Anatomical Record 15 (1918-19)

From Embryology
Embryology - 18 Jan 2020    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic Textbook" and "Historic Embryology" appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms and interpretations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)
Anat. Rec.: 1 - 1906-08 | 2 - 1908 | 3 - 1909 | 4 - 1910 | 6 - 1912 | 7 - 1913 | 8 - 1914 | 9 - 1915 | 10 - 1915-16 | 11 - 1917 | 12 - 1917 | 13 - 1917 | 14 - 1918 | 15 - 1918-19 | 16 - 1919 | 19 - 1920 | 21-22 - 1921 | 23-24 - 1922
Links: Historic Embryology Papers
Other Historic Journals: Amer. J Anat. | Anat. Rec. | J Morphol. | J Anat.


Editorial Board

Irving Hardesty Tulane UnlversUy

Clarence M. Jackson Unlvcralty of MlDnesota

Thomas G. Lee Uclverslty of MiDQesota

Frederic T. Lewis Harvard Unlvenlty

Warren H. Lewis John! Hopkins Unlvenlty

Charles F. W. McClure Princeton University

Wiluau S. Milixr University of Wisconsin

Florence R. Sarin Johns Hopkins University

George L. Streeter University of Michigan G. Carl Huber, Managing Editor

1330 Hill Sueet. Ann Arbor, Michigan






Kaethe W. Dewey. a contribution to the study of the pathways of the cerebrospinnl fluid and the choroid plexus. Nine figures 1

II. E. Jordan. A description of a case of false hermaphroditism. One figure. 27

II. E. Jordan. The histology of lymph, with special reference to platelets 37

William Snow Miller. A differential injection massfor use with stercoroentgenograms . 47

William F. Allen. Surface view of.injected intestinal villi. One figure 49

William F. Allen. An inexpensive .microscopic projection and drawing apparatus. One figure 53


Henry Erdmaxn Radasch. A duplication or branching of the neural canal. Twelve figures 57

Wilhtjr C. Smith. On the process of disappearance of the conua arteriosus in Teleosts. Sixteen figures 65

Wayne Jason Atwell. The development of the hypophysis of the Anura. Eighteen figures 73

HoRTON R. Casparis Lymphatics of the omentum. One figure 93

Matthew Marshall. An unusual right lung 101


Hal Downey. Further studies on the reactions of blood- and tissue-cells to acid collodial dyes. Six figures (two plates) 103

K. D. Conodon. The embryonic structure of avian heart muscle with some considerations regarding its earliest contraction. Ten figures 135

Eliot R. Clark and Eleanor Linton Clark. Phagocytosis of carbon and carmine granules in the transparent tails of tadpoles. One plate 151


Richard E. Scammon and Edoar H. Morris. On the time of the post-natal obliteration of the fetal blood-passages (foramen ovale, ductus arteriosus, ductus venosus). Three figures 166

Wayne J. Atwell and Ida Sitler. The early appearance of the anlagen of the pars tuberalis in the hypophysis of the chick. Five figures 181

F. A. McJuNKi.v. The identification of endothelial leucocytes in human tissue. Third report of studies on the mononuclear cells of the blood. Two figures 189

Frank Blair Hanson. The scapula of Tragulua. Nine figures . 197


Wamu NaKaiiara. Some obsorviitions on the KrowiiiK oncytes of the stoncfly, Pcrla immnrRirinlA Sny, with R|ivoinl regarii to the origin iind function of the nucleolar trurlurt's. Nino figuroa 203

ll\Hiii-.>s U Hu.sT. Variability in the common carotid arteries of the domestic cat. IV. limm-s 217

Makuison It IIVNT. AbBciirc of one kidney in the domestic cat. Two figures 221

AtH)Lr H. S< iirLTX. Ob.ierval ions on the cnnalis basilaris chordae. Three figures 225

Mleanoh I.intdn Ci.akk and Eliot R. Clakk. On the reaction of certain cells in the tadpole's tail toward vital dyes. Four figures 231

Kdward I'iiclps Allih, Jr. On the origin of the hyomandibula of the Teleostomi. One figure 257


RiruARD K. PcAMUoN. On the development and finer structure of the corpus adiposum biircao Nine figures 267

Frank Ulaih IIan.hon. Nerve foramina in the pig scapula. A peeuliar relation existing b«'iworn the dorsali.s branch of several spinal 'nerves and the suprascapula in the pig Twenty-one figures 289

Florence May Aijop. The effect of abnormal temperatures upon the developing nervous system in the chick embryos. Thirteen figures 307

