
SUSANNA PHELPS GAGE

left

a
S
2 v
Sa
S
SW s
' ~ ~
Vote ext!
: \
ae a sal iehnall
Ses |
x we
‘i a
> Vy —
Ey MS
Dy
A?
X
Nd
y
ed
S
eS ee en er nee So
: =
hawt
‘ ~ >
~ y é x, E SN
<< S . s So ~ USS
: : SAL gS Sige age Dig eas ea ie Ns
5 a

te, Om re
“Fe

”
x
waa

melat

euraf ore

9
é

Nron

melalela

AMERICAN JOURNAL OF ANATOMY--VOL IV

THE BLOOD AND LYMPH VESSELS OF THE LUNG OF
NECTURUS MACULATUS.

BY

WILLIAM SNOW MILLER.

Associate Professor of Anatomy, University of Wisconsin.

WITH 3 TEXT FIGURES AND 2 PLATES.

ANATOMICAL RELATIONS oF THE LuNnes.—The lungs of Necturus
maculatus consist of two elongated cylindrical sacs which are situated
one on each side of the body cavity. Anteriorly they meet in the mid-
line and communicate by means of a narrow slit, the glottis, with the
short, wide pharynx. No septa are present in the lungs and both inner
and outer surfaces are smooth except for such irregularities as are
occasioned by the blood vessels.

When fully distended the lungs measure, in an adult animal, from
80 to 100 mm. in length and from 7.5 to 10 mm. in diameter; they are
slightly crescentic in shape (Fig. 1) and enclose an elongated oval
space which is occupied by the posterior portion of the cesophagus, the
spindle-shaped stomach and the anterior portion of the intestine.

Each lung is attached throughout nearly its entire length to a fold of
the peritoneum; the right lung to one which extends from the liver to
the mid-dorsal body wall, the left, to one which extends from the
stomach to the mid-dorsal body wall. In an average sized animal only
15 or 25 mm. of the normally distended lung is free from this peritoneal
attachment. '

Tue Broop Vessets.—Three afferent branchial arteries convey the
blood on each side from the heart to the gills. Each artery runs along
the ventral border of the corresponding gill giving rise to numerous fine
branches which break up into a capillary network. From this capillary
network numerous radicles unite to form on the dorsal border ‘of each
vill an efferent branchial artery. There are thus formed on each side
three efferent branchial arteries (Fig. 1, 2. B.) which form by their union
the right and left aortic roots (Fig. 1, R. A. and R. A.’) these, in turn,
unite in the mid-line dorsal to the anterior end of the stomach to form
the aorta (Fig. 1, A.).

AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

446 The Vessels of the Lung of Necturus Maculatus

The pulmonary artery (Fig. 1, P. A.) arises from the third (fourth)
efferent branchial artery after it has been joined by the second efferent
artery and lateral to the entrance of the anastomosing branch from the
first efferent artery. It runs obliquely towards the lung, giving off along

| ad ogee eRe a H
UGE es

be

a ey AR ACS
‘ gy id

Fig. 1. The blood-vessels of the lung of Necturus. A., aorta; H. B., efferent
branchial arteries; #. C., external carotid; I. C., internal carotid; Ph., cesoph-
agus; P.A., P.A’., pulmonary artery; P.bd., P.oe., P.s., branches of the
pulmonary artery outside of the lungs; P.V., pulmonary vein; R.A., R.A’.,
aortic roots.

its course small branches to the muscles of the shoulder, to the group of
muscles lying ventral to the buccal cavity and to the cesophagus (Fig 1,
P. s., P. b., P. @.). The artery now passes across the dorsal surface
of the lung to gain its dorso-mesial side, along which it runs, gradually
diminishing in size, to its tip (Fig. 1, P.A.’). Throughout the course

William Snow Miller 447

of the pulmonary artery in the lung lateral branches are given off, some-
times from opposite sides of the main trunk, more frequently, alternately
from one side, then, the other. These branches are so arranged that, as
a rule, an arterial radicle lies between two venous radicles (Fig. 1).

Arising from the main arterial trunk, and also from its branches, is
a rich network of capillaries which is spread out over the entire inner
surface of the lung; the venous radicles take their origin from this
network (Pl. II). Not infrequently smail branches of the main artery
extend across one or the other side of the lung and enter directly into
the main venous trunk. This direct union of artery and vein is quite
common in some lungs, while in others it is entirely absent. In some
instances two small arterial radicles will join the end of a venous radicle
giving the appearance of a vein arising from the main artery by a forked
extremity. Examples of each type can be seen in Plate IT.

The pulmonary vein (Fig. 1, P. V) extends along the ventro-lateral
side of the lung. It is formed not only by the capillaries, into which
the pulmonary artery breaks up, entering it directly, but also by venous
radicles which take their origin from the same capillary network. As
the lungs converge to unite on the ventral side of the cesophagus the
pulmonary veins leave the side of the lung and cross it obliquely to meet
and fuse in the mid-line into a single vessel (Fig. 1, P. V., and PI. L.,
Fig. 2). From this point of union the now single pulmonary vein
runs anteriorly towards the heart; passing to the dorsal side of the two
arms of the hepatic sinus it usually continues along the wall of the left
arm of the sinus and opens into the left side of the atrium of the heart
(Fig. 3).

This description of the pulmonary blood vessels differs very materially
from that given by Suchard for Triton and Salamandra. Suchard re-
verses the course and position of the artery and vein from what I have
found to be the case in Necturus; the artery, according to his descrip-
tion, occupying the position of the vein end the vein that of the artery.

In still another important particular does the distribution of the
blood vessels in Necturus differ from that given by Suchard for Triton
and Salamandra. In Necturus there is no interruption of the capillary
network over the pulmonary vein; Suchard describes such an interrup-
tion in Triton and Salamandra. In Necturus the capillary network is
spread out over both artery and vein.

Tue Lympu Vessets.—Broadly stated, the lymphatics follow the
blood vessels, both arteries and veins; their arrangement, however, about
these vessels is different.

448 The Vessels of the Lung of Necturus Maculatus

Along the course of the main artery three lymph trunks are usually
found placed nearly equidistant from one another and connected together
by numerous anastomosing branches (Pl. II). The main lymph trunks
show great irregularity in size; sometimes, by their wide dilatations
coming nearly into contact, the artery is practically hidden from view
(Pl. I, Fig. 1). As the artery approaches the tip of the lung the num-
ber of lymph trunks are, in most lungs, reduced to two, placed one on
either side of the artery. Numerous anastomosing branches connect
the two Jymph vessels.

Each lateral branch of the pulmonary artery as it leaves the main
trunk is accompanied by two lymph vessels which arise from the main
trunks, and like the main trunks they are connected together by fine
anastomosing branches (Pl. II). The lymph vessels accompanying the
branches of the pulmonary artery can be traced across the interval be-
tween the pulmonary artery and pulmonary vein to their union with
one of the main lymph trunks about the pulmonary vein (PI. IT).
When a branch of the pulmonary artery forms a direct anastomosis with
the pulmonary veins there can usually be recognized two accompanying
lymph vessels which join directly one of the main lymph trunks about
the pulmonary vein.

At the root of the lung the network of lymph vessels about the pul-
monary artery usually forms two large trunks which join the exceedingly
rich network of lymph vessels in the wall of the stomach on its dorsal
side (Pl. 1, Fig. 1). In some animals small lymph vessels pass from
the peri-arterial network about the left pulmonary artery along the
dorsal mesogastrium to join this same network.

In some lungs one of the two main trunks above mentioned passes
around to the ventral side of the lung and forms an anastomosis with
the lymph vessels about the pulmonary vein (Pl. I, Fig. 1).

Hoffmann says in regard to the lymph vessels of the lung of Rana:
“Die Lymphgefisse begleiten in der Lunge ausschliesslich die arteriel-
len, nie die vendsen Blutgefiisse.” Suchard, in his description of the
lymph vessels of the lung of Triton, says: “les branches du réseau peri-
arteriel sont moins nombreuses et moins volumineuses que celles du
réseau peri-veineux.” We shall see that neither of these statements
holds true for the lung of Necturus.

The main trunk of the pulmonary vein is accompanied by two large
lymph trunks which are connected together by less numerous anastomosing
branches than is the case with the lymph trunks about the artery (Pl. IT).
In some lungs they are apparently absent or for some reason they fail

’

William Snow Miller 449

to inject (Pl. I, Fig. 2). Each lateral branch of the pulmonary vein
is accompanied by one or more lymph vessels, but the arrangement is not
as regular as that about the branches of the pulmonary artery (Pl. II).

The branches of the peri-venous lymph trunks anastomose freely with
the very irregular network of lymph vessels which is spread out between
the pulmonary artery and vein. This network is formed by anastomosing
vessels which connect the lymph vessels accompanying the lateral
branches of the artery and vein with each other. At the root of the

Fic. 2. Transverse sections through the pulmonary artery and vein, show-
ing the position of the large lymph trunks (L. V.) and the attachment of the
peritoneal fold which is cut through when injecting the lymph vessels. x 40.
(From Bulletin of the University of Wisconsin, No. 33, 1900.)

lungs the main peri-venous lymph trunks join a large lymph sinus which
is situated on the ventral side of the stomach in the angle formed by the
divergence of the lungs and through this with the network of lymph
vessels situated on the ventral side of the stomach (PI. I, Fig. 2).

The lymph vessels of the lung of Necturus hke those of Triton and
Salamandra (Suchard) are more superficially situated than the blood
vessels. In transverse sections taken through the artery and vein, the
position of the large lymph trunks is easily made out even though the

150 The Vessels of the Lung of Necturus Maculatus

vessels are not injected (Fig. 2). In injected specimens the main
blood vessels and their chief branches appear to be surrounded by a net-
work of anastomosing lymph vessels. The lymph vessels themselves
form a system of closed tubes. In none of the numerous preparations
which I have made have I seen any evidence of so-called lymph capilla-
ries, lymph spaces or lymph channels leading out from the lymph vessels.
A distinct wall could be demonstrated for every vessel and the injection
mass did not pass outside of this wall.

LYMPHATICS OF THE WALLS OF THE SToMACH.—The intimate rela-
tion between the lymph vessels of the lungs and those of the stomach
renders it necessary to give a brief account of the latter. Reference to
Plate I, Fig. 3 will give an idea of the exceeding richness of the plexus

eee

Digs

Fig. 8. Diagram of the duct of Cuvier and the principal venous trunks.
D. C., duct of Cuvier; Ji., internal jugular; Je., external jugular and jugu-
lar sinus; Sbc., subclavian; La., lateral; P. C., posteardinal.

of lymph vessels present in the walls of the stomach. Along the dorsal
side (Pl. I, Fig. 1) there is present a large and irregular sinus which
communicates not only with the lymph vessels of the stomach and peri-
arterial lymph vessels of the lungs but also with those belonging to the
remaining viscera in the abdomen. On the ventral side (Pl. I, Fig. 2),
as already mentioned, there is found in the angle formed by the diverg-
ing lungs a lymph sinus of considerable size and owing to the fact that
there are no valves, the entire plexus can be easily injected.
ConNECTION oF LympH VESSELS WITH THE VEINS.—In order that
the connection of the lymph vessels with the veins may be clearly under-
stood it will be necessary to describe briefly the Ductus Cuvieri and the

William Snow Miller 451

venous trunks by which it is formed. Each Ductus Cuvieri (Fig. 3,
D.C.) is formed by the union of the following veins:
jugular,

subclavian,

posteardinal,

lateral.
The jugular is the largest of the venous trunks and appears as a direct
continuation of the Ductus Cuvieri. Its direction is at first dorsal, then
curving anteriorly it passes above the subclavian to divide after a short
distance into the external and internal jugulars (Fig. 3, Ji.,Je.). Con-
nected with the external jugular is a widened expansion, the jugular sinus
(Fig. 3, Je.).

The subclavian (Fig. 3, Sbc.) joins the Ductus Cuvieri just internal
to the jugular and on its ventral side. The postcardinal and lateral
join the Ductus Cuvieri on the caudal side, the lateral being the more
external (Fig. 3, iP. C., Za.). In some animals these last two veins
unite just as they join the Ductus.

We have seen above how the lymph vessels of the lung are connected
with the very rich network of lymph vessels in the walls of the stomach
(Pl. I, Figs. 1,2, and 3). Arising from the anterior and outer margin of
this network of lymph vessels there is found on each side of the stomach a
lymph trunk of moderate size which follows the course of the posteardinal
vein until just before it joins the Ductus Cuvieri; it then passes dorsal
to the postcardinal and lateral veins and enters sometimes directly into
the Ductus Cuvieri, sometimes into the jugular (Pl. I, Fig. 3). Just
before joining the vein this lymph trunk is joined by lymph vessels coming
from the head and anterior extremity.

\

4
Metuops.—The technique of injecting the lymph vessels of the lung of
Necturus is quite simple. The animal is killed with chloroform. If, on open-
ing the abdomen, the lungs are not well distended, it facilitates the injection
to insert a fine glass tube into the glottis and gently fill the lungs with air.
The free tip of one of the lungs (I generally make use of the left lung) is
grasped with a pair of broad pointed forceps and drawn away from the
mid-line; this puts the peritoneal fold on the stretch. A nick is next made in
this fold (Fig. 2), close to the artery, with a pair of sharp scissors, care
being taken not to wound the artery itself. If the nick has been properly
made a probe can now be introduced through this opening into one of the
large lymph trunks which runs along the course of the artery. The cannula
of a small syringe, which has been filled with a thin vermilion starch mass
or Chinese ink rubbed up with normal salt solution, is pushed in beside the
probe; the probe is now withdrawn and the cannula held in place between
the thumb and finger of the left hand. The piston of the syringe is slowly

452 The Vessels of the Lung of Necturus Maculatus

pushed in and the injecting mass can be seen running rapidly through the
lymph vessels. The vessels are of considerable size, the mass flows easily,
and, as there are no valves, but little pressure is necessary. The injection
should always be made toward the head. This procedure should, with a little
practice, give well filled lymph vessels in the lungs and also in the stomach.
Warm masses do not give as good results as cold. I have frequently filled
the lymph vessels of the lungs and stomach with a celloidin mass and have
obtained very instructive preparations by digesting in pepsin.

In my hands granular injecting masses have given the best results, and
thus one may avoid the uncertainty of, e. g., such fluids as aqueous solutions.
of Berlin or Prussian-blue. The lungs and stomach can be dissected out, har-
dened in alcohol, cleared in oil of clove, washed out with xylol, and mounted
in balsam. By cutting the lungs and stomach open before mounting, the
study of the preparation is made easier.

LITERATURE.

HorrMan, J.—Die Lungen Lymphgefasse der Rana temporaria. Dorpat.,
1875.

Mitter, W. S.—Contributions from the Anatomical Laboratory of the Uni-
versity of Wisconsin. Bulletin of the University of Wisconsin, No.
33, 1900.

SucHarp, E.—Des vaisseaux sanguins et lymphatiques du poumon du Triton

crete. Arch. d’Anat. Micros. -T. III.

Structure du poumon du Triton et de la Salamandre maculée. Arch.

d’Anat. Micros. T. VI.

EXPLANATION OF PLATES.
IPVATHe le

Fic. 1. Dorso-lateral view of the stomach and right lung of Necturus.
The heavily shaded vessel is the pulmonary artery; the lighter shaded vessels
are the lymph vessels. The stomach was empty and contracted when the
injection was made. Note the large sinus-like lymph trunk on the dorsal
side of the stomach.

Fic. 2.—Ventral view of stomach and lungs of Necturus. Same prepara-
tion as Fig. 1. The heavily shaded vessels are the pulmonary veins; the
lighter shaded vessels the lymph vessels. Note the sinus in the diverging
angle of the lungs and the large lymph trunk on the mid-ventral surface of
the stomach.

Fic. 3. Lymph vessels of the lung and stomach of Necturus and their con-
nection with the veins. The shaded vessels are the lymph vessels; the
veins are shown white. The plate is best seen from the side. The connect-
ing vessel is indicated by the *. The left lung is reflected to the right and
the stomach is partly dissected free.

PEATE ele

Lung of Necturus with the blood and iymph vessels injected. The lung
was cut open, spread out flat, and mounted in balsam. A, artery; V, vein.
Only the main branches of the artery.and vein are shown. In two places
the capillary network is indicated diagrammatically. The blood and lymph
vessels were drawn by means of the camera lucida and show the exact rela-
tion of both sets of vessels.

THE VESSELS OF THE LUNG OF NECTURUS MACULATUS PLATE |

W. S. MILLER

AMERICAN JOURNAL OF ANATOMY--VOL. Iv

PLATE Il

THE VESSELS OF THE LUNG OF NECTURUS MACULATUS

2 VS SZ

<i aw) |

‘i = SS =
= = 3 i E U
u 5 ; 1 } /
6 7 4 f
q ff)
.
\ Cc
q t \
~ 7 . = : 0
d ‘ i
rt

qd dd

5
aeup-<

WT ¥
1
- 4 y

iS
XY

ve
iS

eS

oS
AN

y .

39

PERG Sy wsKcsg
oe Si Paro

\

<?

\

ee

A CONTRIBUTION TO THE ANATOMY AND DEVELOPMENT
OF THE VENOUS SYSTEM OF CHELONIA.

BY

FRANK A. STROMSTEN,
Fellow in Biology, Princeton University.’

WitH 12 Text FIGURES.

The present series of observations on the anatomy and development
of the venous system of turtles was made by the writer in the Morpho-
logical Laboratory of Princeton University. The material for the adult
anatomy has been, for the most part, collected in the neighborhood of
Princeton, New Jersey. All of the embryological material has, through
the kindness of Professor MéClure, been furnished by the Princeton
embryological collection.

I take great pleasure in expressing here my deep obligations to Pro-
fessor Charles F. W. McClure for his valuable suggestions and helpful
criticism during the preparation of this paper. I am also indebted to
Mr. Charles F. Silvester for the preparation of some beautiful corrosions
of the veins of the liver.

METHODS OF INVESTIGATION.

For the study of the adult venous system, about forty turtles of the
more common species were dissected. These included the following
forms:

CHELYDRID®: Chelydra serpentina (L.).

KINOSTERNIDE: Kinosternon pennsylvanicum (Bosc.), Aromochelys
odoratus (Latreille).

Emypipa: Chrysemys picta (Hermann), Clemmys (Chelopus) in-
sculpta (LeConte), Clemmys guttatus (Schneider), Terrapene carolina
be):

The turtles were killed with chloral hydrate and injected through the
left abdominal vein. Chloral hydrate was used in preference -to either
chloroform or ether because it leaves the animal in an extended con-

1Presented to the Faculty of Princeton University for the degree of D.Sc.

AMERICAN JOURNAL OF ANATOMY.—VOL, IV.

454 Anatomy and Development of Veins of Chelonia

dition and to a large extent gets rid of the muscular contractions. The
best results were obtained when the turtle was killed several days before
the injection was made. ‘A gelatin injecting mass was used in most
cases. In order to prevent the cooling of the gelatin before all the veins
were filled, the specimens were placed in warm water several minutes
before injecting. This is not necessary, however, if potassium iodide is
used to lower the melting point of the gelatin. For the purpose of
studying more exactly the relations of the veins in the liver and the
kidneys, some of the turtles were injected with Huntington’s, 97, wax
injecting mass and corroded with concentrated commercial hydrochloric
acid. By injecting the liver through the precava with one color and
through the abdominal vein with another, it is possible to distinguish
the advehent from the revehent systems even to their finest branches. A
few of the larger snapping turtles were injected with the common starch
injecting mass.

About fifty turtle embryos were studied for the development of the
veins. Kinosternon pennsylvanicum was the species chosen for embryo-
logical investigation because it was possible to obtain a more complete
series of this than of any other form. The more important stages were
also studied in Chrysemys picta. For purposes of comparison several
embryos of the common lizard, Sceloporus undulatus, as well as chick
embryos of various stages, were examined.

Most of the material had been fixed in picro-sublimate or picro-nitric.
The picro-sublimate material was, as a rule, satisfactory both as to fixa-
tion and for the subsequent staining reactions. On the other hand, not
much can be said in favor of picro-nitric either for fixing the tissues, or
for preparing them for the stain. The embryos were dehydrated,
cleared, and embedded in paraffin according to the usual methods, and
serial sections were cut about 20 thick. Several methods of staining
were tried, but the best results were obtained with Delafield’s hema-
toxylin and picric acid. After picro-sublimate fixation this combination
differentiates the blood vessels very clearly. Frontal reconstructions, in
whole or in part, were made of the venous system of all the embryos
studied. Wax reconstructions were made wherever deemed necessary.

HISTORICAL SKETCH.

The literature dealing with the embryology of the venous system of
Chelonia is very fragmentary. Rathke, 48, and Agassiz, 57, are the
only investigators that have hitherto made any attempt whatsoever to
even superficially sketch the development of the veins of this interesting
reptilian form. Agassiz gives as a reason that so little is known about

7?

Frank A. Stromsten 455

the embryology of turtles the fact that “In the Old World, no turtles
are found in the immediate vicinity of the great centers of study.”

Rathke, 48, in his monograph, “Ueber die’ Entwicklung der Schild-
kroeten,” merely mentions the presence of some of the larger veins in a
few of the earlier stages.- His older embryos were so poorly fixed that
it was not possible for him to trace the advanced stages of venous devel-
opment. Agassiz’s remark that, “It is felt, on almost every page of
his work, that he labored under a scarcity of materials which con-
stantly impeded his progress,” is particularly true in respect to his treat-
ment of the venous system.

About ten years later, Agassiz, 57, undertook to “ write anew the em-
bryology of this order of reptiles,” that he might continue the work
which Rathke was unable to complete. His investigations, which are
published in a large volume of nearly two hundred quarto pages, together
with thirty-two plates, are concerned chiefly with the study of the embryo
as a whole without the aid of serial sections. Hence, almost no atten-
tion is paid to the development of the intra-embryonic veins, although
considerable space is devoted to the consideration of the vitelline circula-
tion.

It is thus seen that no serious attempt has hitherto been made to give
a complete history of the development of the veins of turtles. On the
other hand, very careful and accurate observations have been made by a
number of investigators on the embryology of the veins of snakes and
lizards. Rathke, 39, Hoffmann, go, Hochstetter, 93, Grosser and Bre-
zina, 95, and Choronshitzky, 99, have investigated in whole or in part
the development of the venous system in these reptiles so that our knowl-
edge here is quite complete.

The purpose of the present paper is: (1) to confirm the work already
done on reptiles by other writers, and to furnish another link in the
comparative embryology of the venous system of amniota; (2) to inves-
tigate somewhat more in detail than has hitherto been done the changes
which the umbilical veins undergo, and their ultimate relations to the
abdominal veins. . As a basis, therefore, for comparing the various stages
of venous development in Chelonia with those of other reptiles, a brief
résumé of the chief results of Hochstetter’s (93 and 03) papers is given
below.

DEVELOPMENT OF THE VEINS OF LACERTA AND TROPIDONOTUS.

According to Hochstetter, the first veins to appear in reptiles are the
omphalo-mesenterics. These are shortly followed by the precardinals,
the postcardinals, and the umbilicals.

35

456 Anatomy and Development of Veins of Chelonia

The transformation of the paired omphalo-mesenteric veins into a
single vessel to form the anlage of the portal vein.—The two omphalo-
mesenteric veins fuse dorsal to the intestine just caudad of the first
anlage of the pancreas to form the dorsal anastomosis and again further
back ventral to the intestine to form the posterior ventral anastomosis.
A ring is thus formed around the intestine which, however, is soon broken
by the atrophy of that part formed by the right vein. In the meantime
both omphalo-mesenteric veins craniad of the dorsal anastomosis are
more or less broken up into smaller vessels by the liver tubules. In
snakes, the left omphalo-mesenteric vein loses its connection with the
sinus venosus before the posterior ventral anastomosis takes place, while,
on the other hand, in lizards this connection is retained until the for-
mation of the ring mentioned above is completed. In both reptiles it is
the right vein that persists in front of the dorsal anastomosis and the left
vein behind it. There is thus formed a single vessel, arising from the
vitelline veins at the umbilicus, which takes a spiral course around the
intestine and enters the right half of the liver where it breaks up to
form the hepatic network. The vein thus formed constitutes the anlage
of the portal vein. The cardiac end of the right omphalo-mesenteric
vein is retained as a part of the right hepatic revehent vein.

The formation of the postcava, and the reversal of the primitive renal
portal system.—The revehent veins of the primitive renal portal svstem
of reptiles are represented by the postcardinals. In lizards, the post-
cardinal veins arise in the caudal region of the body and extend forward
dorsal to the mesonephrie ducts to join the precardinals and with the
latter form the ducts of Cuvier. They are broken up near the cranial
ends of the mesonephroi into a number of broad blood sinuses (blut-
riumen) and lose their dorsal position. In snakes, the postcardinals
are connected with the caudal vein from the first and do not lose their
integrity as distinct vessels at their cranial ends.

The advehent veins are formed by the bifurcation of the caudal vein.
The two branches thus formed extend forward along the ventromedial
border of the mesonephroi and appear to end blindly a short distance in
front of the origin of the omphalo-mesenteric artery. As a matter of
fact, however, they are connected with the postcardinals by means of the
mesonephric network. All of the blood from the caudal vein must,
therefore, in order to reach the heart, stream through the mesonephroi.

The primitive renal portal system just described soon undergoes a
complete reversal. One of the most important events connected with
this change is the formation of the postcava. This is brought about in
the following manner: first, the two branches of the caudal vein fuse

Frank A. Stromsten 45%

just craniad of the origin of the omphalo-mesenteric artery to form an
unpaired stem. A short branch is then formed in the caval mesentery
which connects the unpaired stem just mentioned with the right hepatic
revehent vein in lizards, but with the common hepatic vein in snakes.
A continuous path is thus formed from the caudal vein direct to the
heart, and the result is that most of the blood from the hinder parts of
the body now reaches the heart by this route. The new path thus formed
is the first anlage of the postcava.

Craniad of the origin of the omphalo-mesenteric artery the postcava is
an unpaired vessel formed from the following parts: the fused portion
of the two branches of the caudal vein, the new formation in the caval
mesentery, sinusoids of the liver, and the right hepatic revehent vein
(lizard), or the common hepatic vein (snakes). Caudad, the postcava
is paired and is formed by the two branches of the caudal vein. The
postcava loses its connection with the caudal vein, however, while at the
same time a strong anastomosis is formed between the caudal vein and
the postcardinals. It thus appears that while the posteava is still con-
nected with the caudal vein, it may function for a short time both as a
revehent and as an advehent vein to the mesonephroi.

Shortly after the postcardinals have joined the caudal vein they lose
their connection with the sinus venosus and the reversal of the primitive
portal system of the mesonephroi is thus completed. The postcardinal
veins which now receive the blood from the caudal vein act as advehent
veins, while the postcava with its branches becomes the revehent system.

Further changes in the postcava.—During the time the above changes
are taking place, the two branches of the postcava (original branches
of the caudal vein) have fused just caudad of the origin of the
omphalo-mesenteric artery so that this artery appears to pass through
a foramen in the unpaired postcava. The unpaired postcava craniad
of the omphalo-mesenteric artery receives a branch on each side from
the cranial regions of the mesonephroi. The left cranial mesonephric
branch then begins to spht away from the main stem backward as
far as the origin of the omphalo-mesenteric artery so that finally this
artery no longer passes through the postcava, but to the left side of it.
The postcava now extends as an unpaired vessel from the sinus venosus
to the origin of the omphalo-mesenteric artery, immediately caudad of
which it divides into two branches. The right branch passes backward
as a direct continuation of the postcava, while the other branch bends
sharply to the left behind the origin of the omphalo-mesenteric artery,
and then proceeds backward. The left cranial mesonephric vein opens

458 Anatomy and Development of Veins of Chelonia

into this left branch of the postcava, while the right cranial mesonephrie
vein opens directly into the main stem of the postcava.

Development of the posterior vertebral veins.—The degeneration of
the cranial portions of the mesonephroi is accompanied by the gradual
atrophy of the postcardinal veins from before backward. As these veins
degenerate, their dorsal intersegmental branches fuse ventral to the ribs
to form a pair of longitudinally running vessels, the posterior vertebral
veins of the adult. They arise in the caudal region of the body and
extend along its dorsal wall, one on each side of the vertebral column,
to open into the subclavian veins. A connection between the liver circu-
lation and the posterior vertebral veins is also made by means of two or
three branches.

The posterior vertebral veins are, at first, symmetrically developed on
each side of the body in snakes just as they are in the lizards. In snakes,
however, anastomoses early develop between this pair of veins and the
veins of the cesophagus, stomach, and portal system, which soon result
in the degeneration of the vertebral veins, so that they are represented
in the adult system only as fine anastomoses between the intercostal
veins. According to Rathke, they may in some cases disappear alto-
gether. Hochstetter attributes the degeneration of these veins to the
peculiar locomotion of snakes.

Advehent veins of the permanent kidneys.—The caudal portion of the
postcardinal veins in snakes, as well as in lizards, is still retained in the
adult system as the advehent veins of the permanent kidneys. Sincé the
permanent kidneys arise dorsal to the mesonephroi, the persisting por-
tions of the postcardinal veins are found on their ventral surfaces, but
still retain their original position dorsal to the mesonephrie ducts. _

The umbilical veins and their transformations—The umbilical veins
of a very young embryo of the lizard arise in the allantois and enter the
body through the umbilicus. In the umbilicus they unite to form a
single vein, but separate again on entering the body. They extend for-
ward in the body-wall and open into the sinus venosus together with the
omphalo-mesenteric vein of the same side. At the umbilicus they receive
a pair of small veins which extend backward along the ventral body-wall
to join the postcardinals. This small pair of veins constitutes the
posterior anlage of the abdominal veins. In a later stage than this just
described, the left umbilical vein connects with the hepatic network.
Soon a direct channel is formed through the liver between the left um-
bilical vein and the hepatic revehent vein. The left umbilical vein
completely degenerates craniad of the point where it enters the liver so
that all of the blood from this vein must pass through the liver in order

Frank A. Stromsten 459

to reach the heart. Later, the portal vein joins the hepatic portion of
the left umbilical, so that there is a direct channel from the portal to
the hepatic revehent vein. The right umbilical vein does not enter the
liver in the lizards, but in snakes it very early fuses with the right
omphalo-mesenteric vein to form a common stem and thus comes to lie
within the liver for a considerable distance. Another important differ-
ence between the snakes and lizards is found in the fact that in the
former there is no direct connection between the left umbilical vein and
the portal, as there is, as we have seen, in the lizards. In both of these
reptiles, however, the abdominal portion of the right umbilical vein com-
pletely degenerates, while the same portion of the left vein continues to
increase in size for some time although it, too, apparently degenerates
after birth.

The abdominal veins—The abdominal veins are developed, as has
already been stated, from a pair of small vessels which extend along the
ventral abdominal wall and connect the umbilical veins with the post-
cardinals. Later they fuse in the mid-lne and connect with the portal
vein. As soon as this latter connection is made the connection with the
umbilicals is lost. According to Hochstetter, the umbilical veins do not
enter into the formation of the abdominal veins in lizards and snakes.

DEVELOPMENT OF THE VEINS OF KINOSTERNON PENNSYLVANICUM.

Unless otherwise stated, the following description will be based entirely
on the embryos of Kinosternon pennsylvanicum. Owing to the rela-
tively advanced stage of development of the youngest embryo studied,
it is not possible to describe the earliest appearance of some of the veins.
Furthermore, no attempt will be made to trace the development of veins
not directly connected with one or the other of the portal systems. The
plan of the present paper will be to take up first a study of the anatomy
and development of the hepatic portal system, and then a similar study
of the renal portal system. It is not the purpose of the writer to give
at this time a detailed comparative description of the adult venous sys-
tem. For this reason only those veins are mentioned whose development
is to be more or less carefully traced.

HEPATIC PORTAL SYSTEM.

Veins oF ApULT KINOSTERNON.—Fig. 1 represents the general ar-
rangement of the liver veins of Kinosternon pennsylvanicum. As shown
by the figure, the hepatic venous system of turtles is made up of two
entirely distinct sets of veins, the revehent and the advehent veins.

460 Anatomy and Development of Veins of Chelonia

The revehent system consists of the posteava, the left hepatic revehent
vein, and several smaller veins, all of which open separately into the sinus
venosus.

The largest of the revehent veins is the hepatic postcava. It passes
directly through the right superior lobe of the liver and, after receiving
a number of branches from this organ, opens into the right side of the
sinus venosus. The largest of its hepatic tributaries is the right hepatic
vein which usually opens into the postcava where the latter joins the
sinus venosus.

The other large vein which opens into the sinus venosus is the left
hepatic revehent vein.

Besides these two large veins, a number of smaller ones open directly

= =PRECAVA

RIGHT HEPATIC ADVEHENT : #+-+-—— SINUS VENOSUS

RIGHT ae ee LER HEPATIC

HEPATIC“

Fie. 1. Hepatic portal system of Kinosternon pennsylvanicum. Ventral
view.

into the sinus venosus from the liver. The number of these is not con-
stant, but depends, as we shall see when we come to study their develop-
ment, upon the relative backward extension of the venous sinus.

The advehent system consists of the two abdominal and the portal
veins.

The abdominal veins (see also Fig. 10) arise as branches of the circum-
flex iliac of each side and extend mediad and craniad for a short dis-
tance along the cranial border of the ventral segment of the pelvic
girdle. Just in front of the apex of the triangle formed by the union
of the two pubic bones, the two abdominals are connected by a short
anastomosing branch. From this point they continue forward along
the abdominal wall as separate vessels and then enter the liver.

The left abdominal vein is much larger than the right. It passes
dorsad through the left inferior lobe of the liver to its gastro-duodenal

Frank A. Stromsten 461

surface where it unites with a vein from the lesser curvature of the
stomach, the gastric vein, to form the portal vein of Bojanus. (Bo-
janus, Anatome Testudinis Europee. Tab. XXV, Fig. 128, sz.)

The right abdominal vein passes through the right inferior lobe of
the liver and may open into portal vein of Bojanus, the mesenteric por-
tal, or, as in Fig. 1, into a large lateral branch of the latter.

The portal vein of Bojanus (which in the adult is really the direct
continuation of the left abdominal vein) arises, as stated above, from
the union of the left abdominal with the much smaller gastric vein. It
passes directly across the body, from left to right, embedded in a groove
on the ventral surface of the “isthmus ” of the liver, apparently increas-
ing very gradually in calibre as it proceeds, and finally joins the mesen-
teric portal where the latter enters the liver. It receives during its
course a number of tributaries from the duodenum, pancreas, and liver.
The largest of these branches comes from the dorsal region of the left
inferior lobe of the liver, and probably functions as the chief advehent
vein for that region.

The mesenteric portal vein arises in the mesentery of the intestines
and enters the right half of the liver where it breaks up into the hepatic
venous network. The chief branches of the mesenteric portal are the
following veins: the descending colic, the ileo-colic, the intestinal, the
splenic, the posterior pancreatic, and the portal vein of Bojanus. Within
the liver it gives off a number of advehent branches, the largest of which
usually receives the right abdominal vein.

DEVELOPMENT OF THE VEINS OF THE HEPATIC PORTAL SYSTEM IN
KINOSTERNON PRNNSYLVANICUM.

The embryonic veins to be considered in this connection are: the
omphalo-mesenteric, or vitelline veins, and the umbilicals, or allantoic
veins.

In the youngest stage studied (7.4 mm. crown-rump measurement
around the curvature of the body),° the anterior limbs are just beginning
to make their appearance as a low ridge involving several muscle seg-
ments on each side of the body. The lungs are not yet indicated. The
liver is represented by a few tubules ventral to and on the right side of
the intestine. The first anlage of the pancreas has just been split off
as a simple tube from the dorsal side of the flat mid-gut. The venous
system is represented by the following five pairs of veins: (1) the

2 All measurements were made in this manner.

462 Anatomy and Development of Veins of Chelonia

omphalo-mesenterics, (2) the precardinals, (3) the posteardinals, (4)
the umbilicals, and (5) the subeardinals of Lewis, 02.

The Transformations of the Omphalo-Mesenterics to form the
Mesenteric Portal and the Right Hepatic Vein.

The omphalo-mesenteric veins in an embryo of about 7.4 mm—In
this stage the omphalo-mesenteric veins are represented by a pair of
vessels arising in a network of the yolk sac and extending forward along
the dorsal wall of the mid-gut near the outer border of the intestinal
epithelium. Cramiad, they gradually approach each other and _ finally
fuse immediately caudad of the first anlage of the pancreas to form the
dorsal anastomosis. Both veins are about equal in size at this point, but
further caudad the left vein is somewhat larger. Craniad of the dorsal
anastomosis, both veins enter the liver. The left vein passes around the
border of the liver and is not broken up by the hepatic tubules. It opens
into the sinus venosus close to the umbilical of the same side. The right
vein, on the other hand, soon after entering the liver is more or less
broken up into a number of smaller vessels which unite again at the
cranial end of the liver and open into the sinus venosus as a single vessel.

In an embryo of nearly the same measurement, but in which the first
indications of the lungs appear, the right omphalo-mesenteric vein is
further broken up to form a rete mirabile. This breaking up continues
to such an extent that all of the blood from the right vein must pass
through a network of sinusoids in order to reach the heart. In the
meantime the left vein also begins to break up somewhat and there is
established a sinusoidal connection between the two halves of the liver.
The clogging up, as it were, of the entire hepatic portion of the right
omphalo-mesenteric vein by the invasion of the liver tubules causes a
large portion of the blood from this vein to stream through the sinusoids
to the much freer channel of the left vein. This eventually results in
the formation of an anastomosing branch through the liver connecting
the two omphalo-mesenteric veins (Fig. 2). The branch thus formed
arises at about the point where the right vein begins to break up and
passes diagonally through the liver to the left vein, ventral to the intes-
tine. The blood of the right vein that does not pass through this anas-
tomosis still streams through the sinusoids and is collected at the cranial
end of the liver by two or three small vessels and thus carried to the
sinus venosus.

Thus we find in turtles as well as in snakes and lizards that a double
anastomosis is very early formed between the two omphalo-mesenteric

Frank A. Stromsten 463

veins. Usually the dorsal anastomosis is the first connection formed
between these two veins and then later a ventral anastomosis is formed
eraniad of this through the liver. Choronshitzky, 00, however, has found
a curious variation from this in Anguis fragilis. In this reptile, accord-
ing to that investigator, a ventral anastomosis is formed threugh the
liver before the dorsal anastomosis and is to a great extent the cause of
the latter, as is seen by the following extract from his paper on “ Die
Entstehung d. Milz, Leber, Gallenblase, Bauchspeicheldriise, ete.” (Ana-
tomische Hefte, XIII, 8. 572):

“Die zwischen der queren Anastomose und hinterem Ende des Sinus
venosus befindlichen und in die Leberanlage eingeschlossenen. Teile der
beiden Venae omph.-mesentericae wurden unterdessen durch die starke

(GHT
PREC AMBINAL LEFT

fem me PRECARDINAL

SINUS VENOSUS

\ LEFT
XS OMPHALOMESENTERIC

RIGHT
OMPHALOMESENTERI

= DORSAL ANASTOMOSIS

Fig. 2. Frontal reconstruction of the liver veins of a 7.5 mm. embryo of
Kinosternon pennsylvanicum. Ventral view.

Zerkliiften ihrer Winde in ein doppeltes Rete mirabile umgewandelt,
wobei aber von der rechten Vena omph.-mes. noch ein die Leberanlage
kaudo-kranial durchbohrender starker Ast nachblieb, waihrend die linke
Vena omph.-mes. sich in ganz schmiichtige Lebervenen aufgeldst und
von der queren Anastomose abgetrennt hat. Der hinter dieser Anasto-
mose befindliche Abschnitt der Vena omph.-mes. sinistra muss jetzt sein
Blut in die gen. Anastomose ergiessen, damit es durch den in die
Leberanlage eingeschlossenen Teil der Vena omph.-mes. dextra zum
Herzen gelangen kann. Aber die queré Anastomose wird, . . . . stark
durch die beiden Leberginge komprimiert. Das Blut muss sich einen
anderen Ausweg suchen und findet einen solchen in der Gestalt einer
anderen Anastomose, welche weiter hinten liegt und, die dorsale Darm-

464 Anatomy and Development of Veins of Chelonia

wand umbiegend, beide Venae omph.-mes. vereinigt.” (Dorsal anasto-
mosis. )

This will not account for the formation of the dorsal anastomosis in
turtles. In the first place, the dorsal anastomosis is the first connection
formed between the two omphalo-mesenteric veins. Also, it is formed
at a time when the left vein has reached its maximum of development
and before the right vein is greatly broken up in the liver. In Kinoster-
non, the formation of the dorsal anastomosis seems to be almost entirely
due to the pinching together of the dorsal wall of the mid-gut to form
the dorsal anlage of the panereas. The two veins are thus brought so
near each other that even a slight obstruction of the blood current might
cause the thin parting walls to break and thus establish a connection

SINUS VENOSUS

wo HEPATIC SHVUSOIDS

NEW PATH FOR
be— CMPH= MES,
(RIGHT)

g—OMPHALOMESENTERIC
y (LEFT

DORSAL
ANASTOMOSIS

OMPHALOMESENTERIC
(LEFT )

Fic. 3. Frontal reconstruction of the liver veins of an 8 mm. embryo of
Kinosternon pennsylvanicum. Ventral view.

between them. In other words, it may be due not so much to the neces-
sity for a connecting path between the two veins that the dorsal anasto-
mosis is formed, as it is to the position of these veins on the dorsal wall
of the mid-gut previous to the formation of the anlage of the pancreas.

This brings us to an interesting comparison of the varying lengths of
time which the left vein retains its connection with the sinus venosus in
the different reptiles. In Anguis, as we have seen, this connection is
lost before the dorsal anastomosis is made. In Tropidonotus and Kin-
osternon, on the other hand, it is retained until after the formation of
the dorsal anastomosis but degenerates before the two omphalo-mesen-
teric veins fuse further back to form the so-called posterior ring around
the intestine. In Lacerta, this connection with the sinus venosus per-

Frank A. Stromsten 465

sists still longer, and is present for a short time after the posterior ring
is formed. f

The rapid development of the liver gradually forces the stomach, to-
gether with the left omphalo-mesenteric vein, from its original median
position. At the same time this vein becomes more and more broken
up into a rete mirabile. This change in position, as well as the breaking
up of the left omphalo-mesenteric vein, causes it to be a less favorable
route for the blood to reach the heart than the sinusoids of the right half
of the liver. As a result, we find in an 8 mm. embryo (Fig. 3) that a
new path has been formed for the right vein which soon serves to carry
most of the blood from the omphalo-mesenteric veins to the heart. The
left omphalo-mesenteric vein then begins to degenerate and eventually
loses its connection with the sinus venosus. In the meantime, the anas-
tomosing branch through the liver connecting the two omphalo-mesen-
teric veins has been gradually breaking up into sinusoids so that later it
becomes entirely lost in the hepatic network. All of the blood from the
liver must now enter the sinus venosus through the newly formed hepatic
portion of the right omphalo-mesenteric vein.

The omphalo-mesenteric veins of an embryo of about 10 mm.—The
two omphalo-mesenteric veins now fuse ventral to the intestine some
distance caudad of the dorsal anastomosis to form the posterior venous
ring around the intestine (Fig. 4). The left vein craniad of the dorsal
anastomosis is now represented by a very slender branch which passes
forward and downward in the wall of the intestine without entering the
liver. The right omphalo-mesenteric vein splits soon after entering the
liver into two branches of nearly equal size which, after passing forward
a short distance, unite again to form a single vessel. Since these two
branches are to form the right hepatic revehent and the mesenteric portal
veins, while the common stem formed by their union cramially is to be
the common hepatic vein, they have been designated by these names in
Fig. 4.

Just as the right omphalo-mesenteric vein enters the liver, it receives
a small vein from the ventral surface of the stomach, the gastric vein
of Fig. 4. As this vein is to play a somewhat important role in the
formation of the portal vein of Bojanus, it is advisable at this point to
notice briefly its earlier development.

The gastric vein is first definitely seen at the time when the ventral
anastomosis between the two omphalo-mesenteric veins is formed through
the liver. It then is a very short branch, situated in the ventral mesen-
tery of the stomach anlage and opens into the anterior ventral anasto-
mosis. Later, as the liver increases in size, the vein becomes longer, and

466 Anatomy and Development of Veins of Chelonia

when the ventral anastomosis is destroyed it opens into the right omphalo-
mesenteric vein. During further development, as we shall see, it comes
to connect with the mesenteric portal vein. Finally, the left umbilical
vein unites with the gastric vein and their common trunk forms the
portal vein of Bojanus. In the meantime the twisting of the stomach
brings the gastric vein to lie on the lesser curvature of that organ.

In the 10 mm. stage, we find the first definite indications of a con-
nection being formed between the right hepatic and the right sub-

SINUS VENOSUS

RIGHT
UMBILICAL

MESENTERIC PORTAL
RIGHT HEPATIC =

LEFT
OMPHALOMESENTERIC

DORSAL ANASTOMOSIS We

VENOUS RING

RIGHT
OMPHALOMESENTERIC =|. OMPHALOMESENTERIC
‘ (LEFT )

VENTRAL ANASTOMOSIS
( POSTERIOR )

Fic. 4. Frontal reconstruction of the liver veins of a 10 mm. embryo of
Kinosternon pennsylvanicum. Ventral view.

cardinal vein (not shown in Fig. +). A small vein is formed in the
ventral border of the caval mesentery which extends from the right
subeardinal vein to the dorsal border of the liver where it apparently
breaks up into a number of finer vessels. Whether these make connec-
tions with the liver or not, it is impossible to determine with certainty.

The omphalo-mesenteric veins of an embryo of about 11 mm.—Fig. 5
is a reconstruction of the liver veins of this stage. For sake of clearness
the smaller branches and the sinusoids are omitted. The posterior ring

Frank A. Stromsten 467

around the intestine is now destroyed by the atrophy of that part formed
by the right omphalo-mesenteric vein. Thus we have the unpaired
anlage of the mesenteric portal vein formed from the originally paired
omphalo-mesenterics.

The splitting of the right omphalo-mesenteric vein to form the right

‘DUCT OF CUVIER

© _|-sueclavian ARTERY
ae

}
u AORTA

—GASTRIC VEIN

MESENTERIC
PORTAL

POSTCAVA_

OMPHALOMESENTERIC
ARTERY

---s—_

LEFT
——OMPHALOMESENTERIC

SUBCARDINAL

Fic. 5.—Frontal reconstruction of the liver veins of an 11 mm. embryo of
Kinosternon pennsylvanicum. Ventral view.

hepatic and the hepatic portion of the mesenteric portal, which was
begun in the last stage, is still more extensive in this stage. While, at
first, the right hepatic and the mesenteric portal veins are united both
eranially and caudally, they now become more and more separated
through the breaking up of these connections into sinusoids, as shown in
Fig. 5, in which this process has already begun.

468 Anatomy and Development of Veins of Chelonia

The omphalo-mesenteric veins of an embryo of about 12 mm.—In this
stage the separation of the right hepatic from the mesenteric portal is
still more complete. The original connection between the two veins is
now almost entirely reduced to sinusoids (not shown in Fig. 6). There
still remains, however, one, sometimes two, direct anastomosing branches
which may represent the remains of the original connections. These,
when present, occur just after the mesenteric portal enters the liver.

We find at this stage, in addition to the original connections between
these two veins, a series of three or four anastomosing branches that

SINUS VENOSUS

UMBILICAL
(RIGHT)
GASTRIC VEIN
COMMON 3 g
HEPATIC 4
we UMBILICAL
(LEFT)

Fic. 6. Frontal reconstruction of the liver veins of a 12 mm. embryo of
Kinosternon pennsylvanicum. Ventral view.

appear to be new formations. As shown by Figs. 6 and 7 “ 4,” they take
a somewhat curved course through the liver and open into the dorsal
side of the right hepatic vein. They appear to be formed through the
fusion of some of the sinusoids ventral to, and on the right side of, the
right hepatic vein. The first indications of these curved veins appear
in the 11 mm. stage.

In the 12 mm. stage these curved veins have reached their maximum
of development. The first one to be given off from the mesenteric portal
on entering the liver is found a short distance caudad of the gall bladder,
while the other two pass dorsal to it, as in Fig. 7. When the fourth is

Frank A. Stromsten 469

present it is found further craniad and opens into the common hepatic
vein. As shown by Fig. 6, one of these branches is given off from the
mesenteric portal almost directly opposite the opening of the gastric
vein. This branch persists in the adult form where it functions as the
chief lateral branch of the mesenteric portal, the right hepatic advehent
branch of Fig. 1. It is possible that the other branches are also retained,
in part at least, as branches of the mesenteric portal of the adult, but it
is not possible to trace their subsequent development with certainty.
The right omphalo-mesenteric vein has split still further craniad than
in the previous stage, so that, as shown in Fig. 6, there is a small portion

VERTEBRAL VEIN

CHORDA

+) POSTCARDINAL

AORTA
POSTICAVA = { Rag ® Os oat A: MESONEPHROS
RIGHT HEPATIC
A —— STOMACH
MESENTERIC PORTAL GASTRIC VEIN

GALL BLADDER -
RIGHT UMBILICAL—

LEFT. UMBILICAL

INTESTINE

Fig. 7. Cross-section of a 12 mm. embryo of Kinosternon pennsylvanicum,
at the level where the postcava enters the liver, showing the curved anasto-
mosis (A) between the hepatic and mesenteric portal veins and the opening
of the gastric vein into the mesenteric portal. The portal vein of Bojanus is
not yet formed.

(Z) of the left branch next to the common hepatic vein that is net yet
broken up into sinusoids. <A part of this short, wide stem is retained in
the adult as the terminal portion of the left hepatic revehent vein, the
other part of this hepatic vein is, as we shall see, developed from the
umbilical vein. A part of the left revehent vein, therefore, as well as
nearly the entire anlage of the right hepatic vein, is split off from the
right omphalo-mesenteric vein.

Further changes in the derivatives of the omphalo-mesenterics—The
right hepatic and the mesenteric portal veins are becoming more and
more separated. The curved anastomosing branches which were so preimi-
nent in the 12 mm. stage begin to break up in the 13 mm. stage at the

470 Anatomy and Development of Veins of Chelonia

point where they open into the hepatic vein, and in a 14 mm. embryo they
no longer connect with the right hepatic except through the general sinu-
=

soidal network of the liver. In the 15 mm. stage the derivatives of the
omphalo-mesenteric veins have practically assumed their adult relations.

THE UMBILICAL VEINS. THE FORMATION OF THE RIGHT AND LEFT
ABDOMINAL VEINS AND THE LEFT HEPATIC VEIN.

It is not possible from the material at hand to determine anything
definite about the origin of the umbilical veins in turtles. Dr. Frederic
T. Lewis, who has made a very careful study of the early development
of the umbilical veins in the rabbit, has very kindly sent the writer a
number of drawings and sketches (not yet published) which undoubtedly
show that, in the rabbit at least, the umbilical veins appear earlier than
the postcardinals and arise in part from the vitelline veins and in part
from the intersegmental arteries, thus: “ The vitelline vein sends sprouts
forward and backward in the somatopleure. The forward branch be-
comes the anterior cardinal vein; the posterior branch, incorporating in
its progress prolongations of the intersegmental arteries, becomes the
umbilical vein.” (Lewis: Am. Jour. Anat., 1904, Vol. III, p. XII.)

In the 7.4 mm. embryo of the turtle, as we have already seen, the
postcardinals and umbilicals are both present. The umbilicals are, how-
ever, much smaller than the posteardinals and are connected with them,
especially near their cardiac ends, and through the region of the anterior
limb buds, by a large number of cross branches. Caudally, the umbilical
veins cannot be followed with certainty.

In the 9 mm.embryo, however, they become very large and may easily

be followed along their entire length. At this stage (9 mm.) they are
of nearly equal size. They arise in the allantois and extend along the
body wall near the somato-amniotic angle and open into the sinus
venosus close to, or in common with, the omphalo-mesenteric vein of the
same side. In most embryos of this age the right umbilical vein appears
to open into the right omphalo-mesenteric vein, while the left opens
into the sinus venosus some distance from the omphalo-mesen-
terie vein of that side. Both umbilicals are still somewhat connected
with the postcardinals especially through the limb anlages. .
Umbilical veins of a 10 mm. embryo.—The umbilical veins are now
increasing very rapidly in size and importance. They begin, at about
this time, a series of changes which is destined to play an important role
in the development of the portal system of the liver. This is brought
about chiefly by the separation of the pericardial cavity.

Frank A. Stromsten 471

In the 7.4 mm. embryo, the pericardial cavity is separated from the
pleuro-peritoneal cavity by the septum transversum. The ducts of
Cuvier and the umbilical veins reach the sinus venosus through the
lateral portions of this septum, that is, by the lateral mesocardes of
Koelliker. Later, two folds are developed in the body-wall just caudad
of the septum transversum. These folds grow mediad, ventrad, and
caudad, separating the liver from the heart from before backward. The
umbilical veins le in the body-wall just at the region where this ingrowth
of connective tissue takes place. They are, therefore, carried mediad
with the ingrowth of these folds, beginning at the point where these
veins open into the sinus venosus and gradually extending caudad. In
the present stage (10 mm.) the left umbilical vein is just beginning to
show a slight deviation mediad, while the right is yet unaffected.

The umbilical veins now no longer connect directly with the post-
cardinal veins in the pelvic regions, but indirectly by means of a small
pair of veins which are formed on the ventral abdominal walls caudal to
the stalk of the allantois. In this stage, these veins are still very short.
They form the posterior anlages of the abdominal veins of the adult,
and are designated in this paper as the postabdominals.

Umbilical veins connect with the hepatic circulation. Embryo of
about 11 mm.—The left umbilical vein has now been carried so far
mediad that for several sections it comes to he between the heart and
the liver, and instead of opening into the sinus venosus laterally as at
first, it now opens into its dorsal side (Fig. 5). Being thus squeezed in,
as it were, between the liver and the heart, the vein becomes more or
less flattened in this region. Instead of being able to increase as rapidly
in size as the caudal portion, it really gradually becomes smaller. The
rapid development of the allantois at this time results in an enormous
increase in the size of the umbilicals. It therefore becomes necessary
for this ever-increasing amount of blood to find some way to the heart
other than the narrow cardiac end of the left umbilical vein. It so
happens that at the point where the left vein leaves the body-wall to
enter the pericardium, it comes to lie very close to the ventral border of
the liver. A connection is soon made with the hepatic sinusoids of this
region and a channel is quickly formed through the liver between the
left umbilical and the common hepatic vein (Fig. 5). In an older stage
(11.5 mm.) the right vein also sends a branch through the liver to the
common hepatic vein.

Thus we see that in turtles as well as in snakes, both umbilicals enter
the liver, while in lizards only one, the left, does so. In snakes; however,
this secondary connection with the common hepatic vein is made much

36

472 Anatomy and Development of Veins of Chelonia

earlier; in fact, it appears to occur, according to Hochstetter, before the
right vein has been broken up by the hepatic tubules. The postabdom-
inals now open into the iliacs instead of into the postcardinals.

The backward shifting of the point where the umbilical veins enter the
liver, and the formation of the first anlage of the left hepatic revehent
vein. Embryo of about 12 mm.—Both umbilical veins, as we have seen,
now send branches through the liver to the common hepatic vein (Fig. 6).
The portion of the left umbilical vein in the body-wall craniad of its
hepatic branch has completely degenerated. The right vein, on the
other hand, still retains for a short time its connection with the sinus

VERTEBRAL VEIN

POSTCARDINAL
AORTA

MESONEPHROS UBCARDINAL

SUBCARDINAL GENITAL GLAND

RIGHT HEPATIC

MESENTERIC PORTAL

GALL BLADDER

RIGHT UMBILICAL

Fic. 8. Cross section of a 13 mm. embryo of Kinosternon pennsylvanicum.

venosus, or, more accurately speaking, with the terminal portion of the
common hepatic vein (Fig. 6). Before, however, the 12.5 mm. stage is
reached there is no longer any direct connection between the umbilical
veins and the sinus venosus, so that all of the blood from these veins
must pass through the liver in order to reach the heart.

The continued backward growth of the folds of the lateral body-walls
that go to form the dorsal pericardium causes more and more of the
umbilical vein of each side to lie between the liver and the heart. In
the case of the left vein, this results in a shifting backward of the point
where this vein enters the liver and a consequent atrophy of the vein
craniad of this point. The left umbilical vein does not form an entirely
new path through the liver; on the contrary, only that portion marked XY

Frank A. Stromsten 473

in Fig. 6 is a new formation. The rest is the original path of the
11 mm. stage. The outer portion of the original path (partly hidden in
Fig. 6 by the part marked X’) is still retained as a tributary of the
hepatic umbilical and is later to form a part of the left hepatic revehent
vein of the adult.

The umbilical veins join the mesenteric portal vein and lose their con-
nection with the common hepatic. Completion of the anlage of the left
hepatic revehent vein. Embryo of 13 mm.—In the 13 mm. stage, the
umbilical veins undergo a still further shifting mediad. Just caudad

CHORDA VERTEBRAL VEIN

AORTA POSTCARDINAL

MESONEPHROS
CAVAL MESENTERY
POSTCAVA

STOMACH

PORTAL VEIN-

2 GASTRIC VEIN
OF BOJANUS

VENTRAL LIGAMENT - — SSS AIL LEFT UMBILICAL

Fig. 9. Cross section of a 13 mm. embryo of Kinosternon pennsylvanicum
through the level where the postcava enters the liver, showing the union of
the left umbilical and gastric vein to form the first anlage of the portal vein
of Bojanus.

of the pericardial cavity the right umbilical vein sends a large branch to
the mid-line of the ventral body wall, where it receives a small branch
from the left umbilical vein (Fig. 8). The median vein formed by the
union of these two branches enters the liver through the primary ventral
ligament and then immediately divides into two branches. The left
branch passes directly dorsad to open into the mesenteric portal, while
the right branch soon becomes lost in the hepatic network.

This short median vein of the primary ventral ligament corresponds
in position to the “veine mediane” described in birds by Brouha, 98.
It does not, however, arise as an independent vessel in the ventral body-

474 Anatomy and Development of Veins of Chelonia

wall, as it does in birds, but is merely the result of the fusion of the two
branches of the right and left umbilicals, as shown by Fig. 8. Further-
more, in turtles, it is only a temporary condition and merely marks one
stage in the gradual shifting of the position of the umbilical veins.

The right umbilical vein craniad of its connection with the median
vein continues forward for a short distance in the body-wall and finally
disappears (Fig. 8).

The left umbilical vein continues still further forward and enters the
liver a little to the left of the mid-lne (Fig. 9). It now forms an
entirely new channel through the liver and opens into the gastric vein
mentioned in the earlier part of this paper. The gastric vein, as we re-
member, opened at first into the ventral hepatic anastomosis between the
two omphalo-mesenteric veins, then into the right omphalo-mesenteric
vein. After the mesenteric portal vein is split off from the right hepatic
the gastric vein is split off with it, so that in the present
stage it opens into the mesenteric portal vein. That part of the gastric
vein which conveys the blood from the left umbilical vein to the mesen-
teric portal becomes very much larger than the other part and may be
considered as the continuation of the umbilical vein. We may, therefore,
now say that the left umbilical vein opens directly into the mesenteric
portal; and that one of its branches is the gastric vein. The small
common trunk formed by the union of the gastric and left umbilical
veins is the first anlage of the portal vein of Bojanus.

The original path of the left umbilical vein through the liver now
becomes a part of the left hepatic vein, the terminal end of which, as we
have already seen (Fig. 6, Z), is split off from the right omphalo-mesen-
teric vein. The left hepatic vein is, therefore, formed from two parts:
one, the terminal portion, is split off from the right omphalo-mesenteric
vein, while the other portion is the new path forméd through the
sinusoids of the liver by the left umbilical vein when it first entered that
organ.

The ground-plan arrangement of the veins of the liver has now been
laid down. The chief events in the further development of the umbil-
cal veins and their derivatives are as follows:

The right umbilical vein makes direct connections with the mesen-
teric portal, or with one of its branches.

The portal vein of Bojanus lengthens as the liver increases in width.

The common hepatic vein continually grows shorter, being absorbed
by the backward extension of the sinus venosus. The result is that the
left hepatic vein gradually approaches the sinus venosus and eventually
opens into it. A still further caudal extension of the venous sinus

Frank A. Stromsten . 4%5

causes an apparent shifting laterad of the left hepatic vein so that this
vein finally comes to open into the extreme left end of this organ as in
Fig. 1. Some of the small vessels which formerly opened into the left
hepatic vein would now open into the sinus venosus, and the number
of these would be related to the extent of the backward prolongation of
the latter.

The two veins earlier described as the postabdominals become more
important as the hinder limbs increase in size. With the degeneration
of the allantois they become the chief tributaries of the intra-embryonic
umbilicals and with the latter form the abdominal veins of the adult.

The abdominal veins are formed from three anlages: (1) the hepatic
portion is the remains of the last channels formed by the umbilical veins
through the liver to the mesenteric portal, or its tributaries; (2) the
portion between the liver and the umbilicus is the remains of the original
umbilicals which have been carried mediad by the ingrowth of the con-
nective. tissue to form the pericardium; and (3) the postabdominals
which connect with the iliacs as in the adult.

RENAL PORTAL SYSTEM.

VEINS OF THE ADULT.

Fig. 10 is a schematic diagram of the larger body veins of Kinosternon
pennsylvanicum.

The afferent veins of the renal portal system of turtles are the anterior
renal advehent, the posterior renal advehent, or hypogastric, and the
external renal advehent, or iliac veins. The efferent veins from the
kidneys, testes (or ovaries) with their ducts, and veins from the supra-
renal bodies open into the two root branches of the postcava.

The anterior renal advehent vein is a large ventral branch of the
posterior vertebral vein of its side. It comes out between the fifth and
sixth ribs and passes over the ventral surface of the kidney to open into
the posterior renal advehent vein by an end to end anastomosis. In its
course it receives an intercostal vein, and gives off numerous branches to
the kidney, testis (or ovary), and suprarenal body. The posterior ver-
tebral veins which supply the blood to the anterior advehents may be
briefly considered at this point. This pair of vessels runs the entire
length of the body on each side of the spinal column dorsal to the ribs.
They receive the first four intercostal veins and veins from the spinal
cord which pass out between the vertebree along with the spinal nerves.
Just craniad of the first thoracic vertebra they unite to form a short
stem which runs forward a short distance dorsal to the vertebral column,

476 Anatomy and Development of Veins of Chelonia

and then divides into three large branches. One of these branches con-
tinues forward just under the skin on the dorsal surface of the neck. It
receives a number of branches from the muscles of the neck, and an
anastomosing branch from the internal jugular, then divides and enters
the skull to open into the occipital sinus. The other two branches pass

EXTERNAL
—- JUGULAR
/

— VERTEBRAL vein

B— BRACHIAL
— SUBCLAVIAN

— INTERCOSTAL

INTERNAL JUGULAR
— SINUS VENOSUS

RIGHT — LEFT HEPATIC

HEPATIC
— —PORTALOF BOJANUS

GASTRIC VEIN

MESENTERIC PORTAL

POSTCAVA

ADVEHENT

_ CIRCUMFLEX
ILIAC

ISCHIADIC

LATERAL
—— CAUDAL

—LATERAL COCCYGEAL
=-SUPERIOR COCCYGEAL

LATERAL CLOACAL—

Fic. 10. Diagram showing the larger body veins of Kinosternon pennsyl-
vanicum. Ventral view.

around parallel to the anterior border of the shell and open into ‘the
external jugular of each side. Caudally they give off branches to the
dorsal surface of the kidneys, as well as the large branch to the ventral
surface.

The posterior renal advehent veins or the hypogastric veins of Bo-
janus, 19, arise from a number of small veins from the inner and hinder

Frank A. Stromsten Aly dre

parts of the pelvis (especially the external genitals, cloaca, etc.) and
pass forward to the ventral surface of the kidneys where they open into
the anterior advehent veins in the manner described above. They give
off afferent veins to the kidneys, testes (or ovaries) with their ducts, and
to the suprarenal bodies. In the large snapping turtle (Chelydra ser-
pentina) they also make direct connections with the roots of the post-
cava. Jourdain, 59, denies that such a direct connection takes place.
The writer has not been able to trace such connections in Kinosternon,
nor, indeed, in all of the snappers. In the smaller turtles these connec-
tions could not be made out except by a successful celloidin corrosion
even if they were present on account of the extreme smallness of the
veins. Bojanus, 19, shows such a connection in his figure 124, plate
XXYV, O*, of Testudo europee, but this is probably the result of his
erroneous idea of the real significance of the veins of the kidney.

The external revehent veins, or the iliacs of Bojanus, arise as dorsal
branches of the circumflex iliacs. Each receives, before reaching the
kidney, a large branch from external border of the shell (not shown in
figure). This border vein receives the terminal ends of the intercostal
veins, and the veins from the fat body.

According to Nicholai, 26, and later Martino, 41, all three of these
veins act as afferents to the kidneys. Jourdain, 59, supposed, however,
that this need not always be the case, but that sometimes the blood
current of the iliacs may be reversed. At all events, it appears that the
blood from the circumflex ilacs is usually divided so that a part of it
must pass through the capillaries of the kidney to the postcava, while
another part must pass through the hepatic network before reaching the
heart. Therefore, it does not seem improbable that at times a larger
amount of blood may pass through the iliacs to the kidneys than at
others, and that under certain conditions the blood current in this short
vein may actually be reversed.

The circumflex iliac vein (vena iliaca circumflexa) is formed by the
union of the ischiadic and the common coccygeal veins. The common
coccygeal vein arises from several veins of the caudal region, the more
important of which are: the superior coccygeal, the lateral coccygeal,
the lateral cloacal, and the lateral caudal. The latter two veins often
unite to form a short single stem just before opening into the common
coccygeal. The circumflex iliac divides in front of the pelvis into a
dorsal and a ventral branch, the iliac and the abdominal veins, respec-
tively.

The abdominal veins extend along the ventral wall of the body from
the pelvis to the liver. They are connected in front of the pelvis by a

478 Anatomy and Development of Veins of Chelonia

small anastomosing branch. As has already been stated, they open into
the portal veins of the liver.

The postcava arises by two short branches between the kidneys and
extends forward as an unpaired vessel ventrolateral to the aorta and,
after passing through the liver, opens into the right side of the sinus
venosus. The two root branches receive the renal, suprarenal, and sper-
matic revehent veins. The unpaired portion receives only the revehent
veins of the liver.

The Chelonia differ from both the Lacertilia and the Ophidia in the
possession of an unpaired postcava.

DEVELOPMENT OF THE RENAL PORTAL SYSTEM.

Development of the postcava.—The veins to be considered in this con-
nection are the postcardinals and the subcardinals. Owing to the fact
that these veins are already laid down in the 7.4 mm. embryo, it will not
be possible to trace their earliest development in this reptile. According
to Rabl, 92, and Hoffmann, 93, who investigated the earlier development
of the postcardinals in the selachians, these veins arise by a longitudinal
fusion of the intersegmental branches of the aorta. Brouha, 98, in his
investigations on the development of the liver in birds, finds that in a
chick embryo of 47 hrs., the postecardinals and umbilicals are not yet
formed. There is found, however, in the body-walls caudad of the
sinus venosus a network of vessels which open into the ducts of Cuvier.
In a 62 hr. chick embryo, two vessels, which are still connected by a
number of cross branches, are formed from this network. The outer of
these vessels is the postcardinal, while the inner represents the umbilical.
It is not until the 72 hr. stage is reached that these cross branches com-
pletely disappear. Lewis, 04, finds that in the rabbit the postcardinal
veins appear later than the umbilicals. “The umbilical vein sends
branches toward the aorta, into the posterior limb and less distinctly into
the anterior limb. A longitudinal anastomosis of these vessels, uniting
with a sprout from the venous end of the heart, produces the posterior
cardinal vein, which then is cut off from the umbilical vein, carrying
with it the veins from the limbs.” (Lewis: Am. Jour. Anat., Vol. ITI,
ps XckL-)

In a turtle embryo of 7.4 mm. the postcardinal veins extend forward
from the caudal termination of the mesonephric ducts along the entire
length of the mesonephroi, dorsolateral to the mesonephric ducts. No
distinct caudal vein is yet formed. At the caudal termination of the
mesonephric ducts the postcardinal of each side receives a number of
small branches from the tail, the last dorsal intersegmental branch, and

Frank A. Stromsten 4%9

a stout anastomosing branch from the subcardinals of the same side.
The portion of the postcardinal between the cranial end of the meso-
nephros and the sinus venosus, that is to say the azygos portion of
Lewis, 04, is broken up into two or three vessels; often one large vein
and one or two much smaller, all of which open separately into the ducts
of Cuvier. While this would seem to correspond with the condition
found in lizards, it is interesting to note that in Chelonia the post-
cardinals retain their identity as distinct vessels throughout their entire
course, and are not broken up cranially into broad blood sinuses as they
are in lizards. The conditions here met probably represent the remains
of the venous plexus described by Brouha in the chick. The postcar-
dinals are still connected with the umbilicals by a number of anasto-
mosing branches.

The postcardinals already receive a series of intersegmental branches
from the dorsal body wall and also a number of branches from the mesen-
tery ventral to the aorta. As soon as the anterior limb ridge is formed, it
sends six or eight vessels to the postcardinal. These lmb veins, as
Hoffmann, go, has already shown in the lizards, comes from between the
successive myotomes of which the limb is developed. They also connect
at this time with the umbilicals.

The name subcardinals was given by Lewis, 02, to a pair of veins
which extend along the ventromedial border of the mesonephroi and cor-
respond to the vene renales revehentes of Hochstetter. According to
Lewis, 02, 04, the subcardinal veins are split off from the mesonephric
portion of the postcardinals (“ Mesonephric azygos”). Miller, 03, claims
that “The subcardinal veins arise in birds as unconnected vessels or
islands, which are without doubt independent structures; and the con-
nections with the postcardinals are formed later and secondarily.” (Mil-
ler: Am. Jour. Anat., Vol. II, p. 286.)

In a 7.4 mm. turtle embryo the subcardinal veins extend from near
the root of the tail forward along the ventromedial border of the meso-
nephroi to a point somewhat craniad of the origin of the omphalo-mesen-
teric artery where they open into the postcardinals. The subcardinal
veins do not appear to be connected with each other in any way at this
stage, although, on the other hand, their connection with the postcar-
dinals is very complete, both through direct connections and through the
venous network of the mesonephroi.

Both the postcardinals and the subcardinals are still very small. The
mesonephroi are very simple, being scarcely more than a series of seg-
mentally arranged tubules opening into the mesonephric ducts.

As the mesonephroi increase in size and complexity, the postcardinal

480 Anatomy and Development of Veins of Chelonia

and subeardinal veins also become more important. The subcardinal
veins unite with the caudal vein as soon as it is formed and they become
the advehent veins of the primitive kidneys. Fig. 11 is a reconstruction
of the postcardinal and subcardinal veins of a 9 mm. turtle embryo.
Only the direct connections between these two sets of veins are shown.
The sinusoidal network of the mesonephroi is very complex at this stage.
The posteardinals have now reached their maximum of development in
their cranial portions, just as have the subcardinals in the caudal re-
gion. In other words the primitive portal system of the mesonephroi
is now at its height of development.

At about this time a series of changes takes
place which ultimately results in the complete
reversal of the primitive portal system of the
mesonephroi.

These changes are made possible by two im-
portant formations: (a) the right hepatic vein,
w. and (b) the caval mesentery. The formation
women of the right hepatic vein has already been dis-

cussed in an earlier part of this paper.

The development of the caval mesentery has
been frequently described by a number of in-
vestigators, Hoffmann, Hochstetter, and others,
so that it is not necessary to take it up at this
time. It arises in the cranial region of the body
as a bud or process from the median mesentery
and grows both ventrally and caudally so that
it becomes attached to the dorsal surface of the
liver further and further caudad. When it

reaches a certain point slightly craniad of the
origin of the omphalo-mesenteric artery, a short
vein is developed along its caudal border, which
unites the right subeardinal with the right
pennsylvanicum. — Ventral enatie vein and thus completes the formation
of the anlage of the postcava. The short vein formed in the caval
mesentery seems to grow from the right subcardinal toward the liver,
since in the 10 mm. stage, as noted in an earlier part of the paper, it may
be traced up to the border of the liver where it apparently breaks up into
a number of smaller vessels. It appears to later form a path for itself
through the sinusoids of the liver and thus join the right hepatic vein.

In the meantime the two subcardinal veins have become connected by

several anastomosing branches which at first are small and of about equal

AVIAN

Fre. 11. Frontal recon-
struction of the postcardinal
and subcardinal veins of a 9
mm. embryo of Kinosternon

Frank A. Stromsten 481

calibre. Later, however, some of these branches increase greatly in size
as in Fig. 12 and eventually the two veins fuse throughout almost their
entire length caudad of the origin of the omphalo-mesenteric artery to
form the unpaired postcava of the adult. In neither Kinosternon nor
Chrysemys do the subeardinals fuse, as a rule at least, in front of the
origin of this artery. In only one embryo has the writer seen the slight-
est indications of such an anastomosis.

The formation of the postcava permits the blood from the caudal
regions to enter the heart through the liver; a path which immediately
becomes very important since the postcardinals are now beginning to
degenerate cranially. Fig. 12 is a reconstruction of the postcardinal and
subeardinal veins of a 12 mm. turtle
embryo, caudad of the point where the
posteava opens into the right hepatic vein.
The postcardinals have now lost their con-
nection with the sinus venosus. Caudally,
on the other hand, they make strong con-
nections with the caudal vein.. The sub-
cardinals are rapidly losing their connec-
tion with the caudal; in fact, they arenow ~
connected with the postcardinals slightly
caudad of the opening of the iliac veins,
instead of directly with the caudal. Even
this connection is soon lost and the sub-
cardinals (or root branches of the post-
cava) now only function as revehent veins
for the mesonephroi and the portal system
is reversed just as it is in lizards and
snakes.

OMPHALOMESENTERIC
ARTERY

LEFT

‘aie
j

SUBCARDINAL
ar ~—— ANASTOMOSIS

RIGHT
SUBCARDINAL

—SUBCARONAL

POSTCARDINAL

\

As the mesonephroi degenerate from
before backward, the subcardinals fuse
further and further caudad as mentioned o¢'the posteardinal and subeardioe
above. When the permanent kidneys are Jernon penneylanteuie Veurel
formed in the pelvic region dorsal to the ‘©’
mesonephroi, they also give off veins to the subcardinals. In the oldest
stage studied, 25 mm., the subcardinals have fused as far caudad as to the
cranial end of the permanent kidneys.

The postcava of turtles is, therefore, formed from the following parts:
(a) usually a small part of the common hepatic vein, (b) the right
hepatic revehent vein, (c) the hepatic sinusoids dorsal to the right
hepatic vein, (d) the vein formed in the caval mesentery, (e) the right

482 Anatomy and Development of Veins of Chelonia

subeardinal craniad of the origin of the omphalo-mesenteric artery, and
(f) the fused subcardinals caudad of this point.

The posterior vertebral veins are formed in exactly the same manner
as in lizards except that the longitudinal fusion of the dorsal interseg-
mental branches of the postcardinals takes place dorsal to the rib anlages
instead of ventral to them. ‘These veins persist in the adult form.

The anterior and posterior advehent veins of the permanent kidneys
represent the remains of the postcardinals. The connection between the
anterior renal advehent vein and posterior vertebral vein is formed
through the persistence of one of the dorsal branches of the primitive
postcardinal.

RESUME.

In Chelonia the unpaired mesenteric portal vein is formed, in the
abdominal region, as in other reptiles, from the originally paired
omphalo-mesenterics. Within the liver it is split off from the right
omphalo-mesenteric vein.

The cranial termination of the left hepatic revehent vein as well as
the greater part of the right revehent of the liver is also split off from
the right omphalo-mesenteric.

Both umbilical veins enter the liver and at first open into the common
hepatic vein; later they join the portals.

During development the umbilical veins are carried mediad by the
ingrowth of the connective tissue folds which are to form the pericar-
dium; and also enter the liver further and further caudad. Eventually
they come to form the pre-umbilical portion of the abdominals. The
portion of the abdominals behind the umbilicus is developed from a
pair of veins which extend forward along the ventral abdominal walls
from the iliac veins and open into the umbilicals at the umbilicus.

The portal vein of Bojanus is not developed from the omphalo-mesen-
terics, but from a venous trunk formed by the union of the left umbilical
with the gastric vein.

The postcava is formed from the following parts: (a) the common
hepatic vein, (b) the right hepatic revehent vein, (c) sinusoids of the
liver, (d) a new venous formation in the caval mesentery, (e) the right
subeardinal craniad of the origin of the omphalo-mesenteric artery, and
(f) the fused subcardinals caudad of this point.

The vertebral veins develop as a longitudinal fusion of the dorsal
intersegmental branches of the postcardinals above the costal anlages.

The caudal portions of the postcardinals are retained, as in other rep-
tiles, as the anterior and posterior advehent veins of the permanent
kidneys.

Frank A. Stromsten 483

In general the development of the veins of the hepatic and renal portal
systems is the same in turtles as in lizards and snakes. The more im-
portant differences are noted below:

Turtles differ from lizards as follows:

1. The left umbilical vein loses its connection with the sinus venosus
before the posterior venous ring around the intestine is formed.

2. Both umbilical veins enter the liver, and both persist, in part, to
form the pre-umbilical portion of the abdominals of the adult.

3. The subeardinals do not form a venous ring around the origin of
the omphalo-mesenteric artery; hence, the right subcardinal, instead of
an unpaired stem in front of the origin of this artery, joins the right
hepatic vein to form the postcava.

4. The postcava is a single vessel, in the adult, back as far as the cranial
ends of the permanent kidneys, where it bifurcates into two branches
which receive the suprarenal, renal, and spermatic veins.

5. The posterior vertebral veins are developed dorsal to the anlages of
the ribs.

6. The postcardinals are not broken up in the early stages into broad
blood sinuses by the mesonephric tubules as they are in liaards, but
retain their integrity as independent vessels along their entire length.

Turtles differ from snakes as follows:

1. The right emphalo-mesenteric vein is, at first, smaller than the left.

2. The right umbilical vein does not enter the liver until after the
left has done so.

3. Both umbilicals eventually open into the portals.

4. The postcava does not open into the common hepatic vein formed
by the union of the right umbilical with the right hepatic, but into the
right hepatic before the latter joins the common hepatic.

5. The postcava, as stated above, is unpaired.

6. The postcardinals are not, at first, connected with the caudal, but
a connection is later and secondarily formed.

7. The posterior vertebral veins are tormed above the rib anlages, and
persist in the adult.

484 Anatomy and Development of Veins of Chelonia

LITERATURE.

AGASsiz, LOUIS, 57.—Contributions to the Natural History of the United
States, Vol. II, Part III. Embryology of the Turtle; with thirty-
four plates. Boston.

Boganus, Lupvicus HENRICUS, 19-21——Anatome Testudinis Europee; with
thirty-nine plates. Vilne.

Brouna, M., 9g8.—Recherches sur le Développment du Foie, du Pancréas, de
la Cloison Mésentérique et des Cavités Hépato-entériques chez les
Oiseaux. Journal de l’Anatomie et de la Physiologie, XXXIVe.
Année, pp. 305-363.

Corti, A., 47.—De Systemate vasorum Psammosauri grisei. Vindobone.

CHORONSHITZKY, B., 00.—Die Entstehung der Milz, Leber, Gallenblase, Bauch-
speicheldrtise und des Pfortader Systems bei den verschiedenen
Abteilungen der Wirbeltiere. Anat. Hefte, Band 13.

GEGENBAUR, CARL, o1.—Vergleichende Anatomie der Wirbelthiere. Leipzig.

GRANT, Rospert E., 41.—Outlines of Comparative Anatomy. London.

GRATIOLET, J., 53.—Note sur le Systéme veineux des Reptiles. Journal de
Institute, T. XXI.

GROSSER, O., and BReEzINA, E., g5.—Ueber die Entwickelung der Venen des
Kopfes und Halses bei Reptilien. Morph. Jahrb., B. XXIII.

HALLER, B., 04,—Lehrbuch der Vergleichenden Anatomie. Jena.

HOCHSTETTER, FERDINAND, 88.—Beitrage zur Entwickelungsgeschichte des
Venensystems der Amnioten. I. Htihnchen. Morph. Jahrb., B. XIII,
Sb GIy.

——— 88.—Beitrage zur vergleichenden Anatomie und Entwickelungsge-
schichte des Venensystems der Amphibien und Fische. Morph. Jahrb.,
B. ST.

——— 87.—Ueber die Bildung der hinteren Hohlvene bei den Saugetieren.
Anat. Anz., B. II.

— 88.—Ueber den Hinfluss der Entwickelung der bleibenden Nieren auf
die Lage des Urnierenabschnittes der hinteren cardinalen Venen.
Anat. Anz., B. III.

—— 88.—Ueber das Gekrése der hinteren Hohlvene. Anat. Anz., B. III.

—— g1.—Ueber die Entwickelung der Extremitaétvenen bei den Amnioten.
Morph. Jahrb., B. XVII.

——— 93.—Entwickelung des Venensystems der Wirbeltiere. Ergeb. d. Anat.
u. Entwick., B. III.

—— 93.—Beitrage zur Entwickelungsgeschichte der Amnioten. II. Rep-
tilien: Morph. Jahrb, B: XDX, S. 428:

——— 93.—Beitrage zur Entwickelungsgeschichte der Amnioten. III. Sauger.
Morph. Jahrb., B. XX, S. 548.

————— 94.—Ueber die Entwickelung der Abdominalvene bei Salamandra
maculata. Morph. Jahrb., B. XXI.

——— 98.—Bemerkungen zu Zumstein’s Arbeit “ Ueber die Entwickelung der
V. cava inferior beim Maulwurfe und bei dem Kaninchen.” Anat.
lake, 16h, 88) 1b, o<

—— o03.—Handbuch der Vergleichenden und Experimentellen Entwickel-
ungslehre der Wirbeltiere (Hertwig), Vol. III, p. II, S. 129.

Frank A. Stromsten 485

HorrMan, C. K., 84.—Beitrage zur Entwickelungsgeschichte der Reptilien.

Zeitsch. f. Wiss. Zoologie, B. 40.

85.—Weitere Untersuchungen zur Entwick. der Reptilien. Morph.

eater os, 18), lal.

go.—Bronn’s Klassen und Ordnungen des Thierreichs. B. VI, Abth.

III, Reptilien. Leipzig.
93.—Zur Entwickelungsgeschichte des Venensystems bei den Sela-
chiern. Morph.. Jahrb., B. XX.

HunTINGTON, GreorGE S., g7.—Corrosion Anatomy, Technique and Mass. Pro-
ceedings of the Association of American Anatomists. Washington,
May 4-6, 1897.

JaKkopson, S. L., 13,—Recherches anatomiques et physiologiques sur un
systéme veineux particulier aux Reptiles. Meckels Archives, T. III,
p. 154.

Jacquart, H., 55—Mem. sur les organes de la Circulation chez serpent Python.

Ann. des Sc. nat., 4 Serie, T. IV.

58.—Sur plusieurs points du syst@me veineux abdominal du Caiman a

museau de brochet. Ann. des Sc. nat., 4 Ser., T. IX, p. 129.

JouRDAIN, S., 59.—Sur le Systéme V. porte renale. Ann. des Sc. nat., 4 Ser.,
Aes Deal

Lewis, FREDERIC T., 02.—The Development of the Vena cava inferior. Ameri-

can Journal of Anatomy, Vol. I, p. 283.

o4.—The Intra-embryonic Blood Vessels of Rabbits from 81% to 13

days. Amer. Jour. Anat., Vol. III, p. XII.

o4.—The Question of Sinusoids. Anat. Anz., 18), 7A5)-

Martin, H. Newer, and Moats, W. A., 81.—Handbook of Vertebrate Dissec-
tion. Part I. How to Dissect a Chelonian. New York. *

MARTINO, A. DE, 41.—Memoire sur la Direction de la circulation dans le
systéme renal de Jacobson chez les Reptiles. Ann. des Sc. nat., Ser.
IO, ike als 10s GOR.

McCuure, CHARLES F. W., 03.—A Contribution to the Anatomy and Develop-
ment of the Venous System of Didelphys marsupialis. Part 1.
Anatomy. Amer. Jour. Anat., Vol. II.

Mitter, A. M., 03.—The Development of the Postcaval Vein in Birds. Amer.
Jour. Anat., Vol. II.

MILNE Epwarps, H., 58.—Lecons sur la physiologie et l’anatomie comparée.
ABS UE

Minor, C. S., 00.—On a hitherto unrecognized form of blood circulation with-
out capillaries in the organs of vertebrates. Proc. Boston Soc. Nat.
History, Vol. 29.

Nuun, A., 86.—Lehrbuch der Vergleichenden Anatomie. Heidelberg.

Owen, RicHarp, 66.—Anatomy of Vertebrates. London.

RATHKE, H., 48.—Ueber die Entwicklung der Schildkroten. Braunschweig.

WIEDERSHEIM, ROBERT, 02.—Vergleichende Anatamie der Wirbelthiere. Jena.

PROCEEDINGS OF THE ASSOCIATION OF
AMERICAN ANATOMISTS

EIGHTEENTH SESSION.

Wistar Institute of Anatomy, Philadelphia, Pennsylvania,
December 27, 28 and 29, 1904.

At its business sessions the Association took the following actions:

The secretary’s mimutes were corrected to the extent of making the fol-
lowing order, Section 2 of Article VIT of the constitution, as was the in-
tention of this Association by the action taken at its Washington meeting
(1902): “ Any change in the constitution of this Association must be
presented in writing at one meeting in order to receive consideration and
be acted upon at the next meeting; due notice of the proposed change to
be sent to each member at least one month in advance of the meeting at
which such action is to be taken.”

Section 2 of Article VII of the constitution as prior to this change thus
becomes Section 3 of Article VII.

The Committee of Codperative Investigation reported as follows:

“Circulars were sent to all members of the Association in September,
1903, requesting that observations be made upon three topics, namely:

“1. The origin, insertion and nerve supply of the brachialis anticus.

“2. Variations in the hepatic artery.

“3. Variations in the external origin of the spinal accessory nerve.

“The circulars also contained a request that the data collected upon any
or all of these topics should be reported to the chairman of this committee
and a similar request was recently included in a notice issued by the
Secretary of the Association.

“The committee regrets to state that but three reports have been
received, too few to be of value in the line of investigation which the
committee had in view. It would seem that for various reasons the
members of the Association are not interested in the object for which the
committee was appointed and the committee therefore requests that it
be discharged.”

Respectfully submitted,

J. Puayrarr McMorricu, Chairman.
AMERICAN JOURNAL OF ANATOMY.—VOL. IV.

19

Il Proceedings of the Association of American Anatomists

On motion this report was accepted and the committee discharged with
thanks.

It was moved by Doctor Burt G. Wilder and seconded by Doctor Edw.
Anthony Spitzka:

1. That there be appointed a Committee on Brain Bequests and
Methods.

2. That said committee be requested to report at as early a meeting as
practicable upon the following points: (a) A General Form of Brain Be-
quest; (b) Methods of removing the brain and of its transportation,
preservation, study, delineation and description.

3. That the committee consist of Doctors Ales Hrdlicka, Edw. Anthony
Spitzka and Burt G. Wilder. .

Motion was carried and Doctor Burt G. Wilder was made chairman of
the committee.

Forty-one new members were elected.

Dr. Charles R. Bardeen was elected member of the Executive Com-
mittee for term expiring in 1909.

TREASURER’S REPORT FOR THE YEAR 1904.

Balanceronsnandebecembert24- 1903s eeaec eee eee $126.94
DUC eeLOTY LO OLE Ae. A Reversal ate tatatenche tsuenovedon anche te hover eye: SCHGIal Sua oNC Reon $00.00

$926.94 $926.94
Expenditures for 1904:
Expenses of Secretary for Philadelphia meeting, 1903... $24.00
To Association’s share of expenses for last Congress of

American Physicians and Surgeons................. 84.80
To American Journal of Anatomy for 162 subscriptions

BE DASD Oi sodans ex ehevaverere it ctiav oncheraveee iain os tne hol Sie ap alcve een C oR ARR Tne 729.00
To American Journal of Anatomy for reprints of Pro-

ceedines7and AbstractsnOr DPApPerS sacs sce ioc ce ccieciate 8.85
Mnvelopess and spostacei cioceicictecre ceeleieeroreieceie eiciotcie rere 18.00

i Sai gah OUI 602 Rees Senet tA eC Pee PAE nreONy ON REIOECRC UTI RPO EG OO Groin 14.25

$878.90 $878.90

Balance on hand December 24, 1904................ $48.04

Dr. Charles $8. Minot, Embryological Laboratory, Harvard Medical
School, Presidential Address.—Genetie Interpretations in the Domain
of Anatomy. (See American Journal of Anatomy, Vol. IV, No. 2.)

Proceedings of the Association of American Anatomists DEE

ABSTRAGTS:OF, PAPERS PRESEN FED:

SOME CURRENT TERMINOLOGIC INCONSISTENCIES THAT ARH
APPARENTLY NEEDLESS. By Burt G. Wiper. Cornell University,
Ithaca, New York.

A NEW FORM OF BRAIN BEQUEST. By Burt G. WiLpEeR. Cornell Univer-
sity, Ithaca, New York.

REPORT OF A STUDY OF THE BRAINS OF SIX EMINENT SCIENTISTS
AND SCHOLARS BELONGING TO THE AMERICAN ANTHROPO-
METRIC SOCIETY; TOGETHER WITH A BRIEF DESCRIPTION
OF THE SKULL OF ONE OF THEM. (Specimens, drawings and
charts.) By Epw. ANTHONY SPITZKA. Department of Anatomy,
Columbia University, New York City.

A report was made on the brains of Prof. Joseph Leidy, Dr. Philip
Leidy, Dr. A. J. Parker, Prof. Harrison Allen, Prof. E. D. Cope and
Dr. William Pepper, together with a brief description of the skull of
Prof. Cope. The specimens belong to the collection of the American An-
thropometric Society, an association of scholars and scientists having for
its object the preservation and study of the brains and other bodily
organs of its members, which was founded in 1889 or 1890 by Drs.
Joseph Leidy, Harrison Allen, William Pepper, F. X. Dercum and E. C.
Spitzka. The study of the specimens extended over a period of two
years, being conducted in connection with similar researches on many
other human brains of -various classes and races of men. Systems of
brain measurement were devised and applied to all the brains to aid
in the comparative study. The entire work, fully illustrated, will be
published from the Wistar Institute, where the brains of the six men
referred to are now deposited.

The brain-weights are:

Philip heidy .. ise eye ere 1415 grams
H.: D. Cope: scars eects ae 1545 “
Harrison’ Allens) ieteepieeenst. 13 ees
Wim... Pepper ss .a see eee 1593 =

The weight of Prof. Leidy’s brain may be estimated at 1545 grams at
least; it may have weighed more. Dr. Parker’s brain probably weighed
about 1475 grams. The main morphological characteristics and contrasts
among the brains were briefly pointed out, the brains of the two Doctors
Seguin, Maj. J. W. Powell, George Francis Train and Maj. J. B. Pond
being included in the comparisons. Among the many interesting results
of the study, a comparative tabulation of the cerebro-cerebellar ratio of
weight showed this to be fully a unit higher than among ordinary men

IV Proceedings of the Association of American Anatomists

(1:7.5). The cross-section area of the callosum among 10 eminent men,
ranging from 5.7 to 10.6 sq. cm. (Joseph Leidy) averaged 7.3 sq. em.;
among the same number of ordinary men, the range was 4.7 to 6.7 sq.
em., averaging 5.6 sq. cm. ‘The significance of this redundancy in its
relations to the greater elaboration of brain structure as expressed by this
commissural system was discussed in detail but cannot be abstracted here.

The skull of Prof. E. D. Cope is in fairly good condition and is re-
markable for the proportionately large size of the cranium as compared
with that of the face, in this respect approaching that notable skull
of Immanuel Kant. The parietal bones are notable for their large ex-
panse. The cranial capacity is 1645 cubic centimeters. Full measure-
ments and stereographic drawings have been recorded.

ON A RACIAL PECULIARITY IN THE BRAIN OF THE NEGRO. By
ROBERT BENNETT BEAN. Department of Anatomy, Johns Hopkins
University. z

In a study of thirty-seven brains of the American Negro, and after
careful measurement and comparison (individually and collectively)
with seventeen brains removed from American Caucasians, it is found
that the anterior part of the Negro brain is smaller than the anterior
part of the Caucasian brain. The difference varies up to ten millimeters
actual linear measurement over the anterior association area in brains of
the same antero-posterior diameter. This is apparently more marked
in women than in men. The uniformity of the difference is striking
when we consider that the American Negro contains various strains of
Caucasian blood.

We believe we have found that the anterior association area with its
connections is smaller in the American Negro than in the American
Caucasian. The posterior part of the Negro brain may be relatively
larger than the posterior part of the Caucasian brain, but this is not so
marked, nor is it constant. We present only a preliminary note and hope
to make a complete report after the examinatioa of a larger number of
brains.

THE MAMMALIAN LOWER JAW. By J. S. Kinestey. Tufts College.

In several works, notably the manuals of human anatomy, the state-
ment occurs that the lower jaw in the mammals ossifies from several
centers, but, except in a brief paragraph in one of Kitchen Parker’s
publications, I have found no attempt to homologise these elements with
the bones occurring in the mandible of the lower vertebrates. The work
described has been carried on by means of wax reconstructions of the
region concerned in embryos of the pig.

Proceedings of the Association of American Anatomists ix

In the lower jaw are two cartilaginous elements, the well-known
Meckel’s cartilage and a second very large cartilage lying outside the
posterior part of the Meckelian and extending downward to form the
angle of the definitive jaw and upwards to form the condyle for articula-
tion with the glenoid fossa of the squamosal. This condyloid cartilage,
which ossifies by endochondrostosis, the author, like Parker, would
homologise with the lower labial cartilage of the elasmobranchs which can
be traced upwards at least as far as the ganoids, where it occurs in
Polypterus (Van Wijhe) and Amia (Allis). In front this condyloid
bone becomes confluent with the dentale, a membrane bone arising on the
external surface of the Meckelian cartilage. On its anterior-superior
margin is a second membrane bone, the coronoid, which forms the
coronoid process of the upper jaw. On the inner side of the dentale
oceurs a third membrane bone in the position of the splenial of the
lower vertebrates, while the inner alveolar margin of the anterior end of
the dentale may prove to be the prosplenial, but this point has not yet
been decided. The articulare is, as has been pointed out in a previous
paper on the ossicula auditus, to be found in,a part of the malleus, to
which a distinct membrane bone, the angulare, also contributes.

NOTES ON THE ORIGIN OF THE CAROTID GLAND AND THE MORPHO-
LOGICAL COMPARISON OF THE TRIGEMINAL AND FACIAL
NERVES IN MAMMALIAN EMBRYOS. By Henry Fox. The Temple
College, Philadelphia.

According to one view, the carotid gland arises by a thickening of
the walls of the capillaries derived from the carotid artery; according to
another view, which receives its strongest support in the careful investi-
gation of Prenant, the organ arises from the endoderm of the third
pharyngeal pouch. The author’s investigations support the latter view.
In a pig embryo of 9 mm., the gland appears as a series of blind follic-
ular outgrowths from the anterior wall of the third pharyngeal pouch.
By the study of several succeeding stages, these outgrowths were traced
into the definitive carotid gland of later stages. Similarities in structure
in the case of the trigeminal and facial nerves are strikingly brought out
in the study of early pig embryos. The trigeminal nerve bears the
same relation to the mouth (stomatodeum) that the facial does to the
first pharyngeal pouch. The Gasserian ganglion lies above the angle of
the mouth; the geniculate, above the first pharyngeal pouch. Hach
ganglion gives off two nerves, one of which passes in front of the cor-
responding pouch, the other behind it. The former corresponds to the
pretrematicus of lower vertebrates; in the mammals it becomes the
superior maxillary nerve in the case of the trigeminal; in the case of

VI Proceedings of the Association of American Anatomists

the facial it becomes the large superficial petrosal nerve. Both nerves,
superior maxillary and petrosal, arise exclusively from the ganglia.
The posterior branch on the other hand, is joined by fibers from the
ventral roots of the nerves and hence in both cases the branch thus
formed—the postrematicus—is a mixed nerve. In the case of the tri-
geminal, it becomes the inferior maxillary nerve; in the case of the facial
it forms the main trunk of the nerve. The latter divides in the hyoid
arch. The anterior branch which receives most if not all of the fibers
derived from the ganglion passes beneath the point of union of the
first pharyngeal pouch with the skin and enters the mandibular arch;
it becomes the chorda tympani nerve. The posterior branch remains in
the hyoid arch, its fibers being distributed to the muscle anlagen ; it cor-
responds to the facial proper of the adult.

The ophthalmic branch of the trigeminal has no counterpart in the
facial nerve in the mammalia.

NOTE ON THE LIGAMENTS OF THE MAMMARY GLAND. By Wi1iAmM
Ker~tter. Medical Department, University of Texas, Galveston.