Proceedings of the American Society of Zoologists. Sixteenth Annual Meeting 333

Proceedings of the American Society of Zoologists. Abstracts 341


JoMKi-ii M. Tni'KiN<iER. The anatomy of a dicephalie pig. Monosomus diprosopus. Ten tigure.i 359

\. (!. I'oHLUAN. Double ureters in human and pig embryos. Three figures 369

A. <i. roHL.MA.v. The Urtc of a simple graphic method of recording the relations in serial sections, particularly for use in teaching embryology. Four charts 375

\. Ct. I'oiiLMAN. The use of bayberry wax in hardening paraffin blocks. One chart.... 385

.\. O. Poiii.Mw. A modification of the Born paper-wax reconstruction plate 389

II. K. JoRDA.s. The histogenesis of blood-platelets in the yolk-sac of the pig embryo... 391


A Contribution To The Study Of The Pathways Of The CereBrospinal Fluid And The Choroid Plexus

Kaethe W. Dewey

The Research laboratory of the College of Dentistry, University of Illinois

Nine Figures

In experimental work for demonstrating the presence of Ij'mphvessels in the dental pulp,' use was made of vital staining as a supplementary attempt to obtain some histological information regarding endothelial lined perivascular or other lymph-channels. There are observations recorded in the literature which are vaguel\- suggestive of such a possibility. Of particular interest from this standpoint are statements made by Brass* and by Evans,' that the endothelial cells lining the sinuses and h-mphchannels of lymph-glamls are always intensely stained, and by Kiyono* that vitaliy stained endothelial cells are found "in the lymph spaces of the interstitial tissue of the lung."

A large number of rabbits and a few dogs and cats were injected intraperitoneally and chiefly intravenously with trypan blue and lithium carmine, and microscopic preparations made from practically all organs. Chief attention was paid to those vitally stained cells, which occur in the connective tis.sue of the various organs and which have been classified as endothelial cells, reticuhmi cells, resting wandering cells, clasmatocytes, rhagiocrines, histiocytes, pyrrol cells. The impression witli me from

' Dewey, Kaethe, and 1". B. Noyes, "A Study of the lymphatic vessels of the dentiil pulp." Dental Cosmos, 1917, 58, 43C.

' Brass, "Cber physioloRisohe PiKmentatilafcerunR in den Kapillarcndothelien des Knochenmarks." .Vrch. f. mikrosk. .•Vnat.,'1913, 81, 61.

'Evans, "The macrophages of mammals," Amer. Journ. of Phvs., 1915, 37, 242.

• Kiyono, "Die vitale Karminspeicherung. Jena, 1914.

thosE studies wjus that tlu\se colls arc intimately related to the iymphatir system. The regularity with which they occur in the same regions within the organs, the apparently systematic arrangement in wliich they present themselves, the definite course wliich so oft^'ii tlu-y seem to jmrsue, an; striking features and tend to «lispel the impression that they are cells indiscriminately dispersed within the connective tissue, resting wandering ceils waiting for a chance stimulus to arouse them to functional activity and l<K-omotion. When the uijections have been continued for some time the cells are large and crowded with stained granules in their cytoplasm and their long processes: the latter not infre<iu«'ntly almost touch one another so that nearly continuous cell tracts are formed, running along the blood-vessels and independently. From these observations, in conjunction with studies of the literature on the lympiiatic distribution throughout the body, I have been led to suggest the possibility that these vitally .stained cells are for the most part endothelial cells lining the l>nnph-channels within organs and that \ital .staining may furnish us the means of demonstrating lymph-channels in organs and regions where they are not yet definitely known or are difficult to demonstrate by the usual methods of injections.

.\s to the central nervous .system it has been and still is a much debated (jue.stion whether the perivascular spaces should be regarde<i as lymph-vessels, the decision hinging on this point whether they have an endothelial lining. It is a general observation that the favorite location of vitally .stained cells within the connective tissue in all organs is the immediate neighborhood of the blood-\e.ssels. It was therefore to be ascertained whether vitally staineil cells are also found about the cerebral ve.s.sels. With this point in view, the brain with its membranes and the choroid plexus were studied. The dura and pia, and the choroid plexus of rabbits were carefully removed and sjiread on glass slides; the choroid plexus of dogs and cats was reduced in thickness by the freezing microtome. Of the membranes the dura always pres<-nts the greatest abundance of vitally .stained cells; it is also one of the tissues in which a ilefiniteness in the arrangement of these cells is pronounced. They form long continuous tracts, niiminK aloiin the l^lood-vo.ssols, but luurc often apparently in entirely independent directions. Two-, three-, and four-cell tracts may run parallel to one vessel, from which they may he entirely separated. Their independent course is also very conspicuous in regions wher<> tlie blood-vessels form Ktops; here the blue- or red-cell tracts sweep across them in continuous straight lines. Compared with the enormous supply of these cells in the dura, those of the pia arachnoid are .scanty and more slender. There are always areas where there are no cells or only a few here and there, and are slender and barely- visible with the low-power lens. But there are always patches where they are in rather dense collections, are larger and more deeply stained, and from these cell groups lines of cells run in all direcJtions. .\fter lithium cannine injections the number of vitally stained cells always seem increased and the individual cells are larger. Frequently they have very long processes which aUnost touch one another, ami thus very definite tracts are formed which ari)i1ranly, it seems, follow some of the blood-ves.sels, while others and especially all tlie larger vessels are unaccompanied by such cells. The irregular distribution and the presence of cell patches is a prominent feature in all preparations of pia. In the choroid plexus the ependymal cells covering the vascular convolutions are all densely tilled with a finelj' granular .stain and form a continuous blue or red layer around the tufts. In the i)ial connective ti.ssue of these and the velum interpositum there are always numerous vividly stained cells like those in the dura and pia. They do not form absolutely definite cell tracts, nor do they follow the blood-vessels closeh'; in fact, they are generally at some distance from them. They are quite evenly distributed throughout and anything resembling the cellular patches of the pia does not occur here. In the intensity of the staining, the large number and the distribution of these cells, the velum interpositum and the pial tissue about the tufts n*semble more the dura than the pia.

Being inclined to consider the vitally .stained cells as endothelial cells lining the lyjuph-ehannels and lymph-spaces I suggest that the cell complexes in the pia arachnoid might be looked upon as thn stnu'turf> ilcsc-ribcd hy W coil :is urachiinid \illi, strurturps normally present in the anic-hnoid of anbniils arul man, and through which, according to hlin, the cerebrospinal fluid is filtered from the arachnoid spaces into the venous sinuses. In view of the presence of such vitally stained cells in the membranes of tlie brain and the choroid plexus, clearly indicating the passjige through and into them of fluid containing the staining solution, it is the more striking that the brain tissue fails to show any vitally stained cells whatsoever. The only other exception from the general observation is the fet-us which also remains unstamed. CJoldmann' explains this fact b}' the staten\ent that the dye derived from the plasma is filtered off by the cubical epithelial cells of the j^lexus and the .syncytial cells of the placenta and thus kept off from the cerebrospinal and the anmiotic fluid. Those two tis.sues form a protecting limiting membrane to the brain and fetus. TrjTian blue, according to him, is a substance extremely toxic to the nervous tissue. He claims that while 50 cc. of a 1 per cent st)lution may be injected in a single into a rabbit intravenously without any unt-oward effect, J cc. of a J per cent solution of the stain injected into the spinal or cerebral subarachnoid space produces tonic and clonic convulsions within ten to twenty minutes and death in coma within one or two hours. In these cases he found that the ganglif)n cells had a dilTu.sely stained cytoplasm and a .stained nucleus, indicating that death of the cells had taken place. But he also (U'.>icribed gemiine vital staining of cells within the connective ti.ssue of the pia and along the vessels of the pial septa. (These cell.s are called in all his publications 'pyrrol cells,' because he first observer! typical granular vital staining of these cells after injections of pyrrol blu*-. ) "The supposition." he writes, "that

• CuiihiiiK. Wrijd iiiul \VL'({i'f«rth, "Studies on the cerebrospinal fluid." Journ. of Med llpurnrrh. 19I4, 2('., .il.

• fioldmann. 1. Die fiusscre und innuro .Sckrction des gcsuiiden und krankcn Organiamua in Lichte vitalcr I'lirbung. Bcitr. z. klin. Chir., 1909, 54, 192, iind 1012, 78, 1.

2 Kxpvrimcntcllc Untersuchungen iiber die iMinktion der Plexus choroidcii und der llirnhnut Arch, f kliii Chir . 1913. 101.,.