Besides the cutaneous ligaments of Cooper, the fibrous tissue of the
mammary gland spreads out peripherally in all directions so as to form
a continuous sheet, which binds the gland to the deep fascia beneath the
clavicle, along the side of the sterum, to the fascia of the lower costal re-
gion and to the fascia of the floor of the axilla. This fascia forms the
true suspensory ligament of the gland. It is rich in lymphatics which
run in all directions and is of the utmost surgical importance. So far
as the writer has been able to inquire, it is not adequately described.

A NOTE ON THE DEVELOPMENT OF THE GiSOPHAGEAL EPITHELIUM.
By R. H. WHITEHEAD. Department of Anatomy, University of North
Carolina.

Within the past decade a number of observers have reported finding in
the upper part of the human csophagus small areas which greatly re-
semble the mucous membrane of the cardiac end of the stomach con-
taining tubular glands similar to the cardiac and fundus glands. In
the attempt to explain their occurrence, some authors maintain that they
are due to developmental disturbances, that they result from a downgrowth
of ectodermal epithelium from the oral cavity, which replaces the original
endodermal epithelium over the entire cesophagus, except in minute areas
which remain and develop like the gastric epithelium. Others hold that
these areas are normal structures and two authors have reported finding
their anlagen in human embryos. The writer finds that in the pig the
stratified squamous epithelium of the cesophagus is developed im loco from

Proceedings of the Association of American Anatomists VII

the original epithelium of the foregut, and that there is no downgrowth
of ectodermal epithelium from the oral cavity. This conclusion is in
accord with the observations of Neumann and of Shaffer of human
embryos. In the pig, however, the details of the process are somewhat
different, chiefly in the absence of ciliated cells at any time; nor were
any appearances noted which might be interpreted as anlagen of the
superficial tubular glands mentioned above.

METHODS OF PREPARING THE SALIVARY GLANDS FOR STUDY AND
MUSEUM PURPOSES. By Cuartes F. Sytvester. Department of
Biology, Princeton University.

MORPHOLOGY OF THE SALIVARY GLANDS. (From Material at Columbia
and Princeton Universities.) By CHuRCHILL CARMALT. Department of
Anatomy, Columbia University, New York City.

With lantern slides from photographs and drawings of specimens
in the Museum of Columbia University and Princeton, was presented
the salivary apparatus of Alligator mississippiensis, Chelonia mydas,
Ornithorhynchus anatinus, Echidna hystrix, Didelphys virginiana,
Petauroides volans, Tamandua bivittatta, Myrmecophaga jubata, Dasypus
sexcinctus, Lepus sylvestris, Sciurus carolinensis, Manatus americanus,
Equus caballus, Bos taurus, Canis familiaris, Arctocephalus, Scalops
aquaticus, Pteropus, Cynocephalus, Macacus inuus and human parotid,
submaxillary, sublingual, labial, pharyngeal, anterior and posterior lin-
gual, palatal and uvular glands. Cross sections, models and corrosions
of these same human glands illustrate relations.

In conclusion there is morphologic constancy and identity of salivary
apparatus throughout the mammalian series.

There is a group of glands along the alveo-buccal borders with com-
pound acinous elements, the parotid at its posterior or caudal union.

There is a group of glands along the alveo-lingual groove, the anterior
end of which becomes compound with one, two or three ducts in different
orders, represented as variations in human subjects.

The glands correspond in size with the digestive function, herbivorous
animals having large, carnivorous small glands. ‘This only applies in
a wider sense, for carnivores like the bear with herbovore habits grow
larger glands.

The submaxillary gland is dependent for its size upon mobility of the
tongue.

Salivary glands increase in size with age and with pregnancy.

TOPOGRAPHY OF THE PANCREAS IN THE HUMAN FQ:TUS. By C. M.
Jackson. Department of Anatomy, University of Missouri.

VIII Proceedings of the Association of American Anatomists

PRELIMINARY REPORT OF EXPERIMENTAL WORK AND OBSERVA-
TIONS ON THE AREAS OF LANGERHANS IN CERTAIN MAMMALS.
By Lypra M. De Wirr. Department of Histology and Embryology, Uni-
versity of Michigan.

I. By wax-plate reconstruction of injected and uninjected areas of
Langerhans in man, rabbit, cat and rat, it is seen that these islands are
spherical or oval or irregular or lobulated in form, vary -greatly in size,
both in animals of the same species and in animals of different species ;
they consist of anastomosing irregular cords of cells forming a sponge-
hike structure. One or several thin-walled blood vessels pass in tortuous
course through the central part of each area, divide several times, the
branches being connected by many anastomosing capillaries. Numerous
capillaries and several larger vessels leave the areas and communicate
freely with the interacinar blood vessels. The large and irregular size
of the vessels, the thin walls and the similarity in structure of the afferent
and efferent vessels indicate that the circulation in the areas is sinusoidal.

IT. The operations on the twenty-eight cats used consisted in hgation
of the duct, division of the gland between two ligatures, cauterization of
the cut ends and sometimes tying a fold of omentum around the cut
ends of the gland. The animals were killed in from three to two hun-
dred days after the operation. Portions of the gland were removed for
microscopic study and the remainder extracted with water or glycerine
and its digestive action on proteids, starches and fats and also its
glycolytic action tested, the results in all cases being controlled by similar
tests with extracts of the normal portion of the same pancreas. ‘The
microscopic changes consisted of atrophy of lobules and acini and increase
of connective tissue, the areas of Langerhans being in most cases well
preserved. In seven of the most successful experiments, the amylolytic,
steatolytic and proteolytic action was absent and in many others it was
much diminished, while the glycolytic action was not in any case
diminished.

My results seem therefore to point to:the conclusion that the areas of
Langerhans are vascular glands having a sinusoidal circulation in inti-
mate relation with the acinal blood supply; that they have a secretion
whose glycolytic action is equal to that of the normal pancreas and which
corresponds in its reactions and behavior to the activator principle of
the pancreas described by Cohnheim.

ON THE ANATOMY OF A 4.0 MM. EMBRYO FROM THE HARVARD
EMBRYOLOGICAL COLLECTION. By J. L. Bremer. Embryological
Laboratory, Harvard Medical School.

Proceedings of the Association of American Anatomists {XxX

THE TOTAL FOLDS OF THE FOREBRAIN, THEIR ORIGIN AND DE-
VELOPMENT TO THE THIRD WEEK IN THE HUMAN EMBRYO.
By Susanna Puetps Gace. Laboratory of Histology and Embryology,
Cornell University.

The first pair of total folds appears in birds and lower forms when
the flat neural plate is molded over the blind end of the enteron. With
the growth of the neural plate and its dipping downward over the end of
the enteron, these folds are carried with it. ‘This early association of
enteron and neural plate is soon lost, the notochord first and the meso-
derm later intervening between the end of the enteron and the part of
the neural plate just mentioned. These first neural folds later be-
come the albicantial folds and in mammals (man and cat) are relatively
most conspicuous when the gill clefts are at their height of development
(third to fifth week in man). In man while the neural plate is growing
forward, but is still widely open (Homo 12, Johns Hopkins University
Collection), these albicantial folds and those giving the first indication
of eyes can be seen. As the plates further coalesce (Homo 714, Harvard
University embryological collection), in addition to the above folds
and nearer the margin of the plate, are two pairs of folds representing
the olfactory and diencephalic regions. With the final closure of the
plate to form the neural tube (Homo 209, Johns Hopkins University col-
lection), the four pairs of folds already formed, albicantial, visual,
olfactory and diencephalic, unite to form corresponding lobes.

At three weeks, there being still a remnant of the neuropore which
marks the point of the closure of the neural tube (Homo 148, Johns
Hopkins University collection), these four lobes show further differentia-
tion: (a) the albicantial being divided into (1) the albicantial and (2)
hypophyseal folds; (b) the visual being divided into (3) the eye vesicle
and (4) the eye stalk; (c) the olfactory divided into (5) the striatal,
(6) the olfactory and (7) the cerebral folds; (d) the diencephalic divided
into two folds (8 and 9). That is, at the end of three weeks the main
folds seen in the forebrain are already outlined.

The human embryos above mentioned were placed at my disposal by
Drs. Mall and Minot for comparison with the large series of mammalian
and immammalian material in the embryological collection of Cornell
University.

THE DEVELOPMENT OF THE PARAPHYSIS AND THE PINEAL
REGIONS OF NECTURUS. By JoHN WarRREN. Embryological Labora-
tory, Harvard Medical School.

The paraphysis appears first in an embryo of 12 mm. as a small
diverticulum from the roof of the forebrain anterior to the velum

Xx Proceedings of the Association of American Anatomists |

transversum. ‘The epiphysis arises from the roof of the midbrain some
little distance posterior to the velum at a slightly earlier stage. At 15
mm. the paraphysis is a slender tube parallel to the velum and opening
just anterior to its lower extremity. The roof of the forebrain just
anterior to it has grown downward to a marked degree so that a very
marked angle has been formed in the roof of this part of the brain. The
epiphysis is attached to the brain by a short stalk which, however, con-
tains no cavity and the posterior commissure appears for the first time
just behind the stalk. From now on the paraphysis begins to give oif
lateral tubules. It les in close relation to two vessels, one anterior
and one posterior. The velum grows rapidly backward nearly to the
hindbrain and another outgrowth arises from the roof of the fore-
brain just anterior to the paraphysis and extends downward and back-
ward towards the infundibular recess. The vessels in relation to the
paraphysis grow into these projections and form two choroid plexuses,
a superior and an inferior, between the origins of which opens the para-
physis. The choroid plexus of the lateral ventricles springs from the
base of the inferior plexus. The superior commissure appears first at
16.5 mm. In the adult the paraphysis extends forward between and
above the hemispheres. From its central cavity, a confused mass of
tubules are given off, which appear to anastomose to a certain extent
and each tubule is in intimate relation with a vessel, no connective
tissue lying between them. This is apparently a type of sinusoidal cir-
culation and the structure is of a glandular character. The choroid
plexuses are very extensive and surround the opening of the paraphysis.
They intermingle to a great extent with each other and extend backward
to the hindbrain. The epiphysis is a narrow tube extending forward
over the superior commissure and connected to the brain by a short
slender stalk. Its cavity is subdivided to a certain extent by incomplete
septa.

THE DEVELOPMENT OF THE CUTANEOUS NERVES OF THE POS-
TERIOR LIMB IN MAN. By CwHartes R. BARDEEN. Department of
Anatomy, University of Wisconsin.

The cutaneous nerves of the posterior limb in man first extend out to
the anterior, distal and posterior margin of the limb-bud and from these
margins they then grow over the ventral and dorsal surfaces of the limb.
The segmental distribution of the sensory fibers of the spinal nerves
which supply the limb described by Head, Sherrington and others, seems
to be due to the fact that each spinal nerve has a path of maximum
directness toward the margins of the limb bud at the period when the
sensory branches of the main nerves of the limb are given off.

Proceedings of the Association of American Anatomists XI

NORMAL PLATES OF THE DEVELOPMENT OF THE RABBIT. By
Ewine Taytor. Embryological Laboratory, Harvard Medical School.

The standard plates of the development of the rabbit will form
one of the series of “ Normentafeln zur Entwickelungsgeschichte der
Wirbeltiere * edited by Prof. Franz Keibel of Freiburg in Breisgau.
The work will have three plates. Two of these will show, on a uniform
magnification of five diameters, the entire series of embryos, extending
from the stage of the blastodermic vesicle with a circular embryonic
shield up to the twenty day condition. The third plate will show the
young stages up to and including the ten and a half day embryo on a
magnification of twenty or ten diameters. A description of the embryos
corresponding to the figures on the plates, will constitute the text of the
Normentafeln. A tabulation of the internal development of the rabbit,
based on a study of the serial sections, will be included in the work.
Three series, cut in three different planes, but measuring approximately
the same, were studied for each period of development, except in the
younger stages of eight and a half days and under. The specimens
for each period were obtained, except as otherwise noted, by selection
from several litters of the same age. The entire work is practically
completed, excepting the bibliography, which now contains more than -
five hundred titles.

THE ANATOMY OF THE INGUINAL AND FEMORAL REGIONS WITH
SPECIAL REFERENCE TO THE ARRANGEMENT OF THE FASCIA.
By ArtrHuur S. VospurcH. Department of Anatomy, Columbia Univer-
sity, New York City.

The subject was presented from two standpoints:

1st. A description of the structures, layer by layer, illustrated by
diagrams, charts and drawings.

2nd. The same structures were shown in a series of sagittal and
transverse sections through the groin and proximal portion of the
thigh.

(Sections and Drawings.

The following points were emphasized:

1. Formation of triangular ligament from fibers derived from medial
pillar of opposite side.

2. Absence of anything that can consistently be described as “ Con-
joined tendon.” The structure usually so described is a small dorsal
lamina of the internal oblique spht off from the main muscular mass
by the passage of the abdominal nerves. The structure often mistaken
for “conjoined tendon” is the transversalis fascia forming the dorsal

XII Proceedings of the Association of American Anatomists

wall of the inguinal canal, here strengthened by scattered fibers derived
from the internal oblique and the transversalis.

3. The impossibility of “conjoined tendon” forming a covering for
“direct inguinal hernia,” it being a thin or thicker portion of trans-
versalis fascia according to point of impact.

+. Formation and description of the fascial compartment in which
the femoral vessels lie. The femoral sheath was described as made up
solely of areolar tissue. This areolar tissue is continuous with the
areolar sheaths of vessels and with the pro- and retro-peritoneal tissue.

5. Femoral canal does not exist. It is sunply a complementary space
mesad to the vein, filled with lymphatic structures and areolar tissue.

GLYCOGEN IN ANIMAL TISSUE. By Srvon H. Gace. Laboratory of
Histology and Embryology, Cornell University.

Assuming the correctness of Bernard’s generalization, that in the
course of development the liver gradually assumes and the rest of the
tissues and organs of the body lose the glycogenic function, it was hoped
that the presence or absence of glycogen in the blind sae, the so-called
liver, of Amphioxus might aid in determining its homology with the
liver of higher vertebrates. Abundant material was obtained directly
from the ocean at the Bermuda Biological Station during the session
of 1904.

For comparison the most closely related fresh-water form, the Ammo-
coetes, obtained from Cayuga lake inlet, was used.

In Amphioxus and the related form, Asymmetron, the blind sac was
found not to possess a specialized glycogenice function; but glycogen was
found in the intestinal and branchial epithelium, also in the epidermis
and the nerve cells of the central nervous system.

In Ammocecetes, glycogen was found in the liver, the intestinal
epithelium, the epidermis, the heart and skeletal muscles, in the central
nervous system, cartilage, the notochord and in fat cells.

In the horse it was found that the glycogenic function was not
limited in the adult to the liver, but persists in the skeletal and cardiac
muscles.

The main results gained are then: (a) In some animals at least,
the glycogenic function of the adult (and the independently feeding
lamprey or Ammoccetes) is not restricted to the liver, but persists as a
function of the tissues generally.

(b) Glycogen is apparently present in the Amphioxus and Asymme-
tron only for a short time after feeding. These animals and Ammocecetes
are constant feeders and with both a brief fast is sufficient to free them
from glycogen.

Proceedings of the Association of American Anatomists XIII

The most important result was the discovery of abundant glycogen
in the nerve cells of the myel (spinal cord) of Amphioxus and Asymme-
tron, and also in the brain of Ammoceetes. All previous observers have
stated that glycogen is absent from the nervous system of vertebrates.

KARYOKINETIC DIVISION IN THE SPINAL GANGLION CELLS OF
TRITON LARVA. By Ross G. Harrison. The Anatomical Labora-
tory, Johns Hopkins University.

In Triton larve (10-13.5 mm. in length) spinal ganglion cells with
long and defined process were found undergoing karyokinetic division.

ON THE HISTOGENESIS OF SPINAL GANGLIA IN MAMMALS. By
Grorce L. Srreeter. Department of Anatomy, Johns Hopkins Uni-
versity.

Spinal ganglia of the chick, pig and man were studied by comparing
teased preparations with preserved material cut in paraffin. The ob-
servations are as follows: |

1. The cells of the neural crest, as they separate off and migrate ven-
tralward between the myotome and neural tube, possess elongated vacuo-
lated cell bodies and appear in well preserved material to fuse into cell
clumps, possibly a syncytium. At this time ganglion cells cannot be
distinguished from capsule and sheath cells.

2. Neural crest cells are differentiated into—

a. Ganglion cells—Formation of processes. Condensation of pro-
toplasm of cell body. Appearance of basic stainable substance on
periphery. Increase and clumping of stainable substance (Nissl
bodies). Nucleus eccentric and comparatively late it retires to cen-
ter of cell and assumes resting appearance.

b. Capsule cells—Cell body converted into branching flat irregu-
lar processes which envelop early the ganglion cells and resemble
neuroglia cells. Capsule cells are identical with sheath cells.

_ 8. The early ganglion cells have a fusiform shape with processes at
the poles. Between these are found cells which develop later as a
secondary growth and are characterized by greater irregularity in shape
of cell body, and in situation, size, and number of processes.

4. Many bipolar cells in pig and man do not undergo a transformation
into the Ranvier T cells. Some unipolar cells seem to be unipolar from
the first.

5. Multipolar ganglion cells are found in considerable variety in adult
human spinal ganglia.

XIV Proceedings of the Association of American Anatomists

MARKED DIFFERENCES BETWEEN THE SKIN OF THE MALE AND
THAT OF THE FEMALE FROG. By A. O. FISHER. Department of
Anatomy, University of Wisconsin.

During the fall the dermis of the female frog is thinner and is less
resistant to acids and alkalies and digestive fluids than that of the male.
These changes seem to be associated with the formation of the ova.

THE BLOOD AND LYMPH VESSELS OF THE LUNG OF NECTURUS. By
WittiaAmM S. Minter. Department of Anatomy, University of Wisconsin.

THE MESENTERY IN AMPHIBIA AND REPTILIA. By Wrt1iam S.
Mitter. Department of Anatomy, University of Wisconsin.

In this note, attention was called to the lesser peritoneal cavity
(bursa hepato-enterica) in Necturus, Cryptobranchus, Amblystoma and
Chrysemys.

In Necturus the ligamentum hepatogastricum is attached to the liver
along the course of the hepatic-portal vein and to the cephalic half of the
stomach, the gastric attachment being less than the hepatic. Between
the caudal margin of this ligament and the entrance of the ductus
choledochus into the intestine there is complete absence of any mesentery
on the ventral side of the stomach. The ligamentum hepato-cavo-
pulmonale extends as an unbroken fold from the liver to the mid-dorsal
body wall. The dorsal mesentery caudal to the splenic blood vessels
has a large perforation. There are therefore two openings into the
bursa hepato-enterica, one ventral to the stomach, the other at the
caudal end of the dorsal mesentery. These two openings are already
present in embryos 25 mm. in length.

In Cryptebranchus the intestine forms a long loop after leaving the
stomach. The liver is connected with the stomach and intestinal loop by
a ligamentum hepato-gastro-duodenale and it is without any perforation.
A ligamentum hepato-cavo-pulmonale is present and it also is intact:
but caudal to the ductus choledochus there is an opening leading into the
bursa hepato-enterica, a true foramen epiploicum. The dorsal meso-
gastrium has a wide opening in it caudal to the splenic blood vessels
and in this respect shows a less complete bursa than Necturus.

In Amblystoma the ligamentum hepatogastricum extends the entire
length of the liver, but like Necturus, its gastric attachment is shorter
than its hepatic; the opening between its caudal margin and the ductus
choledochus is smaller than in Necturus. The ligamentum hepato-cavo-
pulmonale is without any perforation. The dorsal mesogastrium has
undergone a very considerable modification. Only the extreme cephalic
end of the stomach is in connection with this membrane; it is then

Proceedings of the Association of American Anatomists XV

continued along the mesial border of the anterior half of the left lung
where it ends. A second remnant of the dorsal mesogastrium is found
in the ligamentum gastro-lienale connecting the stomach and spleen.
The bursa hepato-enterica is therefore much less complete in Amblystoma
than in either of the two preceding forms.

In Chrysemys, the bursa hepato-enterica, like that of mammals, is
connected with the genera! body cavity in the earlier stages of develop-
ment; but as development goes on, the cavity becomes much enlarged,
takes an irregular H shape and is finally cut off from all communication
with the general body cavity. The spleen is situated in its own pocket
which may or may not communicate with the bursa hepato-enterica.

ON THE SHAPE OF THE URINIFEROUS TUBULES OF MAMMALS. By
G. CArL Huser. Department of Histology and Embryology, University
of Michigan.

American Journal of Anatomy, Vol. IV. (Supplementary number.)

CYTOLOGICAL CHANGES IN THE KIDNEY DUE TO DISTILLED WATER
AND VARYING STRENGTHS OF SALT SOLUTION. By Ferrpinanp
ScuMiTtTer. Department of Anatomy, University of Wisconsin.

Fresh kidney tissue macerated in distilled water shows large vesicles
within the lumen such as are usually seen in chronic interstitial neph-
ritis. Biitschli’s schaumstruktur appears in the epithelial cytoplasm.
Imbrication of epithelium occurs in places. Bowman’s capsule becomes
greatly distended. Maceration in hypertonic salt solution causes ap-
pearance of brush border in convoluted tubules. Strong solutions also
cause vacuoles and canals to appear within the cytoplasm.

RECENT ADDITIONS TO THE HARVARD EMBRYOLOGICAL COLLEC-
TION. By CuHarLes S. Minor. Harvard Medical School.

DEMONSTRATIONS.

1. Dr. Charles R. Bardeen: Models of a, The development of the thoracic
vertebre in the human embryo; 0b, The development of the skeleton of
the leg in the human embryo.

2. Dr. R. B. Bean: Preparations showing a racial peculiarity in the brain
of the Negro.

38. Dr. J. L. Bremer: Models of a 4.0 mm. human embryo from the Harvard
Embryological Collection.

4. Dr. Lydia M. DeWitt: Models of areas of Langerhans of certain mammals.

5. Professor Simon H. Gage: Preparations showing glycogen in animal
tissues.

6. Susanna Phelps Gage: Plates illustrating the anatomy of a three-weeks
human embryo, with especial reference to the forebrain.

XVI ___—~Proceedings of the Association of American Anatomists

-1

. Dr. Ross G. Harrison: Sections showing the development of the peri-
pheral nerves without sheath cells.

8. Dr. G. Carl Huber: Model of uriniferous tubule.

. Dr. George S. Huntington: Types of bronchial and arterial distribution
in the mammalian lung. (Morphological Museum of Columbia Uni-
versity. )

10. Dr. Wm. Keiller: a. Models showing the anatomy of the lymph nodes of

the axilla. b. Casts showing the ligaments of the mammary gland.

11. Dr. Abraham T. Kerr: Specimens from the dissecting-room.

12. Dr. Franklin P. Mall: The chorion and decidua of a human ovum of the

second week. '

13. Dr. Charles S. Minot: A human embryo before the formation of the

medullary plate, and other human embryos.

14. Dr. George L. Streeter: a. Preparations illustrating the histogenesis of

the spinal ganglia. 0b. Apparatus for dissecting small objects.

15. Dr. John Warren: Models illustrating the development of the paraphysis. ©

and pineal region in Necturus.

16. Dr. Burt G. Wilder: Exhibition of a manatee heart .and fetus, and of the

brains of a manatee, macroscelides, ceratodus, chimera, and of a frog

that lived five years after the removal of its cerebrum.

©

CONSTITUTION AND LIST OF OFFICERS AND MEMBERS.

CONSTITUTION.
ARTICLE I.
Section 1. The name of the Society shall be the “ Association of
American Anatomists.”
Section 2. The purpose of the Association shall be the advancement
of anatomical science.

ARTICLE II.

The officers of the Association shall consist of a President, two Vice-
Presidents, and a Secretary, who shall also act as Treasurer. The
officers shall be elected by ballot every two years.

ARTICLE III.

The management of the affairs of the Association shall be delegated
to an Executive Committee, consisting of seven members, including the
President and Secretary, ex-officio. One member of the Executive Com-
mittee shall be elected annually.

ARTICLE IV.

The Association shall meet annually, the time and place to be deter-
mined by the Executive Committee.

Proceedings of the Association of American Anatomists XVII

ARTICLE VY.

Section 1. Candidates for membership must be persons engaged in
the investigation of anatomical or cognate sciences and shall be pro-
posed in writing to the Executive Committee by two members, who
shall accompany the recommendation by a list of the candidate’s pub-
lications, together with the references. The election shall take place in
open meeting, a two-thirds vote being necessary.

Section 2. Honorary members may be elected from those not Ameri-
cans who have distinguished themselves in anatomical research.

ARTICLE VI.

The annual dues shall be five dollars. A member in arrears for dues
for two years shall be dropped by the Secretary at the next meeting of
the Association, but may be reinstated, at the discretion of the Executive
Committee, on payment of arrears.

ARTICLE VII.

Section 1. Five members shall constitute a quorum for the transac-
tion of business.

Section 2. Any change in the constitution of the Association must
be presented in writing at one meeting in order to receive consideration
and be acted upon at the next meeting; due notice of the proposed
change to be sent to each member at least one month in advance of the
meeting at which such action is to be taken.

Section 3. The ruling of the Chairman shall be in accordance with
“ Roberts’ Rules of Order.’

OrDERS ADOPTED BY THE ASSOCIATION.

The election of delegates to the Executive Committee of the Congress
of American Physicians and Surgeons shall take place every three years.

Newly elected members must qualify by payment of dues for one
year within thirty days after election.

The maximum limit of time for the reading of papers shall be twenty
minutes.

The Secretary and Treasurer shall be allowed his traveling expenses
and the sum of $10 toward the payment of his hotel bill, at each
session of the Association.

That the Association discontinue the separate publication of its pro-
ceedings, and that the AMERICAN JOURNAL OF ANATOMY be sent to each

member of the Association, on payment of his annual dues, this journal
20

XVIII Proceedings of the Association of American Anatomists
o

to publish the proceedings of the Association, including an abstract
of the papers read.

Contributors of papers are requested to furnish the Secretary with
abstracts within a fortnight after the meeting.

OFFICERS FOR 1905.

I RAS ORA OUR POS BOSCO OIC O aod e CHARLES S. MINOT.
TMS WICZIETRASTUE NES 66 Goo dado ac GEORGE A. PIERSOL.
Second Vice-President......... J. MARSHALL FLINT.
Secretary and Treasurer............ G. Cart HUBER.

Executive Committee.

Wiis) Hy MARKER .)- cm at Term expiring in 1905.
FREDERIC H. GERRISH........ Term expiring in 1906.
GroRGE S. HUNTINGTON...... Term expiring in 1907.
lnesicdronp dey WU he Aa oo eo oF Term expiring in 1908.
CHARLES R: BARDEEN........ Term expiring in 1909.

Delegate to Executive Committee of Congress of Physicians and Surgeons,
19038-1906.

JOSEPH A. BLAKE.

Alternate.
FRANK BAKER.

Delegate to the Council of the American Association for the Advancement of
Science.

Simon H. GAGE.

Member of Smithsonian Committee on the Table at Naples.
GEORGE S. HUNTINGTON.

For addresses of officers, see list of members.

Honorary Members.

JOEEN EOERUAND chrteyiee cherseeaetocers iets Glasgow, Scotland.
JOHN DANIEL CUNNINGHAM..... Edinburgh, Scotland.
WU SHONG, IDIOM bse ooo aooo obo oocUGaNS Paris, France.
CAMILLO: IGOLGLS J ser ie a3 kolensis toreenerceeted Pavia, Italy.
ALBERT VON KOLLIKER......... Wiirzburg, Germany.
ALEXANDER MACALISTER........ Cambridge, England.
PS EVAINWVIRRIR rata veret ote itoiens! eet cieiete rel taken remrerehets Paris, France.
GUSTAVs RET ZTUS ieee silos creck Stockholm, Sweden.
(CHAT Dyer! CO} CID Benen nneted OA, coi sonetta Geo vo cack Vienna, Austria.
Sm WILLIAM TURNER......... Edinburgh, Scotland.

IWGGHELM: WALDEYER.. 6.2... 00+ - << Berlin, Germany.

Proceedings of the Association of American Anatomists XIX

MEMBERS.

BENNET Mitts ALLEN, Ph. D., Instructor in Anatomy, University of Wisconsin,
310 Murray St., Madison, Wis.

Epwarp Puerps Axis, Jr., Associate Editor of Journal of Morphology, Mil-
waukee, Wis. Menton, France.

FRANK Baker, A. M., M. D., Ph. D., Professor of Anatomy, Medical Depart-
ment, University of Georgetown, 1728 Columbia Road, Washington, D.C.

CHARLES RUSSELL BARDEEN, A. B., M. D., Professor of Anatomy, University of
Wisconsin, Madison, Wis.

LEWELLYS FRANKLIN BarKER, M. B., Professor of Anatomy, University of Chi-
cago, Quadrangle Club, Chicago, Til.

Rospert BENNETT BEAN, B. S., M. D., Assistant in Anatomy, Johns Hopkins
Medical School, The Anatomical Laboratory, P. O. Station O, Baltimore,
Md.

ELexious THompson BEtt, B. S., M. D., Instructor of Anatomy, University of
Missouri, 508 S. 5th St., Columbia, Mo.

Ropert RussELL BENSLEY, A. B., M. B., Assistant Professor of Anatomy, Uni-
versity of Chicago, Chicago, I11.

ARTHUR DEAN BeEvVAN, M. D., Professor of Surgical Anatomy and Clinical Sur-
gery, Rush Medical College, 100 State St., Chicago, Til.

VinrAy Papin Buatr, A. M., M. D., Lecturer on Descriptive Anatomy and
Demonstrator of Anatomy, Medical Department, Washington Univer-
sity, 305 North Grand Ave., St. Lowis, Mo.

JosrpH AUGUSTUS BLAKE, A. B., M. D., Professor of Surgery, College of Physi-
cians and Surgeons, Columbia University, 601 Madison Ave., New York
City.

Joun Lewis Bremer, M. D., Harvard Medical School, Boston, Mass.

GroRGE EmMeRSON Brewer, A. M., M. D., Instructor in Surgery, College of Phy-
sicians and Surgeons, Columbia University, 61 W. 48th St., New York
City, N. Y.

SAMUEL Max Brickner, A. M., M. D., Gynecologist to Mt. Sinai Hospital Dis-
pensary (N. Y.); Special Student in Anatomy, College of Physicians
and Surgeons, Columbia University, 136 W. S5th St., New York City,
INES

J. Futmer Bricut, M. D., Instructor in Anatomy and Assistant Demonstrator
of Anatomy, Medical College of Virginia, 408 W. Grace St., Richmond,
Va.

Max Bropet, Artist, The Johns Hopkins Hospital, Baltimore, Md.

WiILitIAM ALLEN Brooks, M. D., 167 Beacon St., Boston, Mass.

FRANK IrvinG Brown, A. M., M. D., Instructor in Anatomy, Medical School of
Maine, Box 205, South Portland, Me.

WILLIAM BROWNING, Ph. B., M. D., Professor of Diseases of the Mind and Ner-
vous System, Long Island College Hospital, 54 Leffert Place, Brooklyn,
ING

CHARLES HENRY BUNTING, B. S., M. D., Instructor in Pathology, Johns Hopkins
University, Assistant Resident Pathologist, Johns Hopkins Hospital,
Baltimore, Md.

XX Proceedings of the Association of American Anatomists

J. F. Burknowper, M. D., Professor of Physiology, Dental School of the Uni-
versity of Illinois and of the Dearborn Medical College, Professor of
Ophthalmology, Chicago Eye, Ear, Nose and Throat College, 100
State Street, Chicago, Ill.

WILLIAM FRANCIS CAMPBELL, A. B., M. D., Professor of Anatomy and His-
tology, Long Island College Hospital, S86 Green Ave., Brooklyn, N. Y.

CHURCHILL CARMALT, A. B., M. D., Assistant Demonstrator of Anatomy, Col-
lege of Physicians and Surgeons, Columbia University, 600 Madison
Ave., New York City, N. Y.

WILLIAM Puittips Carr, M. D., Professor of Physiology, Medical Department,
Columbian University, 1418 L St., N. W., Washington, D.C.

WILLIAM Guy CHRISTIAN, M. D., Professor of Anatomy, Medical Department,
University of Virginia, Charlottesville, Va.

CoRNELIA Marta Ciapp, Ph. D., Professor of Zoology, Mount Holyoke College,
South Hadley, Mass.

Benson A. Conor, A. B., M. D., Associate in Anatomy, University of Chicago,
Hull Laboratory of Anatomy, Chicago, IIl.

Grorce E. Coauitt, Ph. D., Professor of Biology, Pacific University, Forest
Grove, Oregon.

Wittiam Merritt Conant, M. D., Instructor in Anatomy in Harvard Medical
School, 486 Commonwealth Ave., Boston, Mass.

Epwin Grant ConKkLin, A. M., Ph. D., Professor of Zoology, University of
Pennsylvania, Philadelphia, Pa.

EUGENE ROLLIN Corson, B. S., M. D., 11 Jones St., East, Savannah, Ga.

JosErH Davis Craic, A. M., M. D., Professor of Anatomy, Albany Medical Col-
lege, 12 Ten Broeck St., Albany, N. Y.

WILLIAM DarracHu, A. M., M. D., Assistant Demonstrator of Anatomy, Cornell
Medical School, New York City, 852 Lexington Ave., New York City.

ALVIN Davison, M. A., Ph. D., Professor of Biology, Lafayette College, Easton,
Pennsylvania.

Rospert H. Mackay DAawsurRN, M. D., Professor of Surgery and Anatomy, New
York Polyclinic Medical School and Hospital, 105 W. 74th St., New York
City, N.Y.

BasSHForD DEAN, Ph. D., Professor of Vertebrate Zoology, Columbia Uni-
versity, New York City, Honorary Curator of Fishes, American Museum
of Natural History, 20 West 82nd Street, New York City, N. Y.

FRANCIS X. DEeRCUM, M. D., A. M., Ph. D., Professor of Mental and Nervous
Diseases, Jefferson Medical College, 1719 Walnut St., Philadelphia Pa.

Lypra M. DreWIr?, M. D., B. S., Instructor in Histology, Department of Medi-
cine and Surgery, University of Michigan, 925 N. University Ave., Ann
Arbor, Mich.

FRANKLIN Dexter, M. D., 148 Marlborough St., Boston, Mass.

JoHN Mirton Dopson, A. M., M. D., Professor of Medicine, Rush Medical Col-
lege, 568 Washington Boulevard, Chicago, Ill.

Henry HERBERT DONALDSON, A. B., Ph. D., Professor of Neurology, University
of Chicago, Chicago, I11.

Proceedings of the Association of American Anatomists XXI

Grorce A. Dorsrty, Ph. D., Field Columbian Museum, Chicago, Ill.

Joun Tuomas Duncan, M. B., C. M., Professor of Anatomy, Woman’s Medical
College and Ontario Veterinary College, 45 Bloor St., East, Toronto,
Canada. 7

Tuomas Dwicut. M. D., LL. D., Parkman Professor of Anatomy, Harvard
Medical School, Boston, Mass.

CHARLES RocHESTER EASTMAN, Ph. D., Curator of Vertebrate Paleontology,
Museum of Comparative Zoology, Harvard University, 354 Brookline
St.. Cambridge. Mass.

Rosert G. Eccres, M. D., Professor of Organic Chemistry, Brooklyn College of
Pharmacy, 191 Dean St., Brooklyn, N. Y.

CoRINNE Burorp Ecktrey, Demonstrator of Anatomy in Medical and Dental
Departments, College of Physicians and Surgeons, University of IIli-
nois, 979 Jackson Boulevard, Chicago, Ill.

WILLIAM Tuomas Ecktey, M. D., Professor of Anatomy, College of Physi-
cians and Surgeons, University of Illinois; also Dental Department,
University of Illinois. 979 Jackson Boulevard, Chicago, Ill.

CHARLES LincotN Epwarps, Ph. D., Professor of Natural History, Trinity Col-
lege, Hartford, Conn. ;

Cart H. E1rcENMANN, Ph. D., Professor of Zoology, Indiana University; Direc-
tor of the Indiana University Biological Station, Bloomington, Ind.

Dantev N. Ersenpratu, M. D., Adjunct Professor of Surgery, Medical Depart-
ment of the University of Illinois, 3125 Michigan Ave., Chicago, Ill.

GitBert M. Ettiorr, A. M., M. D., Assistant Demonstrator of Anatomy, Medical
School of Maine, Brunswick, Maine.

CHARLES ANDREW ErRpwtAN, M. D., Professor of Anatomy, Medical Depart-
ment, University of Minnesota, Minneapolis, Minn.

ALBERT CHAUNCEY EYCLESHYMER, B. S., Ph. D., Professor of Anatomy, Medical
Department, University of St. Louis, St. Louis, Mo.

Harry Burr Ferris, A. B., M. D., Professor of Anatomy, Medical Department,
Yale University, 118 York St., New Haven, Conn.

JOSEPH MARSHALL FLINT, B. S., A. M., M. D., Professor of Anatomy, Medical
Department, University of California, San Francisco, Cal.

JoHN PIERREPONT CopRINGTON Foster, A. B., M. D., Lecturer on Anatomy, Yale
School of Fine Arts, 109 College St., New Haven, Conn.

Henry Fox, B. S., A. M., Professor of Biology and Chemistry, Temple College,
Philadelphia, 5603 Germantown Ave., Germantown, Penna.

GinmaANnN Duspots Frost, A. M., M. D., Professor of Anatomy, Dartmouth Medical
School, Hanover, N. H.

Frank R. Fry, A. M., M. D., Professor of Diseases of the Nervous System,
Washington University, 3133 Pine St., St. Lowis, Mo.

JEREMIAH SWEETSER FERGUSON, M. Sc., M. D., Instructor in Histology, Corneil
University Medical College, New York City, 330 W. 28th St., New York
City, Ne Y¥.

Smion Henry Gace, B. S., Professor of Histology and Embryology, Cornell
University, Ithaca, N. Y.

Mrs. SuSANNA PHELPS GAGE, Ithaca, N. Y.

XXII Proceedings of the Association of American Anatomists

Bern Bupp GALLAupDET, A. M., M. D., Demonstrator of Anatomy and Instructor
of Surgery, College of Physicians and Surgeons, Columbia University,
60 W. 50th St., New York City, N. Y.

NorMAN J. Gewrina, A. B., M. D., Assistant Demonstrator of Histology, Medi-
eal School of Maine, 684 Congress St., Portland, Maine.

FREDERIC HENRY GERRISH, A. M., M. D., LL. D., Professor of Anatomy, Bowdoin
College, 675 Congress St., Portland, Maine.

Puitie KInGcswortH GILMAN, B. A., Johns Hopkins Medical School, Balti-
more, Md.

GrorGE LINCOLN GoopALE, A. M., M. D., -Professor of Botany, Director of
Botanic Gardens. Harvard University, 5 Berkeley St., Cambridge, Mass.

Exein Anaus Gray, B. A., M. B., Instructor in Anatomy, Cornell University
Medical College, Ithaca, New York, 12 Osborn Block, Ithaca, N. Y.

Mitton J. GREENMAN, Ph. B., M. D., Director of the Wistar Institute of Anat-
omy, 36th St. and Woodland Ave., Philadelphia, Pa.

EvisHa Hatt Grecory, M. D., Professor of Anatomy, Northwestern University
Medical School, 2431 Dearborn St., Chicago, Ill.

MicuHaetL F. Guyer, Ph. D., Professor of Biology, University of Cincinnati,
Cincinnati, Ohio, 554 Evanswood, Clifton, Cincinnati, Ohio.

Cart A. Hamann, M. D., Professor of Anatomy, Medical Department, Western
Reserve University, 661 Prospect St., Cleveland, Ohio.

Irvinc Harpesty, A. B., Ph. D., Instructor in Anatomy, Medical Department of
University of California, San Francisco, Cal.

JESSE Ratpr Harris, M. D., 1423 Binney St., Washington, D. C.

Ross GRANVILLE Harrison, Ph. D., M. D., Associate Professor of Anatomy,
Johns Hopkins University (Managing Editor of the Journal of Experi-
mental Zoology), Baltimore, Md.

Basin CoLeEMAN Hyatt Harvey, A. B., M. B., Assistant in Anatomy, Hull
Anatomical Laboratory, University of Chicago.

IrvING SAMUEL HAYNES, Ph. B., M. D., Professor of Practical Anatomy, Cornell
University Medical College, 1125 Madison Ave., New York City, N. Y.

CHARLES Morse Hazen, A. M., M. D., Professor of Physiology, Medical College
of Virginia, Richmond, Bon Air, Va.

Joun C. Heister, M. D., Professor of Anatomy, Medico-Chirurgical College,
Philadelphia, Pa., 3829 Walnut St., Philadelphia, Pa.

CHARLES JUDSON Herrick, Ph. D., Professor of Zoology, Denison University
(Co-editor of Journal of Comparative Neurology), Granville, Ohio.
ARTHUR E. Hertzter, A. M., M. D., Ph. D., Professor of Histology and Path-
ology, University Medical College, Kansas City, Mo., 508 Altman Build-

ing, Kansas City, Mo.

GEORGE JuLius Hever, B. S., Johns Hopkins Medical School, 1524 N. Broad-
way, Baltimore, Md.

ADDINELL Hewson, A. M., M. D., Professor of Anatomy of Philadelphia Poly-
clinic for Graduates in Medicine; Demonstrator of Anatomy, Jefferson
Medical College, 1508 Pine St., Philadelphia, Pa.

EBEN CuHar.Les Hitt, A. B., Johns Hopkins Medical School, 2120 N. Charles St.,
Baltimore, Md.

EpMuUND WALES Ho.tmgs, M. D., 2035 Chestnut St., Philadelphia, Pa.

Grant SHERMAN HopxKINs, D. Se., D. V. M., Assistant Professor of Veterinary
Anatomy, Cornell University, Ithaca, N. Y.

Proceedings of the Association of American Anatomists XXIII

Ares Hrpricka, M. D., Curator of the Division of Physical Anthropology,
Washington, D. C., United States National Museum, Washington, D.C.

G. Cart Huser, M. D., Professor of Histology and Embryology, University of
Michigan, 333 BE. Ann St., Ann Arbor, Mich.

George S. Huntrnaton, A. M., M. D., D. Sc., Professor of Anatomy, Depart-
ment of Medicine, Columbia University, 437 W. 59th St., New York City,
INS WY

CLARENCE M. Jackson, M. S., M. D., Assistant Professor of Anatomy and His-
tology, University of Missouri, 1201 Paquin St., Columbia, Mo.

Horace Jayne, M. D., Ph. D., Professor of Zoology, University of Pennsylva-
nia, 36th St. and Woodland Ave., Philadelphia, Pa.

WILLIAM KeiLurr, L. R. C. P., and F. R. C. S. Ed., Professor of Anatomy, De-
partment of Medicine, University of Texas, Galveston, Texas.

Grorce T. Kemp, M. D., Ph. D., Professor of Physiology, University of Illinois,
Hotel Beardsley, Champaign, Iil.

Apram T. Kerr, B. S., M. D., Professor of Anatomy, Department of Medicine,

Cornell University, Ithaca, N. Y.

ALFRED Kine, A. B., M. D., Instructor and Demonstrator of Anatomy, Medical
School of Maine, 610 Congress St., Portland, Maine.

B. F. Krnassury, Ph. D., Assistant Professor of Physiology, Cornell University,
Ithaca, New York.

J. S. Krnestey, Se. D., Professor of Biology, Tufts College, Mass.

Henry McE. Knower, A. B., Ph. D., Instructor in Anatomy, Johns Hopkins
University (Co-editor and Secretary of the American Journal of Anat-
omy), Baltimore, Md.

CHARLES A. Kororp, Ph. D., Assistant Professor of Histology and Embryology,
University of California, Berkeley, Calif.

PRESTON Kyes, A. M., M. D., Associate in Anatomy, Hull Laboratory of Anat-
omy, University of Chicago, Chicago, III.

DANIEL SmitH Lamp, A. M., M. D., Pathologist Army Medical Museum; Pro-
fessor of Anatomy, Howard University, Medical Department, Army
Medical Museum, Washington, D.C.

ApRIAN V. S. Lampert, Assistant Demonstrator of Anatomy, College of Physi-
cians and Surgeons, Columbia University, 126 H. 39th St., New York
CHIE INE 3 &

Tuomas G. Lez, M. D., Professor of Histology and Embryology, University of
Minnesota, Minneapolis, Minn.

JOSEPH Letpy, Jr., A. M., M. D., Late Assistant Demonstrator of Anatomy and
Pathologic Histology, University of Pennsylvania, 1319 Locust St., Phil-
adelphia, Pa.

DEAN D. Lewis, M. D., Associate in Anatomy, Hull Anatomical Laboratory,
Chicago University, Chicago, III.

FreDeRIC T. Lewis, A. M., M. D., Instructor in Histology and Embryology,
Harvard Medical School, 36 Highland Ave., Cambridge, Mass.

WARREN Harmon Lewis, B. 8., M. D., Associate Professor in Anatomy, Johns
Hopkins University, Baltimore, Md.

WILLIAM Evan Lewis, M. D., Professor of Anatomy, Miami Medical College,
409 East 5th St., Cincinnati, Ohio.

XXIV _ Proceedings of the Association of American Anatomists

FrRANK RatruaAy Litiir, Ph. D., Associate Professor of Zoology and Embry-
ology, University of Chicago; Assistant Director of the Marine Bio-
logical Laboratory, University of Chicago, Chicago, I11.

W. A. Locy, Ph. D., Professor of Zoology, Northwestern University, Evanston,
Ill.

Hanan Wotr Lorep, A. M., M. D., Professor of Nose and Throat Diseases,
St. Louis University, 3559 Olive St., St. Louis, Mo.

JOHN Bruce MacCatium, A. B., M. D., Assistant Professor of Physiology at
the University of California, Berkeley, Cal.

JoHN GrorGE McCartny, M. D., Lecturer and Senior Demonstrator of Anat-
omy, McGill University, 61 Drummond St., Montreal, Canada.

GrEORGE McCLeLttAN, M. D., Professor of Anatomy, Academy of Fine Arts,
Philadelphia, Pa., 1352 Spruce St., Philadelphia, Pa.

CHARLES FREEMAN WILLIAMS McCuure, A. M., Professor of Comparative Anat-
omy, Princeton University, Princeton, N. J.

Epwarp JosepH McDonovucu, A. B., M. D., Instructor in Histology, Medical
School of Maine, 624 Congress St., Portland, Me.

R. Tart McKenzir, A. B., M. D., Professor of Physical Education and Director
of the Department of Physical Education, University of Pennsylvania,
121 South 18th St., Philadelphia, Pe.

JAMES PLtayrarr McMurricu, A. M., Ph. D., Professor of Anatomy, University
of Michigan, 1701 Hill St., Ann Arbor, Mich.
FRANKLIN P. Matt, M. D., LL. D., Professor of Anatomy, Johns Hopkins Uni-

versity, Baltimore, Md.

Epwarp LAuURENS Mark, Ph. D., LL. D., Hersey Professor of Anatomy and
Director of the Zoological Laboratory, Harvard University, 109 Irving
St., Cambridge, Mass.

WALTON MarTIN, Ph. B., M. D., Instructor in Surgery, College of Physicians
and Surgeons, Columbia University, 6S Hast 56th St., New York City,
NG eYE

RupotpH Matas, M. D., Professor of Surgery, Medical Department of Tulane
University, 2255 St. Charles Ave., New Orleans, La.

WASHINGTON MatrHews, M. D., LL. D., Major and Surgeon U. S. Army (Re-
tired), 1262 New Hampshire Ave., Washington, D.C.

Epwarp Linpon Metius, M. D. (abroad); address, Anatomical Laboratory,
Johns Hopkins Medical School, Baltimore, Md.

ApotrpH Meyer, M. D., LL. D., Director of the Pathological Institute of the
New York State Hospitals, Ward’s Island; Professor of Psychiatry,
Cornell Medical School, Ward’s Island, New York City, N. Y.

ARTHUR W. Meyer, S. B., Johns Hopkins Medical School, Baltimore, Md.

Cureton Merepiru Miter, M. D., Demonstrator of Anatomy and Adjunct Pro-
fessor of Diseases of the Eye and Ear, Medical College of Virginia,
217 East Grace St., Richmond, Va.

WittraAs Snow Mitter, M. D., Assistant Professor of Vertebrate Anatomy,
University of Wisconsin, Madison, Wis.

CHARLES SEDGEWICK Minor, S. B. (Chem.), S. D., LL. D., D. Se., Professor of
Histology and Human Embryology, Harvard Medical School, Boston,
Mass.

SAMUEL JASON MIxter, B. S., M. D., Instructor of Surgery, Harvard Medical
School, 180 Marlborough St., Boston, Mass.

— ee

Proceedings of the Association of American Anatomists XXV

Mary Briarr Moopy, M. D., Station A, New Haven, Conn.

Rosert OrtoN Moopy, B. S., M. D., Assistant in Anatomy, Hearst Anatomical
Laboratory, University of California, San Francisco, Cal.

JAMES DupLeY Morean, A. B., M. D., Clinical Professor, Georgetown Univer-
sity Hospital and Medical School; Physician to Garfield Memorial Hos-
pital, 919 15th St., McPherson Square, Washington, D.C.

JoHN CuMMINGS Muwnrop, A. B., M. D., Instructor of Surgery, Harvard Medical
School, Professor of Surgery, Tufts Dental School, 173 Beacon St.,
Boston, Mass.

Burton D. Meyers, A. M., M. D., Associate Professor in charge of the Depart-
ment of Anatomy, Indiana University, Bloomington, Indiana.

Harriet ISABEL Nosie, M. D., Prosector and Curator of the Department of
Anatomy and Instructor in Anatomy, Woman’s Medical College of Penn-
sylvania, 1544 Centennial Ave., Philadelphia, Pa.

Henry Farrcnitp Osporn, Se. D., LL. D., DaCosta Professor of Zoology in
Columbia University; Curator of Vertebrate Paleontology, American
Museum of Natural History, 77th St. and Highth Ave., New York City,
NOY

CHARLES AUBREY PARKER, M. D., Instructor in Anatomy and Instructor in
Surgery, Rush Medical College, University of Chicago, 1660 Fulton St.,
Chicago, Ill.

GEORGE HowaArp PARKER, D. Sce., Assistant Professor of Zoology, Harvard Uni-
versity, 6 Avon Place, Cambridge, Mass.

STEWART Paton, A. B., M. D., Associate in Psychiatry, Johns Hopkins Univer-
sity; Director of Laboratory, Sheppard and Enoch Pratt Hospital, 213
West Monument St., Baltimore, Md.

WILLIAM PATTEN, Ph. D., Professor of Zoology and Head of the Department
of Biology, Dartmouth College, Hanover, N. H.

RicHArD M. PEARCE, JrR., M. D., Director of Bender Laboratory, Adjunct Pro-
fessor of Pathology and Bacteriology, Albany Medical School, Albany,
INS Ye

GEORGE A. Priersot, M. D., Professor of Anatomy, University of Pennsylvania,
4724 Chester Ave., Philadelphia, Pa.

Avucust G. Pon~tmMAN, M. D., Assistant Professor of Anatomy, University of
Indiana, Bloomington, Ind.

EUGENE H. Poot, A. B., M. D., Assistant Demonstrator of Anatomy, College of
Physicians and Surgeons, Columbia University, 57 W. 45th St., New
York City, N. Y.

Peter Porrer, M. S., M. -D., Associate Professor of Anatomy, St. Louis Uni-
versity, Marion-Sims-Beaumont Medical College, Grand Ave. and Caro-
line Sts., St. Louis, Mo.

H. J. Prentiss, M. D., M. E., Professor of Anatomy, University of Iowa, Iowa
City, lowa.

ALEXANDER PRIMROSE, M. B., C. M. Ed., M. R. C. S. Eng., Professor of Anatomy,
University of Toronto, 100 College St., Toronto, Canada.

ALBERT Moore Reese, A. B., Ph. D., Associate Professor of Histology and Em-
bryology, Syracuse University, Syracuse, N. Y.

Emory WILLIAM REISINGER, M. D., Assistant in Anatomy at the Georgetown
Medical and Dental Schools, 1209 13th St., N. W., Washington, D. 9.

XXVI Proceedings of the Association of American Anatomists

DANIEL GRAISBERRY Reveti, A. B., M. B., Associate in Anatomy, University of
Chicago, 667 E. 62d St., Chicago, I11.

SyLvAN RosenneEm™, A. B., M. D., Assistant in Laryngology, Johns Hopkins
Medical School, 522 North Charles St., Baltimore, Md.

Avaustus Henry Rotu, A. B., M. D., Instructor in Anatomy, University of
Michigan, Ann Arbor, Mich.

NELSON G. RuSSELL, M. D., Assistant in Anatomy, Medical Department, Uni-
versity of Buffalo, 448 Franklin St., Buffalo, N. Y.

FLORENCE R. Sapin, B. S., M. D., Associate in Anatomy, Johns Hopkins
Medical School, Baltimore, Md.

JOHN ALBEATSON SAmpson, A. B., M. D., Instructor in Gynecology, Johns Hop:
kins University, Johns Hopkins Hospital, Baltimore, Md.

MARIE CHARLOTTE SCHAEFER, M. D., Lecturer and Demonstrator of Biology,
Normal Histology and Biology, Medical Department, University of
Texas, Galveston, Texas.

FERDINAND SCHMITTER, A. B., M. D., Instructor in Anatomy, University of
Wisconsin, 41/4 N. Henry St., Madison, Wis.

Magsor G. Seeric, A. B., M. D., Assistant in Anatomy, Medical Department of
St. Louis University, Vanol Building, 3908 Olive St., St. Lowis, Mo.

Haroup D. Senior, M. B., F. R. C. S., Demonstrator of Anatomy, Medico-Chi
rurgical College, 10 South 18th St., Philadelphia, Pa.

GEORGE E. SHAMBAUGH, Ph. B., M. D., Instructor in the Anatomy of the Ear,
Nose and Throat, University of Chicago; Associate in Osteology, Rush
Medical College, 100 State St., Chicago, Ill.

FRANCIS JOHN SHEPHERD, M. D., C. M., M. R. C. S. Eng., Professor of Anatomy
in McGill University, 152 Mansfield St., Montreal, Canada.

DANIEL Kerroot Suuts, A. B., M. D., Professor of Anatomy, Columbia Univer:
sity, 1719 De Sales St., Washington, D.C.

RICHARD DRESSER SMALL, A. B., M. D., Instructor in Anatomy, Portland Medi:
cal School, 154 High St., Portland, Me.

CHARLES DENNISON SmitH, A. M., M. D., Professor of Physiology, Bowdoin
College; Instructor in Physiology and Clinical Microscopy, Portland
Medical School, 126 Free St., Portland, Me.

EUGENE ALFRED SmitH, M. D., 1018 Main St., Buffalo, N. Y.

JoHN Hotmes SmitruH, M. D., Professor of Anatomy, University of Maryland,
27 W. Preston St., Baltimore, Md.

EpMonpD Soucnon, M. D., Professor of Anatomy and Clinical Surgery, Tulane
University, 2408 St. Charles Ave., New Orleans, La.

Epwarkp ANTHONY SpirzKa, M. D., Fellow in Anatomy, Columbia University,
66 East 73d St., New York City, N, Y.

JoHN ANDERSON SPRINGLE, M. D., C. M., Lecturer on Anatomy, McGill Univer-
sity, 1237 Dorchester St., Montreal, Canada.

GEORGE Davin Stewart, M. D., Professor of Anatomy and Clinical Surgery,
University and Bellevue Hospital Medical School, 143 Hast 37th St.,
New York City, N. Y.

Grorce L. STREETER, M. D., Instructor in Anatomy, Johns Hopkins University
Baltimore, Md.

Proceedings of the Association of American Anatomists XXVIT

MerRVIN T. SupterR, M. D., Ph. D., Instructor in Anatomy, Cornell University,
Ithaca, N. Y.

Ewine Taytor, A. B., M. D., Austin Teaching Fellow in Histology and Embry-
ology, Harvard Medical School, Boston, Mass.

Ropert JAMES Terry, A. B., M. D., Professor of Anatomy, Medical Department,
Washington University, 1806 Locust St., St. Lowis, Mo.

WitiiAmM C. Turo, A. M., Instructor in Histology and Embryology, Cornell
University, Ithaca, N. Y.

WALTER E. Toptr, M. D., Assistant Demonstrator of Anatomy, Medical School
of Maine, 126 Free St., Portland, Me.

Paut Yorr Tupper, M. D., Professor of Applied Anatomy and Operative Sur-
gery, Medical Department, Washington University, Linmar Building,
St. Lowis, Mo.

ALBERT HENRY TuTTLE, M. Sc., Professor of Biology, University of Virginia,
Charlottesville, Va.

ARTHUR S. VospureH, A. B., M. D., Assistant Demonstrator of Anatomy, Medi-
cal Department, Columbia University, 40 West SSth St., New York City,
INE Y.

FREDERICK CLAYTON WAITE, A. M., Ph. D., Assistant Professor of Histology and
Embryology, Western Reserve University, St. Clair and Erie NSts.,
Cleveland, Ohio.

GEORGE WALKER, M. D., Instructor in Surgery, Johns Hopkins University, Cor.
Charles and Centre Sts., Baltimore, Md.

CHARLES HoweELt Warp, Director of Anatomical Laboratory for the Prepara-
tion of Osteological Specimens and Anatomical Models, 47 Hope Ave.,
Rochester, N. Y.

JOHN WARREN, M. D., Demonstrator of Anatomy, Harvard Medical School,
Boston, Mass.

ALDRED Scott WARTHIN, Ph. D., M. D., Professor of Pathology and Director of
the Pathological Laboratory, University of Michigan, Ann Arborae
Michigan.

JOHN CLARENCE WEBSTER, B. A., M. D., F. R. C. P. Ed., Professor of Obstetrics
and Gynecology, Rush Medical College, University of Chicago, 706 Re-
liance Bldg., 100 State St., Chicago, Ill.

FANEvUIL D. WEIssE, M. D., Professor of Anatomy, New York College of Den-
tistry, 46 West 20th St., New York City, N. Y.

CHARLES IGNATIUS WEsT, M. D., Lecturer on Topographical Anatomy, Howard
University, and First Assistant Surgeon, Freedman’s Hospital, 924 W
St., N. W., Washington, D. C.

RicHARD HENRY WHITEHEAD, A. B., M. D., Professor of Anatomy, Medical De-
partment, University of North Carolina, Chapel.Hill, N. C.

Burt G. Wixper, M. D., B. S., Professor of Neurology, Vertebrate Zoology and
Physiology, Cornell University, Ithaca, N. Y.

Harris HAWTHORNE WILDER, Ph. D., Professor of Zoology, Smith College, 72
Dryads Green, Northampton, Mass.

SAMUEL WENDELL WILLISTON, M. D., Ph. D., Professor of Paleontology, Univer-
sity of Chicago, Chicago, II].

XXVIII Proceedings of the Association of American Anatomists

J. Gorpon Witson, M. D., Associate in Anatomy, University of Chicago, 5842
Rosaline Court, Chicago, Il.

JAMES MerrREDITH WIson, Ph. B., M. D., Assistant Professor of Embryology
and Comparative Anatomy, St. Louis University, Medical Department
St. Louis University, St. Louis, Mo.

WILLIAM POWELL WILSON, B. S., Se. D., Director of the Philadelphia Commer-
cial Museum, 233 South 4th St., Philadelphia, Pa.

THOMAS CASEY WITHERSPOON, M. D., Professor of Operative and Clinical Sur-
gery, St. Louis University, 4318 Olive St., St. Louis, Mo.

Rospert Henry Wotcott, A. M., M. D., Associate Professor of Zoology in
charge of the Anatomical Laboratory in the Industrial and the Medical
College, University of Nebraska, Lincoln, Neb.

FREDERICK ADAMS Woops, M. D., Instructor in Histology and Embryology,
Harvard Medical School, 936 Beacon St., Brookline, Mass.

GrorRGE Wootsery, A. B., M. D., Professor of Anatomy and Clinical Surgery,
Cornell University Medical College, 117 East 36th St., New York City,
Nays

Smuon M. Yutzy, M. D., Senior Demonstrator of Anatomy, University of Michi-
gan, Ann Arbor, Mich.

INDEX

(See also Contents for Authors and Full Titles)

Roman numerals refer to pages in the Proceedings of Assoc. of American Anatomists.

PAGE

ASSOCIATION OF AMERICAN ANATO-
MISTS, PROCEEDINGS OF........
APPENDIX I-XXVIII
ASSOCIATION OF AMERICAN ANATO-
MISTS, List OF MEMBERS.......
APPENDIX XIX-XXVIII
Allolobophora Foetida, see matura-
tion of egg.
Amphibia, see Necturus, and Mesen-
tery.
erural flexorsiofa.ss. ser 34
Anatomy, see genetic interpretation.
Angle, see elbow.
Ankle, development of ........... . 290
Arteries, of human embryo’s brain,... 6
Artery, see subclavian.
Articular surfaces, size of in long

IDONEBIOL, WO SEXEB... oo. s etter 19
Axone, sheath cells of......... Booogie
BDELLOSTOMA, see excretory organs.
Betz cells, see giant cells.

Blood vessels, see Brain, injections.
of lung, see Necturus.
of three weeks’ human
NVA MVE aaa oc 0055 417
Bones, see articular surfaces, spinal
column, leg, skeleton.
Brain, development of blood-vessels
IOLA CMU LYON sire levonstn= 1
of Negro, Proc. Assoc. Amer.
INTRO MORRO OS OAC ciab c IV
of three weeks’ embryo....... 425
folds Of Ime mMbry Ol. «icye o1-/stsie 428

and Proc. Assoc. Amer. Anat.. IX

of six Scientists, Proc. Assoc.
Amer. Anat...... -

also see giant cells of cortex.

CaRoTID GLAND, Proc. Assoc. Amer.

AM AE). Sn aca cvAreceralel aly! oie einer Regeteystets V

Central nervous system, see Brain,
axone.

framework in........ 343

Chelonia, venous system of. ........ 453

Claigtoy ect) hor late ome Hig DRY OICIOcN Se cic,” 213

Classification, of Mesenchyma...... 252

of Glandsy i 2. =... 253, 256

Connective tissues). .-6 sn. 2. weezer

Cortex, see giant cells.

Cranial nerves, see occipital.

OVROM MIE Ondo g oonaH ono sc oosdecconr 38

Crus, see leg.

Cutaneous nerves of leg, Proc. Assoc.
AV JMNM I 5 bOco Co oaUGmatOUd = x

Cy tomoOnrphe steers rcscsists oss oeerateree 247

EARTHWORM, see maturation.
Egg, see maturation.

Elastic tissues, see larynx.
Elbow, angle of .....
Embryology, see genetic.
Excretory organs, of Bdellostoma ...117

FaciaAL NERVE, Proc. Assoc. Amer.

I TLAT scien oyeae wiadoles euctera eee eee Vv
Femoral region, Proc. Assoc. Amer.
DNS) Geer ERE ALCO OCs ora Oi ae XI
Fifth nerve, Proc. Assoc. Amer.
INN OU. Weyetsasnsreress's-sseunyaraetete enePeiecens A
Flexors, see crural.
Folds of embryo’s brain ............ 428
Foot, development of ...... ernie papi aeeste 290

GANGLIA, Proce. Assoc. Amer. Anat.. XIII
Genetic, interpretation in Anatomy. .245
Giant cells, location of in Cortex....405
Glands, see lymphatic nodes.

Glosso-pbaryngeal nerve,... ....... 102
Glycogen, Proc. Assoc. Amer. Anat. . XII

HAEMOLYMPH GLANDS......-......- 384
His, Wilhelm, relations to institutions

Ob JOarninor creo cansteunere sic eres eveneere 139

iy poclossally NetWielareciristisiclelrcty sie 108
INGUINAL REGION, Proc. Assoc. Amer.

TU Lies eutiey toy etreneteel spay sneratics sterete orale XI
Injections of vessels of embryonic

le Ut ea DAmOkg nes Mone mans 2

OL lympNAViCS erie oterl- 357
Interstitial cells (of Leydig) in Testis

OLPE I Oo rercetelevonshersisieietsnere oi fehelouees 1s 193

PATI AD Uhe ca sie ecateis, siete Misha saiekeferave usr eretees IV

488 Index

, PAGE

= ‘

KIDNEY, shape and development of
tubules, SUPPLEMENT to

Vig ei ueter cn etere 11-98
cytological changes in, Proc.
Assoc. Amer. Anat...... XV
Kinosternon, wens Of. ce)... eases 459
TAAGHR DAN VGUNStOR erties) volte lensis1eco ce 455
CHUA MERON OL aye yeia.n-telie oF 41
Langerhans’ islands, Proc. Assoc.
VASE @ Wiig AMI Eip ee react foiofcvajcoseuresar eters VEU
Larynx, elastic tissue of........... 175
Leg, development of skeleton of. 279,286
nerves of mammalian......... 60
also see cutaneous nerves, and
crural.
Leydig, see interstitial.
Limb, see leg.
PERVOE a, beron bates cs tyere ts ciel ecg ancaskeweaeseiat = 258

Lumbar vertebrae, development of...268
Lung, see Necturus.

Lymph Hearts of embryo pig,....... 358
Lymphatic nodes, development of in
PIG. CHUDEVOSterc te ce ee yes se 355

Lymph vessels of lung, see Necturus.
Lymphatics, see injection.

MAMMALIA, crural flexors of....... 49, 64
Mammary gland, Proc. Assoc. Amer.
INTE oh ogeoooadbennbagdneUe uno VI
Maturation of egg of Allolobophora
JOYS GOIN) Alteig coe coe Bo cacmicn oe orc 199
Meckel’s cartilage, Proc. Assoc. Amer.
IMIM NS Seon bob oonSae en ooneododD Vi
Medullary sheath of nerves ........ 3338

Mesentery, Proc. Assoc. Amer. Anat. XIV
Mesoderm of three weeks’ embryo...415

classification of ..... -=.. 252
Metanephros, SUPPLEMENT to Vol. IV. 11

Necturus, Blood and lymph vessels
Oi BINGO Grey oi sowsaraa = ine) snes nis e)= 445
see also paraphysis, and pineal.
Negro, see brain and elbow.

Nephric system of three weeks’
Gia i) Sua ie RIO ana se aCe Oe 418

Nerves, see cutaneous, occipital,
cranial, spinal, medullary, leg,
axone.

Nodes, see lymphatic,

Nf Vesey ty Sie soteeratonenenercl sale) svaja.ctls yao esas 205

OccipiTaAL NERVES, development of
in: human-emryo 2...) cee 83

| Skeleton, see

Oesophagus, epithelium of, Proc.
ASSO CHANG T eae heen «sss ieiee VI

PANCREAS, see Langerhans.
Paraphysis, development of, Proc.

INN UNM INA Gono oon oss IX
Parathyroid, framework of.......... (aa
Pelvis) developmenttnot areata 280
Pineal region, development of, Proce.

NSO, TNE NENG ee ene aoa male Ix

RABBIT, development of, Proc. Assoc.
Amer AN@te. .taccmeieeereieeee XI
Reptilia, see mesentery, and Lacerta.

SACRAL VERTEBRAE, development of .270

Salivary glands, Proc. Assoc. Amer.
IMC Snooa Con Ged bdacdcadde€ VI

Sex, see elbow, articular surfaces.

Sheath cells of nerve, see axone.

thoracic, lumbar,
sacral, coccygeal.

Skull of Prof. Cope, Proc. Assoc.

INVER AVIENG ooo OooD GD Odo oad CC - IV
Spinal column, see lumbar, sacral,
coccyx.
column, length and curves,
during development......... 275

ganglia, see ganglia.
nerves, see occipital.
Spindle, maturation spindle of egg. .202

Sibelavianvarteny 505546 rice acres 303
TESTIS, see interstitial cells.
INTE GK Gado aedas, caooposiodcs 231
Thoracic vertebrae, development of. .163
Three weeks’ human embryo........- 409
Trigeminal nerve, see fifth nerve.
Tropidonofius veins) Of... 3... 455

Tubules of kidney, see uriniferous.
Turtle, see chelonia.

URINIFEROUS TUBULES, shape and de-
velopment of, SUPPLEMENT of

WOMEN caso oocosoGs GacUenoo0C 1-98
Urodeles, crural flexors of .... .... 34
VAGUS NERVES. 72 < ec Ais crore aeenetene 104
Veins, of human embryo..... ..... 417

of human embryo’s brain...... 9

also see chelonia, turtle, Rep-
tiles, Tropidonotus, Kinoster-
non.
Vertebrae, see thoracic.
Vocal cords, elastic tissue of....178-188

Se he

THE AMERICAN JOURNAL

OF

ANATOMY

EDITORIAL BOARD

LEWELLYS F. BARKER,

University of Chicago.

THOMAS DWIGHT,

Harvard University.

JOSEPH MARSHALL FLINT,

University of California.

¢.MON H. GAGE,

Cornell University.

G. CARL HUBER,

University of Michigan.

GEORGE 8. HUNTINGTON,
Coiumbia University.
FRANKLIN P. MALL,
Johns Hopkins University.
J. PLAYFAIR McMURRICH,
University of Michigan.
CHARLES 8. MINOT,
Harvard University.

GEORGE A. PIERSOL,
University of Pennsylvania.

HENRY McE. KNOWER, SECRETARY,
Johns Hopkins University.

SUPPEEMEN |

VOLUME IV

THE AMERICAN JOURNAL OF ANATOMY
BALTIMORE, MD., U. S. A.

JUNE 1, 1905

a

TAVSIUOLP

©
s

ON THE DEVELOPMENT AND SHAPE OF URINIFEROUS
TUBULES OF CERTAIN OF THE HIGHER MAMMALS. |
BY
G. CARL HUBER.
From the Histological Laboratory of the University of Michigan.

WitH 24 FIGURES.

As is well known, the excretory system of the amniota develops as a
series of distinct organs, the pronephros, the mesonephros, and the meta-
nephros. The pronephros, which is the first excretory organ to differ-
entiate, and is also phylogenetically the oldest, disappears in all amniota;
its consideration is here dispensed with. ‘The mesonephros, which func-
tionates throughout life as the chief excretory organ of anamnia, is an em-
bryonic organ in amniota, in which it disappears as an excretory organ
and is replaced by the permanent or true kidney, the metanephros. The
development of the metanephros is, however, so closely related to that of
the mesonephros, in both its phylogeny and ontogeny, that a consideration
of the development of the former will to some extent necessitate a con-
sideration of the development of the latter. This will be done only so far
as necessary, as a consideration of the development of the mesonephros
will not form a part of this contribution.

Our present day conception of the anlage and development of the meta-
nephros dates from Kupffer’s contribution based on observations made on
sheep embryos. In an embryo 8 mm. in length he found that from the
dorsal wall of the Wolffian or mesonephric ducts near their posterior ter-
mination, there is formed an evagination which he designated as the
“ Nierenkanal.” These buds, one of which appears in connection with
each Wolffian duct, grow dorsally and cephalad and, as observations on
older embryos revealed, become associated with the development of the
permanent kidney. Anticipating these observations, we find those of
Remak and Ko6lliker, who recognized these buds, but traced their origin
to the cloaca or bladder, and of still earlier date the observations of
Rathke, who recognized a blastema situated between the dorsal wall of the
embryo and the mesonephros in which the kidney had its origin and from
which the ureters were supposed to grow toward the bladder. That the

AMERICAN JOURNAL OF ANATOMY.—SUPPLEMENT TO VOL. IV, JUNE 1, 1905.

2 Development and Shape of Uriniferous Tubules

permanent kidneys have their origin from buds which arise from the
Wolffian ducts is now very generally accepted, Kupffer’s observations hay-
ing been widely confirmed and extended to embrace the different classes
of the amniota. While, as just stated, there is a wnanimity of views
concerning the anlage of the permanent kidney, the views concerning the
mode of development and the histogenesis of the tubular and other struc-
tural elements are still at variance.

According to Remak and somewhat later Kolliker, the development of
the permanent kidneys, after their anlage in the buds arising from the
Wolffian ducts, proceeds in a manner similar to that observed for other
tubular and for alveolar glands, or, as stated by these observers, after the
manner of the development of the lungs. The epithelial kidney anlage
was said to grow forward and to present an anterior vesicular enlarge-
ment, which soon elongates, the anlage differentiating into ureter and
primitive kidney pelvis. On this, solid buds and ampulle make their
appearance, which in their further growth enlarge and elongate and
obtain a lumen, thus forming anlagen for the papillary ducts. These in
their further growth undergo division and give origin to hollow buds
which form the anlagen for the convoluted (secretory) portions of the
uriniferous tubules and the epithelial portions of the Malpighian cor-
puscles. According to this view the collecting ducts and coiled urin-
iferous tubules are developed by direct budding from the epithelial renal
anlage derived from the Wolffian duct.

On the other hand, Kupffer described a discontinuous origin for the
uriniferous tubules. In a sheep embryo of 10 mm. length, he observed
a group of cells, clearly differentiated from the surrounding tissue and
in close relation with the renal anlage—Nierenkanal—in which he
recognized the anlage of the kidney (figured in Fig. 4, Pl. XV of his
article). In this group of cells Kupffer recognized three zones—a zone
consisting of compactly arranged cells in close relation to the renal anlage ;
a middle zone of less compactly arranged cells, and an outer zone in
which connective tissue fibers were observed. In a sheep embryo of 15
mm. length, a differentiation in the middle zone was observed in that
the cells were arranged in twisted cords (“ Zellen sich in gewundene
Streifen ordnen”) which were as yet not clearly differentiated. The
cells of these twisted cords were not connected with the renal anlagen.
These cords were interpreted as the anlagen of the coiled uriniferous
tubules. The darker zone of cells immediately surrounding the renal
anlage was thought to contribute to their further development. Kupffer
could not exclude the possibility that perhaps later generations of
tubules had an origin which differed from that here given. According to

G. Carl Huber 3

this view, the coiled uriniferous tubules have their origin in a tissue
which is distinct from that of the renal anlage. Kupffer’s observations
though faulty in many respects, as later investigations have shown, must
be recognized as of fundamental importance, as concerns both the phylo-
genetic and the ontogenetic development of the permanent kidney.

Nearly all investigators who, since the appearance of Kupffer’s observa-
tions, have considered the question under discussion, have adopted one
or the other of the two views above hastily sketched and, even taking into
consideration the most recent contributions to this subject, the statement
seems warranted that the problem under discussion is still awaiting final
solution. Schreiner, in a recent most admirable contribution to this
subject, especially as concerns the earliest stages of the development of the
permanent kidney, has grouped under respective heads all the more
important contributions dealing directly or indirectly with the develop-
ment of the permanent kidney, and, as it is not my purpose to enter
extensively into a discussion of all the literature bearing on the topic
under consideration since this has been undertaken by Riickert, Herring’
and Schreiner, I have adopted and extended the above mentioned classi-
fication of Schreiner. So far as accessible to me (the exceptions I have
noted), I have critically reviewed the literature to which reference is
here made. It should, however, be stated that, while adopting such
a classification, I must regard it as somewhat forced. A number of the
contributions here referred to and especially certain of the earlier ones,
require interpretation through a view-point gained by familiarity with
more recent investigations and with actual preparations before a classi-
fication of them can be made.

Among investigators who adhere to the view that the tubules of the
permanent kidney are developed by a direct budding from the epithelial
renal anlage derived from the Wolffian duct after the manner of other
tubular and of alveolar glands may be mentioned :—Remak, 55; Kdlliker,
61; Colberg, 63; Waldeyer, 70; Toldt, 74; Pye, 75; Lowe, 79; Ribbert,
80; Hortoles, 81; Kallay, 85; Janosik, 85; Nagel, 89; Golgi, 89; Minot,
92; Haycraft, 95; Schultze, 97; Kollmann, 98; V. Ebner, 99; Gerhardt,
or; Stoerk, or; Strahl, 02; Disse, 02; Stoerk, 04. Stoerk’s fuller publi-
cation gives no additional data as to his view of the origin of the
uriniferous tubules and I assume that he still adheres to the views
expressed in his earlier contribution.

The investigators who have followed Kupffer in assuming a separate
and distinct origin for the coiled uriniferous tubules may be arranged
as follows:

Bornhaupt, 67; Thayssen, 73; Riedel, 74; Balfour, 76; Braun, 78;

a Development and Shape of Uriniferous Tubules

Fiirbringer, 78; Emery, 83; Wiedersheim, 90; Hamburger, 90; Weber,
97: Chievitz, 97; Ribbert, 99; Herring, 00; Schreiner,’ 02; Haugh,
03; Keibel, 03; Felix, 04. These observers, while expressing a unanimity
of view concerning the separate anlage of the tubular system of the
permanent kidney, are not in accord as to the histogenesis of the tissue
from which the tubules are differentiated.

Certain observers, among whom may be mentioned Sedgewick, 80;
Riede, 87; Hoffmann, 89; Gregory, 00, may be separately grouped, since
they assume an intermediate position inclining toward a separate anlage of
the tubules but assuming a histogenetic relationship between the tissue
from which the tubules are developed and the epithelial renal anlage.
This brief array of the contributors to the literature under consideration
may serve as an introduction to a fuller discussion of certain of the more
recent contributions; these and others will receive further mention in

presenting the results of my own investigations.

Of the recent contributions to our knowledge of the development of the permanent
kidney, the article of Schreiner deserves special mention. The results presented are
based on observations made on representatives from the different classes of amniota
and embrace a study of the origin and development of the tubules of the Wolffian body
and of the anlage of the different constituent elements of the permanent kidney. I find
it somewhat difficult to give a brief summary of this article, which is accompanied by
numerous illustrations of sections, of profile reconstructions and of reconstructions after
the Born method. The discussion of his own observations, he begins for each type
investigated with a consideration of the histogenesis of the tubules of the Wolffian
body, following this by a discussion of the origin and development of the permanent
kidney in each type. Schreiner traces the origin of the tubules of the Wolffian body
to a cell-mass which he designates as the nephrogenic tissue (a term suggested by Rabl),
which is identical with the blastema of other authors. This nephrogenic tissue has its
origin in the intermediate cell-mass and extends as an unsegmented cord along the mesial
and dorso-mesial side of the Wolffian ducts to their termination in the cloaca. From the
cells of this cord of cells, which increase in number by division, are differentiated the
tubules of the Wolffian body. These appear first in the anterior segments and, as
development proceeds, also in the posterior segments. The differentiation of the tubules
of the Wolffian body does not, however, extend to the posterior limits of the nephrogenic
tissue, but ceases a number of segments anterior to the termination of the Wolffian
ducts and nephrogenic tissue. The permanent kidneys have their origin in evaginations
(one for each side) from the dorso-mesial wall of the Wolffian ducts not far from their
termination in the cloaca. These evaginations grow dorsally into the posterior portion
of the nephrogenic tissue and-in doing so become capped with the nephrogenic tissue.
The evaginations, or as they are known, the renal anlagen, renal ducts or Nierengiinge,
at first present a bulbous extremity. This elongates in an antero-posterior direction
and develops buds and ampulle. In the meantime, the nephrogenic tissue, which at first
surrounds the bulbous extremity of the renal anlagen as a compact cell-mass, breaks
up into several cell-masses, each of which caps one of the primary branches of the
renal anlagen, at the same time losing its connection with the nephrogenic tissue from
which are developed the tubules of the Wolffian body. This portion of the nephrogenic
tissue Schreiner now terms the metanephrogenic tissue in contradistinction to the meso-
nephrogenic tissue. The metanephrogenic tissue, in the majority of the forms studied
by Schreiner may be differentiated more or less clearly into an inner zone composed of
cells more compactly arranged and presenting other characteristic features and an
outer zone, the cells of which are less compactly arranged and approach in appearance
mesenchymal tissue. Schreiner further shows that the epithelium of the ureters and
pelvis of the kidney as also the collecting tubules to their terminations are developed
from the renal anlagen (Nierengiinge), while the secretory portion of the uriniferous
tubules from their termination in the collecting tubes to and including the epithelial
portion of Bowman’s capsule are differentiated from the inner zone of the metanephro-
genie tissue, the interstitial tissue and the capsule of the kidney developing from the
outer zone. Schreiner’s observations, accompanied by a very full discussion of the
literature and substantiated by numerous figures, appear to argue conclusively for a
separate anlage of the tubules of the Wolffian body and the secretory portion of the

G. Carl Huber 5

uriniferous tubules. I regard it as a distinet achievement on his part to be able to
trace the close relationship between the mesonephrogenic tissue from which are devel-
oped the mesonephric tubules, the separate anlage of which is generally accepted, and the
metanephrogenie tissue from which are developed the secretory portions of the uriniferous
tubules of the permanent kidney. As is no doubt evident, the nephrogenic tissue here
referred to has long been known as the renal blastema: there are, however, various opin-
ions regarding its origin and its share in the development of the kidney. I shall have
occasion to make further reference to this excellent article of Schreiner.

Herring, in a very creditable contribution based on observations made on material
derived from human embryos, was first led to believe that the kidney tubules were
branches of the collecting tubes, but “could never find the early stages, which should
have been easily seen if that view were correct.” He further states that “in the thick
layer of the capsule, cells are seen which show a gradual transition in appearance from
embryonic connective tissue cells to a character resembling that of the epithelial cells of
the early convoluted tubules. Between these cells and those of the ampulle”’ (dilated
ends of the collecting tubes) ‘there is always a distinet line of separation and in the
majority of instances there is a space where the tubule has shrunk during the process
of hardening. The careful examination of serial sections has convinced me that these
masses of cells which appear under the capsule and in the interlobular septa, are quite
independent of the ureter branches and give rise to the Malpighian bodies, convoluted
tubules and Henle’s loop, uniting always with short branches from the ampulle.”

Hamburger’s paper, based largely on investigations made on the kidneys of mouse
embryos and young mice, contains important data concerning the later stages of the
development of the uriniferous tubules and the development of the Malpighian pyramid.
His observations are based on serial sections and on what must be regarded as very suc-
cessful preparations made by maceration and teasing. He makes only incidental mention
of the origin of the uriniferous tubules, which he regards as developing from small
bodies (K6rperchen), the smallest of which are spherical cell groups which have only an
apparent connection with the enlarged distal ends of the collecting ducts.

Haugh, in one of the most recent contributions to the subject, in which he presents
observations made on an extensive material derived from human embryos of all ages,
gives a full account of the development of the pelvis of the kidney and of the shape of
the pelvis of embryonic and adult kidneys, based on metal injections and on wax recon-
structions. He also presents data concerning the lobulation of the human kidney and
on the development of the collecting tubules. He treats, however, very superficially—to
use his own words—the earlier stages of the development of the uriniferous tubules,
allying himself with those authors who defend a separate anlage of the uriniferous
tubules.

Stoerk, on the other hand, defends the older view of the development of the uriniferous
tubules. According to this observer, the renal anlage, derived from the Wolffian duct,
grows dorsally and, after repeated dichotomous division, meets the anlage of the renal
capsule which retards the further radial growth of the tree-like branches of the renal
anlage. Each of the branches now develops an ampullar enlargement at its extremity,
which later assumes a heart-shape. Each half of the heart-shaped enlargement buds out
laterally and forms a short tube, the whole structure now assuming the appearance of
a Y. Each arm of the Y-shaped termination then divides dichotomously, the one branch,
which for a time remains small, passing toward the periphery, the other, which grows
more rapidly, budding and growing toward the stem of the Y. The Y-shaped structure
now presents the appearance of an inverted anchor. The points of this anchor-shaped
structure now grow in such a manner as to assume the shape of an 8. Such an S-shaped
structure is an anlage of a uriniferous tubule and of Bowman’s, capsule. Stoerk’s more
recent contribution contains only a brief statement concerning the anlagen and early
developmental stages of the uriniferous tubules and this is merely a reiteration of the
views presented in his earlier paper. In this second paper, Stoerk gives the results of
observations made on serial sections of material derived from human embryos and
numerous wax-reconstructions made therefrom. These reconstructions represent mainly
the earlier developmental stages of the uriniferous tubules, a fact which is no doubt re-
sponsible for certain errors which he has committed, in using the data gained by a study
of the earlier stages to interpret the shape and structure of more fully developed
tubules. As I shall have occasion to refer frequently to this paper in connection with a
discussion of my own results, its further consideration may here be dispensed with.

Gerhardt, who has recently published, from O. Hertwig’s laboratory, also defends the
view that the uriniferous tubules are developed by budding and that the permanent
kidney is an organ sui generis. As material, he used largely mouse embryos. but also
those of chicken, dog, and pig. Gerhardt’s article is concerned largely with a discussion
of the literature and with theoretic speculations. His own observations are very
briefly given and, as illustrations are wanting, may be summarized by giving one of his
conclusions: ‘*‘ The peripheral portions of the uriniferous tubules arise through a con-
tinuous growth of the collecting tubes. It cannot be demonstrated that in the cortex
formed tubes make secondary connection with the tubules of the medulla.” The capsule
of the glomeruli he regards as an invagination of the distal end of the uriniferous
tubules.

In Chapter IV (Entwickelungsgeschichte und Missbildungen der Nieren) of Vol. 52b,
Deutsche Chirurgie (IKiister, Krankheiten der Niere), written by Strahl, this observer

6 Development and Shape of Uriniferous 'Tubules

states that so far as his own observations go, he is inclined to accept a continuous
epithelial anlage of the uriniferous tubules as given by Minot and Nagel; he does not,
however, desire to commit himself until he himself has made further investigations. <A
number of excellent illustrations are found in this chapter.

Disse, who discusses briefly the development of the kidney in his account of the
anatomy of the urinary organs as given in Bardeleben’s Handbuch, states that as a result
of his own observations made on guinea pig and pig embryos, he is led to conclude that
both the straight and the coiled uriniferous tubules have the same anlage, namely from
the hollow buds of the primitive kidney pelvis. The selection of the section which forms
the basis for his Fig. 77, showing the anlage of the coiled uriniferous tubule seems to
me unfortunate. The S-shaped portion (anlage of the coiled tubule) is not cut through
the center of its lumen, only a small part of which is seen in the figure. The appar-
ently continuous structure as represented might, it seems to me, be simulated by an
overlapping of the two bevelled surfaces of two discontinuous structures (anlage of
coiled tubule and collecting duct), especially as the section from which the drawing
was made seems to have been relatively thick.

Hayeraft’s contribution to the development of the kidney of the rabbit may yet
receive consideration as his observations have been widely accepted. He states that
“the tubules and Malpighian bodies arise as buds of solid cells from the wall of the
primary renal vesicles’’ (ampullar enlargement of the primary divisions of the renal
anlage) ‘close to the cortex. The bud makes a double bend like an 3S, first turning
with a large sweep away from the vesicle, then turning toward it and sharply away
again. The basement membrane can be traced along it and a central lumen forms almost
to its tip.’ Haycraft also traces the development of these S-shaped buds into the
uriniferous tubules and gives his conception of the form of such a tubule as present in
the adult kidney of rabbits. The figures accompanying this article are of special interest
to me, since my series of sections of rabbit embryos of various ages as also numerous
reconstructions of uriniferous tubules made from the same, afford a basis for their inter-
pretation. His Fig. &, * A high power view of the first formation of a urinary tubule
from a primary_ renal vesicle’’ represents, I fear, a much older stage, only a portion
of this tubule being shown in the section sketched. His Fig. 10, I interpret as embracing
parts of at least two tubular anlagen, there sketched as one tube, owing to the position
given to the glomerular anlage with reference to the other portions of the tube. Further
reference to this article will be made later.

Brief mention may yet be made of the conclusions reached by a number of observers
who have investigated the development of cystic kidneys. Hildebrand suggested that
the want of union between the collecting tubules and the anlagen of the uriniferous
tubules was the cause of the cysts of the congenital cystic kidney, in that, when this
condition obtains, as ‘‘ the glomeruli begin to functionate there is no opportunity for
the outflow of the secretion and consequently the tubules are expanded into cystic
structures.’ Ribbert, who describes observations made on a cystic kidney of a new-born
male child, was able to show that many of the cysts were developed from Malpighian
corpuscles which showed no connection with collecting tubules, the ends of certain of
which were also cystically enlarged. Meyer found in microscopic sections of the kidneys
obtained from a 9-weeks-old female child, which showed other congenital defects, cer-
tain regions where the kidney parenchyma was normally developed and other regions
where collecting tubules and Malpighian corpuscles were observed, but where the tubular
portions of the uriniferous tubules which normally unite these structures are wanting
and in place of which there was observed ‘‘an undifferentiated tissue rich in cellular
elements in which the cells only in certain regions were arranged in the form of rings
or cords.” These observations, it appears to me, confirm in a satisfactory way the con-
clusions of observers, who recognize separate anlagen for the collecting tubules and the
uriniferous tubules proper.

From this brief summary of the literature, it may be seen that the controversy found
in the early literature concerning the origin and development of the coiled portion of
uriniferous tubules of the permanent kidney is still found in the more recent literature,
the weight of evidence, however, being on the side of a discontinuous origin of the
coiled portion of the tubule, as Felix has very recently correctly stated.

This investigation was begun in the spring of 1902. In considering
the possible sources of error of other investigators who have dealt with this
subject, it occurred to me that the thickness of the sections studied
might alone be responsible for a certain per cent of the discrepancies
found in the literature. The average thickness of serial sections of
embryos used in embryological investigations is seldom below 10 », more
often perhaps 20 p. My own observations soon convinced me that

G. Carl Huber yi

sections having a thickness of 10 to 20 » were not suitable for interpreting
the early stages of kidney development, as in some sections fields were
found which might be used respectively for demonstrating either a con-
tinuous or discontinuous origin of the coiled uriniferous tubules. It was
further hoped that a solution of the question might be obtained by making
free use of reconstruction methods, especially of the Born wax-plate
method. The wax-plate method of reconstruction had previously been
used in the study of the development of the kidney by Chievitz. He gives
three figures, which are, however, somewhat difficult to interpret, owing to
the fact that the models were not completed by smoothing over the
irregularities which always result when the cut-out portions are placed
together. Since my own investigation was begun and in part completed,
there have appeared three contributions based in part or as a whole on
data gained by reconstructions. Mention has been made of the work of
Schreiner, who, however, studied, among other questions, only the anlage
of the coiled tubules and of this portion of his work he gives no recon-
structions. Haugh’s reconstructions are confined almost wholly to such
of the pelvis and straight collecting tubules of kidneys of relatively young
human embryos. It is rather to be regretted that Haugh did not
devote more time to the finishing of his wax models, which would have
enabled him to give clearer illustrations of the models made, than
accompanying his otherwise excellent article. The most recent con-
tribution to this subject, that of Stoerk, is based largely on results obtained
by means of wax-plate reconstructions, the value of which method he
clearly recognizes. The contributions of Schreiner, Haugh, and Stoerk
which, as stated, have appeared since my own work was in progress, have
been of material assistance to me in formulating my own views. They
do not, however, as appears to me, make superfluous the publication of my
own results. These are confirmatory, as will appear, of that portion of
Schreiner’s work which deals with the anlage of the metanephros and
its tubules; Haugh touches only incidentally on the development of the
coiled uriniferous tubules; and finally my own observations have led me to
conclusions which differ materially in a number of particulars from those
reached by Stoerk.

The investigation, as projected, included not only a study of the
anlage of the uriniferous tubules and their mode of development, but also
a study of the form of the more fully developed uriniferous tubule, if
possible by means of wax-plate reconstruction. That this seemed desirable,
requires, I believe, no argument. A portion of the results here given were
presented at the December meeting of the Association of American

8 Development and Shape of Uriniferous Tubules

Anatomists in 1903, accompanied by a demonstration of a portion of the
models here figured.

MATERIAL.

In selecting the material to be used, it seemed to me desirable to in-
clude in this study developmental stages of both simple and lobulate
kidneys; the embryos selected depended largely on the accessibility of the
material. The embryological material used in the investigation was as
follows:

HUMAN EMBRYOS.

= | LmneTH on AcE oF FIXATION. REMARKS,
iS EMBRYO.
Z
1 | About 15 days old Formalin Received from Dr. Chivers of
Jack., Mich. (Not well preserved).
2 | 10-mm. neck-breech Alcohol Received from Dr. Steiner of
Lima, Ohio.
3 | 15-mm. neck-breech Miiller’s fluid Received from Dr. Roberts, Buf-
TEM, If YC
4 | 18-mm. head-breech Formalin Received from Dr. Peterson, Ann
Arbor.
5 | 20-mm. head-breech Formalin Keceived from Dr. Peterson, Ann
Arbor. Model, Fig. 1 (B)
6 | 22-mm. head-breech io} 01th eo Sromisncorgncokd Gus coe
7 | 38-cm. head-breech Formalin Received from Dr. Hickey,
Model, Fig. 9.
8 | 4-cm. head-breech Formalin Received from Dr. Hickey.
9Rie5:8-cmhead-=breechy a sHoOrLm alin eye smn nee escjeeisirstese terres eee
10 | 6-cm. head-breech Wiwilicres alemiel I eco SISO AUR IAD OS O
11 | 6.5-cm. head-breech | Formalin Received from Dr. Hickey,
Model, Fig. 10.
12 | 18-cm. head-breech Formalin Received from Dr. F. C. Wright,
then of Calumet, Mich.
13 | 14.5-cm. head-breech | Formalin Received from Dr. Darling, Ann
Arbor.
14 | 14.5-cm. head-breech | Formalin Received of Dr. EF. (G2 Wrights
then of Calumet, Mich.
15 | 18.5-cm. head-breech | Formalin Received of Dr. F. C. Wright,
then of Calumet, Mich.
16 | Early part of 7th mo. | Bichloride Dr. Warthin, Ann Arbor, only
the kidneys obtained. Models,
Fig. 4.
17 | 27 cm. in Length Zenker’s and Orth’s | Dr. C. 8. Minot, Boston. Only
fluids the kidneys available.
Ss cChilaiessithanwweels| Hormalin yes | mere ceete eset ere ier ner
old

19 | Child 10 days old Bichloride

G. Carl Huber 8)

CAT.

Fs :

5 a ergy ie ‘FIxarTion. REMARKS.
a

1 | 9-mm. neck-breech Bichloride mn mre paw ye ate we riccstetstecste) sie cnemena' yey te ehexek

2 | 13-mm. head-breech Bichloride Model, Fig. 1 (A).

3 | 20-mm. head-breech Miiller’s fluid Forma GOOD ea Dood Stas

4 | 22-mm. head-breech Bichloride Models, Fig. 6.

5 | 38-cm. head-breech Formalin S500 | pmoAduupoDSDTD.OuS

6 | 4-cm. head-breech Formalin Model, Fig. 16 (A).

7 | 4.5-em. head-breech Novae GUE ye ee ek ell Mee A Seta ee arciiccny ooDiGIdaS

8 | 6-cm. head-breech Formalin Model, Fig. 16 (B).

9 | 9-10-em. head-breech

estimated as just be-
fore birth Zenker’s fluid Models, Fig. 18 (A) and Fig. 20.

10 | Kitten 1 month old Zenker’s and Tellye-
SMO ae) whinGlsy 7 | Sooo sendcndandaodug odes

12 | Kitten 214 months | Zenker’s and Tellye-
old POMOC YS THC i! Epo occudsoooauneccsccsc0s

Full grown cat Zenker’s and Tellye-
oO QP EV NICKS 1) So mako docoossmoouucdsHdS

RABBIT.

1 | 9-mm. neck-breech Belong = lB buscoosnecuoodHoD GoodC

2 | 14-mm. head-breech | Zenker’s Fluid Models, Fig. 5, A, EB, F.

3 | 18-mm. head-breech | Bichloride Models, Fig. 5, B, C, D.

4 | 20-mm. head-breech | Tellyesnicky’s fluid | ......-....--.++++..---

6 | 22-mm. head-breech | Alcohol-acetic = = | ........--+-seeeeeeeoeees

7 | 29-mm. head-breech | Alcohol-chloroform-

| Eanes Nl) db ooeonnccosonnGo moor

8 | 3.5-cm. head-breech | Zenker’s fluid Models, Fig. 14.

9 | 4-cm. head-breech AW GretsithiGl) Nap econbocboopbooOcUGooDC
10 | 4.3-cm. head-breech | Tellyesnicky’s fluid | = ........--.++-2ee++-ees
11 | 4.5-cm. head-breech WihvibereewilpnGl so Gass andoooocdnccUbODe
12 | 4.8-cm. head-breech Mogae | WW sod>c¢anoospocenuceooDooG
13 | 6.5-cm. head-breech Formalin Models, Fig. 18 (B) and Fig. 19 (A).
14 | Rabbit one day old Zenker’s fluid Model, Fig. 19 (B).

PIG.

1 | 5-mm. neck-breech | Sublimate-acetic | .....-.....++.-20++se.0e-

2 10-mm. neck-breech Sublimate-acetic | sce eicieewrctnceesseses

3 | 14-mm. neck-breech | Chromic acid I Basconmonepacrosdg bons:

4 | 18-mm. head-breech Graaf’s fluid I Swi pert avoschican fokelotevavenel aveter oeyapetets

5 | 20-mm. head-breech Picro-sulphuric N) er naca te Ore tie aed ae

6 | 23-mm. head-breech | Chromic acid Ht Store dau tekst suuvaeate a mibiatctstaeer frais

7 | 24-mm. head-breech | Alcohol-acetic Model, Fig. 17, B.