3 Die Vitnlfarlmnn des Centriilnirvcnsystems Berlin, 1913.

the poriviiscuhir prolongations of tho pial funnels functionate as 'lyinj)lispH('es' finds a further support in the fact that the cells which line these spaces, as well as the fine connective-tissue trabecuia trn versing these spaces, store up our vital stains in fine granules in exactly the same way as the reticulum cells in the lymph glands." These statements by Golilmann induced me to also use subarachnoid injections of trypan blue solutions not only for the demonstration of endothelial lined perivascular lymph-channels in the brain, but as well for some further infonnation about the choroid plexus. However plausible his explanation of its function as a protecting memlirane, some questions evolving from such a conception remain still obscure; for example, how is it possible that the filtering jjrocess bj' the cuboidal epithelial cells of the choroid plexus can protect the entire brain against injurious substances from the blood plasma when other vessels beside the choroidal arteries, in fact, the larger number of the cerebral vessels, pass directly into the brain substance? In what manner is the spinal cord protected which has no choroid plexuses?

The clinical results in twenty-four rabbits which I injected intraspinally and intracranially with tr>'pan blue, did not present the grave symptomatic picture which (Joldmann describes as characteristic for the rabbit. Sjinptoms occurred in a number of these anunals: but there were no fatal cases, and in several rabbits even repeated injections failed to elicit any sj'mptoms. In the absence of all grave conditions of convulsions and of coma, the results res(>mble(l much more observed in dogs. Concerning the striking dissimilarity of results observeti byGoldmann in these two animals, this author remarks hunself: "It seems to me very remarkable that in the dog the symptoms of motor irritation elicited by the dye cannot be compared in any way with those in the rabbit above all, those grave conditions

of convulsions may be lacking in the dog which are so (>X('(>edingly characteristic of the toxic effects of the dye in the ral)bit." I am inclined to believe that the sjiiiptoms do not arise so much from any toxicity of the staining substance, as that they are due to mechanical and physical factors and direct injuries of the ncrvdus tissue in the process (if the injections, which jire not so eiisily coiitrollei! in the rabbit. The most severe paralysis oceurntl in n control rabbit after the first injection of distilled wiitor: the pjiralysis persistetl, i)ut the anunal recovered otherwise and remained well in spite of a second injection of distilled watiT and two injections of trypan blue.

The injections were made, as a rule, in the lumbar and lower thoracic region: eight rabbits were injected into the cerebral subarachnoid space. After having anaesthetized the aniinal with ether, the needle was jiasseil through the skin and juuscles into the spinal cavity by way of the interspinous foramina, which are rather large in young rabbits. The cranial injections were nuide through the occipito-atlantoid ligament. I found soon that it was best to use a very fine needle and pass it quickly through the cord in an oblique direction and inject the fluid into the .space opposite the point of entrance. The needle's striking the bony wall makes it sure that the Huid is not forced into the nervous tissue, an accident which is not so easy to avoid if the fluid is injected wliere the needle passes through the interspinous ligament. In injections into the cerebral subarachnoid space some of the fluid was always withdrawn before injecting the staining solution. In larger anunals the drops would flow out spontaneously through the medium-sized needle; in very young rabbits a very fine needle was u.sed and here the fluid was aspirated with a medicine dropper. As to the .spinal cord, attempts to let some of the fluid escape before injecting the solution were given up as rather difficult. For this reason I cannot claim, in .spite of all practical considerations anil careful .study of the anatomic conditions of the spinal cord in the rabbit, to furnish a positive proof that the staining solution actually passed into the subarachnoid space and not into the epidural space. But the clinical and histological observations in these injections being largely the same as in intracranial injections, the controlling tests of the latter should be sufficient to justify my general statements.

In the following table the distinguishing feature for the results, niarke cells indicates a grave impainnent of their vitality, and since I claim that this is not the result of any toxic action of trvf)aii blue, but arises from other causes incidental to the injections, the presence of dilTusely stained cells is applied to represent 'negative results,' and the presence of vitally stained cells 'positive results.' There is evidence that the cells may recover from the condition in which they arc susceptible to dift'use staining. Those rabbits which were killed when serious sjanptoms arose and persisted for several hours were nearly all 'negative' cases; on the other hand, in instances where the animals presenting grave symptoms were allowed to li\o and to recover and sub.sequently were again injected with trjpan blue without any untoward effects, the results were generally 'positive.'


Inlroipinous injections Nervous symptoms of a more or less serious nature with negative results

a <

a o




b. O

z a

















Paralysis of hind legs; spnsms of muscles of mastication, excitement






Severe spinal symptoms; paralysis, motion in a circle from







right to left Severe symptoms after the fourth injection due to injury







in injecting Paralysis of hind legs and slight








spinal symptoms Transitory slight symptoms







Remark : injection was made when the animal was sucoumhing to ether anaesthesia



Intraspinal injrcltonK

Grave nervous symptoms with positive results






a j



5 »

• 2


'; l^














20 hours

Paralysis, excitement, spasms of muscles of mastication







5 days

Paralysis and spinal symptoms after the 4th injection







5 days

Paralysis and spinal symptoms after the 4th injection



1 .. .