8 | 28-mm. head-breech | Formalin | Models, Fig. 17, A and C.

9 | 3-cm. head-breech jac hie WW BA ange oe oa siggnaGo Dood He
10 | 8.5-cm. head-breech | Formalin =|] nee e cere eeeeeereees
11 | 4.4-cm. head-breech | Formalin 42 || ee vvseeseeeeeeeeees 5
12 | 5-cm. head-breech Wopoenevbiey PP has enianGr didn Sillominc ciao

10 Development and Shape of Uriniferous Tubules

Mernop.

Of the various fixing fluids used, fixation by Zenker’s fluid proved most
satisfactory.

The wax-plate reconstruction method was largely used in gaining the
data which form the basis for this contribution. It was found that the
material could be most readily manipulated in this way by using serial
sections 5 w in thickness and a magnification of 400 diameters, necessitat-
ing the use of wax-plates 2 mm. in thickness. As already stated, it is
often exceedingly difficult in sections more than 5, in thickness to
interpret with any degree of clearness the earlier stages in the develop-
ment of the kidney tubules; the same may be said of sections from more
fully developed kidneys, where the tubules are already well differentiated.
The difficulty here met with is that even in sections having a thickness

of not more than 10 p, the tracing of a tubule through a series of sections
is often made difficult by the fact that portions of two tubules lying over
each other—with reference to the plane of the section—and in contact
are both included in the same section and appear as one tubule, thus
easily leading the observer astray. For the earlier stages—embryos
having a length of about 1.5 em.—the entire embryo was cut in sagittal
serial sections; for embryos measuring from 1.5 to 2.5 em., only the
posterior half was thus cut, and for older stages, the kidneys were
removed after fixation and cut in series either longitudinally or trans-
versely. The difficulty of obtaining unbroken series of sections 5 in
thickness, without wrinkling or distortion when mounted, was met for
many of the series by cutting the sections one by one on an ordinary
sliding microtome—Altmann pattern—with the knife at an angle of about
45° and with the knife blade covered with a layer of distilled water while
cutting. This latter procedure I have found very useful in cutting thin
paraffin sections. The sections are transferred as cut to an Esmarch dish
filled with distilled water in which they float and flatten out. They are
arranged as cut on a slide covered with albumin fixative placed at an
angle in the same dish. When the slide has been filled with sections, it
is placed on the warm oven until the water evaporates. The sections are
then ready for staining. This procedure, though much more time-
consuming than when the sections are cut with an automatic microtome,
seemed to present advantages which compensated for the additional time
necessary for its prosecution; especially was this true for the older stages,
in which the kidneys alone were cut. The sections were stained on the
slide in Ehrlich’s hematoxylin and counterstained mainly in eosin or
erythrosin, the latter giving the better differentiation and sharper con-
trast than the other protoplasmic stains used.

——————— ee CO

G. Carl Huber ah

The outline drawings used in making the models were sketched with
the aid of the camera lucida, a comparatively easy procedure for the
earlier stages, very time-consuming for the later stages, where constant
shifting of the field was necessary. The steps used in making the models
were those generally used when the Born wax-plate method of recon-
struction is employed. Special orienting planes or lines were generally
dispensed with, as it was found practicable to use for this purpose certain
prominent structures within the sections. The reconstruction of the
earlier stages of kidney development—the primary renal anlage and renal
pelvis with its primary branches, also the earliest developmental stages
of the uriniferous tubules—is a comparatively simple procedure. The
same can, however, not be said of the older stages and more fully
developed tubules. The complicated and varied arrangement of the
proximal and distal convoluted portions of the more fully developed
uriniferous tubules, above all the length of the loop of Henle make the
tracing of a single tubule through a series of sections a matter of difficulty.
However, on careful search through a series of sections, instances would
be found where one or the other arm of Henle’s loop was cut through
nearly its entire length. By using this as a starting point, it was generally
possible to trace the remainder of the respective tubule. In doing so, it
was found helpful to make profile or temporary wax reconstructions of
certain parts of the tubule or to make temporary reconstructions of all the
tubules of a given area supposed to contain the portions of the tubule
sought and in this way separate it from surrounding tubules. After thus
mapping out a given tubule in its entirety, drawings were made which
could be used for reconstruction of the whole tubule.

ANLAGE AND EARLY STAGES OF DEVELOPMENT OF THE METANEPHROS.

The earlier stages of the development of the permanent kidney, the
anlage and early development stages of the renal evagination and of the
nephrogenic tissue (kidney blastema) shall here be considered very
briefly, since the appearance of Schreiner’s contribution obviates the
necessity of a fuller discussion of this portion of my investigation.
Neither does it seem necessary to reproduce by way of illustrations the
numerous models made in the earlier part of this work showing shape,
size, and relationship of the renal evagination as found in the eat, rabbit,
pig, and man, nor to discuss the appearances presented by them during
the earlier developmental stages, since such figures would in nearly
every instance duplicate the figures of similar stages given by Schreiner
(rabbit, text-figure 15-21; human, text-figure 26-27 ; pig, text-figure 29-30
of his article) based on profile and wax.reconstructions, except for the

12 Development and Shape of Uriniferous 'Tubules

eat; this animal he did not study, but it presents no special features.
Schreiner considers very fully the anlage and early developmental stages
of the kidney of the rabbit. A brief review of his account may here be
given. In the rabbit, the renal evagination arises in the 31st segment
(about the 11th day, Kolliker, Minot) from the dorso-median wall of the
Wolffian duct near its termination in the cloaca and grows into a tissue
known as the nephrogenic tissue, which soon surrounds its blind end as
a cap of hemispheric shape. As concerns the renal evagination,
Schreiner’s and my own observations confirm those of other investigators,
who, since the appearance of Kupffer’s contribution, have studied the
anlage of the kidney in amniota. Opinions differ, however, as concerns
the histogenesis of the tissue generally known as the renal blastema here
designated as the nephrogenic tissue and as to the role it plays in the
development of the kidney. The general statement may be made that
observers who regard the kidney tubules as developing by a process of
budding, interpret the nephrogenic tissue as of mesenchymal origin,
destined to form the capsule, interstitial connective tissue and vascular
sheaths. Concerning it Minot states that “the histogenesis of the
mesenchymal portion of the kidney is almost unknown.” The majority
of those observers who accept a separate origin for the kidney tubules have
traced their origin to a special tissue (nephrogenic tissue, renal blastema)
the histogenesis of which may here be considered, although in doing so
it will be necessary to anticipate certain facts which will be considered
more fully later. This seems justified, since the tissue in question
makes its appearance before the renal evaginations are present.

Braun, in his account of the development of the urogenital system of reptiles (lizards),
described irregular cords of cells (Nierenzellenstrang, generally known as Braun’s
cords), whose origin he traced to the cclomic epithelium and which were thought
to be the anlage of the uriniferous tubules. Minot regards these cords “as merely the
beginning of the condensed mesenchyma of the renal anlagen,’’ and Schreiner regards
them as identical with the cell-mass designated by him as nephrogenic tissue. Wie-
dersheim, in his account of the development of the urogenital apparatus of crocodile
and turtle embryos, confirms in part Braun’s observations, differing from him in
that he regards the caudal end of the mesonephros as the seat of origin of the cell-
masses or cords from which the kidney tubules are developed. He recognizes further
a condensed mesenchyme which surrounds the metanephric duct and which he char-
acterizes as staining more deeply than the surrounding mesenchyme with which it
blends. This proliferating cell-mass, which, as he states, is also found in other
amniot embryos can not alone be regarded as the ‘‘metanephros blastema,’” as_ it
in itself does not give origin to the strictly glandular parts of the kidney, but the
interstitial connective tissue and the vessel sheaths. Sedgewick, who gave a very clear
account of the anlage of the mesonephros and metanephros as observed in chick
embryos, traces the origin of the Wolffian tubules to a cell-mass designated by him the
Wolffian blastema, which has its origin in the intermediate cell-mass. As described by
him, the Wolffian blastema extends as far back as the 34th segment, that is, to the
opening of the Wolffian duct into the cloaca. This Wolffian blastema breaks up into
the Wolffian tubules as far back as the 30th segment. He states further that “ be-
hind this point it remains for some time almost or quite passive and ultimately gives
rise to the epithelium of the permanent kidneys. In consequence of this, I have
ealled that part of the Wolffian blastema between the 31st and 34th segments the kidney
blastema,’ and again ‘‘ that it is important to notice that this kidney blastema develops
in exactly similar manner to the Wolffian blastema. It is not until the fourth day,
when the ureters have appeared that it is possible to draw a line between the two.”
“In yet older embryos, in which the ureter is more fully developed and overlaps
the hind end of the Wolffian body, the kidney blastema has quite broken off from the
Wolffian body and invests the anterior end of the ureter.’ I have quoted thus fully

.
ee ae

G.. Carl Huber as

from Sedgewick’s paper, because he is the first to give a clear account of the his-
togenesis of the tissue into which grows the renal evagination. Ribbert, in a recent
eontribution, also discusses the origin of the tissue (which he characterizes as a dif-
ferentiated cell-layer), which surrounds the blind end of the epithelial renal anlage,
but gives no definite conclusions. - Certain it is, he states, that the ureters grow
into a dense cell-cord, which may be regarded as the distal end of the mesonephros,
as is shown in his Fig. 4, combined from three sections of a series of sections of a
Guinea-pig embryo. The histogenesis of the nephrogenic tissue (renal blastema) has
been very fully considered by Schreiner, as has been stated in a brief review of his work
given on page 4. For all types studied by him, he was able to show very conclu-
sively that the nephrogenic tissue of the metanephros in common with that of the
mesonephros had its origin in the intermediate cell-mass and extended as an unsegmented
cord of cells through the regions in which is developed the mesonephros to where
the Wolffian duct terminates in the cloaca. In the region of the mesonephros, this
tissue gives origin to the mesonephric tubules; posterior to the mesonephros, it be-
comes associated with the development of the metanephros, forming the nephrogenic
tissue, which surrounds the blind end of the epithelial renal evagination, as was
shown for the chick by Sedgewick. The ahove quotations were selected with a view
of presenting the various views held concerning the histogenesis of the nephrogenic
tissue.

We may now resume a consideration of the anlage and early develop-
mental stages of the metanephros as observed in the rabbit. The renal
evagination, which, as stated, buds dorsally into the nephrogenic tissue,
in its further growth extends in a dorsal direction (the account here given
follows Schreiner) pushing with it the nephrogenic tissue. This now
assumes a dorso-median position with reference to the Wolffian duct,
instead, as before, of lying in contact with it. In its growth, the renal
evagination develaps an enlarged blind end, which may be known as the
primary renal pelvis and a stalk which may be designated as the ureter.
The nephrogenic tissue, which surrounds the primary renal pelvis
differentiates into an inner zone composed of epithelioid cells, which
immediately surround it and an outer zone which presents the appearance
of a condensed mesenchyme, which shows no distinct demarcation toward
the surrounding mesenchyme; the two zones form the metanephrogenic
tissue which has become separated from the mesonephrogenic tissue.
As development proceeds, the primary renal pelvis grows in a dorsal and
cephalad direction and elongates in an antero-posterior direction; at the
same time, it begins to rotate on its axis in such a way as to make its
dorsal surface have a more lateral position. The ureter likewise elongates
and curves cephalad. The metanephrogenic tissue with its inner and
outer zones is clearly recognized and surrounds the primary renal pelvis
as a continuous layer. In slightly older embryos, the primary renal
pelvis reaches a position dorsal to the posterior end of the mesonephros.
In shape it has become still more elongated and presents two slightly
enlarged ends with a narrower middle portion from which arises the
ureter; it shows thickenings and buddings in its wall which in slightly
older stages are clearly recognized as evaginations, the anlagen of the
primary branches of the primary renal pelvis. The primary renal
pelvis now consists of a narrower middle portion, from which arises
the ureter and two enlarged ends and from it and more particularly from

14 Development and Shape of Uriniferous Tubules

the narrowed middle portion, there spring three pairs of branches, each
of which, as soon as clearly differentiated, presents a short stalk and a
bulbous end. The entire metanephros now has a position which is dorsal
to the hind end of the mesonephros. The metanephrogenic tissue, which,
during the earlier stages, formed a continuous layer entirely surrounding
the primary renal pelvis, presents a different disposition with the appear-
ance of the primary branches and this applies more particularly to its
inner zone, which, while it shows the structural appearances observed in
earlier stages, presents a characteristic distribution. On the appearance
of the primary branches, the inner zone of the metanephrogenic tissue no
longer surrounds the entire primary renal pelvis, but only its enlarged
extremities and the bulbous ends of its branches, the narrowed middle
portion and the stalks of the branches being free from it. The inner
zone of the metanephrogenic tissue is clearly recognized both by its
staining reactions and by its structural characteristics. The primary
renal pelvis and its branches, as also the inner zone of the metanephro-
genic tissue, are surrounded by a tissue that stains less deeply than the
inner zone and presents fewer nuclei, the outer zone of the metanephro-
genic tissue. The entire renal anlage is surrounded by mesenchymal
tissue, containing relatively few nuclei, these having a more or less con-
centric arrangement. For a fuller discussion of the anlage and early de-
velopmental stages of the renal evagination and the accompanying changes
presented by the nephrogenic tissue as observed in the rabbit, the reader
is referred to Schreiner’s paper (pages 98-121). My own observations
confirm his results in full except as concerns the anlage of the meta-
nephrogenic tissue, where the necessary stages are wanting in the ma-
terial at my disposal. A separate discussion of the anlage and early
developmental stages as observed in cat, pig§ and human embryos does
not seem necessary here, as this would involve unnecessary repetition,
since the types studied present great similarity in the shape of the renal
anlage and of the earlier stages of its metamorphosis. This is true, not
only when considering types with simple non-lobulated kidneys as the
cat and rabbit, but also when comparing these with types having lobu-
lated kidneys as man and pig. In each type a renal evagination grows
in a dorsal direction into the nephrogenic tissue and in its further growth
differentiates into a primary renal pelvis and ureter, the former being
enclosed and surrounded by the nephrogenic tissue, which in rabbit and
human embryos permits of a differentiation into an inner and outer
zone of nephrogenic tissue, as described by Schreiner, while in pig and
cat embryos of corresponding stages, a differentiation of nephrogenic
tissue into two zones is at this stage not made with certainty. (Compare

G. Carl Huber 15

Schreiner, pages 121 to 128 and pages 138 to 146.) The further devel-
opment of the primary renal pelvis and nephrogenic tissue is for cat,
pig, and human embryos much the same as above described for the
rabbit. For each type, with the appearance of the primary renal branches
of the primary renal pelvis, the inner and outer zones of the nephrogenic
tissue may be clearly made out, the inner zone breaking up into cell
groups which enclose and surround the ends of the branches as these
become differentiated. At this stage, the primary renal pelvis for each
type studied and modeled presents a middle portion of somewhat irregu-
lar shape, the primary renal pelvis proper, from which arises the ureter
and an anterior and caudal end, shghtly enlarged and more or less lobu-
lated (depending on the stage of development) and three pairs of branches
connected with the middle portion; these are not, however, made out
with the same degree of clearness and do not present quite the same
arrangement in the different types studied. These three pairs of branches
are quite readily made out in cat embryos and present an arrangement not
unlike that seen in the rabbit (text-figure 21 B, Schreiner), while in pig
embryos the grouping of the primary branches is not so regular and only
after they have attained some size is it possible to make out clearly what
may be regarded as three pairs of branches. This, though to a less extent,
is true also of human embryos. Attention may, however, here be called to
the fact that models of the primary renal pelvis and its branches, in
embryos of the same genus and of about the same age, even of the two
kidneys of the same embryo (irrespective of size) present slight variation
in the form of the primary renal pelvis and in the arrangement of the
branches. A difference is also noted in the different types in the shape of
the ureter as it enters the primary renal pelvis. In rabbit and cat em-
bryos, this presents only a slight funnel-shaped expansion, while in pig
and human embryos the expansion is much greater and is compressed in a
dorso-ventral direction with its long axis parallel to that of the renal pel-
vis. The enlarged anterior and caudal ends of the primary renal pelvis,
soon after the anlage of the three pairs of branches to which reference
has been made, present, as development proceeds, more and more marked
irregularity of form and develop each two, sometimes three, branches,
which in appearance are like the other branches, showing a slightly nar-
rowed stalk and enlarged ends; at the same time, the inner zone of the
nephrogenic tissue which enclosed and surrounded the enlarged ends of
the primary renal pelvis, breaks up into masses which surround only the
bulbous ends of the resulting branches. In Fig. I, A and B, are repro-
duced two models, obtained by wax reconstruction, of the ureter, renal
pelvis and branches of a cat and a human embryo for a stage slightly older

16 Development and Shape of Uriniferous Tubules

than above discussed, in that the primary branches had already under-
gone secondary division. In these drawings, the nephrogenic tissue, for
the sake of clearness, is not included. The models are purposely shown
from different aspects, the one (A) giving a lateral view, the other (5)
a dorsal view with reference to the position of these structures in the
respective embryos. In both figures, in A more clearly than in B, the
three pairs of primary branches are still discernible. The bulbous ex-
tremities of the primary renal pelvis described for earlher stages show
here a division into branches, more clearly seen in B than in A, the former
representing a slightly older stage.

Fie. 1. Wax reconstruction of the ureter, primary renal pelvis, and
branches. x 50. <A, cat embryo (No. 2), 13 mm. in length;B, human embryo
(No. 5), neck-breech, 16 mm., head-breech, 20 mm. In each figure only about
one-third of the length of the ureter is represented.

The primary branches of the renal pelvis, or, as they may hereafter be
known, the primary collecting tubules (Nierengangiiste—Schreiner), ex-
_tend to near the periphery of the renal anlage. Each shows, as has been
stated, a distal enlargement, which is known as the ampulla (primitive
renal vesicle—Haycraft), and which is surrounded by a cap of tissue
which we have described as the inner zone of the metanephrogenic tissue.
Each ampulla, in its further development, enlarges, the distal wall of the
enlargement becoming somewhat flattened and two lateral buds are de-
veloped; the primary collecting tubule now presents the appearance of a
Tor a Y when seen in longitudinal section or in a reconstruction. The
distal end of each bud, soon after its beginning, develops a new ampulla.
The inner zone of the metanephrogenic tissue which surrounds the pri-
mary ampulla as a continuous cell-mass, begins to show a separation into

G. Carl Huber ; . lrg

two. parts as the ampulla becomes flattened and the separation becomes
complete as the buds or lateral branches develop, each becoming in this
way capped by an inner zone of metanephrogenic tissue. The stage here
reached is the one shown in the reconstruction reproduced in Fig. 1.

RENAL VESICLES.

In Fig. 2 is shown a portion of a longitudinal section of a developing
kidney of a human embryo 18 mm. in length and presenting essentially
the same development of primary renal pelvis and branches as that in B of
Fig. 1; this shows a longitudinal section of a primary collecting tubule, a,
the distal end of which shows an ampulla, slightly flattened and present-
ing on each side a lateral extension recognized as the anlage of a branch.
The section from which the sketch was made does not pass through the

Fie. 2. From a longitudinal section of a developing kidney of human
embryo (No. 4), 18 mm. in length. xX 233. A, primary collecting tubule
with ampulla; b, b’, inner zone of metanephrogenic tissue; c, outer zone of
metanephrogenic tissue; d, anlage of capsule; e, renal vesicle.

center of the tubule. It is owing to this that its wall has the appearance
of being composed of stratified epithelium. In other sections of this
tubule, as also in sections of primary collecting tubules at corresponding
stages from cat, rabbit, and pig embryos, it may be seen that their wall is
composed of a single layer of cylindrical cells, varying somewhat in length,
the nuclei of which are not always in the same plane. In the tissue sur-
rounding the tubule and ampulla here figured, the following may be seen:
Immediately surrounding the lateral extensions of the ampulla and ex-
tending for a short distance along the tubule there are recognized groups
of cells, b, b’, which may be separated from the surrounding cells. These
groups of cells we may know as the inner zone of the metanephrogenic
tissue. The cells of this tissue are indistinctly bounded and_ possess

18 Development and Shape of Uriniferous Tubules

relatively large oval nuclei which stain rather deeply and have a radial
position with reference to the ampulla and tubule. In the preparation
figured they are arranged in two rows quite clearly defined. This inner
zone of the metanephrogenic tissue presents a sharp demarcation toward
the ampulla and tubule, emphasized in the preparation sketched by reason
of the fact that the ampulla and tubule were slightly contracted during
the process of hardening. Herring has called attention to similar appear-
ances presented by his preparations. In the relatively thin sections into
which my entire material was cut, the distinct demarcation between the
inner zone of the metanephrogenic tissue and the epithelium of the pri-
mary collecting tubules and their ampullar enlargements may nearly
always be readily made out, if the plane of the section passes through the
lumen of the tubules and their ampullar enlargements, thus giving a cross
section to their epithelial walls.

Such appearances as described by Sedgewick, when he states ‘‘Some of the larger
columnar cells of the kidney tubules become branched, the processes being continuous
with the processes of the branched cells of the kidney blastema. In fact, every
stage between a columnar lining cell of the tubule and a branched cell of the blastema
is visible’”—I have not observed. The same may be said of the observations of
Riede, who regards the epithelium of the renal evagination and of the primary col-
lecting tubules as contributing to the formation of the renal blastema. It is only in
relatively thick sections or in sections which pass obliquely through the epithelial
walls of a primary collecting tubule without including the lumen, in which the nephro-
genic tissue overlaps the tubular epithelium that a distinction between the two tis-
sues is not readily made. Herring, who speaks of a cap or mantle of cells lying
over the end of each ampulla, speaks of a sharp line of demarcation between the
tube and the cap or mantle of cells. Ribbert describes groups of cells which surround
ecap-like the ends of collecting tubules. These groups consist of two or three layers
of epithelium-like cells which may be separated from collecting tubules on the one
side and from “renal blastema’”’ on the other. Schreiner, for all the types studied by
him, describes and figures a distinct demarcation between the inner zone of the
metanephrogenic tissue and the primary collecting tubules (Nierengangiiste).

Immediately surrounding the inner zone of the metanephrogenic tissue
there is a tissue in which the cell boundaries are also indistinct, with rela-
tively large, round or oval nuclei, which show no definite arrangement, c,
and which do not stain so deeply as the nuclei of the inner zone. This
constitutes the outer zone of the metanephrogenic tissue. This zone
blends with a tissue which is recognized as mesenchymal tissue. In the
mesenchymal tissue, there is recognized a layer in which the majority of
the nuclei show an elongated oval form and fairly regular arrangement,
their long axis being parallel to the outer boundary of the kidney anlage,
entirely surrounding it. In this layer is recognized the anlage of the
capsule of the kidney, d. External to it, there is seen a mesenchyme
containing relatively few nuclei, not sketched in the figure. The entire
renal anlage measures in the embryo from which Fig. 2 was taken almost
exactly 1 mm. in length. In the stained sections it is readily recognized
with the naked eye as a small area, staining more deeply than the sur-
rounding tissue and having a bean or kidney shape. The appearances
presented in sections of the developing kidney of rabbit, cat, and pig

G. Carl Huber . 19

embryos for corresponding stages are in all essentials as here described
and figured for the human-embryo. In each type the inner zone of the
metanephrogenic tissue consists at this stage of generally two, sometimes
three, layers of cells, which, as other observers have stated, have the ap-
pearance of epithelial cells, always surrounding the ampulle of the pri-
mary collecting tubules or of their branches much as shown in Fig. 2.
The same may be said as concerns the structure and disposition of the
outer zone of the metanephrogenic tissue and of the capsule anlage.

If the inner zone of the metanephrogenic tissue found surrounding the
right ampullar extension as shown in Fig. 2 (b’) is traced in sections
preceding and following the one shown in the figure, it will be seen that
its border is not for its entire circumference an even one, but that it
presents a bud-like prolongation which extends for a short distance along
the side of the collecting tubule. In this prolongation, which is here cut
through its center, the cells show an arrangement in two distinct layers,
continuous at the end of the prolongation. I was led to recognize this
fact by Schreiner’s description of similar stages. -Such bud-like prolonga-
tions of the inner zone of the metanephrogenic tissue early acquire a
narrow lumen, around which the cells assume a radial position, certain
cells of the bud in the region of its junction with the main mass of the
nephrogenic tissue turning with their inner ends toward the lumen. In
this way the bud becomes separated from the main mass of the inner zone.
In the figure this process of separation is shown at its very beginning;
a lumen is made out with difficulty and one cell, the fourth in the inner
Tow, appears to have been fixed in the act of turning. In the further
differentiation of such a cell-mass, the lumen increases in size and the
cells surrounding it increase in length, becoming more distinctly columnar
in shape. The whole presents now the appearance of a small vesicle, the
wall of which is formed by a single layer of columnar cells, which during
its differentiation has completely separated from the nephrogenic tissue.
Such a vesicle is shown to the left in Fig. 2 (e). A number of investi-
gators have observed such vesicles and have interpreted them as represent-
ing anlagen of uriniferous tubules. They were first described and quite
correctly figured by Emery, who speaks of them as “ vésicules rénales,”
later by Riede, very briefly by Hamburger, Chievitz, and Ribbert, and
quite accurately by Herring, whose excellent photographic reproductions
are worthy of study, and finally by Schreiner, who gives a minute and ac-
curate account of their origin and structure in the different types of
animals studied by him. The account here given, based on sections of
human embryos, is readily verified in sections of rabbit, cat, and pig
embryos of corresponding stages. In each type small vesicles, which we

20 Development and Shape of Uriniferous Tubules

shall designate as renal vesicles (following Emery, not to be confused
with the primitive renal vesicles as described by Haycraft) are differ-
entiated from the inner zone of the metanephrogenic tissue in the man-
ner above described. The first renal vesicles differentiate in the types
of embryos studied by me simultaneously with the anlage of the two
branches into which the branches (primary collecting tubules) of the
primary renal pelvis divide. That the renal vesicles have no connection
with the primary collecting tubules may generally be made out by tracing
these structures through a series of relatively thin sections. wax recon-
structions of the primary collecting tubules and of the renal vesicles give,
however, conclusive evidence of their independence.

«

Investigators who hold that the uriniferous tubules are developed by a process of
budding from the primary, secondary and further branches of the epithelial renal
evagination do not recognize the renal vesicles as here described. As they can not
have escaped their notice entirely, it must be assumed that they are regarded by them
as continuous with the developing collecting tubules at all stages of their development
and that they were thus interpreted as solid or hollow buds of the collecting tubules.
Few of the contributions to which reference is here made contain satisfactory accounts
of the earliest stages of the development of the uriniferous tubules nor are they
accompanied by figures which clearly portray these stages. The figures presented
are, in the great majority of instances, of stages where a direct continuity between
collecting tubule and uriniferous tubule can not be questioned. Certain of the
figures to which reference is here made showing earlier stages of kidney develop-
ment deserve brief consideration. We may mention Fig. 581 of Kélliker’s Entwickelungs-
geschichte des Menschen und Hodheren Thiere (1879), which shows a longitudinal
section of the kidney of an embryo rabbit 16 days old. In this figure are shown
several primary collecting tubules (‘* Mndsnrossen des Ureters oder Ampullen’’),
showing T or Y-shaped division of their distal ends and at least seven renal vesicles,
if I may be allowed to interpret the figure in the light of observations made on my
own sections of rabbit embryos of about corresponding age. In the accompanying text,
K6lliker states that by further growth of the branches of the collecting tubules, these
form the uriniferous tubules; no mention is made of the renal vesicles. In Figures
1, 2, 3, and 6 (of Plate XXIV) of Léwe’s article, sketches of the entire kidney, as
seen in longitudinal section of rabbit embryos of 5, 10, 20, and 50 mm. length
respectively are shown numerous renal vesicles in various stages of development in
close relation with a tissue, which from its form and position must be regarded as
the inner zone of the nephrogenic tissue. Lowe, who recognized the cell-masses,
has interpreted them as renal vesicles and traced their origin to Braun’s cords, stating
that from them is developed the endothelium of the capillaries of the Maipighian bodies,
while all parts of the uriniferous tubules with the epithelium of Bowman’s capsule
are derived from the primary ureter branches. Again in Figs. 40 and 41 of Strahl’s
communication (kidneys of human embryos of 28 mm. and 35 mm. length respectively)
the inner and outer zones of the nephrogenic tissue and renal vesicles are recognized.
Stoerk’s series of figures grouped under Fig. 2, giving in a ‘‘ semi-schematic’’ way the
development of the ends of the “ureter branches” and the anlage of the Malpighian
bodies (Nieren-kérperchen) show that he has not observed the renal vesicles, but, as
No. 4 of this series of fitures would indicate, he has interpreted the renal vesicles as
buds of the ureter branches, as the position of the downward growing buds, as there
figured, coincides with the relations shown by renal vesicles and primary ureter
branches at a corresponding stage of development. Mention was made of Stoerk’s view
of the anlage of the uriniferous tubules on page 5. Toldt, whose observations are fre-
quently quoted as giving conclusive evidence of the development of the uriniferous
tubules by a process of budding or further growth of the ureter branches, removed the
kidneys from young embryos, stained them in hematoxylin or carmine, cleared them in
turpentine and placed them on a slide as a whole. In such preparations, the primary
branches of the ureter could be traced without interruption to the periphery of the organ
where they terminated in enlargements. A further examination of the ureter branches
designated by Toldt as ‘“ Zellsprossen’’ with higher magnification showed that they
grew forward first as solid buds which later attained a lumen. In Fig. 1 of the
plate accompanying his article (kidney of cat embryo 13 mm. in length) is reproduced
such a preparation. This shows however only ureter, primary renal pelvis and its
primary branches, surrounded hy kidney blastema and gives no evidence of the anlage
of the uriniferous tubules. It should be remembered that it is conceded by the majority
of observers that the epithelium of the primary renal pelvis and its branches is derived
from the renal evagination by growth and budding. The figures given by Toldt
and obtained from sections and teased preparations to show the anlage of the urini-
ferous tubules are of more advanced stages. It must, therefore, be stated that the
method selected by Toldt to show the anlage of the uriniferous tubules is not a suit-
able one. The figures given by Pye, Nagel, Golgi, Minot, Hayeraft, Von Ebner, and
Disse show that they are dealing with developmental stages that do not represent

G. Carl Huber _ 21

the anlage or the beginning of the uriniferous tubules but more advanced stages
where union between tubular anlage and collecting tubule had already taken place
or that when earlier stages are represented as for instance by Disse (Fig. 77), +it
seems apparent that a continuity of the discontinuous structures is simulated by an
overlapping of the two structures. In sections passing through the epithelium of an
ampullar enlargement, without passing through its lumen and through the inner zone
of the nephrogenic tissue and a renal vesicle, it often seems as if these structures were
continuous, presenting the appearance of a bud growing from the ampullar enlargement
of the primary collecting tubule. This is also true if the section passes obliquely
through these structures. In such sections, especially if relatively thick, the inner
zone of the nephrogenic tissue, which, as has been stated, resembles epithelial tissue,
appears as a continuation of the epithelium of the collecting tubule. Waycraft makes
the statement that the most conclusive evidence in favor of the view that the epithe-
lium of the kidney tubules arises from the blastema ‘‘ would be the presence of
isolated masses of cells and their subsequent junction with the tubules, but of this,
as we have seen, there is no proof.’ ‘‘ What we have is an appearance at the tips
of the collecting tubules which suggests to the minds of many observers that the
cells of the blastema gradually transform themselves into epithelium.’’ He further
states that such appearances are due to “nothing more nor less than oblique sections
of tubules, not through their extremities, but through bends in their course.’ ‘To
this it may be answered that reconstructions show conclusively the presence of isolated
cell-masses—renal vesicles, the anlagen of uriniferous tubules—which subsequently
join the collecting tubules and further that the very argument which Haycraft uses
to point out the sources of error of those who hold to the discontinuous origin of the
uriniferous tubules from special cell masses may be used with equal pertinence to
explain the sources of error of observers who hold that the uriniferous tubules are
developed by a process of budding from the collecting tubules, namely that the
latter have been led to interpret as continuous structures certain discontinuous
structures—primary collecting tubules, inner zone of metanephrogenie tissue and
renal vesicles—which by reason of their close relation and a certain resemblance in
structure of their constituent cells appear in oblique sections as forming a continuous
whole. My own observations, pertaining to human, cat, rabbit and pig embryos, as
may be apparent from statements here made, have enabled me to confirm the observa-
tions of investigators who recognize a discontinuous origin for the uriniferous tubules
and who have described small vesicles, variously known as Bliischen, acini, or renal
vesicles, as the anlagen of the uriniferous tubules proper.

DEVELOPMENT OF THE COLLECTING TUBULES AND THEIR RELATION TO
THE RENAL VESICLES.

The further growth of the kidney is accomplished by a further
division of the branches of the primary renal pelvis and their differentia-
tion into the straight collecting tubules, by a constant new formation of
renal vesicles and by a differentiation of the renal vesicle into the urin-
iferous tubules and their union with the collecting tubules. Each pri-
mary collecting tubule early divides, as has been stated, into two branches,
each of which shows a distal enlargement, an ampulla, capped with a
layer of nephrogenic tissue. Each of these branches divides again into
two branches, the division beginning with the ampullar enlargement. With
the formation of this second series of branches, the zone of nephrogenic
tissue which surrounded the ampulla of each primary branch separates
into two parts, each of which surrounds the end of one of the resulting
branches. Several similar divisions follow in relatively quick succes-
sion and with the division of each end branch the nephrogenic tissue
which surrounds it separates into parts, which in each instance surround
the ends of the resulting branches. The end branches from each division
develop ampullar enlargements, which are found at the periphery of the
organs under the developing capsule, which enlarges as the kidney
develops. (In the developing kidney of human embryos, the end branches
are scen to end under the capsule and at the boundaries of the primary

co
eo

Development and Shape of Uriniferous Tubules

lobules, as will be stated more fully later.) The repeated division of
the branches of the primary renal pelvis, as here briefly sketched, results
in systems of branched, relatively straight tubules, which extend from the
pelvis of the developing kidney to the periphery of the organ and are
recognized by authors generally as the anlagen of the straight collecting
tubules. Such tubules are lined throughout by epithelium having rela-
tively clear protoplasm and round or oval nuclei, the arrangement of
which differs somewhat in the earlier branches from that found in the
later branches. In the former, the larger tubules, those nearer the
pelvis, the epithelium is for the greater part of embryonic life of a
pseudostratified type, while in the smaller tubules, those nearer the per-
iphery, the epithelium is simple columnar. The division of the end
branches of these tubular systems continues without much variation from
the manner here described to about the time of birth or until the final
number of collecting tubules has been formed. Hamburger states that
the division of the collecting tubules terminates in the human embryo in
the fifth month of foetal hfe, in the pig in embryos about 14 cm. in
length, in the rat and mouse at about birth. The details of the devel-
opment of the collecting tubules will not be entered upon here, as the
necessary data are not yet in my possession. “This requires a much more
extensive series of reconstructions at different stages of development
than I have been able to make, and not for one animal, but for a series
of animals, as my limited observations indicate that, while the general
plan of development of the straight collecting tubules is the same for
different mammals, variations in detail are to be looked for. _ This may
be seen also from Hamburger’s account, who briefly considers this ques-
tion, basing his observations on teased preparations. For the purpose of
this account, which is concerned with the development and shape of the
uriniferous tubules proper, the brief statement will suffice that the
straight collecting tubules are developed by a process of continuous
growth and of repeated divisions of the ends of the tubules which grow
from the primary renal pelvis. These divisions take place at the per-
iphery of the organ (or of the primary lobes, when present) immediately
under the capsule. It has been stated that with each division of the
ampullar enlargement of an end branch of a system of developing col-
lecting tubules, the inner zone of the metanephrogenic tissue likewise
separates into parts which surround the ends of the resulting branches.
During the development of the kidney, this tissue proliferates by mitotic
division. At all stages of development it is found surrounding the
ampullar enlargements of the end branches of the developing collecting
tubules as a layer consisting of one or two rows of cells, possessing oval

G. Carl Huber . 23

nuclei which stain relatively deeply, the layer being in contact with epi-
thelial cells which form the wall of the ampullar enlargement. This
nephrogenic tissue is therefore found at the periphery of the developing
kidney immediately under the capsule. From this tissue, with each
successive division of the ampullar enlargements of the end branches of
the collecting tubules, new generations of renal vesicles are differentiated
in a manner similar to that previously described for the first generation
of renal vesicles. While new renal vesicles are thus differentiating
about the ends of the terminal branches of the collecting tubules, those
previously formed are developing into uriniferous tubules. Those first
formed show the greatest degree of development, the various generations
of renal vesicles undergoing essentially the same metamorphosis in devel-
oping into uriniferous tubules. Beginning with a relatively early stage
in the development of the kidney through the entire period when new
tubules are forming, there may therefore be observed a peripheral zone
containing the end branches of the collecting tubules showing ampullar
enlargements surrounded by nephrogenic tissue, forming and formed
renal vesicles, and of these others which show the earlier stages of
development leading to the formation of uriniferous tubules. This sub-
capsular zone stains somewhat more deeply than other portions of the
developing kidney and may therefore be recognized with the naked eye.
It surrounds the entire kidney except the place where the ureter enters.
It has repeatedly been recognized as the zone where the earliest stages
of the developing uriniferous tubules are to be found and is described by
Hamburger and Stoerk as the neogenic zone (“die neogene Zone ”—
Hamburger). In this way new generations of uriniferous tubules de-
velop outside of those previously formed, the latter thus coming to lie
deeper down in the parenchyma of the kidney and showing according to
their age more or less advanced stages of development. The period when
the new formation of renal vesicles ceases varies somewhat for different
animals. In the rabbit, I have observed their formation as late as the
first week after birth, although the number of newly forming renal ves-
icles is for the later period of embryonic life relatively small. In the
cat, newly forming renal vesicles were observed in a kidney removed
from an embryo a short time before birth. Herring states that no more
tubules are formed in the human kidney after the 8th month of feetal
life. Stoerk, who has considered this question quite fully, agrees with
Herring. Toldt, on the other hand, states that new Malpighian cor-
puscles form in dog and man in the entire peripheral part of the kidney
for 8 to 10 days after birth. My own human material is too limited to
be of any value in determining this question.

24 Development and Shape of Uriniferous 'Tubules

DIFFERENTIATION OF THE RENAL VESICLES.

We may now consider the further development of the renal vesicles.
As has been stated, these are differentiated from the inner zone of the
metanephrogenic tissue and, when completely separated, consist of a
single layer of columnar cells with oval nuclei which stain relatively
deeply and surround a central lumen. At this stage of their develop-
ment they are generally quite definitely circumscribed and have a shape
which varies from that of an irregularly spherical body to that of an
egg. They le in close relation with the collecting tubules and their
ampullar enlargements, these being often slightly flattened in the region
of their contact with the vesicles. At least two renal vesicles develop
in connection with each ampullar enlargement; these do not, however,
present the same degree of differentiation and development, as may be
seen from Fig. 2, and is more clearly seen in renal vesicles which de-
velop in connection with the ampullar enlargements of the end branches
of collecting tubules which appear after the first and second division of
the branches, which arise from the renal pelvis. The branches resulting
from each successive division of a collecting tubule form with this soon
after their anlage a T-shaped structure; in their peripheral growth, the
angle between such branches becomes smaller and they form with the
collecting tubule a Y-shaped structure, each branch developing an am-
pullar enlargement surrounded by nephrogenic tissue. A renal vesicle
develops from the nephrogenic tissue on the outside of each branch
(therefore nearer the pelvis of the kidney) earlier than from the nephro-
genic tissue between the branches. (See also Schreiner, especially text-
figure 34.) The renal vesicles, soon after their separation from the
nephrogenic tissue, increase in size by active proliferation of their cells
by mitotic cell division, the vesicle elongating somewhat and coming in
close contact with the ampullar enlargement of the collecting tubules.
At the same time it may be observed that the outer wall of the vesicle,
that away from the collecting tubule, and especially its upper portion,
becomes thicker than the wall of the vesicle in contact with the collect-
ing tubule. This thickening of the outer wall of the vesicle was first
described by Schreiner and fully discussed by him. He states, and in
this my own observations confirm his account, that the cells of the middle
and upper portions of the outer wall of the vesicles elongate, the cells
here appearing longer when the vesicles are in sagittal section’ than in

1 By a sagittal section of a renal vesicle is here meant a section which cuts a col-
lecting tubule in longitudinal direction and parallel with its lumen and passes at the
same time through a renal vesicle which may develop by its side, thus cutting the renal
vesicle in a plane which passes through the middle or nearly the middle of its outer
wall. The renal vesicles seen in Fig. 2 and A of Fig. 3 are cut in sagittal section.

The plane of a frontal section of a renal vesicle passes perpendicularly through it and
would intersect the plane of a sagittal section at right angles.

G. Carl Huber

aoe
or

other parts of the vesicles and that evidences of cell division are often
seen.

In such a section it will be seen that the lower inner ends of these
elongated cells, which no longer hold a radial position to the lumen of,
the vesicle, but are inclined to its long axis, slightly overlap the inner
ends of the cells which form the lower portion of the outer wall of the
vesicle, so that a portion of the outer wall of the vesicle presents the
appearance of a two-layered epithelium. As development proceeds, the
outer wall of the vesicle further thickens and encroaches on its lumen,
which in sagittal sections now presents the appearance of a hook-shaped
space. At about this time there appears on the outer wall of the vesicle
over its thickest portion a slight depression, beneath which a cleft makes
its appearance; this begins to separate the thickest portion of the wall
of the vesicle into two layers. This cleft deepens as the two layers of
cells become more distinctly separated. In A of Fig. 3 is shown a sagit-
tal section of a renal vesicle at this stage of development. The section
sketched does not pass exactly through the center of the vesicle, nor is it
quite parallel to its long axis. This needs to be remembered in exam-
ining this figure, as the arrangement and size of the cells forming the
different parts of the wall of the vesicle are not quite what might be
expected: from the foregoing description. The reproduction gives the
most typic section of the series of sagittal sections of the vesicle from
which the reconstruction was made which is shown in A of Fig. 4. In
A of Fig. 3 is shown an ampullar enlargement of a terminal branch of a
collecting tubule surrounded by nephrogenic tissue, a renal vesicle
and the surrounding mesenchyme and a _ portion of the capsule.
In the wall of the vesicle away from the collecting tubule may be
observed a slight depression and beneath this a, cleft. The nuclei of
the cells adjacent to the cleft are stained a little more deeply than the
other nuclei of the vesicle. Such an appearance is now and then seen,
though not characteristic of this stage. In the reconstruction, this cleft
appears on the surface as a narrow depression placed nearly transversely
to the long axis of the vesicle, a little above its center, though not ex-
tending across its entire wall, and is therefore scarcely seen in the ex-
posure of the model as sketched. That portion of the vesicle, which in
sagittal section shows the narrow cleft, is in reality a small pocket, the
upper and lower walls of which are nearly in contact at this stage of the
development. Similar appearances I have observed in a: number of
other wax reconstructions of this stage. This pocket is not developed
primarily by an infolding or invagination of the outer wall of the vesicle,
but differentiates in a thickening of its outer wall. This thickening

26 Development and Shape of Uriniferous Tubules

results from a slight readjustment of the cells constituting this part of
the vesicle as well as by their proliferation. In the further develop-
ment, this cleft enlarges by deepening and by extending laterally so as
to involve more than the outer wall of the vesicle. In this way the lower
and outer part of the original vesicle becomes separated from the main
part of the vesicle and presents the appearance of a lip, which projects
upward toward the periphery of the kidney. Into this lp-shaped por-
tion the lumen of the vesicle extends. While these changes in form and
structure which affect primarily the lower part of the vesicle are taking
place, the upper part of the vesicle elongates, sometimes more and some-
times less, growing toward the respective collecting tubule, more cor-
rectly its ampulla; at the same time, a slight depression appears in the
uppermost part of the inner wall of the vesicle, that in apposition with
the wall of the collecting tubule, which emphasizes the curvature of the
upper part of the vesicle toward the collecting tubule. In A of Fig. 3
is shown the first indication of this second depression in the wall of the
renal vesicle at a point marked by a cross. The end result of these
changes in form and structure on the part of the renal vesicle are rep-
resented in B of Figs. 3 and 4. It may be seen that by a deepening of
these clefts, the more prominent one in the outer wall, the less promi-
nent one in the upper portion of the inner wall of the original vesicle,
and by a growing of its upper end toward the collecting tubule, the
vesicle has been altered in its shape so as to present in sagittal sections
and at times also in reconstructions the form of an S. In the sagittal
section of this renal vesicle, or, as we shall now know it, the tubular anlage
(the word uriniferous being understood) is cut longitudinally and
through its middle in the section sketched for nearly its entire length,
the upper portion of the anlage deviating a very little from this plane
of section. 'The reconstruction of this tubular anlage, B of Fig. 4, does
not show this S-shape as clearly as certain of the sections of it. The
upper bend of a tubular anlage of about this stage presents a circular
and very narrow lumen. In the tubular anlage figured, the lumen of its
upper bend appears clearly only in the section sketched. In frontal sec-
tions of tubular anlagen of this stage, in which this upper bend would
appear in cross or slightly oblique section, the existence of a small lumen
in this portion of the anlage may be clearly made out, the part itself
being round or nearly so. It constitutes, therefore, a tubular structure,
its wall being made up of a single layer of columnar cells with oval
nuclei. Frontal sections of a tubular anlage of this stage further show
that its lower bend does not represent a structure having a tubular form,
but is flattened from above downward, presenting in such sections the

ras)
~

G. Carl Huber |

appearance of a crescent with concavity directed upwards toward the
periphery of the kidney. Reconstructions of tubular anlagen of this
stage show that this portion presents the shape of the bow] of a spoon or of
a saucer with the concavity turned toward the other parts of the anlage.
In making this comparison, it is to be understood that this structure is
not solid, but double-walled, the enclosed space, which is continuous with
the lumen of the other parts of the anlage, having a similar shape. At
this stage in the development, the upper concave wall of the structure is
composed. of columnar cells, while the lower or convex wall is composed
of cubical cells. Both in the series of sections of the tubular anlage
shown in B of Fig. 3, and also in the reconstruction of the same, B,
Fig. 4, it may be clearly seen that it is not continuous with the ampullar
enlargement of the collecting tubule. It is in close contact with it but
not continuous with it. In C of Figs. 3 and 4 is shown a tubular anlage
which is only slightly further advanced in its development than the one
shown in B of the same figures. The section sketched shows, however,
a continuity of the lumen of the tubular anlage and that of the ampulla
of the collecting tubule, although, as is evident from the sketch, the
place of union of the two previously discontinuous parts may yet be
made out, partly by reason of the position of the nuclei of the respective
cells, also because there is as yet a slight constriction at the place of
fusion, this being not as yet complete. The modus operandi of this
union between tubular anlagen and collecting tubules has been discussed
by Schreiner, who states that he has observed in the tubular anlagen or
in the ampullar enlargement of the collecting tubule at a stage which
he has recognized as just previous to a stage in which there is union
of the two, a cell in mitotic division or cells which by their structure and
staining show that they have just completed division. This division of
cells at the place of contact of the upper end of the tubular anlage and
the ampulla of the collecting tubule, Schreiner associates with the fusion
of these parts and with the formation of a continuous lumen. This
stage is rather difficult to find, as it is not readily recognized unless the
respective tubular anlage is seen in a favorable sagittal section. In the
tubular anlage, a section of which is here shown, there is no evidence
which would indicate that one or more of the cells situated at the place
of its junction with the collecting tubule had just completed its division.
In my own preparations I have not found cell division at the place of
contact between tubular anlagen and collecting tubules more frequent
than in other parts of these structures in a stage which might be inter-
preted as just prior to a fusion of these structures. My own observa-
tions would, therefore, confirm only in part those of Schreiner as con-

Fig. 3. A series of figures, A to J of sagittal sections of tubular anlagen
and uriniferous tubules in early stages of development, showing successive
stages in the development of the uriniferous tubules, from a human embryo
of the seventh month (No. 16). 160. In each figure of this series, there is
reproduced the most typical section of the several series of sagittal sections
used in making the reconstructions shown in Fig. 4. (The corresponding
tubules in Figs. 3 and 4, are designated by the same letter.)

~G. Carl Huber 29

Fic. 4. A series of models A to J, showing successive stages in development
of tubular anlagen and early stages in the development of uriniferous tubules,
with a portion of the collecting tubule to which each is attached; from a
human embryo of the seventh month (No. 16). X 160.