TABLE 3 Absence of symptoms with positive results










5 5













fv ant








Membranes light








Membranes very light





1 fr-1.5



Meinbrances very light. Grave paralysi.s followed first unsuccessful attempt to inject, without any fluid entering the spinal canal. Complete recovery from this in course of following 16 days of experiment








Membranes light

+ 4




1 AHI.K 4

into the cinlerua magna




Si Si












p<r Mill







Grave symptoms due to impeded outflow from eisternu

15 16

550 620

2 2

5 0.5

1 1

3 4


Grave symptoms after the second injection

Disturbance of equilibrium, transitory



17 18

730 950

1 1

5 75


3 3

Slight transitory symptoms NystuKmus; motions of the head

+ +

19 20 21

1060 1530 1500

3 5 1


1 5


, i


5 8 4

S'o symptoms

Xo symjjtoms.

Direct injury of medulla oblongata; lastinR disturbance of equilibrium




il .


TABLE 5 Tests for the comparative offeel o( distilled water and trypan blue solutions in the

same animal


a 2




Ik o


o o

a <

z o



H u Z












ptr ctnt




2 aq. dost


Serious symptoms and


3 trypan




complete and pcrsisting paralysis of hind loRs after first injection of distillod water. Otherwise (gradual recovery



2 aq. dest.


Paralysis of hind legs af

1 trypan




ter the 3d injection. Xo other symptoms



2 aq. dest.


Xo symptoms except dul


2 trypan











Kiiwinnrul' (.. ,i.^r.rl,uu tchfthtT 'KdrtichetUfllcn' in lesioHS such as may occur

, ,. < lake up tri/pan blue when this is injected intravenously






KCMBaa or





i I















6 intrav.


No nervous man

No vitally stained cells

7 intraa.





in nervous tissue at

aq. deal.

site of injection or otherwise. Blue Kornchenzellen in membranes of brain.



6 intrav.



No symptoms

No vitally stained cells

12 introap.




in nervous tissue ex

aq. deat.

cept a few along vessels in one restricted region of brain.



3 intrav.


Complete persist

No vitally stained

4 intraap.




ing paralysis of hind legs after 1st injection of distilled water apparently due to direct injury

cells or Kornchenzellen in region where needle had been passed, but enormous numbers of such in the membranes of the region.

In taliulating my rosvilts I may say: IT) rabbits gave 'positive' results, 8 rabbits 'negative' results. That the greater number of these were obtained during the latter part of the experunent is very probably due to the greater skill in injecting acquired by practice. Three rabbits were killed after one injection because grave symptoms dc\el(tped; all gave negative results. One rabbit was injected in a dying condition of the effect of ether; the was made half an hour after death. Microscopically all the cells and libers of the membranes were diffusel}' stained. Control rabbits were among the last in the series and had the benefit of the better techni(|ue. One of was paralyzed in both hind legs by the first injection of distilled water. The


paralysis rcjiiitincd, l)Ut the animal recovered ()ther\vis(* and remained well in spite of another injection of distilled water and throe injections of trypan blue. It was killed nine daj's after the first injection, Microsc()|)ically the results were positive. Two oth(T control rabbits had each two injections of distilled water and only one injection of trypan blue. There were no unu.sual symptoms from either of the injections; both ^ave negative re.sults. In two rabbits the deep cervical glands grossly showed bluish areas and microscopically revealed rather large regions with vitally stained cells. In two f)ther rabbits the kidneys revealed a distinct bluish tint throughout the cortex: microscopically the epithelial cells of the contorted tubules in ilisseminate areas contained extremely fine bluish granules; only by most careful study with the oil inmiersion lens could the granular nature of the stained be detected. One of these two cases was a rabbit which suffered from an experimentally produced nephritis. Only in two of the negative cases was intense staining of nerve cells of the cord observed comparable to that described by (loldmaim. It was due to the direct injection of the staining solution into the nervous tissue. In remoter regions of the cord there was no such intense staining and it was absent in the brain.

The gross appearance of the brain and cord is characteristic and practically the same when the injections have been succe.ssful. The membranes on the convex surface of the brain are always light and in the best cases appear almost entirely colorless. The coloration begins at the olfactory prolongation. On the base there is always more or less intense staining, most markedly over the olf.ictory lobe, the optic and acoustic nerve, the pons and the medulla oblongata. The membranes of the cord are uniformly tinged. After intracranial injections the coloring gradualh' downwards.