30 Development and Shape of Uriniferous Tubules

cerns this point. I have, however, accepted Schreiner’s account of the
processes of union of the tubular anlagen and collecting tubules as the
most satisfactory that can be given. In Fig. 5 are reproduced a series
of wax reconstructions of the anlage and early developmental stages of
uriniferous tubules obtained from rabbit embryos; in Fig. 6, a similar
series obtained from cat embryos, and in Figs. 9 and 10, a primary col-
lecting tubule, with its branches, the renal vesicles, tubular anlagen, and
uriniferous tubules in early stages of development associated with them.
These were obtained from kidneys of human embryos of 3 cm. and
6.5 em. length respectively. In A and B of Fig. 5 are represented in
each instance a collecting tubule with prominent ampullar enlargement
and a renal vesicle. In A the renal vesicle is of spherical shape, in B
of ege shape, the latter representing a slightly older stage, showing,
however, as yet no characteristic differentiation; both are distinctly
separated from the respective collecting tubule. In C of Fig. 5 is repro-
duced a tubular anlage, which had united with the collecting tubule at a
relatively early stage, before presenting the S-shape to which attention
has been called. The lumen of this tubular anlage, as seen in sections
of it, shows this shape more clearly than does the reconstruction. The
same may be said of the tubular anlage shown in A of Fig. 6, the earlier
stages of the series being here not reproduced. These two figures have
been introduced to show that the tubular anlagen do not always present
the same degree of development and differentiation at the time of their
union with the collecting tubules. Characteristic and constant differ-
ences in this respect I have not observed for the different animals studied.

Variations in the time of fusion of the tubular anlagen and the col-
lecting tubules within the limits shown in these figures may be observed
in a series of sections from developing kidneys obtained from embryos of
the same species of animals. Such fusion does not take place, so far
as my observations go, prior to a time when the renal vesicles show a
distinct indentation of their outer wall, the lumen thus presenting the
shape of a hook, but may take place before the inner upper portion of
the vesicles or tubular anlagen show an indentation leading to the for-
mation of the upper curvature; in this case the curvature develops after
fusion has taken place. A study of the tubular anlagen shown in Fig. 9
and a comparison of these with those shown in Figs. 4, 5, and 6, will
prove instructive, especially the two tubular anlagen seen to the left of
the upper ends of the two prominent diverging branches of the collecting
tubule. Since the latter are sketched from a different view than those
of the previous figures, they represent tubular anlagen just after fusion

G. Carl Huber 31

with the collecting tubules and show clearly the saucer-shaped expansion
of the lower bend of the S-shaped structure characteristic of this stage.

S-sHAPED STAGE IN THE DEVELOPMENT OF THE URINIFEROUS TUBULES.

To characterize the S-stage of the development of the uriniferous
tubules more clearly, one may speak for convenience of description of an
upper S-curve, with concavity toward the collecting tubule; an S-middle
piece or middle S-segment, and a lower S-curve with convexity toward
the collecting tubule. In considering the tubular anlage as a whole,
one may speak of its inner face or side, that turned toward the collecting
tubule, and an outer side, that turned away from the collecting tubule
and further of a front and back side or face, these with reference to a
plane passing through the middle of a tubular anlage as in a sagittal
section of the same, and finally of an upper and lower portion of the
anlage, with reference to its attachment to the collecting tubule, which
would mark its upper portion. In the same way we may speak of a
portion of a tubular anlage as curving or growing inwards or outwards,
upwards or downwards, and forward or backward. In the further de-
scription I shall make use of these terms without further comment. In
this connection it needs to be recalled that renal vesicles and tubular
anlagen develop in connection with each of the two end branches result-
ing from the successive divisions which occur in the course of the devel-
opment of a collecting tubule. Favorable sections now and again show
a collecting tubule with two more or less clearly differentiated end
branches connected with two tubular anlagen seen in sagittal section.
The whole structure presents the appearance of a Y or of an anchor
with the arms bent toward the stem of the Y or shaft of the anchor (see
Figs. 10 and 12). Assuming that the two tubular anlagen are in the
S-stage, only the left one would in reality have the shape of a Roman 8,
while the right one would show a mirror picture of the same. This fact
enables one readily to state whether any particular tubular anlage is
placed to the right or left of a given collecting tubule. By right or left
is meant the relations shown by these structures to the collecting tubule
to which they are attached in any given section. In the series of figures
showing reconstructions of tubular anlagen and early developmental
stages of uriniferous tubules, as given in Figs. 3, 4, 5, and 6, I have
for each stage of development selected a tubular anlage placed to the
right of a collecting tubule. It seemed to me that this would assist ma-
terially in tracing the successive stages and would facilitate a comparison
of similar stages. ‘Tubular anlagen placed to the left of the collecting
tubules show the same developmental .stages which would appear in

32 Development and Shape of Uriniferous Tubules

Fic. 5. A series of models A to F, showing successive stages in the devel-
opment of uriniferous tubules in rabbit embryos. 160. A, E, and F, from
rabbit embryo (No. 2) of 14 mm. length; B, C, and D from rabbit embryo
(No. 3) of 18 mm. length. A and B show each a renal vesicle and a part of a
collecting tubule with prominent ampulla; C and D early and late S-shaped
tubule.

G. Carl Huber : 33

Fie. 6. A series of models, A to H, showing successive stages in the devel-
cpment of uriniferous tubules in a cat embryo 22 mm. in length. x 160.

34 Development and Shape of Uriniferous Tubules

mirror pictures to those here given and show the same relations to the
collecting tubules. ‘They show similar inner and outer, front and back
surfaces as above described, these designations having reference to the
relations of any tubular anlage to the collecting tubule to which it is
attached, independent of the fact that such anlage may be placed to the
right or left of the collecting tubule. With this definition of the ter-
minology to be used, we may now proceed to a fuller consideration of
the S-shaped stage and state that the upper S-curve represents a tubule
with narrow lumen haying an outward curvature and extends without
definite boundary into the S-middle piece, which is also of cylindrical
shape, representing a short tubular segment also with narrow lumen and,
like the upper S-curve, is composed of a single layer of cells with oval
nuclei. When first recognized, it has a nearly horizontal position; but
as development proceeds and the S-shape of the tubular anlage becomes
more pronounced, it loses this horizontal position and inclines upward
toward the collecting tubule. This portion of the anlage is not so clearly
defined as the other parts and the relative time of its differentiation
varies somewhat, as will be more fully stated later. The lower S-curve,
as has been stated, is not of cylindrical shape, but from the time of its
anlage is flattened from above downward and presents the form of a
double-walled saucer or shallow bowl with concavity directed upward,
the cavity which extends into this portion being likewise flattened and
presenting a similar shape. The cells forming the ‘upper wall of this
structure are of columnar shape, those forming the lower wall of cubical
shape or even more flattened, depending on the stage of development.
The saucer-shaped structure becomes narrowed as the region where it
joins the S-middle piece is approached and here gradually changes (in
the most typical anlagen) from a flattened to a nearly cylindrical struc-
ture. The place of junction of the two parts is indicated by a sharp
bend in the lumen of the tubular anlage as this is seen in sagittal sections
and may be indicated by a nearly equally sharp bend when the anlage
is seen in reconstructions. The lower end of the upper S-curve, with
a portion of the S-middle piece, rests in the concavity of the saucer-
shaped portion representing the lower S-curve so that only a narrow
space separates these parts. In sagittal sections this space appears as a
curved cleft and contains, soon after its anlage, a delicate strand of
mesenchymal tissue. (See C of Fig. 3.) In D of Figs. 3 and 4 is
shown a slightly older tubular anlage, which differs from the former in
that the place of union between the tubular anlage and the collecting
tubule is no longer recognized and in the more pronounced curvature
of the anlage as a whole. This is especially clearly seen in the section

G. Carl Huber

Oo
or

of this tubular anlage shown in PD of Fig. 3. In the reconstruction it
will be seen that what appears in section as the lower S-curve represents
in reality a shallow bowl, the edges of the saucer-shaped structure recog-
nized in the former stage and figure having grown upward, thus deep-
ening the concavity; also the space which separates this part of the
anlage from the lower end of the upper S-curve and S-middle piece has
increased somewhat and is now occupied by a vascularized mesenchyme,
the section reproduced showing two capillaries, the one cut cross-wise,
the other seen in longitudinal section. The mesenchyme contained
within the concavity of the lower S-curve has long been recognized as
the anlage of a glomerulus, this portion of the tubular anlage, as may
be stated now, forming Bowman’s capsule. My own observations lead
me to think that the first capillary loop found within the mesenchyme
occupying the concavity of the lower S-curve of a tubular anlage grows
into this from without, as I have generally been able to trace a connection
between it and the capillaries outside of the cleft occupied by the mesen-
chyme. To give definite answer to this question, however, it would be
necessary to confirm the above statement on injected preparations, but
the fresh material of the proper stage of development, obtained during
the course of this investigation, was too limited to attempt this. This
question has received only incidental mention by other investigators; of
the more recent of these we may mention Schreiner, who simply states
that “into the cleft there grow from without connective tissue and
vessels,” and Stoerk, who, in speaking of the development of the S-shaped
stage of the tubular anlage, states that “there is formed between the
upper curve and the middle piece on the one side and the concavity of the
lower curve on the other side a downward curving cleft, which takes up
vascularized interstitial tissue, from which later the glomerulus is de-
veloped.” Herring, on the other hand, states that when the S-shaped
tubule is formed, the space between the lower and the middle limb is
occupied by some connective tissue cells; they are few in number and
resemble the other cells which are found in the matrix of the cortex. He
further states that “ From these cells I believe the capillaries of the
glomerulus are formed im sitw and not as an ingrowth.”

The S-shaped stage in the development of the uriniferous tubule, especially charac-
teristic when seen in sagittal sections, has long been recognized and was first correctly
described by Toldt and is clearly brought out in Golgi’s semidiagrammatic figures of
developing uriniferous tubules of mammals. The series of developmental stages leadin
to the formation of the S-shaped tubular anlage have, I believe, been correctly described
only by Schreiner, whose account my own observations, as given here, in the main
confirmed. > ; ;

Ribbert gives in a series of semidiagrammatic figures (Fig. 3 of his article) his
view of the development of the S-shaped stage of the tubular anlagen. By means
of these figures, he aims to show that the lower portion of the cell-mass from which
a tubule is developed, in its growth turns sharply upward, toward the periphery of the
kidney. In this way, a cleft is formed between the elongation and the cell-mass
from which it develops; the latter now makes union with the collecting tubule, the

36 Development and Shape of Uriniferous Tubules

whole structure acquiring a lumen and the S-shaped stage of the tubular anlage is
reached. Herring states that ‘‘ By active proliferation of cells, the vesicle grows in
length, but not as a straight tube; it becomes comma-shaped, the bend of the comma
being always away from the collecting tubule with which it eventually joins.’ ‘‘ The
comma is now in close contact with the ampulla, but the lumens are not continuous.
As the comma-shaped body grows, it elongates, but the tail remains in the same posi-
tion; further growth is upward toward the capsule along the end of the collecting
tube. It grows more quickly than the latter and takes on another curve at its upper
end, this time toward the ampulla. The result is an S-shaped body, with centra!
lumen. When the §S is well developed, its lumen becomes continuous with the lumen
of the collecting tube.’’ In both of these accounts, the anlage of the lower curve of
the future S-shaped structure, as above given, is missed. Hamburger states that
when the anlage of the coiled uriniferous tubule has reached a diameter of 40 « to 50 nu,
a small cavity develops in its center (renal vesicle) and shortly after this, a union
between it and the ampulla is established and a short solid cord of cells, which extends
from the peripheral end of the ampulla to the lateral wall of the vesicle develops.
In its further development, the tubular anlage obtains a depression, always seen on
the side of the anlage which is turned away from the ampulla with which it is con-
nected; its wall sinks in somewhat, thus giving the cavity the form of a “ half-moon
shaped cleft;’’ this depression deepens and develops into a cleft and there is formed
a ‘“‘wing-like process,’’ which lies against the side of the anlage and terminates with
a sharp free border, which projects toward the surface of the kidney. When the
anlage of the coiled uriniferous tubule has attained the size of 50 » to 60 vp» it still
has a spheric shape” but consists of 1, a bowl (Schale), the concavity of which
is turned toward the surface of the kidney, 2, a short thick tubule of S-form, which
unites this bowl with: the ampulla.’’ This stage of the development corresponds with
the S-shaped stage of other authors. Such early fusion of the anlagen of uriniferous
tubules with collecting tubules as described by Hamburger I have not observed.
This error of observation on his part (for such I take it to be) is, I believe, accounted
for by the thickness of the seetions which he examined (15 to 20 »). In sections
of this thickness, it is often exceedingly difficult to state whether or not renal
vesicles or tubular anlagen are merely in contact with the collecting tubules or
continuous with them, unless the parts are most favorably sectioned, a point to which
I have previously called attention. That the depression on the outer side of the renal
vesicle and especially the cleft which opens into it is not primarily formed by an
infolding of the wall was brought out in describing its formation; the solid cord of
cells uniting the ampulla with the renal vesicle, as described by Hamburger I have not
observed. The upper portion of the renal vesicle retains a lumen as it elongates
toward the collecting tubule and this becomes continuous with that of the collecting
tubule at the time when fusion between these structures takes place.

Authors who regard the uriniferous tubules as developed directly
from the collecting tubules are quite unanimous in their accounts of the
formation of this S-shaped stage of the uriniferous tubules, stating that
the end branches of the uriniferous tubules, or buds from these, grow for
a short distance toward the pelvis of the kidney and then again toward
the periphery, the lobule or bud thus becoming flexed, presenting first a
concavity, then a convexity toward the collecting tubule. I have pre-
viously quoted Haycraft, Gerhardt, and Stoerk as concerns this point.
Minot considered the capsule as playing an important réle in the forma-
tion of the tubular anlagen. He states: “The capsule seems to prevent
the further elongation of the branch in its line of growth, and to force
the end of the branch to curl over, thus by simple mechanical condition
causing the formation of the anlage of the Malpighian corpuscle.” Ger-
hardt also states that the capsule offers a mechanical obstacle to the
further growth of the straight tubules and thus gives the first impulse to
the coil formation. Assuming for the,moment that the uriniferous
tubules are developed by a growth and budding of the collecting tubules,
it must be evident that other factors than the one found in the resist-
ance offered by the capsule to the further elongation of the branches of
the collecting tubules, play a part in causing these branches to curl

G. Carl Huber - 37

over and form the anlage for the uriniferous tubules, as in the develop-
ing kidneys of human embryos of from 3 cm. to 6 em. in length the
neogenic zone is found, not only at the periphery of the organ, imme-
diately under the capsule, but forms septa which extend nearly to the
pelvis of the kidney and are found at the line of union of 4 to 6 primary
lobules into which the parenchyma of the organ may at this early stage
of its development be separated. 'These septa, which Haugh has termed
“the primary column Bertini,” are in reality composed of the neogenic
zones, separated at the time when they are first recognizable and for
some little time after by only minimal traces of mesenchymal tissue,
which occurs here and there in small areas and does not present at this
stage of development the relatively dense structure of the capsule, which
dips down at the superficial boundaries of these primary lobules for a
variable though always short distance. Uriniferous tubules develop on
each side of these septa in the same way as under the capsule and it is
evident from a study of sections that the end branches which reach these
septa or primary columne Bertini do not meet resistance in their further
growth by the capsule or by a tissue of like density, since such tissue is
not found in the primary columne Bertini, when first formed and for
some little time after.

In Fig. 7 is reproduced a portion of a section from a series of longi-
tudinal sections of a kidney obtained from a human embryo measuring
6.5 em. in length (No. 11), showing the junction of two primary lobules.
In this figure is represented a primary columna Bertini, which extends
from the capsule nearly to the pelvis of the kidney and the renal paren-
chyma developing on either side of it. As will be observed, collecting
tubules grow toward the septum from either side and renal vesicles and
tubular anlagen develop on either side of it in a manner similar to those
which develop under the capsule. For the greater part of its extent
this columna or septum contains at the stage of development shown in
the figure very little mesenchymal tissue. This is well developed only
near the pelvis of the kidney, from which it extends outward for a short
distance and here clearly separates the two lobules. For a fuller con-
sideration of the primary lobulation of the developing kidney and of
the primary column Bertini, the reader is referred to Haugh’s paper,
who considers these questions at some length.

Stoerk, in his second paper, discusses and figures (Figs. 3-8) wax reconstructions
of tubular anlagen showing the S-shape. I desire to consider somewhat more fully
his observations relating to this stage of the development of the uriniferous tubules,
as these observations are not in harmony in all particulars with my own. Stoerk in
this paper begins his description of the development of the uriniferous tubules with a
stage, which, in profile view, presents the appearance of an inverted question mark.
This twisted structure simulates, as he states, in sections the appearance of an
S-shaped tubule and is s» recognized by authors. With reservation, he adopts the same
designation. In this he recognizes, for convenience of description, 3 parts (and in

38 Development and Shape of Uriniferous Tubules

this I have followed him), an upper S-curve, an S-middle piece and a lower S-curve.
The upper S-curve passes without definite boundary into the S-middle piece. The lower
S-curve grows upward toward the lateral bend of the upper S-curve; in this way,
there is formed between it and the S-middle piece on the one hand and the concavity
of the lower S-curve on the other hand, a curved cleft, into which vascularized
mesenchyme grows and forms the anlage of the glomerulus. From his reconstruc-
tions, he was able to see that the lower S-curve is in reality not a tubule, but is spread

Fic. 7. A portion of a longitudinal section of the kidney of a human
embryo measuring 6.5 cm. in length (No. 11), showing a primary columna
Bertini, which extends from the capsule to near the pelvis of the kidney.
x 80. Pcb, primary columna Bertini; c, capsule; pe, epithelium of pelvis; d,
end branches of collecting tubules.

out, beginning with the place where the S-middle piece passes into it ‘into a thin,
spherically curved and downward bellying structure with a sharp border, having the
form of a triangle with rounded angles; the base of this triangle represents the up-
curved, blind ending of the S-shaped tubule, as seen in sections. Over the middle
of this structure with upward facing concavity is found just a little above it the
junction of the upper S-curve with the S-middle piece.’ The lower curve of the §,
which is not solid but hollow has therefore the shape of an oyster shell (‘‘ Muschel-
schale”’) and, to continue the comparison. we may say with one haying a deeper
concavity with edges rolled in somewhat, the more fully developed the structure is.

G. Carl Huber 39

The account of the further development of the lower S-curve as given by Stoerk I
shall give in his own words, as it is particularly with this portion that I desire to take
exception. He states (page 297): ‘‘ Die Muschel nimmt einerseits im Sinne der Flaiche
an Grosse zu, andererseits gewinnt sie an Wd6lbung, sodass sie zuniichst der Form
einer ein wenig verzerrten halben Kugelschale zustrebt (s. Fig. 3, 4 und 5); diese
umwo6lbt die Uebergangstelle yom Mittelstiick in die obere S-Hiilfte und wird von dieser
Kaniilechenpartie fast bis zum Kontakt erfiillt ; allmiihlich wird dann auch das S-Mittel-
stiick zur Kugelschalenbildung herangezogen und geht in diese auf (Modell B), so
dass nunmehr der obere S-Bogen iibrig bleibt, der, aus der Kugelschale aufsteigende, von
deren Innenseite durch eben jenen schmalen von Zwischengewebe erfiillten Raum
getrennt wird, der sich, wie erwiihnt, im Schnittbilde als abwirts gekriimmter enger
Spalt priisentiert.”

This double-walled structure, developing from the lower S-curve and
the S-middle piece, with the enclosed mesenchyme, develops into Bow-
man’s capsule and glomerulus. While these are developing, and very
soon after the stage is reached to which reference is made in the quota-
tion, the upper S-curve elongates and becomes flexed and for a time
again has the shape of an 8. This Stoerk designates “secondary 8,”
in contradistinction to the primary S, of which, after Bowman’s capsule
is formed, there remains only the upper S-curve. In reality, as he states,
there is no primary §, this being seen only in sections of this structure
at this stage. He further states that the want of recognition of these
two S-stages and their time relation, on the part of former observers,
accounts for the contradiction which is found between his view and the
views expressed by other observers as to the genesis of the different parts
of the uriniferous tubules, with reference to the different parts of the
S-figure. In discussing these observations of Stoerk, I desire in the
first place to call attention to the fact that, while the various anlagen of
uriniferous tubules from a time when they may be differentiated from
the renal vesicles to a time when they present a distinct S-shape show
at different stages of their development certain typic forms which are
again and again met with, they after all vary in certain details both as
to size and form and also as to the relative time at which certain changes
in form occur. This is true not only when comparing tubular anlagen
obtained from different species of animals, but also when comparing such
obtained from embryos of the same species. With some reservation, the
statement may be made that tubular anlagen which develop from the
first few generations of renal vesicles are larger and present a slightly
different form from those which develop from later generations. In
the earlier stages of the development of the kidney, the branched collect-
ing tubules are relatively far apart and the renal vesicles and tubular
anlagen which develop in connection with any one system of end-branches
are separated by a relatively large amount of interstitial tissue. This
enables the tubular anlagen which develop at this period, when these are
considered as a whole, to assume a more rounded shape than can the
tubular anlagen which develop in the neogenic and subneogenic zones in
the later periods of kidney development at a time when the tubular struc-

40 Development and Shape of Uriniferous Tubules

tures are more crowded and separated by a relatively small amount of
interstitial tissue; consequently the tubular anlagen which develop at
this period, when they are considered as a whole, do not present a rounded
form, but a form which is elongated from above downward. This may
be seen when comparing C and D of Fig. 4, and B and C of Fig. 6, with
the two anlagen shown to the left of the two prominent branches as
shown in Fig. 9, and no doubt explains, in part at least, the difference in
form of early stages of tubular anlagen as given in Stoerk’s figures and
the majority of those here represented. JI am led to assume, both from
a study of his figures and from the statements which he makes, that it
is advisable to study the anlage and early developmental stages of the
uriniferous tubules in as young embryos as possible, because the ele-
ments of the neogenic zone are here farther apart; that his reconstruc-
tions are made from very young embryos, while my own were made from
relatively older stages—the series shown in Fig. 4 from a human embryo
of the seventh month. Besides such differences in shape as here indi-
cated, there are observed in tubular anlagen which show essentially the
same stage of development, especially when these are seen in sagittal
section, minor differences in shape involving the extent of the curvature
of different parts and the relative size and shape and degree of differ-
entiation of the lower S-curve, the saucer-shaped expansion of the
S-shaped tubular anlage, when this is compared with its other parts.
Of the series of models in my possession showing the S-stage of develop-
ment of the uriniferous tubules, no matter whether we speak of those
made from different species or of those from the same species, no two
are exactly alike or even nearly so. They all show certain character-
istics of form which enable a classification as to stage of development,
but differ when compared in detail. This is more particularly true as
concerns the form and size of the lower S-curve and its relation to the
S-middle piece. The lower S-curve is developed primarily, as will be
remembered, not by an invagination of the outer wall of the renal
vesicle, but by a cleft which develops in a thickening of the outer wall.
The extent of this cleft varies. It may not, as has been stated, extend,
when first recognizable, on to the front and back surfaces of the renal
vesicle and may be quite deep before the lower part of the vesicle, that
part which will develop into the lower S-curve, has by extension of the
cleft to any appreciable extent been separated from the part of the renal
vesicle just above it, the part which will form the S-middle piece. Such
a structure in reconstruction does not present the form of an S, but may
do so in sagittal sections. On the other hand the cleft which appears in
the outer wall of the vesicle may, almost from its beginning, extend for

G. Carl Huber 41

a distance on its front and back surfaces, in which case the lower S-curve
and S-middle piece would be recognized in a reconstruction about as
soon as in sections. In both cases there develops from the renal
vesicle, by a deepening of the cleft and by a growth of the structure, a
part which we have known as the S-middle piece and a portion known
as the lower S-curve. The former develops as a short tubular segment
of irregular cylindrical shape, the latter as a double-walled saucer-
shaped structure, with concavity upwards and presenting, as seen from
above or below, either an oval or more rounded or irregular triangular
shape, with a distinct border at which the two layers of the structure
are continuous. In the former case these parts are differentiated at a
relatively later period than in the latter, when the form of the anlage is
compared with its internal structure, as seen in sections. Before the
S-middle piece is completely differentiated, the border or edge of the
lower saucer or shell-shaped structure extends for a variable distance
on to the S-middle piece and recedes gradually from this as it be-
comes more clearly defined. So far as my observations go, from a
time when the S-middle piece may be considered as fairly well defined,
it can be traced as such into the older developmental stages and is not
taken up into the lower S-curve as this changes from a saucer-shaped
structure to one having a deeper concavity and presenting the shape of
an irregular hemisphere, as is stated by Stoerk. This appears to be
the case in tubular anlagen in which the border of the lower S-curve or
saucer-shaped structure extends for an appreciable distance on to the
S-middle piece, but this I interpret as due to a relatively late differ-
entiation of the S-middle piece and not as showing that it is being taken
up into the lower S-curve as in tubular anlagen, where the S-middle
piece differentiates relatively early, this appears to form a definite
tubular segment, which is not lost in its further development. As a
matter of fact, not all of the lower S-curve, as seen for instance in a
tubular anlage such as is shown in D of Figs. 3 and 4, differentiates
into the double-walled concayvo-convex structure, which forms, in its
further development, a Bowman’s capsule; but this point shall be dis-
cussed more fully in presenting the further stages in the development.
I see, therefore, no reasons for recognizing a primary S and a secondary §,
as is done by Stoerk, my observations, as stated, leading me to think that
the different parts of a tubular anlage which would correspond to his pri-
mary S-stage,so far as differentiated, continue as such into the stage which
he has designated as the secondary S-stage, the various parts—upper
S-curve, S-middle piece, and lower S-curve—being in the latter more
clearly defined than in the former; especially is this true of the S-middle

42 Development and Shape of Uriniferous Tubules

piece, which becomes more clearly defined and does not disappear in the
transition from the primary to the secondary S-stage, and this supposed
disappearance of the S-middle piece, as the one stage passes into the
other, appears to me as the crucial point in Stoerk’s argument for the
recognition of these two S-stages. In making these statements, I am
more especially guided by observations made on reconstructions of
tubular anlagen, obtained from developing kidneys of cat embryos vary-
ing between 20 mm. and 30 mm. in length and from a human embryo
of the seventh month rather than on those made of tubular anlagen de-
veloped from the first few generations of renal vesicles of relatively
young human and rabbit embryos, as in the former, more generally than
in the latter, the S-middle piece differentiates relatively early and con-
sequently such tubular anlagen show at a relatively early stage an S-shape
when seen in reconstruction.

Before leaving the consideration of the S-shaped stage in the development of the
uriniferous tubule, brief mention may yet be made of a term which was introduced
by Colberg and which has now and again been used to designate this stage. Col-
berg, in an article in which he attempts to controvert some statements made by Henle
concerning the shane and structure of uriniferous tubules, which need not be entered
upon here, gives also some observations made on the kidneys of a human embryo of the
seventh or eighth month and in the course of his remarks states that the “ open
uriniferous tubules’’ (straight collecting tubules) divide in the cortex, after having
given off several branches in the medulla, into four to six side branches, which
terminate in the periphery of the kidney either in club-shaped enlargements or with
ends which are coiled up several times; these coiled up ends have about the same
size as a fully developed Bowman’s capsule and glomerulus of a coiled uriniferous
tubule. In the coiled up ends of the straight or open tubule, he was not able at first
to detect capillary loops and therefore spoke of them as pseudoglomeruli. Toldt, in
discussing a stage in the development of the uriniferous tubule which corresponds to
the S-shaped stage, makes use of this term, although in a foot-note, he calls attention
to the inappropriateness of it. Hamburger, after discussing the mode of formation
of the S-shaped stage, speaks of it, as a stage designated by Colberg as a pseudoglo-
merulus, a term which, as Hamburger states, has been adopted by the majority of later
writers and is retained by him. The term is misleading and inappropriate and need
not be retained, as the mesenchyme and capillary loops found in the coneavity of the

lower S-curve and recognized as the anlage of the glomerulus, form at this stage only
a small part of the entire tubular anlage.

THE GENESIS OF THE DIFFERENT PARTS OF A URINIFEROUS TUBULE
FROM THE TUBULAR ANLAGEN OF THE S-STAGE.

The developmental changes undergone by the anlagen of uriniferous
tubules, following the stage at which these present the S-shape, affect
simultaneously the different parts of the S-shaped structure. In their
further growth the tubular portions, embracing the upper S-curve and
S-middle piece, elongate, while the bowl-shaped portion, the lower
S-curve, enlarges and obtains a deeper concavity; this is accompanied
by a proliferation of the vascularized mesenchyme contained within this
concavity. We may consider first the changes affecting the tubular por-
tion. ‘This, as may be seen, is relatively fixed at its two ends, on the one
hand by its attachment to the collecting tubule, on the other by its
attachment to the bowl-shaped structure which contains the vascularized
mesenchyme and which early forms a relatively fixed structure. Thus

} G. Carl Huber ety:

in its elongation it is forced to acquire secondary curvatures. Of these
one is fairly constant both as to the time of its appearance and as to its
location, appearing soon after the S-stage is reached and involving about
the middle of the upper S-curve. The convexity of this curve is often,
though not always, directed inwards toward the collecting tubule. Soon
after the appearance of this curvature or coincident with it, the region of
the junction of the S-middle piece and lower S-curve becomes rounded
out into an arched tubule with convexity turned upwards, the lower
S-curve contributing to its formation. Hence, as previously stated, not
all of the lower S-curve goes to form the bowl-shaped structure. A por-
tion of variable length in the region of its junction with the S-middle
piece assumes a tubular form and with a portion of the S-middle piece
forms the arched tubular segment to which reference is here made. This
tubular segment may as a whole be bent forward or backward or may early
acquire secondary curvatures. The region of the junction of the upper
S-curve with the S-middle piece now forms a distinct loop. This region
in the S-stage lies in the sagittal plane of the tubular anlage, as may be
seen in D and # of Fig. 3. Coincident with the developmental changes
above referred to, this portion of the tubular anlage becomes turned on
its axis to a variable extent, the arm of the loop formed by the lower por-
tion of the upper S-curve being moved backward, so that the plane of
the loop now intersects a plane passing as in sagittal section through the
portion representing the lower S-curve. The angle formed by these two
planes varies in the different anlagen from an acute-angle to nearly a
right angle. These changes in form and relative position of the upper
S-curve and S-middle piece seen in various degrees of development are
shown in the models reproduced in H#, F, and G of Fig. 4, D of Fig. 5,
and D and # of Fig. 6. #H, F, and G of Fig. 3 show the appearance pre-
sented by tubular anlagen in these stages when seen in sagittal sections ;
E shows a tubular anlage when the tubular portion deviates but shghtly
from the sagittal plane, the tubule being cut through its entire length,
its lumen appearing, however, only here and there, while in D the curva-
ture of the tubule has progressed to such an extent that.only a part of it
falls within the plane of a sagittal section passing through the middle of
the anlage; this is more marked in G of this figure. A comparison of
these figures with 7, F’, and G of Fig. 4, reconstructions of the respective
tubules, will assist in their interpretation. While the above mentioned
developmental changes, which affect the tubular portion of the anlage,
are in progress, changes affecting the lower S-curve are also to be ob-
served. These changes consist in an enlargement as a whole of this
portion and in a growing upwards of its border, thus deepening its con-

dt Development and Shape of Uriniferous Tubules

cavity and changing it from a structure having the shape of a segment
of a sphere to a hemisphere, the mesenchyme found in the concavity of
the shallow structure of the earlier stage proliferating as this obtains a
greater concavity, especially noticeable being an increase in the number
of the capillary loops. That the epithelial portion of this structure is
double-walled is evident from what has been said of its anlage and further
development; its border represents the edge along which the outer wall
becomes continuous with the inner wall. At about this stage of its de-
velopment, its inner layer on the side toward the collecting tubule and
in the region where it is continuous with the tubular portion of the
anlage and at about the level of its border, develops a fold which grows
outward, away from the collecting tubule, which fold in its further growth
becomes continuous with the border. In sagittal sections, this fold ap-
pears as a spur and is seen in F’ and G of Fig. 3 and also in certain of
the tubular anlagen as shown in Fig. 11. It assists in narrowing the
wide opening by means of which the space (occupied by mesenchyme
and capillaries) enclosed within the double-walled structure communi-
cates with the outside. This fold further assists in differentiating more
clearly this structure of hemispherical shape from the tubular portion
of the anlage, in that the former now obtains a continuous border along
the entire length of which the inner wall is reflected into the outer wall.
This brings the attachment of the tubular portion in connection with its
outer wall. The attachment of the tubular portion is always on its inner
side, on the side turned toward the collecting tubule.

In a tubular anlage developed to the extent here described, there may
be recognized the anlagen of the different parts of a uriniferous tubule,
as these are known after its development. Briefly stated, the genesis
of the different parts of a uriniferous tubule is as follows: The double-
walled hemispherical structure, which develops from the greater part of
the lower S-curve, forms the inner and the outer layer of Bowman’s
capsule (more correctly stated, the outer layer forms the epithelial lining
of Bowman’s capsule, while the inner layer forms the glomerular epithe-
lium) ; the space between the two layers, continuous with the lumen of
the tubules, forms Bowman’s space; the vascular mesenchyme enclosed
within this double-walled structure forms the anlage for the glomerulus,
the two structures taken together forming a stage in the development of
a Malpighian body or corpuscle. The region of the junction of the
lower S-curve and the S-middle piece, which, as stated, differentiates
into an arched tubular segment, represents the anlage of the proximal
convoluted tubule. Between this arched tubular segment and Bow-
man’s capsule there is described a short segment, which is known as

G.CarlEtuber AS

the neck of the uriniferous tubule. At this stage of development this
portion of the tube is not clearly defined and is not fully differentiated
until the epithelium of the portion which will form the proximal con-
voluted tubule shows structural differentiation. The region of the
junction of the S-middle piece with the lower part of the upper S-curve
forms the anlage for the loop of Henle, that portion of the S-middle
piece which does not participate in the formation of the proximal con-
voluted tubule forming the proximal arm of the loop, the lower part of
the upper S-curve to the region where appears the secondary curvature,
about its middle, as above stated, forming the distal arm of the loop,
the junction of the two parts, the loop itself. The secondary curvature
of the upper S-curve first to appear forms the anlage of the distal con-
voluted portion. The upper portion of the upper S-curve from its
attachment to the collecting tubule to the region where the distal con-
voluted portion has its anlage may be recognized as the beginning of
the junctional tubule. The epithelium of the entire tubular portion
of the uriniferous tubule presents at this stage of development essen-
tially the same structural appearances. The cells throughout are of
columnar shape, with a relatively small amount of protoplasm, which,
unless the bleaching has been very thorough, retains to a certain extent
the nuclear stain, for instance staining lightly in hemotoxylin if this
stain has been used. The nuclei are relatively large, of round or oval
shape, and stain relatively deeply. The lumen throughout is of about
the same size and is narrow. The statements here made as to genesis
of the different parts of a uriniferous tubule are based on observations
made on reconstructions of developing uriniferous tubules, of stages
beginning with the S-stage to a time when the metamorphosis has pro-
gressed until the different parts of a uriniferous tubule may be recog-
nized. Attention should, however, be again called to the fact that,
while models which show these stages of development present a certain
similarity of form and arrangement of parts, they differ when studied
in detail, as was stated in discussing earlier stages of development.

In the further elongation of the tubular portions of the uriniferous
tubules, as these proceed in their development, beginning with a stage
as above discussed, the portion which was above mentioned as forming
the anlage of the loop of Henle needs first consideration. This portion
of the tubule is found in the S-stage and some little time after, imme-
diately above the concavity of the lower S-curve, the anlage of Bow-
man’s capsule, as may be seen in reproductions of reconstructions show-
ing this stage. As this loop elongates, it either grows a little to one
side (forward) and then elongates downwards toward the pelvis of the

46 Development and Shape of Uriniferous Tubules

developing kidney, as shown in F’ of Fig. 6, or grows downward from
the beginning and in doing so appears to push aside the anlage of Bow-
man’s capsule, which in being pushed aside is turned on its axis so that
its posterior border attains a higher level than its anterior border. This
is very well shown in @ of Fig. 6, also, though not quite so clearly, in #
of Fig. 5. While the loop of Henle is thus developing, the portion which
is to form its ascending or distal limb remains in close proximity and
on its inner side to the developing Bowman’s capsule, after crossing the
place where this is joined to the tubular portion which forms the anlage
for the proximal convoluted tubule, the descending or proximal limb
of the loop lying to the inner side or slightly in front of the distal limb.
The two arms of the loop are generally, from the time when they may
be clearly recognized, parallel or nearly parallel; now and then the de-
scending limb may be twisted slightly over the ascending limb, as in
F of Fig. 5. In the models here reproduced, which show the early
stages in the development of the loop of Henle, it will be observed that
the loop of Henle grows down in front of the anlage of the Malpighian
corpuscle or in front of the tubular segment attached to it. This may
be accepted as a general law and holds good also for the tubules develop-
ing to the left of the collecting tubules, the ones shown in the figure all
developing to the right of the collecting tubules. This statement I
shall desire to amplify somewhat in discussing more fully the relations
of the developing uriniferous tubule to the collecting tubules when I
shall also consider Stoerk’s observations on this point. Coincident with
the formation of the definite anlage of the loop of Henle, as here given,
the portions of the tubule which I have designated as destined to form
the proximal and distal convoluted portions also increase in length,
especially the former, which grows upward and acquires two or three
secondary curves. It begins at the developing Malpighian corpuscle and
extends upward behind the anlage of the loop of Henle, then arches for-
ward between the collecting tubule and the other parts of the respective
uriniferous tubule, then passes downward to become continuous with the
descending limb of Henle’s loop. Just before the loop is reached, it
often shows a distinct flexure, which is fairly constant and character-
istic, appearing also in older stages. The curvature which forms the
anlage of the distal convoluted portion does not grow in length as rap-
idly as that part which forms the proximal convoluted portion. The
curvature becomes more pronounced as development proceeds and may
obtain one or more secondary bends, or may show in much older stages
than here discussed a single bend. It may here be added that there
is observed a considerable variation as to the relative length of this

G. Carl Huber - 47

portion of a developing uriniferous tubule. Of two uriniferous tubules
of apparently the same stage of development, the distal convoluted
portion of one may be as long again as that of the other. This is to
a certain extent true of uriniferous tubules reconstructed from embryos
of the same species, more clearly seen when those reconstructed from dif-
ferent species are compared. As arule in reconstructions of uriniferous
tubules showing early stages of development obtained from human
embryos, the distal convoluted portion is relatively longer than in such
obtained from cat and rabbit embryos. Compare for example J of Fig.
4 with H of Fig. 6, two tubules showing about the same stage of
development. The distal convoluted portion of a uriniferous tubule is
generally found just above the Malpighian corpuscle of the respective
tubule, a relation which is fairly constant even for later stages of
development, and will receive further attention in considering these.
During the time in which tubular portions of uriniferous tubules de-
velop as above described, the anlagen of the Malpighian corpuscles of
hemispherical shape change to corpuscles of spherical shape. This
is accomplished by a growing upwards and a turning in of the border
of the double-walled epithelial structure representing the anlage of
Bowman’s capsule and by a growing outwards of the fold which arises
from its inner wall in the region of its attachment to the tubular
portion, this fold being continuous with its border. In this way the
opening into the double-walled structure is gradually narrowed until
only a relatively small opening remains, at which, as in earlier stages,
the inner wall becomes reflected into the outer wall. This opening into
the double-walled spherical structure as thus developed and which may
now be known as a Bowman’s capsule, is situated in its outer and upper
portion. During this time, the mesenchyme and _ vessels recognized
as the anlage of the glomerulus differentiate into a definite glomerulus,
the small opening leading into the cavity of Bowman’s capsule serving
for entrance and exit of the afferent and efferent vessels of the glomerulus,
these being accompanied by a small amount of connective tissue. A
Malpighian corpuscle of this stage of development presents essentially
the same shape as a fully developed corpuscle and differs from those
found in post-foetal life, aside from certain details in cell differentiation,
only in being smaller. The relative degree of development of a Mal-
pighian corpuscle and the tubular portion of a uriniferous tubule, when
the extent of the development of the former is compared with the extent
of the development of the latter, varies somewhat for different tubules
reconstructed from embryos from the same species of animals; more so
when developing uriniferous tubules obtained from embryos of different

48 Development and Shape of Uriniferous Tubules

species of animals are compared. In human and pig embryos the Mal-
pighian corpuscles develop relatively early, when considering the extent
of the development of a given uriniferous tubule and may assume a
nearly spherical form before the portion of the tubule which forms the
anlage of the loop of Henle is clearly differentiated. This, H and I of
Fig. 4 may serve to illustrate. On the other hand in cat and rabbit,
developing uriniferous tubules are often met with in which the anlage of
the loop of Henle is clearly made out, before the Malpighian corpuscles
have progressed in development beyond the stage in which they are of
hemispherical shape, with the epithelial portion in the form of a double-
walled cup, as for instance shown in # of Fig. 5, and F and G of Fig. 6.

Attention may yet be called to the fact that Malpighian corpuscles
develop in about the location in which they have their anlage. Through-
out the period in which new tubules are formed, new generations of
tubules and Malpighian corpuscles develop outside (toward the periphery
of the kidney) of those previously formed, which makes it appear as
though the Malpighian corpuscles of the older generations of uriniferous
tubules sank down into the deeper parts of the parenchyma of the kidney
as this developed. Herring, who has called especial attention to this
fact, expresses himself as follows: “The Malpighian corpuscles seem to
be fixed structures at an early period and do not move their position
as do the tubules during their further growth. The early fixation seems
to be due to the density of the connective tissue around the Malpighian
body and it is likely that a certain part always remains constituting the
framework.” The size of these structures when compared with the other
portions of the tubular anlagen, the compactness of the glomerular anlage,
when compared with the surrounding mesenchyme, but especially the
early development of a definite vascular supply to the glomerulus, all
appear to me to form factors which assist in the fixation of the developing
Malpighian corpuscle.

At about the time when the developing Malpighian corpuscles have
reached a stage of development in which they resemble in shape and to
a certain extent in structure, fully developed Malpighian corpuscles,
there is observed a cellular differentiation in that portion of the tubules
which we have designated as the anlage of the proximal convoluted
portion. Beginning with the region of attachment of this portion of the
tubule to the Malpighian corpuscle and proceeding for the remainder
of its extent it may be observed that the protoplasm of the epithelial cells
increases in quantity and becomes clearer and now takes more readily a
protoplasmic stain, so that in sections stained with hematoxylin and
eosin or erythrosin, the protoplasm of the cells appears tinged with red.

G. Carl Huber 49

At the same time, the nuclei take a position in the basal portion of the
cells, so that the lumen becomes surrounded by a distinct margin of
clear protoplasm. The nuclei also stain less deeply than in earlier stages
of development and present a vesicular appearance. The lumen of this
portion of the tubule, which, before the cell differentiation is observed,
is relatively narrow, becomes prominent. ‘This cell differentiation extends
the entire length of the proximal portion and for a variable distance along
the descending arm of Henle’s loop, as developed at this stage. This cell
differentiation is shown in J and J of Fig. 3. In the former is shown
about the distal one-half of the proximal convoluted portion of the tubule
represented in reconstruction in J of Fig. 4, and its junction with the re-
mainder of the tubule, the epithelium of which shows as yet no differentia-
tion. In the latter is shown a loop of Henle cut through its entire length
about the upper half of the descending limb of which shows an epithelium
with clear protoplasm and nuclei with basal position. The proximal
convoluted portion of the tubule is further seen in cross section just to the
right of the upper end of the descending limb of Henle’s loop as seen in
the figure; this also shows the characteristic cell differentiation. In the
immediate vicinity of the Malpighian corpuscle, where the tubule becomes
continuous with the outer layer of Bowman’s capsule, this cell differentia-
tion is not so pronounced as in other parts of the proximal convoluted
portion, nor do the cells of this region attain the same size as in other
parts of the tubular segment; the lumen likewise is not so well developed.
Consequently the tubule presents here for a short though a variable
distance a smaller diameter than in other parts of the proximal convoluted
portion and may now be recognized both by reason of its smaller size and
by its structure as what is known as the neck of the uriniferous tubule.
At this stage of development, the remaining portions of the uriniferous
tubule, the greater part of the loop of Henle, the distal convoluted portion,
and the junctional tubules present essentially the same structural appear-
ance seen in the earlier stages of development.

The so-called junctional tubules of the uriniferous tubules elongate
as these develop with the growth of the kidney. As has just been stated,
a Malpighian corpuscle develops in about the location of its anlage; the
proximal and distal convoluted portions of a tubule of which such a
corpuscle forms a part, differentiate in its immediate vicinity. As the
developing kidney grows in size, the collecting tubules elongate, their
ends being always found just under the capsule, and as the junctional
tubules are inserted at the peripheral ends of the collecting tubules
(the exceptions I shall note) the collecting tubules, as they elongate, carry
with them the ends of the junctional tubules; those for the uriniferous

50 Development and Shape of Uriniferous Tubules

tubules first developed must then traverse the greater part of the cortex,
so far as developed, in order to reach the peripheral ends of the collecting
tubules. (See Stoerk, page 304.)

A detailed description of the several models, on which is based the account
here given of the early developmental stages of uriniferous tubules, is

4

Fie. 8. Semidiagrammatic figures of the anlage and differentiation of renal
vesicles and early developmental stages of uriniferous tubules of mammals.
1 and 2, anlage and successive stages in the differentiation of renal vesicles,
as seen in sagittal sections; 3, section and outer form of tubular anlage before
union with collecting tubule at the beginning of S-shaped stage; 4 and 5,
successive stages in the development of the tubules, Bowman’s capsule and
glomerulus beginning with a tubular anlage showing a well developed S-shape.

obviated, it seems to me, by the number of illustrations of these stages
here presented. My own observations and the conclusions reached con-
cerning the anlage and early developmental stages of the uriniferous
tubules and the genesis of the different parts of these tubules I have
summarized by way of a series of diagrammatic figures grouped under
Fig. 8.

G. Carl Huber ~.- 51

Brief mention may yet be made of the conclusions reached by former observers who
have considered the genesis of the different parts of the uriniferous tubules. Toldt,
who was the first to recognize clearly the S-shaped stage in the development of the
uriniferous tubule, a stage in which the anlage is spoken of as a pseudo-glomerulus,
does not give a clear account of the anlage of the tubular portion. I find in his
account the following statement, which I shall give in his own words: ‘ Bemerkens-
werth ist ferner, dass die Windungen des Caniilchens constant den der Nierenoberfliiche
zugewendeten, iiusseren Theil des Pseudoglomerulus einnehmen, wiihrend das _ schalen-
formige Ende des Caniilchens stets in dem Abschnitte desselben gelegen ist, welcher
nach dem Inneren der Niere zu sieht.” (Page 134.) ‘This statement taken in con-
nection with Toldt’s account of the development of the Malpighian corpuscle, I in-
terpret as meaning that he regards Bowman’s capsule (and glomerulus) as developed
from the lower curve of the S-shaped structure while the remaining parts of the
uriniferous tubules are developed from the tubular portion of the ‘ pseudoglomerulus ”’
and not, as is now and then stated, that Toldt regards the S-shaped tubules (pseudo-
glomeruli) as merely the anlagen of the Malpighian corpuscles. MHaycraft, in his
account of the anlage of the different parts of the uriniferous tubule, makes reference
to a figure (Fig. 8 of his article), which he describes as *“‘a high power view of the
first formation of a urinary tubule from a primary renal vesicle,’ which figure I have
interpreted as showing only a portion of a tubular anlage. He states that ‘“ each
little sprout from a renal vesicle will, in other series of older embryos, be seen to
elongate, the bulging portion marked H will grow down toward the primary pelvis to
form the convoluted tubules and the loop of Henle, the part at m marks the formation
of the future Malpighian body.’ The bulging portion marked H as shown in his
figure refers to the region of the junction of the S-shaped tubular anlage with the
ampulla of the collecting tubule, m marks the end of the tubule as seen in the figure.
Golgi’s account of the development of the uriniferous tubules, elucidated by his well
known semidiagrammatic figures of developing xenal tubules of mammals is here given as
presented by Minot (Page 510), leaving out the reference letters. ‘The different
parts of the S-shaped tubule have each their fixed destiny. The end of the S (in the
diagrams the lower part) receives the vascular loop, which gives rise to the blood-
vessels of the future glomerulus; the lower limb of the S elongates enormously and
forms the first division of the convoluted tubule including the loop of Henle; the upper
limb of the S also elongates very much—though less than the lower limb—and is the
anlage of the second division of the convoluted tubule; where the two join, the tubule
passes close to the Malpighian corpuscle and seems to be intimately attached to it.’
Except perhaps for the fact that I have differentiated more clearly the different parts
of a tubular anlage of an S-shape and am thus able to give more explicitly the genesis
of the different parts of a uriniferous tubule, my own account may be considered
as confirming Golgi’s account as here given. Hamburger states that the first portion
of the coiled uriniferous tubule to differentiate (leaving out of consideration Bowman’s
capsule) is the loop of Henle, which has its anlage in that portion of the S-shaped
tubule which is taken up by the bowl-shaped anlage of Bowman’s capsule (‘‘ welche
eigentlich durch die von der Schale aufgenommene Biegung des S-férmigen Caniilchens
vorgebildet ist’’). The portion which extends into the concavity of the lower S-curve
in the region of the junction of the upper S-curve and S-middle piece, which I
have also regarded as the anlage of the loop of Henle (I am not certain that I have
interpreted the above statement of Hamburger correctly, as the account is not quite
clear to me). He also states that the distal convoluted portion (Schaltstiick) is
differentiated early, before the proximal convoluted portion (Tubulus contortus)
has acquired any coils. The latter portion, he states, is developed from the third,
the most distal limb of the S-shaped tubule. It should be recalled that Hamburger
recognizes in the S-shaped stage (pseudoglomerulus) a bowl-shaped structure and an
S-shaped tubular portion (see page 36 of this article) much as described by Stoerk
for his secondary S-stage. Hamburger, in speaking therefore of the third or distal
limb of the S-shaped tubule refers to the curved tubular segment which is in continuity
with the anlage of Bowman’s capsule; this would represent the tubular segment which
develops from the region of the junction of the lower S-curve and S-middle piece,
which I have also regarded as the anlage of the proximal convoluted portion. Schreiner
recognizes in the S-shaped tubular anlage essentialivy the same regional differentiation
leading to the formation of the different parts of the uriniferous tubule as given by
me, as may be seen by a study of his Fig. 114, to which especial reference is made in
his account. Stoerk’s observations on the mode of formation of the different parts of
the uriniferous tubules need to be considered somewhat more fully, as he is the only
one of previous workers who has made extensive use of reconstruction methods in
the study of the subject. It will be remembered that he recognizes a primary and
a secondary S-stage in the development of the uriniferous tubules. As the Malpighian
corpuscle completes its development, the S-shaped tubule of the secondary S stage
elongates and acquires from four to five ‘“‘short, plump windings.” The epithelium
of the tubular portion. which up to this stage. has shown the same structure throughout,
now shows a differentiation, beginning with the attachment of the tubule to the outer
layer of Bowman’s capsule and extending for one to two windings of the tubule.
The cells here obtain more protoplasm, which becomes clearer and the nuclei assume
a basal position. This cell-differentiation, as described by Stoerk, both as to the time
of its appearance and as to the region of the tybule affected is about as given by me.
About this time, there develops from the coiled tubular portion a loop which grows
downward over the Malpighian corpuscles to form the loop of Henle. Concerning the
location of this loop, Stoerk has this to say; I use his own words: ‘Es lisst sich
von vornherein durchaus nicht sagen, welche von den urspriinglich untereinander
ganz gleich aussehenden Windungen zu diesem Anwachsen zur Schlinge bestimmt ist,
solange in allen das Epithel ein dunkles ist; erfahrungsgemiiss ist es meist die zweite
nach dem Abgang vom Malpighischen Kérperchen.’’ That the portion of the tubule

a2 Development and Shape of Uriniferous Tubules

which is to form the anlage of the loop of Henle can be more definitely located
than this statement would lead one to think I believe my models will show, especially
those obtained from the cat and rabbit. In models of developing uriniferous tubules
obtained from human embryos, in which the Malpighian corpuscles develop relatively
early and the loop of Henle at a correspondingly later period, the portion of the
tubule which is to form the loop of Henle is often not easily located; yet even from a
study of these models, I have gained the conviction that the loop has its anlage in
what has been termed the region of junction of the upper S-curve and S-middle
piece. It is of interest to note that Stoerk states that experience shows that it is
generally the second loop after the tubule leaves the Malpighian corpuscles, which forms
the anlage of the loop of Henle; this corresponds, in_the majority of the models,
with the region of the junction of the upper S-curve and S-middle piece and confirms,
therefore, the view I have expressed.

The majority of the observers who have considered the development of the Malpighian
corpuscle have in whole or in part followed Toldt, who considered Bowman’s capsule
formed by an invagination, from one side, of the blindly ending spherical dilation of the
developing uriniferous tubule, mesenchyme and capillaries participating in the invagina-
tion and forming the anlage for the glomerulus. 'Toldt compares this process of in-
vagination to the pressing in of one side of a hollow rubber ball until the side pressed
in comes in contact with the other side. The glomerulus develops in the concavity
thus formed. Kd6lliker, Pye, Janosik, Nagel, Minot, Schultze, Gerhardt, and Strahl
may be mentioned as accepting this view, although it is not always clear from
their accounts to what extent they consider the formation of the double walled cup,
the anlage of Bowman’s capsule, due to an actual invagination rather than to a
process of growth on the part of the anlage of Bowman’s capsule by means of which
the anlage of the glomerulus becomes surrounded. Gerhardt, for instance, states that
it is difficult to say whether the blind end of the uriniferous tubule plays an active
or a passive role in the formation of Bowman’s capsule, most likely both, in that, while
the capillary loops grow inward, the tubule takes an active part in the formation
of Bowman’s capsule. A more correct interpretation of the process of development
of Bowman’s capsule and the glomerulus is given by Herring, although he missed
the anlage of the part which is destined to form Bowman's capsule, as has been pre-
viously stated. He states that ‘‘the first appearance of the cavity of Bowman's
capsule is a narrow slit and never a vesicle in the human kidney. Toldt’s description
of the driving in of one wall by capillaries is not an accurate one. At this stage
there are no capillaries in the glomerulus.” ‘‘The further growth in size of the
glomerulus may be described as an invagination, but it is not an invagination of the
kind usually supposed. None of it takes place at the expense of the other side as
oecurs in Toldt’s illustration of the rubber -ball. It is rather an increase in size by
proliferation of its own constituents; the base remains in the same position and is wide
at first but gradually narrows, the narrowing being brought about by the formative
action of large cells covering it.’ 'The formation of the lower S-curve, the anlage of
Bowman’s capsule, has been correctly given by Schreiner, as has been stated, and in
this my own observations confirm him. By way of summary it may be stated that the
two layers of the lip-shaped fold, which is separated from the outer, lower part of the
renal vesicle, and which forms the anlage of the lower S-curve, are from the beginning
nearly in contact, the fold presenting a_ slight coneavity above and a corresponding
convexity below. As the anlage of the lower S-curve proceeds in its development, it
increases in size and assumes the shape of the bowl of a spoon, with a distinct
border except where attached to the tubular portion of the anlage; the handle of the
spoon, which must be thought of as bent so as to form the greater part of an S. At
sbout this time, a small amount of mesenchyme may be found in the concavity of the
lower S-curve, at first without capillaries, these growing in from without. In its
further development, this double walled structure, which represents the segment of a
sphere, changes to a hemisphere by a growing upwards of the border, the mesenchyme
proliferating and the capillaries increasing in number so as to fill the concavity as
this develops. By a further growing upwards and a turning in of the border and by
the formation of a fold on the inner layer, the hemispherical structure changes to a
spherical structure, Bowman’s capsule and the glomerulus being thus completed. That
Bowman’s capsule with its outer and inner layer is developed by growing upwards
and ultimately by a turning in of the border of the structure with shallow concayity
as seen in the Iower S-curve when this is first recognized, both sections and recon-
structions seem to me to show. In sections of a developing Bowman’s capsule, both
the inner and outer layer for a short distance from the border and along its entire
length show undifferentiated embryonic cells indicating the growing zone. This may
be seen in the sketches shown in Fig. 3. This view of the development of the Mal-
pighian corpuscle is the one taken by Stoerk, who has given correctly the main fea-
tnres of its development, leaving out of consideration the anlage of the lower S-curve,
the formation of which he hardly considers, and the formation of the fold which
assists in transforming the hemispherical into a spherical structure; this fold he
recognized but believes it to be developed from the middle S-piece of the primary §,
which, as has been stated, he regards as being taken up into the anlage of Bowman’s
capsule, while the primary S-stage is changing into the secondary S; in my own account,
I have endeavored to show that it develops as a fold of the inner layer of the anlage
of Bowman’s capsule independent of the S-middle piece.

Throughout the period in which new tubules are formed, new genera-
tions of uriniferous tubules develop outside (toward the periphery of the
kidney) of those previously formed, the latter thus coming to lie relatively

G. Carl Huber -. 53

deeper down in the parenchyma of the kidney, as this increases in size ;
the tubules showing the greatest development are therefore formed nearest
the pelvis of the developing kidney. To show the relative size and the
shapes of uriniferous tubules and tubular anlagen developing from differ-
ent generations of renal vesicles as also the relative size, shape and extent
of branching of collecting tubules and their relation to the former, for a
developing kidney of which the most fully developed tubules present
about the size and shape of those shown in J of Fig. 4, and H of Fig. 6,

Fic. 9. A model of a large or primary collecting tubule, cut just above its
origin from the pelvis of the kidney with the collecting tubules resulting
from four successive dichotomous divisions, with three renal vesicles, five
tubular anlagen in the S-shaped stage and two uriniferous tubules in early
stages of development; from the kidney of a human embryo (No. 7), measur-
ing 5 cm. in length. 133. rv, renal vesicles.

these representing the oldest stages thus far discussed, I have inserted
Figs. 9 and 10. In each of these figures is reproduced a model of a
primary collecting duct, beginning with a region a little above its place
of origin from the pelvis of the kidney and showing the successive divisions
of the same to the ampullar enlargements of the end branches, together
with the renal vesicles, tubular anlagen and uriniferous tubules which had
developed therewith. Fig. 9 represents these structures as observed in a
human embryo 3 cm. in length (embryo No. 7). In this figure, for the
sake of clearness, only a part of the -model is reproduced. The cut

54 Development and Shape of Uriniferous Tubules

surfaces recognized by their shading represent the origin of two branches,
similar to those reproduced. The collecting tubules are at this stage
of development relatively large with wide lumen. The one here shown

Fic. 10. A model of a primary collecting tubule, cut just above its origin
from the pelvis of the kidney, showing six successive dichotomous divisions.
Only one of the four main branches of the second generation of branches is
here shown in full, together with renal vesicles, tubular anlagen and urini-
ferous tubules developing in connection with this branch. From the kidney
of a human embryo (No. 11). measuring 6.5 cm. X 133.

presents four successive dichotomous divisions, the end-branches showing
ampullar enlargements. The pelvis of this kidney and the collecting
tubules here shown are lined by a single layer of columnar cells. The
figure shows further three renal vesicles, rv, five tubular anlagen in the
S-stage, seen from different aspects and showing different degrees of

G. Carl Huber . 55

development and two uriniferous tubules of a stage of development in
which the different parts of a uriniferous tubule may be readily made

Fig. 11. A portion of a longitudinal section of a kidney of a human
embryo (No. 11) measuring 6.5 em. (This section is the most typical of the
series of sections used in making the preliminary drawings from which was
made the model shown in Fig. 10.) 133. ©, capsule; d, collecting tubule;
pl, proximal or descending limb of Henle’s loop; m, Malpighian corpuscle with
neck and a portion of proximal convoluted tubule; bc, anlage of Bowman’s
capsule and a glomerulus showing fold of inner wall of capsule; dc, distal
convoluted tubule; a, end branch of collecting tubule and ampulla; iz, inner
zone of metanephrogenic tissue; pc, proximal convoluted tubule.

56 Development and Shape of Uriniferous Tubules

out and the respective Malpighian corpuscles are nearly developed; one of
these is on the rear side of the model as sketched and shows only in part,
between the two prominent diverging branches. The relation of these
structures to each other and to the collecting tubule is shown with sufficient
clearness in the figure to obviate fuller discussion. In Fig. 10, is shown
a similar reconstruction for a human embryo of 6.5 cm. length. The
long diameter of the kidney from which this reconstruction was made
measures 5mm. after fixation. The collecting tube reconstructed presents
five and in some branches six successive dichotomous divisions. Of the
four branches resulting from the second division, only one was recon-
structed in full, the others are represented as cut soon after they are
separated. That the collecting tubules grow in length along their entire
extent as the kidney develops is shown by the fact that a greater distance
intervenes between the successive branches than is shown in the former
stage; the angle at which the branches meet is for the older stage more
acute than for the earlier stage. During this growth in length of the
collecting tubules, they obtain a smaller diameter and consequently a
smaller lumen; especially is this true of the later generations of branches.
In connection with the one large branch here fully reconstructed, there are
found in the model five renal vesicles and 14 tubular anlagen and urinifer-
ous tubules in various stages of development, only a portion of which
could be represented in a sketch presenting one view of the model. The
complexity of this model is such that a reproduction of it can give only
in a general way the shape, size, and relations of the various structures
modeled and this can be shown in one figure quite as clearly as in several
figures drawn from different aspects of the model. In Fig. 11 is repro-
duced one of the most typical sections of the series from which the model
shown in the former figure was made; it falls in a plane which passes
through about the middle of the model. It will serve to show the struc-
ture and cellular differentiation of certain representative tubules and
tubular anlagen shown in the reconstruction; a comparison of the
two figures will enable the identification of the respective parts, as
represented in each figure, as they are drawn to the same scale. In the
section reproduced, a collecting tubule was cut through its entire length
and may be seen to end in characteristic ampullar enlargement. The
epithelium lining the collecting tubules is here shown as a single layer of
columnar cells which in the primary collecting duct presents the appear-
ance of a pseudostratified epithelium. In each of the two models (Fig.
9 and 10) all of the renal vesicles, tubular anlagen and uriniferous
tubules, with the exception of one uriniferous tubule, are associated with
or attached to the ampullar enlargement of the end branches of the

Or
=p

G. Carl Huber |

collecting tubules. The one tubule of each model having a different
attachment is attached to the collecting tubule; in the model showing the
younger stage of development, in the region of the second division, in the
other model, in the region of the third division. I have observed this
condition only in the human embryos. In the kidney of the cat and
rabbit of about the same stage of development, as shown in these models,
all the uriniferous tubules and tubular anlagen are attached to the
ampullar enlargements of the end branches of the collecting tubules.
Hamburger, who has made observations on the mode of attachment of
the “coiled uriniferous tubule” to the “straight collecting tubule” on
certain animals, has reached the following conclusions: He speaks of
the branches of the straight collecting tubules which result from the last
division of these (as seen in the fully developed kidney) as the “ terminal
collecting tubules” and finds that in the beef and the mouse and
generally also in the rat these terminal collecting tubules take up the
ends of the coiled uriniferous tubules and only seldom (in the rat) do
these terminate in the straight collecting tubules proximal to the terminal
branches, while in the pig a greater number of the coiled uriniferous
tubules end in the straight collecting tubules and that these further
form arcades, with convexity upwards, which also serve for the insertion
‘of the coiled tubules. The fact that at a relatively early stage in the
development of the human kidney certain uriniferous tubules end in the
collecting tubules proximal to their terminal branches indicates that this
mode of termination obtains in post-uterine life, while the fact that in
eat and rabbit embryos of about 2.5 cm. length, all the tubules end in
the terminal branches would go to show that even in the fully developed
kidneys of these animals, the great majority of the uriniferous tubules
end in this way. Positive statements concerning this point I am at
present unable to make, as I have not reconstructed collecting tubules
throughout their whole extent for stages older than given in Fig. 10.
Figure 9 will also serve to show that the collecting tubules in their
growth and successive divisions resulting in the formation of new
branches, do not in these successive divisions divide in the same plane,
but with some degree of regularity in alternate planes. The configu-
ration resulting from the last two divisions of a collecting tubule during
the time when these form new branches, would resemble two Y’s the stems
of which are joined to form a single stem, the four arms of which project
into four quadrants of a circle. The diagram given in Fig. 12 may
serve to make this clear. Uriniferous tubules develop in connection
with each of the four end branches thus formed. In each of the two
tubules developing in connection with the two anterior end branches, as

58 Development and Shape of Uriniferous Tubules

given in the figure, the loop of Henle grows down in front of the
Malpighian corpuscle and the first portion of the proximal convoluted
tubule belonging to each tubule. Thus these conform to the general
law as previously stated. In each of the two tubules which develop in
connection with the posteriorly placed end branches, while these hold
the position given in the figure, the loop of Henle grows down behind the
Malpighian corpuscle and the first portion of the proximal convoluted
tubule. If now the whole configuration as shown in the figure be turned
through an are of 180°, thus bringing the two posteriorly placed tubules
in front, these will show the same arrangement and relations of parts

Fic. 12. Semidiagrammatic figure given to show the relations the urini-
ferous tubules bear to the collecting tubules and end branches as observed a
short time before the middle of embryonic life. Not all the uriniferous tubules
would show in such a configuration about the same degree of development as
here represented. c, collecting tubule; e, end branch of collecting tubule; a,
ampulla; m, Malpighian corpuscle; n, neck; pc, proximal convoluted tubule;
pl, proximal or descending limb of Henle’s loop; al, ascending or distal limb
of Henle’s loop;dc, distal convoluted tubule; 7, junctional tubule.

as was shown by the two tubules which previously occupied the anterior
position; that is to say, the loop of Henle now passes down in front of
the Malpighian corpuscle, while in the other two tubules, those which
previously held the anterior position and now occupy a posterior position,
the loop of Henle now passes down behind the Malpighian corpuscle.
In stating, therefore, the general law that the loop of Henle in its
development and elongation passes down in front of the Malpighian cor-
puscle and the first part of the proximal convoluted tubule, cognizance
must be taken of the relationship of a given uriniferous tubule to the
collecting tubule in which it ends. The first part of a proximal con-
voluted tubule, beginning with its attachment to the Malpighian corpus-

G. Carl Huber.” 59

cle and embracing a tubular segment of variable length, passes between
the loop of Henle-and the collecting tubule to which said tubule is
attached. If one, therefore allows the collecting tubule to which a
uriniferous tubule is attached, irrespective of its other relations, to mark
the posterior aspect of said tubular configuration, it may be said that the
loop of Henle passes down in front of the first part of the proximal convo-
luted portion of the tubule, near its origin from the Malpighian corpuscle
and this would also assume a position in front of the corpuscle. I can not,
therefore, agree with Stoerk when he states, in connection with the state-
ment which I have previously quoted (see page 51) in which he refers
to the fact that it was not possible to state which of the four or five
loops of the secondary S formed the anlage for the loop of Henle, that—
and here I use his own words—“ Der Mangel an Gesetzmiissigkeit ihrer
Bildung aiissert sich auch darin dass sie an zwei zum gleichen Sammelrohr
gehorigen Bildungen an der einen vor, an der anderen hinter dem
Malpighischen Korperchen herabfallen kann (Model L und F auf der
Tafelabbildung).” Reference is here made to the downward growth
in its development of the loop of Henle. In model Z of this: plate,
the loop of Henle is well developed and conforms in its relations with
those give in my own descriptions of this portion of the uriniferous
tubule. The tubule shown in model F of this plate is so little developed
that an interpretation of it is not justifiable without seeing the model
itself; I will, therefore, not attempt a discussion of it. In all the urinifer-
ous tubules reconstructed by me, the loop of Henle is seen to pass down
in front of the Malpighian corpuscle or the proximal convoluted tubule
in the vicinity of its attachment to the corpuscle—and this applies to the
older as well as to the younger developmental stages modeled—if by in
front is meant that side of the tubular complex turned away from the
collecting tubule to which a uriniferous tubule is attached as explained
above and as may be seen from the diagrammatic figures given.

DIFFERENTIATION OF THE EPITHELIUM OF THE Loop oF HENLE.

We may now return for brief consideration to the stage of development
represented by the tubule shown in J of Figs. 3 and 4, representing the
oldest stage thus far discussed. This tubule represents one in which the
Malpighian corpuscle shows a spherical form and may be regarded as
fully developed; the loop of Henle is clearly recognized and presents the
relations described for this portion of the uriniferous tubule, the proximal
convoluted portion and about one-half of the proximal or descending
limb of Henle’s loop show an epithelium with clear protoplasm and
nuclei with basal position ; the cells lining the remaining portions of the

60 Development and Shape of Uriniferous Tubules

tubule show as yet no specific differentiation, but present the appearances
shown by the epithelium lining the entire tubule at an earlier stage of
development. In their further development, such tubules increase
greatly in length. This growth in length affects_all parts of the tubule,
though not to the same extent, the proximal convoluted portion elongat-
ing more than the distal convoluted portion, the loops of the former
becoming: more pronounced and new ones forming. ‘The chief increase in
length is, however, observed in the two arms of Henle’s loop; this, in
elongating, grows toward the pelvis of the kidney. While thus elongat-
ing, a characteristic cellular differentiation is observed in the proximal
or descending limb of Henle’s loop, beginning with that region of this
portion of the tubule which is marked by the termination of the cells with
clear protoplasm and extending to near the loop itself—the region where
the descending limb becomes continuous with the ascending limb—though
never involving the loop itself. The epithelium lining this portion of the
descending limb of the loop, differentiates into a flattened epithelium with
flattened elliptical nuclei. The protoplasm of these cells shows very little
granulation. The tubule obtains in this portion a smaller diameter
than in other parts; this is more clearly seen at a later stage of develop-
ment, some little time after the epithelial differentiation here mentioned
may be clearly recognized. The size of the lumen of this portion of the
tubule remains about the same as that of other parts. Coincident with
this change of cell structure, as observed in the descending limb of
Henle’s loop, the cells lining the loop itself, the ascending limb, the distal
convoluted portion and for a variable distance in the junctional tubule,
undergo a change in structure. The protoplasm of the cells lining these
portions of the tubule increases in quantity and acquires a granular,
often faintly striated appearance and shows an affinity for protoplasmic
stains (eosin and erythrosin), these portions then staining more deeply
than the proximal convoluted tubule and the proximal arm of Henle’s
loop. ‘The cells may be described as being of cubical or low columnar
shape, with nuclei of spherical or slightly oval form and placed centrally
in the cells. The diameter of these portions of the uriniferous tubules
is greater than that of the proximal arm of Henle’s loop, showing the
flattened epithelium, though less than in the proximal convoluted portion.
The epithelium lining the different parts of a uriniferous tubule will
receive further consideration in discussing more fully developed urinifer-
ous tubules. In Fig. 13, 4, is shown a model of a uriniferous tubule in
which the epithelial differentiation of the different parts of the tubule, as
here described, may be clearly recognized. The tubule shown represents
one which measures 2.5 em., of which just about one-half falls to the loop

G. Carl Huber 61

of Henle. (This measurement and others giving the length of urinifer-
ous tubules are obtained from models, the length of the tubule as repre-
sented in the model being divided by the magnification at which it was
made — 400). In B of this figure is shown one of the sections of the
series of sections of this tubule from which the model was made; it
presents with other parts, a section of the distal part of the proximal
convoluted portion and its continuity with the proximal arm of Henle’s
loop, showing the flattened epithelium, which is cut through its entire
length to the immediate vicinity of the loop itself. The transition from
the clear epithelium with basal nuclei of the proximal convoluted portion
and the first part of the descending limb of Henle’s loop to the flattened
epithelium of the remainder of the descending limb of Henle’s loop, is
here clearly shown. The ascending limb of Henle’s loop shown in the
figure does not belong to the tubule shown in the reconstruction, but to an
adjacent tubule also in part reconstructed and of about the same shape
and stage of development as the one shown in A of this figure, here
added to enable a comparison betwen the epithelial lining of the two arms
of Henle’s loop as observed at this stage of development. It may be
emphasized for purposes of further discussion that the parts here desig-
nated as representing the proximal and distal arm of Henle’s loop were
thus designated after reconstruction of the two tubules of which they
form a part. .

The account here given of the differentiation of the epithelium of the
two arms of Henle’s loop differs very materially from that given by
Stoerk in his description of the development and cellular differentiation
of this portion of the uriniferous tubule. Stoerk states that at a time
when the loop of Henle is as yet very short, the clear epithelium which
characterizes that portion of the uriniferous tubule destined to form the
proximal convoluted portion extends to about the middle of the bend,
which is to form the loop of Henle, lining therefore its proximal half,
while the other half of the loop anlage is lined by the darker epithelium,
which lines the remainder of the tubule. (The terms clear and dark
epithelium are here used in the sense given them by Stoerk who speaks of
“hellem und dunklem Epithel”). As the loop elongates, the part lned
by dark epithelium grows more rapidly than the part lined by clear
epithelium, so that the dark epithelium takes in the loop and extends for
a variable distance into its proximal arm the clear epithelium ending,
therefore, a little above the loop. He further states that “ The descending
arm of Henle’s loop is therefore to be regarded as representing genetically
and morphologically the end segment of the tubulus contortus of the first
order, while the ascending limb represents genetically and morphologically

Fic. 13. A, uriniferous tubule from the kidney of a human embryo of the
seventh month (No. 16). x 160.

B, a portion of one of the sections of the series from which the preliminary
drawings used in making the model shown in A were made. X160. WM,
Malpighian corpuscle; dc, distal convoluted portion; pc, proximal convoluted
portion; di, descending or proximal limb of Henle’s loop, showing flattened
epithelium; c, collecting tubule; (all parts of the tubule shown in recon-
struction); al’, ascending limb of Henle’s loop of a second tubule; m’, dc’, dl’,
portions corresponding to those designated with like letters without the accent,
belonging to a second tubule.

G. Carl Huber 63

the beginning segment of the tubulus contortus of the second order.”
The descending or proximal arm of Henle’s loop to near its end is there-
fore, according to Stoerk, lined by an epithelium which is like the epithe-
lium which lines the proximal convoluted portion and presents a diameter
of tubule and lumen, which is also like the proximal convoluted portion,
while the ascending Jimb, which is of smaller diameter, is lined by cells
having a darker protoplasm and this not only in earlier stages of
development, but also in later stages and in post-fcetal life, as is apparent
from his description and also his diagram given in Fig. 27 of his article.
To state clearly Stoerk’s position, I shall quote from his summary as
follows:

“Das Protoplasma des Kanilchenepithels, von der Insertionsstelle am _ aitisseren
Blatt der Bowmanschen Kapsel angefargen bis zur Umbiegungstelle der Henleschen
Schleife, wird ziemlich gleichzeitig mit dem Auswachsen der Schleife hell, das lumen
dieses Kaniilchenabschnittes erweitert. Der helle und weitere Schenkel der Henleschen
Schleife ist der absteigende und nicht, wie bisher angenommen wurde, der Aufstei-
gende.’’ Stoerk also states in discussing his observations on this point that ‘‘ Dieser
vollig einwandsfreie und ausnahmslos konstante Befund steht in direktem Gegensatze
zu der Darstellung im Schweiger-Seidelschen Schema und, dem seinerzeit Gesagten
gemiiss, auch im Gegensatze zu dem, was seither zur allgemein giiltigen Anschauung
iiber das Verhalten der Henleschen Schleife geworden ist. Dass die falsche Lehre
bisher keine Richtigstellung erfahren hat, ist um so verwunderlicher, als eine Reihe von
allgemein bekannten Beobachtungen so whol der normalen wie der pathologischen His-
tologie auf das Wiedersinnige in der Sache hiitten hinweisen kénnen.”’

I have quoted thus fully from Stoerk, as his observations (based on
reconstructions) on the development and structure of the loop of Henle
have led him to conclusions which are so at variance with the generally
accepted views concerning the structure of the different parts of this
portion of the uriniferous tubule that a clear statement of his position
seemed necessary and this could best be given by making free use of his
own words. That the older and generally accepted view of the size and
structure of the descending or proximal limb of Henle’s loop—namely
that it represents that part of the uriniferous tubule which shows the
smallest diameter, as generally given in the diagrams of uriniferous
tubules, beginning with the well-known one of Schweiger-Seidel, and that
it is lined by a flattened epithelium—is the correct one and that Stoerk is
in error when he states that this portion of the uriniferous tubule presents
essentially the same diameter as the proximal convoluted tubule and is
lined by a similar epithelium, is shown by those of my models which
show advanced stages in the development of the uriniferous tubules.
The source of his error can, I believe, be readily shown. If one may be
allowed to judge from his figures, he has reconstructed only relatively
early stages in the development of the uriniferous tubules. The tubule
shown as model ZL, as found in his plates, presents the most advanced
stage figured by him. Figure 18 shows in very favorable section a tubule

of about the stage of development as that shown in model L. In the

64 Development and Shape of Uriniferous Tubules

tubule figured as seen in section, as is also apparent in references made to
it in the text, the clear epithelium found in the proximal convoluted por-
tion extends for about half the length of the descending or proximal arm
of Henle’s loop. This portion of the tubule has also a greater diameter
than is shown by the greater part of the remainder of the tubule figured
by him, the latter part being lined by the darker epithelium. The tubule
shown in Fig. 18 of his article presents about the same stage of develop-
ment and cellular differentiation as that shown under J of Fig. 3 of this
article, and this tubule was selected as showing a stage of development
just prior to that in which the epithelium of the proximal and distal
arms of Henle’s loop shows the specific differentiation described for them.
In A and B of Fig. 13, in which is represented a tubule showing a stage in
which the epithelium of all its parts shows characteristic differentiation
and represents a stage a little more advanced than the oldest stage figured
by Stoerk, it may be clearly seen that that portion of the descending limb
of Henle’s loop which in J of Fig. 3 or in Fig. 18 (Stoerk’s) shows the
darker epithelium, differentiates as the loop elongates, into the flattened
epithelium characteristic of the greater part of the descending limb of
Henle’s loop. Stoerk, in formulating his conclusions relating to the
shape and structure of the descending and ascending limbs of Henle’s
loop makes use of data gained from a study of uriniferous tubules
representing stages of development in which the epithelial differentia-
tion is not as yet complete in all parts of the tubules, notably in the
descending and ascending limbs of Henle’s loop. It may readily be seen
how from such insufficient data, he would be led to the conclusions
drawn. If the loop of Henle in either of the tubules shown in J of
Fig. 3 or in Stoerk’s Fig. 18 were drawn out to form a long loop of
Henle, retaining the structure given them in the figures, the result would
be a tubule in which the descending limb of the loop would be lhned by
a clear epithelium to near its end, one which would show a greater
diameter than the ascending limb, which would be lined by a darker
epithelium. This is what he has done and was thus led to the error he
has committed. That he is not justified in assuming that the uriniferous
tubules presenting the stages represented in his Figure 18 and model Z
show the structure and cellular differentiation of fully developed urinifer-
ous tubules, may be seen, I believe, in A and B of Fig. 13 and is also
shown by my other models showing more advanced stages of development.
It may, therefore, again be stated that reconstructions of proper stages in
the development of uriniferous tubules show that the descending limb of
Henle’s loop is lined by a flattened epithelium and presents a smaller
diameter than the ascending limb, which is lined by a cubical or short

G. Carl Huber 65

columnar epithelium, confirming thus the generally accepted view of the
size and structure of the two arms of Henle’s loop.

PECULIARITIES OF ForM PRESENTED BY THE First DEVELOPED
UrRINIFEROUS TUBULES.

Before leaving the discussion of the earlier stages in the development
of the uriniferous tubules, mention may yet be made of differences in
development presented by tubules which develop from the first few genera-
tions of renal vesicles, when compared with those which develop from
the later generations. Such differences of development are expressed
mainly in the relative degree of development shown by the different
parts of the tubules which are first formed, when these are compared
with those which develop later. It should, however, at the beginning be
stated that such differences in development as shall here be mentioned
are not of such a character as to form exceptions to the statement that
the uriniferous tubules which develop from the various generations of
renal vesicles present essentially the same developmental stages as have
been given in the preceding pages. As has previously been stated,
tubules which develop from the earlier generations of renal vesicles
present in the S-stage and for some little time after, when their con-
figurations are taken as a whole, a more rounded form than do those which
develop from later generations of renal vesicles. This difference in form
when the one type is compared with the other, is also observed in tubules
of the respective types in which the anlagen of the different parts of the
uriniferous tubule may be clearly made out in that the first developed
tubules present at this stage, an irregularly spherical mass when the
developing Malpighian corpuscle and the tubular complex are taken
together and considered as a whole. (See # and F of Fig. 5). The
tubules of similar stage of development derived from later generations of
renal vesicles, likewise considered, form a more elongated mass. (See
G of Fig. 6). This, as has been stated, is in part at least due to the
fact that during the earlier stages of development of the kidney, the tubu-
lar structures are relatively far apart and are surrounded by a relatively
large amount of mesenchymal tissue and may thus in growth and elonga-
tion expand in all directions, while in later stages of development the
tubular structures are in much closer proximity, separated by only a
relatively small amount of mesenchymal tissue; this juxtaposition to
surrounding tubules influencing the direction of their growth. In their
further development, tubules which are developed frém the first few
generations of renal vesicles are characterized by a relatively early and
marked elongation of those portions of the tubule destined to form the

66 Development and Shape of Uriniferous Tubules

proximal convoluted portions, so that this portion forms a prominent part
of the entire tubule for a certain period of its development, as may be
seen in reconstructions and in sections. The epithelium lning this portion
of the tubule as this elongates differentiates into one showing cells with
clear protoplasm and nuclei in basal position as described for the proxi-
mal convoluted tubule. The loop of Henle of these tubules for a certain
period of their development remains relatively short, when their length
is compared with that of the proximal convoluted portion. These tubules
generally show throughout, but especially in their proximal convoluted
portions, a greater diameter than tubules which develop later, the lumen
of the proximal convoluted portion being especially wide. Their Mal-
pighian corpuscles are also relatively large. In tubules which develop
from renal vesicles that follow the first few generations of these, the loop
of Henle elongates at a relatively early period of their development so as
to form a prominent portion of the tubule at a time when the proximal
convoluted portion shows only a few relatively short coils; the proximal
and distal convoluted portions taken together form thus a smaller mass
than is formed by similar portions of the tubule first formed at a time
when each type presents a loop of Henle of about the same length. This
gives the two types of tubules when seen in reconstructions a characteristic
form, and enables a distinction between them. To characterize these
differences of form more clearly, reference is made to the models repro-
duced in Figs. 14 and 16. In C and D of Fig. 14, are shown two tubules
reconstructed from the kidney of a rabbit embryo measuring 3.5 em.
They are from the layer of tubules situated nearest the pelvis of the
kidney and are therefore of those which are first formed. Attention is
called to the length and thickness of the proximal convoluted portion of
each of these tubules and to the relative shortness of the loop of Henle.
In A and B, are shown two tubules, presenting different stages of devel-
opment, reconstructed from the same kidney and are representative of
tubules the Malpighian corpuscles of which are situated nearer the peri-
phery of the kidney and were therefore differentiated later than the
Malpighian corpuscles and tubules of which C and PD are types. In the
second type of tubules, the loop of Henle is relatively long, when com-
pared with the proximal convoluted portion; in A, of about the same
length as the loop of tubules C and D, while the proximal convoluted
portion of tubule D measures 1.5 mm., of C, 1 mm., and of A only 0.5
mm. The difference in shape presented by the two types of tubules is
apparent from the figures. The Malpighian corpuscles and coiled por-
tions of the uriniferous tubules situated nearest the pelvis of the kidney
form a distinct layer, just above the mesenchyme which surrounds the

G. Carl Huber 67

pelvis, which is readily recognized in kidneys of embryos of about the mid-
dle of fcetal life; the sections of proximal convoluted portions of these
tubules form the greater part of this layer in such preparations. The ap-

Fic. 14. Tubules from the kidney of a rabbit embryo of 3.5 cm. length
(No. 8). 100. m, Malpighian corpuscle; pc, proximal convoluted portion;
dc, distal convoluted portion; dl, descending limb of Henle’s loop; al, ascend-
ing limb of Henle’s loop; j, junctional tubule; C and D are representative
of the first few generations of tubules. A and B of later generations.

pearance presented by a section of the kidney at this stage of development
is shown in Fig. 15, in which is given a portion of one of the sections of the
series of cross sections from which the models shown in Fig. 14 were made.
The sections of the tubules and Malpighian corpuscles designated by the

68 Development and Shape of Uriniferous Tubules

letter a, belong to a tubule of the type shown in C and PD of the previous
figure ; tubule } is the one shown in A of that figure. In the remainder of
the section may be seen renal vesicles and tubular anlagen in various stages
of development. This portion of the sections which constitutes the neogenic
and subneogenic zones, stains more deeply than the portion containing the
well developed proximal convoluted tubules; these, by reason of the fact
that they are lined by a differentiated epithelium (as previously de-

Fic. 15. <A portion of a cross section through the kidney of a rabbit embryo
of 3.5 cm. length (No. 8). 100. a, sections of proximal convoluted portion
and Malpighian corpuscle of a uriniferous tubule representative of the first
formed tubules of this kidney; a’, section of ascending limb of Henle’s loop of
the same; b, tubule shown in A of Fig. 14, representative of a later generation
of tubules; ct, collecting tubule; c, capsule.

scribed), the protoplasm of the cells of which stains faintly in eosin
and erythrosin, form a zone which stains less deeply than the more
peripheral portions containing the tubular anlagen with undifferentiated
epithelium of an embryonic character. In Fig. 16, are shown two urinif-
erous tubules, 4, from the kidney of a cat embryo of 4 cm. length; B
from the kidney of a cat embryo of 6 cm. length; both are representative
of types of tubules found nearest the pelvis of the respective kidney and
show more advanced stages of development than are shown by tubules C
and D of Fig. 14. Tubule A measures 3.75 mm., of which 1.9 mm.
falls to the proximal convoluted portion, the entire loop (ascending and

G. Carl Huber . 69

descending limbs) measuring .75 mm. The length of tubule B is 5 mm.,
of this 2.65 mm. falls to the proximal convoluted portion and 1.5 mm..
to the entire loop of Henle. In these tubules, as is no doubt apparent
from the figures and measurements given, the proximal convoluted por-

Fie. 16. Two uriniferous tubules, A, from kidney of cat embryo of 4 cm.
length (No. 6); B, from kidney of cat embryo of 6 cm. length (No. 8).
x 100. A’, B’, figures showing the course of the proximal and distal con-
voluted portions of the two tubules; the tubule representing the distal con-
voluted portion is shaded; m’, Malpighian corpuscles.

tions form the more prominent part, a little over half the length of the
entire tubule. Similar observations may be made on the kidneys of °
human embryos of about the third and fourth month; although the
differences of size and shape of the tubules which develop first when com-
pared with those which develop later are not so marked in the human

70 Development and Shape of Uriniferous Tubules

embryo as in those of the cat and rabbit; they are, however, of similar
character and of sufficient degree to merit recognition.

In Fig. 17, are represented three stages in the development of urinifer-
ous tubules as observed in the pig embryo. My observations on the

I'ic. 17. Three uriniferous tubules from kidneys of pig embryos. X 133.
A and C, from embryo of 2.8 em. length (No. 8). The latter is representative
of the first generation of tubules, situated nearest the pelvis and is shown from
the Malpighian corpuscle to the first coil of the distal convoluted portion; the
former represents a much less developed tubule, one situated nearer the
periphery of the kidney; B, one of the most fully developed tubules from the
kidney of an embryo of 2.4 cm. length. m, Malpighian corpuscle; pe,
proximal convoluted portion; dc, distal convoluted portion; J, anlage of loop
of Henle; di, of descending limb; al, of ascending limb; d, collecting tubule.

development of uriniferous tubules in pig embryos are not as yet complete,
owing to the lack of suitable material showing the older stages of develop-
ment. They have for this reason not received more than casual mention
in the preceding pages. A fuller treatment is reserved for another
contribution, in which the development of the collecting tubules, more

G. Carl Huber ih

particularly the development of the so-called arcades of the collecting
tubules discussed and figured by Von Ebner (p. 1105) will receive con-
sideration. The tubules presented in Fig. 17 have been added, as they
illustrate very clearly the fact that the first few generations of tubules of
kidneys of pig embryos show a relatively very early development of the
proximal convoluted portions and of the Malpighian corpuscle, while the
loop of Henle develops at a relatively late period; tubule B, for instance,
has attained a length of 1.5 mm. and presents a fully developed Mal-
pighian corpuscle, at a time when the bend which is destined to form the
loop of Henle (recognized by the fact that the differentiated epithelium
of the proximal convoluted portion extends to nearly the top of the
bend,) may just be made out. In tubule C, the portion which is recog-
nized as the anlage of the loop of Henle presents a number of coils
instead of being a relatively straight hmb—an appearance which I have
observed only in pig embryos. It may here be stated that the formation
of the renal vesicles and the differentiation of the S-shaped stage of the
uriniferous tubules is for pig embryos essentially the same as for cat,
rabbit, and human embryos, as discussed and figured in these pages.
Hamburger’s observations are of interest in this connection. In dis-
cussing the differences in form presented by the anlagen of uriniferous
tubules in younger and older stages, he states that “the size of the
anlage is different in younger than in older embryos and for the same
stage of development it will generally be found that they are larger in
the former than in the latter; especially is this true of the pseudo-
glomeruli, which are especially large in young embryos.” This, as may
have been seen, I have confirmed, adding that they differ, not only in size,
but also in shape. He further describes characteristic differences in the
development of uriniferous tubules in simple kidneys when compared with
those of lobulated kidneys. To quote further: “In animals with simple
kidneys (mouse, rat,) the loop of Henle attains considerable develop-
ment before the Bowman’s capsules become closed so as to form a sphere
and this is as true of the first formed anlagen as for those which develop
later. In animals with lobulated kidneys (‘ zusammengesetzten Nieren ’)
I have found a different condition; in them, the coiled tubules first
formed attain considerable length and the Malpighian corpuscles, full
development, before a loop of Henle may with certainty be discerned ; the
later development, however, is as in simple kidneys.” His description of
the development of the earlier generations of uriniferous tubules, as
observed in embryos of animals with lobulated kidneys (human, pig,)
coincides with that here given. In embryos of the cat and rabbit (animals
with simple kidneys), I have also observed, as has been stated, a difference

2 Development and Shape of Uriniferous Tubules

between the tubules first formed and those which develop later. As I have
not reconstructed the uriniferous tubules of embryos of the mouse and
rat, | am not in a position to judge Hamburger’s observations as concerns
these; he is, however, in error when he generalizes from such observa-
tions. He further states that “ In young embryos of species with lobu-
late kidneys (human, pig, beef), the most fully developed Malpighian
corpuscles are absolutely larger than in older embryos and this is also
true of the diameter of the tubuli contorti. In animals with simple
kidneys (mouse, rat), this not the case.” In this also, my own observa-
tions do not confirm him, as in embryos from both types of animals I
have found the Malpighian corpuscles and the proximal convoluted por-
tions (tubuli contorti) of the first formed tubules larger than of those
which develop later; in the rabbit, this is especially so. (See Fig. 14.)
In a discussion of his observations, Hamburger states that it may be possi-
ble that in tubules showing early development of the coiled portions and
Malpighian corpuscles with absence of a distinct loop of Henle, these may
develop as the kidney development proceeds. “ Since without doubt, how-
ever, the first generation of coiled tubules of lobulated kidneys degenerates,
the supposition is permissible that it is particularly the atypical tubules
that disappear; conclusive evidence of this, I am however unable to
present.” Riedel (I quote from Hamburger) who also recognized the
large size of the Malpighian corpuscles and tubuli contorti of the first
formed tubules, states that these become smaller as development proceeds.
Kolliker, whose description is based largely on rabbit embryos, quotes
Riedel (page 952) as describing a degeneration of the Malpighian corpus-
cles and tubuli contorti first formed, for which he states there is no
evidence. Emery, on the other hand, who studied the development of the
kidney in goat embryos, states that he did observe evidence of a breaking
down of the first formed tubules. Hamburger attempts to harmonize
the conflicting views of Kélliker and Emery by stating that in embryos
of animals with simple kidneys, there is no evidence of a breaking down
(zu Grundegehen) of the first formed tubules, while in embryos of
animals with lobulated kidneys, it is the atypical tubules which atrophy
and disappear. My own observations lead me to say that neither in
embryos of animals with simple kidneys nor in those with lobulated
kidneys have I obtained evidence which would lead me to conclude that
any of the uriniferous tubules atrophy and disappear in the course of
development. Tubules with relatively short loop and well developed
Malpighian corpuscles and proximal convoluted portion, I have observed
in embryos of animals with simple and with lobulated kidneys; in such
tubules, as will be stated in presenting the more advanced stages of devel-

~
ou)

G. Carl Huber

opment of uriniferous tubules, the loop of Henle elongates at a relatively
late period, developing and differentiating as do the loops of other tubules.
I agree, therefore, with Minot when he states: “Some authors have
maintained that there is an atrophy of some of the tubules of the feetal
kidney, but I agree with Golgi in believing that of this there is no valid
evidence.” In offering an explanation for the difference in the develop-
ment of the first generation of uriniferous tubules, when these are com-
pared with those which develop later, which difference may be character-
ized as consisting of a relatively early development of the Malpighian
corpuscles and the proximal convoluted portions and a relatively late
development of the loop of Henle of the former when compared with the
latter, attention may be called primarily to the simple mechanical con-
dition which prevents an elongation of the loop of Henle of the first
formed tubules. The Malpighian corpuscle and tubular portion which
are first formed develop near the developing pelvis of the kidney, the
former being separated from the latter by a relatively narrow zone
of mesenchyme. As the tubules proceed in their development, their
loops of Henle soon reach the epithelium of the growing pelvis and the
denser mesenchyme immediately surrounding it. The loop of Henle of
tubules C and D of Fig. 14 and A and B of Fig. 16 reach to the pelvis
of the respective kidneys. The ends of such loops are often seen bent
to one side or the other, as though attempting to avoid the obstruction
which prevents their further elongation, and they are often slightly
folded and twisted, the end of the loops being bent so as to project
toward the periphery of the kidney, as for instance in tubule A of Fig. 16.
Sections of developing kidneys of pig embryos of about 2.5 to 5 em.
length (the oldest stage available for me) show this in a very characteris-
tic manner. In the series of sections from which the model shown in
C of Fig. 17 was made, the Malpighian corpuscle was separated from the
pelvis of the kidney by a distance which is about two-thirds the length
of the diameter of the corpuscle; a large collecting tubule which opens
into the pelvis by a funnel-shaped expansion is nearly in contact with
the coiled tubular segment designated in the figure as the descending
limb of the loop. There is obviously here no opportunity for the forma-
tion and the elongation of a loop with relatively straight and nearly par-
allel limbs. The fact that so little mesenchyme separates the tubules
and the Malpighian corpuscles from the pelvis of the kidney, and the
shortness of the loop of Henle at a time when the proximal convoluted
portions are well developed, also the fact that the loops of Henle, so far
as developed, are not present in the form of straight tubules, but as coiled
tubules, so that only cross or oblique sections of them are obtained, give

74 Development and Shape of Uriniferous Tubules

to sections of kidneys of pig embryos of these stages a very characteristic
appearance, resembling somewhat the appearance presented by sections of
the mesonephros, and making it easy to distinguish them from sections of
kidneys of cat, rabbit, and human embryos of corresponding stages of
development.

The fact that the anlagen of tubules which are first formed are rela-
tively far apart, the growth of the tubules developing from them being
thus for a time not materially interfered with, may be mentioned as a
possible reason for the early development and elongation of the proximal
convoluted portions of these tubules, certainly an explanation of the
fact that these tubular portions are coiled more in the horizontal plane
than similar portions of tubules which develop later.

Malpighian corpuscles of a stage of development in which they present
the form and cellular differentiation shown by these structures in post-
foetal life, and proximal convoluted portions of tubules in which the
epithelium has become differentiated so as to present the structural ap-
pearances presented by it in more fully developed tubules, even though
such corpuscles and tubular segments are found in early stages of kidney
development, may be thought to have assumed functional activity. If
may be suggested, therefore, that in the uriniferous tubules which are
first formed, those portions which are especially concerned with the
function of excretion are rapidly differentiated to perform this function,
while the other portions which are not especially concerned with the
function of excretion are developed more slowly. Although there is as
yet no unanimity of opinion as to the special functions to be ascribed to
the different parts of a uriniferous tubule, assuming that the differences
of shape and structure observed in the epithelium lining the different
parts of the uriniferous tubules postulates a difference of function for
the several parts lined by the especially differentiated epithelium, the
generally accepted view is that the Malpighian corpuscles and the proxi-
mal convoluted portions of the uriniferous tubules are more particularly
concerned with the function of excretion. The assumption, therefore,
seems justified that it is for this reason, and not alone owing to mechani-
cal conditions, that the Malpighian corpuscles and especially the convo-
luted portions of the first-formed uriniferous tubules are developed and
differentiated relatively early, so as to enable the permanent kidney, at
an early stage in its development, to assume an excretory function.

LATE STAGES IN THE DEVELOPMENT OF URINIFEROUS T'UBULES.