The histological lindings reveal peculiarities. If we take those in intracranial injections as a standard, we hnd, on the average, that the dura is less supplied with vitally stained cells than the pia. Towards the olfactory lobe the lunnber incre^uses enormously; at the same tijne the cells are larger. In the pia there


may Ix- an ovorwIifliiiiiiK aimuint of cells in this roRion. Generally thoro are also throunliout more cells in this membrane than in the dura. In case of marked liypercmia or pial hemorrhage there is sometinu's an amazing number of granular cells and 'Kornclienzellen' in the convex portion of the pia. The thin membrane over the cisterna magna is well suj)plied with vitally stained cells. The more surprising is the observation that the tissue about the tufts of the chorioid plexus is totalh' devoid of \it:illy stained cells. There is also no trypan blue in the epithelial c»^lls them.selves. As to the cord, diffu.sely stained cells are absent, unless the staining solution has been accidentally injected directly into the nervous tis.sue. There are no vitally stained cells along the intracerebral or intra.spinal vessels in nobu;dt,ger they may occur in regions here and there. Such foci are, as a rule, small in luunber: sometimes careful search for them will fail to revejd any.

In a few rabbits lesions had been produced in the spinal cord which were followe<l by destruction of the nervous tissue and immigration of 'Kornchenzellen' in enormous numbers: these cells are all vitally stained. Those which occupy the region of necrosis and necrobiosis lie close together, are large, more or less rounded and besides blue granules contain diffusely stained cell inclusions. The chemotactic influences emanating from this focus have been juade visible, so to speak. Throughout the .sections cells marked by blue granules swarm from all directions towards this center: they are slender, increasing in size as they apjiroach the field of phagocytic action; they are all in the immediate neighborhood of the ves.sels and follow the course of their divisions so that through them the distribution of the vessels is well illu.strated. There cannot be any doul)t in this case but that these "Kornchenzellen' are derived from endothelial cells lining the perivascular spaces. A singular observation, however, is that as we get away from this region of necrosis, blue stained cells along the vessels are no longer .seen, (iokhnann also found such phagoc.vtic vitally stained cells in inflanunations of the meninges at the place of the injury. They were not only in the subarachnoid meshwork, but could be followed from the pial


funnels fur int(j tho cortex. Ho fonsidcrs thorn us normal constituents of the meninges and their intracerebral, intraspinal, and intraneural prolongations.

.Some i)nj)()rtaiit observations from intraspinal injections of trypan l)luo are then Ijriofiy these: \ita.lly stained i)erivascular cells, endowed with the phagocytic properties of 'Kornchenzellen,' are abundant in a wide range about encephalitic foci. Vitally stained endothelial cells about ves.sels in the cord and brain are not generally in e\iilence; they may be present about a scanty number of vessels in the cord and are scarcely, if at all, seen in the brain: apparently there is less tendency in such cells to take the stain as we get away from the region of the injection. The staining .solution, however, is early carried with the cerebrosinal fluid to the optic, olfactory, and all peripheral nerves demonstrated by the early staining of their dural sheaths; it is conveyed to the deep cervical glands. May we not conclude from these observations that the endotheUal cells of the perivascular spaces in the brain and spinal cord have less affinity for the stain than similar cells in the meninges and in other tissues of the body? that some stimulus is needed to arouse them, so to speak, to reaction to the stain? and that the stimulus in these cases come from the local effects of the injections, not from the stain itself as a 'nerve toxin,' but from direct and indirect mechanical injuries to the nervous tissue, from pressure, from disturbances in the circulation? Such an hj'pothesis is not without a parallel. Taking it for granted that these particular cells are endothelial in nature or origin as admitted by various investigators, it is a well-known fact that this property of vital staining is peculiar only to a certain })r()p()rtion of this class of cells. The endothelial cells lining the blood-vessels as a whole do not take the stain. \n exception to this rule are those of the capillaries and venules of the spleen, of the blood-vessels of the bone-marrow and the hemal glands, juid the Kupfifer cells, supposed to belong to the capillaries of the liver. .Vs to the Ipnphatic system vital staining has been ob.served in the endothelial cells lining the lyjnphatic sinuses of the lymph glands. .Vn interesting observation in vital staining concerning these organs has been emphasized bj'