The further growth and development of uriniferous tubules, after a
stage in which all the parts show a characteristic differentiation (as for

G..’Carl Huber -. Td

instance seen in the tubule of Fig. 13) consist primarily in an elonga-
tion of the tubular portion, this affecting all its parts, but especially the
loop of Henle. The data gained by me concerning the later stages in
the development of uriniferous tubules are confined to a large extent to
those obtained from observations on older embryos of cat and rabbit, the
kidneys of which are of relatively simple type, with only one Malpighian
pyramid. The disposition of the collecting tubules and other tubular
elements of the medulla of such a kidney is such that both in series of
cross and longitudinal sections a certain few will be found in which the
plane of section is parallel or very nearly parallel to certain of the col-
lecting tubules and loops of Henle. Nearly every such series will show
here and there a loop, one or the other arm of which is cut for nearly
its entire length, and, by using this as a starting point, I have generally
succeeded in tracing out the remainder of the tubule in question. In the
human kidney of the later periods of foetal life, the conditions are com-
plicated by reason of the fact that the cortex is divided into a number
of primary and secondary lobules, and even when one lobule is used for
sectioning I have found it quite impossible to orient the block in such a
way as to obtain long segments of either of the two arms of the loop of
Henle, especially for the upper regions of the medulla. In this region the
tubular elements very generally appear in oblique section and are sepa-
rated in the kidneys of human embryos of the 8th and 9th months by a
relatively small amount of interstitial tissue, which makes the tracing of
a single tubule through this region and, as is often necessary, through a
long series of sections a matter of difficulty. From a number of partial
reconstructions which were made, I am led to believe that the observations
made on uriniferous tubules of the older embryos of cat and rabbit (espe-
cially the former) may be accepted as presenting the conditions shown
by uriniferous tubules of human embryos of the later months of fcetal
life. My observations on the later stages of development of uriniferous
tubules, I shall present by giving a brief description of certain tubules
showing these, which have been reconstructed in full and which are here
figured. In Fig. 18 are shown two uriniferous tubules, the Malpighian
corpuscles of which are situated in about the middle zone of the cortical
portion of the kidneys from which they were reconstructed and may,
therefore, be taken as representative of tubules which were differentiated
after the first formed tubules of the respective kidneys. Tubule A is
from the kidney of a cat embryo obtained a few days before birth. The
length of this tubule is 4.1 mm., of which 1.75 mm. falls to the proxi-
mal convoluted portion, 1.65 mm. to the entire loop of Henle (the measure-
ment here beginning and ending with the level of the lower border of the

76 Development and Shape of Uriniferous Tubules

Malpighian corpuscle) and .? mm. to the distal convoluted portion and
the junctional tubule. As may be seen from the figure, more clearly
from the key, the proximal convoluted portion presents one prominent

Fig. 18. A, uriniferous tubule from kidney of cat embryo obtained a few
days before birth (No. 9); B, tubule from kidney of rabbit embryo 6.5 mm.
in length (No. 13). X 100. A’, B’, keys giving course of coiled portions
of tubules, distal convoluted portions shaded. m, Malpighian corpuscles; pe,
proximal convoluted portions; es, end segment; dl, descending limb, al,
ascending limb of Henle’s loop; j, junctional tubule.

G. Carl Huber

~~
~

primary loop, which extends toward the periphery, and numerous sec-
ondary loops. The distal convoluted portion is practically enclosed
within the coil formed by the proximal convoluted portion. The proxi-
mal or descending arm of the loop, as the sections show, is lined nearly
throughout—from a little below the level of the Malpighian corpuscle to
near the loop proper—by a flattened epithelium, while the distal or
ascending arm, with the loop itself, presents a slightly larger diameter
and is lined by a cubical epithelium, and on reaching the coiled portion
passes in front of the first part of the proximal convoluted portion, near
the corpuscle. Tubule B of this figure is from the kidney of a rabbit em-
bryo of 6.5 em. length. Its length is 4.2 mm., of which 1.6 mm. falls
to the proximal convoluted portion, 1.8 mm. to the entire loop of Henle,
and .8 mm. to the remainder of the tubule. The disposition of the various
parts of this tubule is similar to that shown by tubule 4, except that the
distal convoluted portion is more exposed. Attention should be called
to the proximal arm of Henle’s loop. About the upper half of this
presents the same diameter (and structure) as the proximal convo-
luted portion. This portion represents the so-called spiral portion of
Schachowa or the end piece (end segment) of Argutinski, and is in
reality the distal segment of the proximal convoluted portion. It does
not, so far as I have been able to observe, form a spiral; the term spiral
tubule is therefore inappropriate. About the lower half of this loop
presents the characteristic shape and structure of the descending arm—
small diameter and lined by flattened epithelium. The transition from
one to the other is clearly shown in the figure. The loop itself, as well
as the ascending limb, presents a larger diameter than the thin portion
of the descending limb and is lined by a cubical epithelium. It may
here be stated that this tubule presents in a very characteristic way the
shape, arrangement, and structure of the several parts found in tubules,
the Malpighian corpuscles of which are situated in a zone of the cortex,
above that occupied by corpuscles and coiled portions of the tubules
first formed, which are nearer the medulla.

Tubule A, shown in Fig. 19, represents one of the most fully devel-
oped tubules from the kidney of a rabbit embryo 6.5 cm. long. It has a
length of 5.8 mm., of which 2.2 mm. forms the proximal convoluted
portion, 2.75 mm. the entire loop of Henle, and a little less than 1 mm.
the remainder of the tubule. This tubule I regard as one which in
earlier stages of development would have shown a relatively long proxi-
mal convoluted portion, with a short loop of Henle—a so-called atypical
tubule (Hamburger)—and for the following reasons: The loop of
Henle reaches to near the pelvic epithelium, being separated from it by

78 Development and Shape of Uriniferous Tubules

only a small amount of interstitial tissue; its Malpighian corpuscle is
situated in the deepest portion of the cortex, just above its junction
with the developing medulla; the prominent coils of the proximal con-
voluted portion have a more horizontal position and are, therefore,
spread out more in a lateral direction than would be true for tubules
which develop later, in which the prominent coils of the proximal con-
voluted portion are more prone to grow in a perpendicular direction,
toward the cortex. As I have seen no evidence of the atrophy of the
first formed tubules—so-ca]led atypical tubules—with relatively short
loops of Henle—it is assumed that this loop elongates as the medulla
develops. The distal convoluted portion of this tubule lies upon—in
front of, as shown in the figure—the coil complex formed by the proxi-
mal convoluted portion, the course of. which may be ascertained by a
study of the key A’. The descending limb of the loop is lined almost
throughout by a flattened epithelium and presents a comparatively
small diameter. The end segment is relatively short, forming only a
small part of the limb. The loop itself and the distal limb have a diam-
eter which is just about three times that of the greater part of the
descending limb. The sharp bend shown by the ascending limb, just
before the Malpighian corpuscle is reached, is due to the fact that a
relatively large arterial branch occupies the space just beneath the
corpuscle, the distal hmb arching partly over this to reach the vicinity
of the corpuscle. For this reason also the upper end of the proximal
limb is separated by a greater distance from the Malpighian corpuscle
than may be considered typical. With these exceptions, this tubule
may be regarded as presenting in a very characteristic manner the shape
and arrangement of the different parts of a uriniferous tubule, the Mal-
pighian corpuscle of which is situated in the deepest portion of the cor-
tex, thus of the first few generations of uriniferous tubules. Tubule B
of Fig. 19 represents one of the most fully developed tubules of the kid-
ney of a rabbit killed the first day after birth. Three exceedingly for-
tunate sections of the series of cross sections 5 » thick, into which one
of the kidneys was cut, contained nearly the entire length of the proxi-
mal limb of its loop. The model of this tubule measures from the tip
of the loop to where it ends in the collecting tubule four feet and four
inches. On completing the reconstruction, the tubule proved to be one
presenting a not quite typical arrangement of the coils of the proximal
convoluted portion, these forming a configuration which is too open
and too much in one plane. The cause of this is not readily made out
at the stage of development here presented. The tubule shown in (
of Fig. 14 presents almost the same relations of its parts, and for that

G. Carl Huber 719

tubule it is evident that the somewhat atypical form assumed is due to
the fact that its juxtaposition to other and slightly larger tubules is
such that for purely mechanical reasons, on growing and elongating
presumably in the direction of least resistance, its coiled portion was

Fic. 19. Tubule A, from kidney of rabbit embryo (No. 13) measuring 6.5
em. X 80. Tubule B, from kidney of rabbit one day old (No. 14). x 40.

80 Development and Shape of Uriniferous Tubules

forced to assume the form presented by it. The Malpighian corpuscle
of tubule B is found in the second tier of corpuscles, while its loop of
Henle extends nearly to the pelvis of the kidney. Its entire length is
7.8 mm., of which the proximal convoluted portion forms 1.85 mm.,
the entire loop 4.8 mm., the distal convoluted portion 0.55 mm., and
the remainder of the tubule 0.57 mm. The arrangement of the differ-
ent parts of the tubule is such that its entire course may be followed in
the figure; a key is therefore dispensed with. The tubule presents only
a short end segment, the greater part of the descending tube to near the
loop itself forming a tubule of small diameter (12, to 15), lined
throughout by a flattened epithelium, while the ascending mb with
the exception of the lower one-sixth is lned by a cubical epithelium
with faintly striated protoplasm. It is evident from the character of |
the epithelium which lines the loop itself and the lowest part of the
two limbs (cells of embryonic character with relatively large, deeply
staining nuclei) that this portion of the loop is elongating to keep pace
with the enlarging and elongating Malpighian pyramid.

The tubule shown in Fig. 20 was reconstructed from the kidney of a
eat embryo obtained a few days before birth; the model presents a
length of four feet. The tubule itself measures almost exactly 9 mm. ;
2.4 mm. belonging to the proximal convoluted portion, 4.85 mm. to the
entire loop of Henle, 1.1 mm. to the distal convoluted portion, and
0.55 mm. to the junctional tubule. This tubule I regard as showing in
a very typical manner the shape, arrangement, and relations of the
different parts of a uriniferous tubule, especially one the Malpighian
corpuscle of which is situated in the deeper portion of the cortex. The
course of the proximal convoluted portion may be ascertained from the
key (unshaded tubule). The curvatures of this part, as shown in the
key, are, for the sake of clearness, not so pronounced as in the actual
model, but are drawn with sufficient accuracy to give correctly the course
of this part of the tubule. The distal convoluted portion presents two
prominent loops and in its upper portion two smaller loops, in which
portion there may also be observed irregularities in the diameter of the
tubule. The shaded portion of the key gives the course of this portion
of the tubule. The coils formed by it lie over—or in front of, as shown
in the figure—the coil complex formed by the proximal convoluted
portion, the prominent loop being found a little above the Malpighian
corpuscle. The proximal limb of Henle’s loop presents almost through-
out a small diameter and is lined in this portion by the characteristic
flattened epithelium. The end segment is short and presents in a typi-
eal manner the transition in shape and structure of the tubule as found

G. Carl Huber _ 81

in the proximal convoluted portion to that seen in the descending limb.
The loop itself and the ascending arm have a diameter which is about

Fic. 20. Uriniferous tubule from kidney of cat embryo obtained a few
days before birth (No. 9). X30. m, Malpighian corpuscle; n, neck; pe,
proximal convoluted portion; es, end segment; dl, descending limb; al, ascend-
ing limb of Henle’s loop; dc, distal convoluted portion; 7, junctional tubule.

82 Development and Shape of Uriniferous Tubules

twice that of the greater part of the descending limb and are lined by a
cubical epithelium. The ascending limb reaches the coil complex in
the immediate vicinity of the Malpighian corpuscle with which it is
practically in contact for about one-third of the latter’s circumference.
The proximal limb leaves the coil complex in close relation with the
upper end of the distal limb, therefore very near to the Malpighian cor-
puscle. The lower end of the loop of this tubule reaches to the pelvis
of the kidney, being separated from its epithelium by only a small
amount of interstitial tissue.

In the tubules shown respectively in B of Fig. 19 and in Fig. 20, the
Malpighian corpuscle is fully developed and the proximal and distal con-
voluted portions may be regarded as also fully developed and as having
attained the size and for all practical purposes the length to which the
tubular portions grow. This statement is based on the measurements
made on uriniferous tubules—especially those which are developed earli-
est—at different stages of their development. The measurements given
below show that the proximal convoluted portions at a relatively early
stage in their development attain a length which is about that presented
by these tubular segments in later stages of embryonic development or
at birth. The following summary of certain of the measurements pre-
viously given presented in the form of a table may serve to substantiate
this:

iB aS 3 ES ipa
<a) Org o- BY S a Z
oie Me HaSa eo) Ses
TUBULE. | OBTAINED FROM. 225 5 Oe 25 Loop
eee. Hie sede eects
2 fia ee
Fig. 16, 4 | Cat embryo, 4 cm. Buchs) Meo, WNIGG) ganja |) Sale 75 mm,
Fig. 16, B Cat embryo, 6.5 cm. 5 mm. P05) Wala | Ae oon. 1.50 mm.
Fig. 20 | Cat embryo just before 9 mm. 2.4 mm. j1.1 mm. |4.85 mm.
| birth |
Fig. 14, D Rabbit embryo, 3.5 em. Peuyabing, Wilan aye W ee soe |) ono cn.
Hilo OSA) Rabbit embryo, 6.5 cm. 5.8 mm. 2:21 gam o'|)- ahr 2.75 mm.
Fig. 19, B | Rabbit at birth 7.8 mm. (1.85 mm.) .55 mm. 4.8 mm

The data given in the table will show, I believe, that the elongation
of the uriniferous tubules, after a certain period in their development,
is to a large extent due to a growth in length of the loop of Henle, the
proximal convoluted portion of each tubule attaining at a relatively
early period in its development approximately the length shown by this
tubular segment in fully developed tubules.

To show how much of a single uriniferous tubule, representing a stage
in which the Malpighian corpuscle and proximal and distal convoluted

G. Carl Huber : 83

portions may be regarded as fully developed, is contained in a selected
section of a series of sections containing the whole tubule, I have in-
serted Figs. 21 and 22, in each of which the several parts of a single
uriniferous tubule are sketched darker than the remaining portions of
the sections reproduced. In Fig. 21, A and B, are reproduced portions
of the cortex of two sections from the series of sections from which
tubule B of Fig. 19 was reconstructed. A comparison of this figure with
that showing the reconstruction will show that the plane of section was at
right angles to that in which the model is reproduced. A line drawn
through the middle of the Malpighian corpuscle, and the proximal conyo-
luted portion (see B, Fig. 19) shows the location of the section, a por-
tion of which is shown in B of Fig. 21. In this section the Malpighian
corpuscle, the proximal convoluted portion soon after leaving the cor-
puscle and two arms of a loop of it as it comes back over the corpuscle
are met with. Their relation to each other and to the surrounding
structures may be seen from the figure. In A of the figure is shown
a portion of the 13th section further on in the series. This section
passes practically through the entire length of the junctional tubule
(leading to the periphery of the cortex) which will indicate its position
in the model; also through the coils of the distal convoluted portion,
which is cut in cross section four times and through a long and a short
loop of the proximal convoluted portion, escaping the Malpighian cor-
puscle. In Fig. 22 is shown a portion of the cortex—about the lower
one-half—and the uppermost portion of the medulla of one of the sec-
tions from the series from which the model was made, which is shown
in Fig. 20. The plane of the section is nearly at right angles to the
model as placed on the page. The Malpighian corpuscles and tubules
belonging to this uriniferous tubule are sketched more deeply than other
parts of the figure. The section passes through about the center of
the Malpighian corpuscle and through the upper end of the ascending
limb, which may be observed as in close proximity to the corpuscle. Of
the five cross sections of tubules arranged nearly in a row above the
Malpighian corpuscle the first two and the last two are sections of the
distal convoluted portion, the middle one and the three nearly longi-
tudinal sections of tubular segments are of the proximal convoluted
portions, the lowest one just to the left of the end of the ascending
limb marks the end of the proximal convoluted portion, beyond which
the tubule becomes smaller to form the descending limb. No part of
the loop of Henle nor of the junctional tubule is shown in this section.
I have purposely selected for figures sections of these two uriniferous
tubules, since they represent tubules of very different form; the arrange-

84 Development and Shape of Uriniferous Tubules

ment of the sections of the tubules and their relation to the surround-
ing tubules are therefore very different. These figures may serve to
show graphically the difficulties met with in attempting to predict what
particular portions of any given section through the cortex of the kidney

Fic. 21. A portion of the cortex of two sections of the kidney, from the
series from which tubule B, of Fig. 19, was reconstructed. The portions of
the tubule represented in these sections are sketched more deeply than the
other parts. 100. For fuller description, see text.

belong to one uriniferous tubule; a question to which I shall return
later.

From what has been said concerning the various stages of develop-
ment of the uriniferous tubules, it may be seen that from a relatively
early period in their development—from a stage in which the various
parts of the tubules may be considered as differentiated to the extent

G. Carl Huber . 85

that the parts are clearly recognized—the different portions of a urin-
iferous tubule present certain definite relations which are not materially
altered as development proceeds and are maintained in full develop-
ment. In the foetal as also in the post-fcetal kidney the proximal and
distal convoluted portions belonging to a uriniferous tubule form a
coil complex which shows certain relations to the Malpighian corpuscles
of such a tubule, likewise the beginning and ending of the loop of
Henle. The coil complex formed by the proximal and distal convoluted
portions of a uriniferous tubule is generally situated just above its

Fic. 22. A portion of the cortex and uppermost part of the medulla from
one of the series of sections from which the model shown in Fig. 20 was made.
x 100. The parts of this tubule represented in this section are sketched
darker than the other parts. For fuller description, see text.

Malpighian corpuscle, forming a more or less compact mass of convolu-
tions, the main coils of the proximal convoluted portion lying to the
inner side—toward the collecting tubule in which the uriniferous tubule
terminates—the coils of the distal convoluted portion lying more to the
outer side—laterally—although often partly embedded in the coils of
the proximal convoluted portion or nearly completely surrounded by
them. The proximal convoluted portion leaves the Malpighian cor-
puscle on its inner side and passes first upwards—toward the periphery
of the kidney, as was also shown by Golgi—then forms coils in various
directions, ultimately to return to the neighborhood of the Malpighian
corpuscle and to pass toward the medulla. The main coils of the distal
convoluted portion are usually found a little above the Malpighian cor-

86 Development and Shape of Uriniferous Tubules

puscle, one coil or loop generally resting upon it. The upper end of
the distal arm of the loop of Henle remains from the time when it may
be first recognized in close relation to the Malpighian corpuscle, either
passing over it at a variable distance from the origin of the proximal
convoluted portion or crossing this near the corpuscle. Hamburger,
who has also observed this close relation of the ascending limb to the
corpuscle, states that in a number of his preparations showing various
stages of development, he found the upper end of the ascending limb of
the loop of Henle in close relation with the Malpighian corpuscle and
in the region where the vessels enter and leave it. This, he states, is
readily explained when it is remembered that the loop of Henle is orig-
inally in the bowl of the “ pseudoglomerulus,” therefore in contact with
the developing glomerulus; and as the loop grows in a central direction,
the Malpighian corpuscle is fastened to this region by means of capil-
laries from the vas efferens. Golgi is quoted as saying (Stoerk) that it
is the thin arm of the loop which is fastened to the corpuscle in the
region of vessel exit. I agree with Stoerk in saying that neither the
proximal nor distal arm shows any definite relation to the vessel porta
of Bowman’s capsule. That, however, especially the ascending hmb
of the loop returns for each tubule to the coil complex formed by proxi-
mal and distal convoluted portions into the immediate vicinity of its
Malpighian corpuscle is clearly shown by all my reconstructions as
well as those of Stoerk, as also the tubules obtained by maceration by
Golgi and Hamburger, retaining thus in later stages of development and
in post-feetal life the relations shown in early stages of development. I
have not satisfied myself that the relations become fixed through the
agency of arterioles or capillaries, branches of the vas efferens, nor by
an especial development of interstitial tissue in this region, although
the latter factor seems now and then to play a role. The descending
limb of Henle’s loop in the region where it leaves the coil complex is
generally in close relation with the upper end of the ascending limb,
sometimes lying to the inner side of it, again over it, therefore also near
the Malpighian corpuscle, showing in this respect also in later stages
of development the relations borne in early stages of its evolution. The
two limbs of Henle’s loop are generally quite parallel and take quite a
direct course toward the apex of the Malpighian pyramid. Certain of
the loops—those belonging to the tubules which are first developed—
extend to near the pelvic epithelium, the apex of the pyramid, retaining
in this respect the relations shown in early stages of development, elon-
gating as the medulla develops. Tubules of the several generations
which develop later terminate at various levels in the medulla and, in

G. Carl Huber _ 87

general terms it may be stated, higher up in the medulla for each suc-
cessive generation of tubules, the loop of those latest formed extending
only to the boundary zone of the medulla and even for the adult human
kidney, remaining entirely within the medullary rays of the cortex.
The descending limb of the loop is the thinner of the two and is lined
by a clear, flattened epithelium; the ascending limb, showing the greater
diameter, is lined by a cubic epithelium with striated protoplasm; the
transition from the thin to the thicker portions of the loop occurs in
the tubules reconstructed, at the lower end of the descending limb, at a
variable distance from the loop itself, though generally near it. This
portion of the descending limb and the loop itself show the same diam-
eter and epithelium as the ascending limb. Schweiger-Seidel found the
transition from the thin to the thicker epithelium in about an equal
portion of the tubules—1, at the lower end of the descending limb; 2, in
the loop itself; 3, at a variable distance from the loop in the ascending
limb. His observations have received very general acceptance. Von Ebner,
one of the more recent writers, who has slightly modified Schweiger-Seidel’s
diagram, states that “the place of thickening is inconstant; now it hes
in the descending mb, now in the loop itself, often and quite regularly
for loops which extend deep into the pyramid, in the ascending limb.”
Hamburger finds for the mouse that the loop itself and the ascending
limb are lined by a granular epithelium to a few days before birth; as,
however, the loop grows in length with the increase in length of the
papilla, the elongation of the loop is obtained through a growth in
length of the descending limb, lined by the flattened epithelium, so
that the loop is formed by it, and only in the basal portion of the pyra-
mid in the boundary zone are there found loops with the dark epithe-
lum. In making the statement, based on observations made on my
models, that the transition from the thinner to the thicker part of the
loop of Henle occurs at the lower end of the descending limb, I do so
with some reservation, since the models carry the development of the
tubules only to the time of birth. I am aware that such statement is
open to the criticism that I have applied to Stoerk’s observations relat-
ing to the size and structure of the descending limb, namely, that they
were made on tubules not fully developed. Yet I have observed a
number of loops in series of sections of kidneys from half-grown and
full-grown cats and rabbits, situated near the apex of the Malpighian
pyramid, the epithelial lining of which consisted of cubic cells with
granular protoplasm, while I have not found clear evidence of a loop
lined with flattened epithelium. Complete reconstructions of a number of
uriniferous tubules from the kidneys of. adult animals are necessary to

88 Development and Shape of Uriniferous Tubules

give conclusive answer to this question. My own observations, based
on reconstructions, are therefore a confirmation of Piersol’s, who states,
“that the relative length of the narrow part of the loop—‘ descending
limb *—and the broader portion vary considerably, but that almost with-
out exception the transition from the conspicuously narrow tube, lined
with the peculiar spindle epithelial cells, takes place in the descend-
ing limb, often at a considerable distance before the loop itself is reached,
so that both limbs in the vicinity of the loop itself are of the same
diameter and lined with the same kind of epithelium.”

The junctional tubule, the continuation of the distal convoluted por-
tion, after leaving the coi] complex, may pass as a relatively straight
tubule or one showing a number of irregularities, for a shorter or greater
distance through the cortex, ultimately arching toward the collecting
tubule. The concave side of such an arch is turned toward the coil
complex formed by the proximal and distal convoluted portions, this
lying in fact between the junctional tubule and the collecting tubule to
which a given uriniferous tubule is attached. This has also been ob-
served by Stoerk.

In giving thus an outline of the course and the relations shown by
the different parts of uriniferous tubules, it should be understood that
this can be given only in a general way. For as concerns the form and
relations of the proximal and distal convoluted portions and their rela-
tion to the Malpighian corpuscle and its relation to the descending and
ascending limbs of Henle’s loop, each tubule presents certain slight
differences and variations. This the models reproduced will show. No
two of the tubules reconstructed are exactly alike or even very nearly
so. The several parts of the various tubules present certain character-
istic relations, which; though varying more or less, are still to be recog-
nized. The relation of the entire uriniferous tubule to the collecting
tubule to which each is attached also varies, and this is quite naturally
more true when considering tubules showing later stages of develop-
ment. In the earlier stages of development, as may have been seen
from the figures given of these, the proximal convoluted tubule leaves
the Malpighian corpuscle from its inner side and passes behind the
upper end of the loop of Henle and, after forming several curvatures,
comes forward between the collecting tubule and the coil complex to
reach again the neighborhood of the Malpighian corpuscle and to pass
toward the medulla. The loop of Henle, as it leaves the coil complex,
is thus in front of the Malpighian corpuscle or the first part of the
proximal convoluted portion. As the proximal and distal convoluted
portions elongate and form more coils, the whole tubule is often turned

G. Carl Huber 59

on its axis to a greater or less extent, so as to bring the upper end of
the loop of Henle more to the inner side of the tubule, nearer to the
collecting tubule to which it is attached—in closer relation to the me-
dullary rays. This brings the Malpighian corpuscle from a lateral to a
more anterior position and the first part of the convoluted portion from
a posterior to a more lateral position. This is shown by a number of
models presenting the later stages of development.

Stoerk has called attention to the fact that the Malpighian corpuscles
and uriniferous tubules are generally described as presenting in cross
sections a round form. ‘This, he states, is, however, not true (at least
for foetal life), as his models show that they present many irregulari-
ties. This is to a certain extent true. The Malpighian corpuscles are
rarely of exactly spherical form, but show here and there more or less
marked depressions and elevations, or they may present a distinctly
oval form. The proximal convoluted portion, especially of older stages,
is often of a quite regular cylindrical form, now and then for a distance
somewhat flattened, but presents as a rule throughout a quite uniform —
thickness. The tubule of the distal convoluted portion presents often
quite marked irregularities of diameter, segments of relatively large
diameter being separated by such as show a much smaller diameter; as
the tubule presents now and again a sharp bend at one or the other
region showing a relatively small diameter, this portion of the tubule
may present a quite irregular form. Stoerk has also described and fig-
ured small cecal appendages attached to the distal convoluted portion;
these I have not observed in my reconstructions.

In giving the position of the different parts of the uriniferous tubule
in the kidney substance, it is generally stated that the proximal and dis-
tal convoluted portions, with the Malpighian corpuscle, are situated in
the cortex between the medullary rays, forming the greater part of the
cortex. This is confirmed by reconstructions. Further that the distal
end of the proximal convoluted portion—the end segment of Argu-
tinsky—enters a medullary ray and with a portion of the descending
limb of the loop remains in the ray until the medulla is reached. This
is true only for certain tubules, for those the Malpighian corpuscles of
which are situated in the cortex above the lowermost two or three tiers
of corpuscles, that is, for tubules which present a distinct end segment.
The descending limb of the tubules, the corpuscles of which are situ-
ated in the lowermost two or three tiers which present indistinct end
segments, passes almost directly into the medulla and cannot be said to
enter a medullary ray. The loops of Henle for the greater part are
found in the medulla, some, as stated, barely reaching the boundary

90 Development and Shape of Uriniferous Tubules

zone, others (the proportions I am not able to give) remaining entirely
within the medullary rays. What has been said of. the upper end of
the descending limb of the loop of Henle applies also to the upper end
of the ascending limb. The junctional tubule passes through the cortex
between the rays until it arches over to reach the collecting tubule in
which it terminates.

I have in connection with Figs. 21 and 22 spoken of the relations, as
seen in sections, shown by the coil complex and the Malpighian cor-
puscles of any given tubule. A study of the sections from which the
models were made has not enabled me to formulate any general law
concerning this relation. This depends, as may readily be seen, even
in what may be regarded as very favorable sections, entirely on the
direction of the sections. Frequently the greater part of the coil com-
plex formed by the proximal and distal convoluted portions of a given
tubule may be cut in a series of sections without cutting the Malpighian
corpuscle of said tubule. The relations of coils of juxtaposed tubules is
such that one or more of the loops, especially of the proximal convoluted
portion of one tubule, may penetrate for a longer or shorter distance
into the coil complex of another tubule, so that in sections the former
might readily be interpreted as belonging to the latter. Stoerk gives in
Fig. 13 of his article a portion of a section of a kidney showing nephritis
hemorrhagica, certain tubules of which contained red blood corpuscles;
this enabled him to select the sections of Malpighian corpuscle and
tubules belonging to one uriniferous tubule. The section reproduced
shows a characteristic relation of corpuscles and tubules cut, such as is,
however, not frequently met with, if I may judge from the sections
from which my reconstructions were made. The relations there shown
are more nearly approached in Fig. 22 and also in the sections from
which model A of Fig. 19 was made than in other series of sections used
in modelling the other older developmental stages presented.

It is somewhat hazardous to attempt an estimate of the length of a
fully developed uriniferous tubule, without the possession of models
showing them in full development. Since, however, the proximal and
distal convoluted portions of certain uriniferous tubules attain approxi-
mately their full development before the birth of the embryo (cat and
rabbit), and since the length of the loop of Henle of these tubules de-
pends on the width of the medulla, as they extend through or nearly
through the medulla, an estimate of the length of a fully developed
tubule seems justifiable. The length of tubules must necessarily vary,
since the loop of Henle, which for the majority of tubules forms their
greatest part, must vary in length, since they terminate at different

G. Carl Huber ~ 9]

levels in the medulla. An approximation of the length of uriniferous
tubules may, therefore, be most readily made for those tubules the Mal-
pighian corpuscles of which are situated in the deepest portion of the
cortex, the loops of Henle terminate presumably near the apex of the
Malpighian pyramid. The most fully developed tubule, showing a typi-
cal arrangement of parts, that I have reconstructed, is the one shown in
Fig. 20. This is from the kidney of a cat embryo, obtained a few days
before birth, and measures 9 mm. In one of the sections of this kidney
containing nearly the entire length of the distal limb of the loop of
Henle of this tubule, the width of the medulla is approximately 2.22 mm.
(the junction of the cortex and medulla does not form a straight line),
very nearly one-half the entire length of the loop of Henle (4.85 mm.).
In sections passing through the middle of the Malpighian pyramid of
the kidney of a full-grown cat, the medulla has a width of approximately
12 mm. Assuming, therefore, that the loops of Henle of certain of the
uriniferous tubules of this kidney, the Malpighian corpuscles of which
are situated in the deepest portion of the cortex, traverse the entire
width of the medulla, and one is justified in doing this, since loops are
seen In section near the very apex of the pyramid; such a loop would
have a length of approximately 24 mm.; add to this the length of the
proximal convoluted portion, which may be given at 2.5 to 3 mm., the
distal convoluted portion about 1 mm., and the junctional tubule about
2 mm. (the width of the cortex in the sections from which the measure-
ments for the medulla were given varies from 2.5 mm. to 3 mm.), the
entire length of such a uriniferous tubule would be approximately 3 cm.
Other tubules, the loops of Henle of which do not pass so deep down in
the medulla, would show a correspondingly shorter length, the shortest
ones measuring, if I may be allowed to estimate, 10 mm. to 12 mm.

V. Ebner quotes Schweiger-Seidel as stating that the proximal convo-
luted portion of a uriniferous tubule forms about one-fourth to one-
fifth of the entire uriniferous tubule, while the distal convoluted por-
tion, which is shorter than the proximal portion, is about one-seventh
of the latter’s length. As concerns the relative length of the proximal
convoluted portion, this may be regarded as correct for tubules with
relatively short loops of Henle. In tubules with relatively long loops of
Henle it forms a relatively shorter portion—from one-eighth to one-
tenth of the entire tubule. So far as I am able to judge from my own
models and from other observations, the distal convoluted portion pre-
sents a length which is about one-third to one-fourth that of the proxi-
mal convoluted portion. If other estimates of the length of uriniferous
tubules or of parts thereof are to be found in the literature, they have
escaped my notice.

92 Development and Shape of Uriniferous Tubules

It is not my purpose to enter here on a discussion of the epithelium
lining the different parts of a uriniferous tubule. For this the reader
is referred to Disse’s quite recent account, he himself having made note-
worthy contributions to this point. The observations made on the sec-
tions from which the models were made lead me to recognize three dis-
tinct varieties of epithelium in a uriniferous tubule:

1. The epithelium lining the entire proximal convoluted portion, to
the region in the upper part of the descending limb, where the tubule is
quite rapidly reduced in size, where it assumes the shape and structure
of the narrow portion of the limb. The epithelium of the proximal
convoluted portion is sufficiently known to obviate an especial description
of it.

2. The clear, flattened epithelium of the descending limb of the loop
in the region in which it presents a small diameter. I have gained the
impression that too much emphasis is put on the statement that the
cells of this epithelium are thicker in the region of the nucleus, thus
bulging into the lumen of the tubule, and as the nuclei do not lie oppo-
site, but alternate in the course of the tubule, its lumen is not of the
cylindrical form, but irregular—wavy as seen in sections. In sections
from material which I have regarded as showing good fixation, the
lumen of this portion of the uriniferous tubule presents a much more
regular outline than in sections from material not so well fixed, the cells
presenting throughout a more uniform thickness. The suggestion is
therefore made that the description, as generally given, refers to prepa-
rations which do not present the correct structure of this portioh of the
uriniferous tubules. This point is reserved for further investigation.

3. The last portion of the descending limb, the loop itself, the ascend-
ing limb, the distal convoluted portion, and the junctional tubule to
near the collecting tube present an epithelium which shows great simi-
larity of form as well as structure. The lining cells show a form which
is irregularly cubical or short columnar. Disse describes the epithe-
lium of the ascending limb as showing indistinct cell boundaries, while
these are more distinct in the epithelium of the distal convoluted portion ;
the protoplasm of the latter, although granular, is also described as
being clearer. In both regions the cells show an outer dark zone which
is finely striated and an inner zone which is lighter, the nuclei being
placed at the junction of the two zones. In my own preparations the
epithelium lining the ascending limb and the distal convoluted portion
do not respectively present a structure which differs to a degree making
it necessary to group them separately, and I have, therefore, placed them
under one head. The uriniferous tubule, for a short though variable

“(16 “SLA FOGL) WIBASBIP 8.410098 VWs (OL STA “PH ITA “Teu 19981q ‘ouBs1OszyoOo[qosoy pun U1By ‘Ueqosuey sap WeBuUY ‘p Yonqpuewy s,ueqejopaeqg)

UIBIFPIP SASSI “MW ! (1S “SLA “FOGL “{410098) orNsy 8.15[0H ‘7 SE “BLA ‘Waqea BJO a[nqny sNodojTULAIN B JO WIRATVIP S1JBIOLVH “OD (ZO “FIA “PA ‘ITI ‘OI TRH
o181q ‘UaTOSuEeW Sep e1ys[PQOMOH "p YONGPuUBH{ SAOFTI9M) WeiseIp s19euqgy UOA ‘_ ! (Bl “SILA ‘FOGL “A19079 AQ UDALS SB) TIBISBIP S,[aplog-—aSIOMYoS ‘pr

“SINGNL SNOYSSINIYN SAHL SO WHOS SH1L SO Q14SH SNOILdSONOO LN3Y355IG JHL ONIMOHS SWVHDVIG €2 ‘Pld

O4 Development and Shape of Uriniferous Tubules

distance, before reaching the collecting tubule, presents an epithelium
which is hike that of the small collecting tubules. This short segment
may be spoken of as belonging to the collecting tubule or with equal pro-
priety as forming part of the uriniferous tubule proper, as it is difficult
to state with any degree of certainty whether it develops as an outgrowth
of the collecting tubule or is differentiated with the other parts of the
uriniferous tubule from the renal vesicle.

Before giving by way of a diagram or scheme the shape of a urinifer-
ous tubule and its relation to a collecting tubule, as this presents itself
to me, brief mention may yet be made of a number of diagrams of urin-
iferous tubules found in the literature, each of which presents certain
characteristics which make it different from the others selected. These
I have grouped under Fig. 23. The different conceptions of the form of
a uriniferous tubule held by the authorities presenting the diagrams here
given and their departure from the conception of the form of a urinifer-
ous tubule as presented in Fig. 24 may be most readily expressed in this
graphical manner.

In A, Schweiger-Seidel’s diagram, is given inaccurately the relative
position of the Malpighian corpuscle and proximal and distal convoluted
portions, more especially the distal convoluted portion which is separated
too much from the proximal convoluted portion. In B, Von Ebner’s
diagram, an attempt is made to bring the distal convoluted portion in
relation with the Malpighian corpuscle (accepting Hamburger’s obser-
vations) and to show that the first part of the proximal convoluted por-
tion extends toward the cortex (Golgi). The diagram presents inaccura-
cies in the relations given to the Malpighian corpuscle and the proximal
and distal convoluted portions. In C, Haycraft’s diagram, of a urin-
iferous tubule of a rabbit, he presents, as is obvious, deductions drawn
from the study of a tubule showing an early stage of development, shown
in his Fig. 10. This, as I have stated, very probably presents parts. of
two tubules, there sketched as one. His diagram is wrong in the position
given to the Malpighian corpuscle and as to the length, shape, and
relative position of the proximal and distal convoluted portions. D, one
of Golgi’s figures, shows a uriniferous tubule at a relatively early stage of
development. This is not a diagram, but the figure is introduced, since
it shows quite correctly the relations of the different parts of a uriniferous
tubule. #H, Disse’s diagram, one of the most recent at hand, gives incor-
rectly the relations shown by the ascending and descending limbs of
Henle’s loop to the Malpighian corpuscle and the course of the first part
of the proximal convoluted portion. /’, Stoerk’s diagram, based on recon-
structions of uriniferous tubules, presents a number of inaccuracies, as is

G. Carl Huber 95

Fig. 24. Diagram of three uriniferous tubules and their relation to a collect-
ing tubule; A, of a tubule, the Malpighian corpuscle of which is situated in the
lowermost portion of the cortex; B, about the middle of the cortex; C, in the
outer portion of the cortex. m, Malpighian corpuscle; v, vessel porta; n,
neck; pc, proximal convoluted portion; es, end segment; dl, descending limb,
al, ascending limb of the loop of Henle; dc, distal convoluted portion; j,
junctional tubule; c, collecting tubule (Huber).

96 Development and Shape of Uriniferous Tubules

shown by my own observations on models of uriniferous tubules showing
much older stages of development than the oldest tubule reconstructed by
him. In this diagram, the Malpighian corpuscle is relatively too large, the
course given to the first part of the proximal convoluted portion is too reg-
ular; the proximal convoluted portion is relatively too short when com-
pared with the length given to the distal convoluted portion, but especially
wrong is the diagram in the presentation given of the descending limb
of Henle’s loop. This, as given by Stoerk, shows essentially the same
diameter as the proximal convoluted portion (in his description he gives
it a similar epithelial lining), disregarding entirely that portion of the
descending limb of the loop which shows a flattened epithelium.

In the diagram of Fig. 24 the loop itself of the loop of Henle is given
as formed by a tubule the same as that of the ascending limb, the transi-
tion from the thinner to the thicker portion of the hmb as occurring in
the lowermost part of the descending limb. This is in conformity with
my models. In discussing this point, I have stated that this observation
was given with some reservation. I have, therefore, not called especial
attention to certain of the loops shown in the diagram A, B, and EF of
Fig. 23, which are formed by the thinner portions of this structure.

In conclusion, I desire to thank Julius H. Powers and Edward G.
Huber for assistance given me in the reconstructions, and Elton P.
Billings and Frederic G. Johnson for assistance rendered in making the
illustrations,—students of medicine in the Department of Medicine of the
University of Michigan.

LITERATURE CITED.*

Batrour, F. M.—On the Origin and History of the Urinogenital Organs of
Vertebrates. Jour. Anat. and Phys., Lond., Vol. X, 1876.

* BORNHAUPT, TH.—Zur Entwickelung des Urogenitalsystems beim Htihnchen.
Inaug. Diss., Dorpat, 1867.

Braun, M.—Das Urogenitalsystem der einheimischen Reptilien, entwicke-
lungsgeschichtlich und anatomisch bearbeitet. Arbeiten a. d. zool.-
zoot. Inst., Wiirzburg, Bd. IV, 1878.

CHIEVITZ, J. H.—Beobachtungen und Bemerkungen iiber Sdugethiernieren.
Arch. Anat. u. Phys. Supplement Band, 1897.

CotsBerc, A.—Zur Anatomie der Niere. Centralb. med. Wissenschaft., Vol. I,
18638.

Disse, J—Harn und Geschlechtsorgane. V. Bardeleben’s Handbuch der Anat-
omie des Menschen, Bd. VII, Erster Teil, Jena, 1902.

v. Epner, V.—KOlliker’s Handbuch der Gewebelehre des Menschen. Bad. III,
1 Halfte, Leipzig, 1899.

Emery, C.—Recherches embryologiques sur le rein des mammiféres. Arch.
ital. Biol., Tome IV, 1883.

* Contributions not seen in the original are marked with an asterisk.

G. Carl Huber ; 9%

FEeLix.—Entwickelung der Nachniere. Hertwig’s Handbuch der vergleich-
enden und experimentellen Entwickelungslehre der Wirbelthiere,
Page 302, Part 20, 1904. (Not yet completed.)

FURBRINGER, M.—Zur vergleichenden Anatomie und Entwickelungsgeschichte
der Excretionsorgane der Vertebraten. Morphol. Jahrb., Bd. IV,

1878.
GERHARDT, V.—Zur Entwickelung der bleibenden Niere. Arch. mikr. Anat.
LVII, 1901.

*GoLa1, C.—Annotazioni intorno all’ istologia dei Reni dell’ Uomo e di altri
Mammiferi. Rendiconti delle R. Accad. dei Lincei, Vol. V, Serie I,
1889.

GreGory, E. R.—Observations on the Development of the Excretory System in
Turtles. Zool. Jahrbticher Abth. f. Anat. u. Ontog. d. Thiere, Bd.
XIII, 1900.

HAMBURGER, O.—Ueber die Entwickelung der Saugethierniere. Arch. Anat. u.
Phys. Supplement, Bd. 1890.

Havueu, E.—Ueber die Anatomie und Entwickelung der Niere. Anat. Hefte,
Bd. XXII, 1903.

Haycrart, J. B—The Development of the Kidney in the Rabbit. Internat.
Monatsschr. Anat. u. Phys., Bd. XII, 1895.

HERRING, P. T.—The Development of the Malpighian Bodies of the Kidney
and its Relation to the Pathological Changes which Occur in them.
Journ. Path. and Bact., Vol. VI, 1900.
HILDEBRAND.—Weitere Beitrage zur pathologischen Anatomie der Nierenge-
schwilste. Arch. f. klinische Chirurgie, Vol. XLVIII, 1894.
HorrMann, C. K.—Zur Entwickelungsgeschichte der Urogenitalorgane bei den
Reptilien. Zeitschr. f. wissensch. Zool., Bd. XLVIII, 1889.

HortroLtes, Cu.—Recherches histologiques sur le glomérule et les épithéliums
du rein. Arch. de physiol. normale et pathol., 1881.

JANOSIK, J.—Histologisch-embryologische Untersuchungen tiber das Urogen-
italsystem. Sitz. Ber. Akad. Wiss. Wien., math.-nat. Kl., Bd. XCI,
III Abth., 1885.

*KALLAY, A.—Die Niere im frtihen Stadium des Embryonallebens. Mitth. a. d.
embryol. Inst. d. K. K. Univ. in Wien, Neue Folge, Heft I, 1885.

KEIBEL, F.—Ueber Entwickelung des Urogenitalapparates von Echidna. Anat.
Anz. Erganzungsheft, Vol. XXIII, 1903.

KOLLIKER, V.—Entwickelungsgeschichte des Menschen und der héheren Thiere.
Leipzig, 1879.

KoLLMANN, J.—Lehrbuch der Entwickelungsgeschichte des Menschen. Jena,
1898.

Kuprrer, C.—Untersuchungen tiber die Entwickelung des Harn- und Ge-
schlechtssystems. Arch. mikr. Anat., Bd. I, 1865. :

Lowe, L.—Zur Entwickelungsgeschichte der Sdugethierniere. Arch. mikr.
Anat., Bd. XVI, 1879.

Meyer, Ertcu.—Ueber einige Entwickelungshemmungen der Niere. Miinchen.
med. Wochenschr., May, 1903.

Minot, C. S.mHuman Embryology. New York, 1892.

Nacet, W.—Ueber die Entwickelung des Urogenitalsystems des Menschen.
Arch. mikr. Anat., Bd. XXXIV, 1889.

98 Development and Shape of Uriniferous Tubules

Prersot, G. A.—Note on Henle’s Loops of the kidney. University Medical
Magazine, March, 1889.

Pyr, W.—Observations on the Development and Structure of the Kidney.
Journ. Anat. and Phys., Lond., Vol. IX, 1875.

*RemMAK, R.—Untersuchungen tiber die Entwickelung der Wirbelthiere. Berlin,

1855.
Riesert, H.—Ueber die Entwickelung der Glomeruli. Arch. mikr. Anat., Bd.
XVII, 1880.

Ueber die Entwickelung der bleibenden Niere und iiber die Entstehung
der Cystenniere. Verhandl. d. Deut. path. Gesellsch., Berlin, 1900.

RieprE, K.—Untersuchungen zur Entwickelung der bleibenden Niere. Inaug.
Diss. Miinchen, 1887.

*RIEDEL.—Entwickelung der Saugethierniere. Unters. aus dem anat. Inst. zu
Rostock, 1874.

Riicxert, J—Entwickelung der Excretionsorgane. Ergebnisse Anat. u. Ent-
wickelungsgesch., Bd. I, 1891.

ScHREINER, K. E.—Ueber die Entwickelung der Amniotenniere. Zeitschr.
wissensch. Zool., Bd. LX XI, 1902.

Scuuttze, O.—Grundriss der Entwickelungsgeschichte des Menschen und der
Sadugethiere. Leipzig, 1897.

SEDGEWICK, A.—Development of the Kidney in its Relation to the Wolffian
body in the Chick. Quart. Journ. mikr. Sc., Vol. XX, New Series,
1880. .

SrorrK, O.—Ueber Nierenveraénderung bei Lues congenita. Wiener klin.

Wochenschr., XIV Jahrg., 1901.
Beitrag zur Kenntniss des Aufbaus der menschlichen Niere. Anat.
Hefte, Bd. XXIII, Heft 2, 1904.

SrrRAHL.—Entwickelungsgeschichte und Missbildungen der Nieren, in Kiister,
Chirurgische Krankheiten der Nieren. Lieferung 52 B, Stuttgart,
1902.

THAYS} A —Die Entwickelung der Niere. Vorlaufige mittheilung. Cen-

f. med. Wissensch., Bd. XI, 1873.

Totpt, C. wuntersuchungen iiber das Wachsthum der Nieren des Menschen
und der Sidiugethiere. Sitz. Ber. Akad. Wiss. Wien, math.-nat. KI].
Bd. LXIX, III Abth., 1874.

WALDEYER, W.—HBierstock und Ei. Leipzig, 1870.

Weser, S.—Zur Entwickelungsgeschichte des Uropoetischen Apparatus bei
Sdiugern, mit besonderer Beriicksichtigung der Urniere zur Zeit des
Auftretens der bleibenden Niere. Morphol. Arb., Bd. VII, 1897.

WIEDERSHEIM, R.—Ueber die Entwickelung des Urogenitalapparatus bei Cro-

codilen und Schildkréten. Arch. mikr. Anat., Bd. XXXVI, 1890.

fs
Mh

‘

ay
\ My
:

Mobis ii
Hania
40

NG

MN a

i

t
rin

Misael
RANI
enh |

J

2

ath

om i
iN
+) ‘ ‘\
Ta 1 La A OM
Brin Ny oa

iN LAR

‘TRS

AN

OH sul
iN } nan ra

MBL WHOI Librar - Serials

WANT

WHSE 04732

Ee, ot RS EX re ere WG
SE a Ce ee ee

4 Bi F x
tS Fk nt er ign tine Si

te oe Se Sn,

lise

LSAT

Sen

ae

Pg,

ee ena.
ee

Ne OE A NE Pe oe rete Pee REE OD OPA PLOT OFA TOR Ai ARAB GIS

eee Pe

wears

Ae rae nad ol gant ces me
oo gm eR EE RR Ne ep eC Fear a ore

pases oe
ae

seca aie ge

5 mi
oe
See
ee wien oS
Oe eco opt
POE
ee