14 KAKTIIK \\ . I) i:\VKV

Kvans,' who writes: '"This purticipation on the part of thopndofhehujii. however, is sharply Hmitod to those definite tracts of it well within the heiiial and lymphatic glands. The entering or draining trnnks. in the of the hanph ghmd do not show any |x»culiarity on the part of their lining cells, whereas, in the case of the heniolyniph glands, nothing is more striking than the abrupt assiunption of brilliant dye granules by the entlotheliiun of a venule just as it enters and resolves itself within the gland." I have ventured to suggest' that this difference in the behavior (tf the «'ndothelial cells towards the stain may correspond to a diflference of function on the part of the Ijinph-channels. Those within the organs are collecting vessels, those outside conducting ve.s.sels. .Vs to the Kupfler cells, it cannot be said that a decision as to whether thej' belong to the blood capillaries or to perivascular hinph-spaces has been absolutely agreed upon. There are a number of writers who assign them to the latter. The other organs in which vitally stained endotheUal cells of the blood-vessels occur, namely, the spleen and bone-marrow, are curiously enough those in which blood and l>^^^ph are in most intimate relationship to each other: the spleen and the bonemarrow are the birthplace of both elements. The hemal glands may be of a similar nature; they are essentially Ijinphatic structures.

From the differences in the staining phenomena observed in the central nervous system, we may perhaps draw some conclusions regarding the flow of the cerebrospinal fluid and the relatitmship of the general lymphatic circulation to the central nervous .sy.stem. It is certainly ver\' singular that the specific connectiveti.ssue cells about the epithelial tufts of the choroid plexus, which invariably are intensely stained after intravenous injections, are ab>o!utely free from any stain after subarachnoid injections. Since this tissue is a continuation of the pia, we should expect to find vitally stained cells in this location, when tJiey ar(> present in abundatice within the pia. It wovdd seem from this that the cerel)rospinal fluid containing the staining solution does not

' I.e. "The iiiHcrophuicrs of mnnimiils." .\m. Journ. of Plus., 1915, 37, 242. •I.e.


enter chauiiels witliiii the connective tissue of the plexus, but that sucli channels receive the dye with a fluid from another source, i.e., the plasma or lymph.

It is stated that diflferenoes exist in the character of the ventricular and the subarachnoid fluid. Investigations by several authorities also seem to show that there are lymph systems within the cerebral and spinal h-mph system as a whole, independent of one another. D'Abundo,» (niillain,'" Marie and (luillain," Orr.'aud Homen," from the n^sults of injections by various methods, came to practically the same conclusions, namely, that Ij-mph from the general circulation, wliich passes along the perineural Kanphspaces of the periplieral nerves and enters the nervous system, follows paths within the limits of definite territories. Thus bacteria or their toxins, experimentally or pathologically inoculated into peripheral nerves, are carried along the IjTnph-stream to the sj)ine ami brain in an ascending direction, capable of producing lesions in regions of the central nervous system quite remote from the seat of the original infection. All these writers agree iu the statement that the cun-ent of Ijanph ascending the posterior roots and columns is independent from that of the anterolateral columns, (luillain and Marie showed that the lesions in tabes, always limited to the posterior roots and columns, are the result of tlie direct propagation of the syphilitic virus along the perineural hinph-vessels. They go so far as to call tabes a l>^nphangitis of the posterior Ijiiiphatic system of the spine.

D'.Vbundo, Sullc vie linfatiche del Sistema Nervosa Centrale. .A.nnali di Nevrologia, 1S90, 14, -'M.

" Guilliiiii. Ln circulation dc la lyniphe dans la moelle (?pinierc. Revue Neurologiquo, 1899, N». 2.'}. 796.

" Marie et Uuillain, Ia's li^sions du systdme lymphatique postc^rieur de la moelle sont I'oriKine du processus anatomo-pathologique du tabes. Revuo NeuroloRiquc, 190:j, U. t'l, 103 and 106.

" Orr, .\ contribution to our knowlcdpoof the course of thelyinph-strcaniin the spinal roots and cord Ucv. Neurol, and I'sychiatr.. Kdinburgh, VMS, 1, tj39.

" Homcn, Die Wirkuntc riniRcr anaoroben Haklericn narnontlich bei Synibiose und aeroben H.aklcrion, sowic ihrer Toxinc auf poriphcre Nerven, Spiualnanglien und da.i lUickoninark. .Vrbeiten aus dem pathologischen Institute der Universitat zu llt-lsingfurs. 190.'>, 1.1.


Accept iiiR such division of l>in|jh systems, we do not necessarily d*>dun' the cxistonro in tl>(« brain and spinal cord of conditiniis rntin'ly ('xccptional from those in tht> rest of the body. Apparent deviations from the general laws of anatomy and physiol'>j5>- are not fundamental differences, but variations conditioned by the structural and functional nature of the organs involved. The meningeal spaces arc equivalents of l3Tnph-vessels adopting the fomi most suitably adjusted to the particular requirements of such an organ with such surroundings. Perivascular lymph-sheaths do not only exist in the brain tissue, which is excee<lingly sensitive to variations in the blood-pressure, but are also found in such an unyielding tissue as bone. The cerebrospinal fluid differs from ordinary Ijanph to the same extent that lymph in other parts of the body is modified by the nature of the tissues which it drains. .\s to tlie independence of restricted lymph systems within the central nerv'ous system, parallels can be found in other anatomic units. Sicard, in discussing this question, refers to the independence of the subserous lymphatic network of the stomach from such a subserous network of the duodenum.

Following these lines, we may very well assume that after intravenous and intraperitoneal injections the dye enters the cerebral and spinal membranes anrl the choroidal connective tissue along re^il l>inph-channels, and the endothelial cells lining these channels take up the vital stain as the fluid containing it reaches them, .\fter intracranial and intraspinal injections, the .staining solution follows the paths of th(> ventricular and sul)arachnoid fluid, and in leaving the cranial and spinal cavities is not only taken up by the venous drainage, but is conveyed along true lymph-channels.

The pronounced affinity for vital stains exhibited by the specific connective-tissue cells of the meml)ranes over the olfactory tract and at the exits of the cranial and .spinal nerves probably ci»incides with a difference in the nature and function of these cells. .\t the olfactory lolx-, where this change is especially striking, we have a transition from the subarachnoid spaces to the IjTnph-channels of the na.sal apparatus. Sunilar transitions


we nmy :issuine to exist at the seat of the arachnoid villi, where the piii-aruohnoid channels more or less directly conununicate with the IjTiiph-channels of the dura.

The choroid plexus is, accordiiiK to (loldjnann, a protective limiting nK'jnl)riine filtoring off from the blood-plasma a substance which acts as a violent toxin to the nervous tissue. He also suRRoststhat there may be protective substances inherent in the granules of the ganghon cells in the dog, which he observed to be sometimes vitally stained, while in the rabbit he only found diffuse staining of the protoplasm and the nucleus of such cells. Providing the necessity f)f protection to the nucleus of these nervecells, we should reasonably assume that the ganglion cells throughout would possess and manifest such protective powers, which, however, was not observed. It would be ec|uall}' reasonable to suspect that this is a property common to all cells staining vitally, for it is a characteristic of the phenomenon of vital staining that only the granules take up the dye, while the nucleus remains unstained. On the other hand, we know that diffuse .staining of the nucleus and protoplasm, the indubital^le sign of injury or death of the cell, takes place not only in the cells of the central nervous system, but in those of all organs of the body. It would rather seem that the ability to take up the stain is dependent on definite chemical or physical properties of the constituents of the granules in certain types of cells. These cells have welldefined characteristics in common, whicii they do not share with other cells provided with granules which are refractive to vital stains. Investigations of the morphological and histological occurrence of lipoids tend to show that the cells capable of storing cholesterol are the same which react to vital staining, and in eitliercasetheability is confined to their protoplasmic granules. In my own experiments with cholesterolized rabbits,'* I have observed that cholesterol is stored not oidy in the more or less coarse granules of interstitial cells classified as macrophages, histiocytes, endothelial cells, etc., but also in the exceedingly fine granules of certain epitheUai cells, notably those of the kidney.

'* Dewey, Experimental hypercholesterolemia. Arch. Int. Med., 1916, 17, 757. TUB ANATUUICAL RKCORD, VOL. 15, NO. 1.


In tlit« clioroid plexus tli<^ lipoid .sul)st;inc('.sare perhaps best demi>ii.strate«l by tlic pctlariziiiK Jiiicroscopo. The vessels appear as if invested in an outer silver lining; the doubly refractive substances occur in excet>dingly fine divisions; they are distributed all through the tissue.

The presi'iice of granules within the protoplasm of certain types t)f cells has been associated by some writers with secretory functions. It has als<j been claimed that these granules are surrounded by a lipoid membrane. In the choroid plexus the extracellular droplets of secretion, which are plainly visible in or fonnal in-fixed specimens, are manifestly enveloped in such a membrane. Nothing is more striking than Sudan III stain(Ml preparations of choroid from cholesterolized rabbits. The tissue is riddled with orange-red spheres and circles. Also in rabbits not injected with cholesterol, these yellow and orange colored tiroplets jnay be verj' conspicuous Evidently the lipoid of this membrane is present in a form other than that within the cells; it does not appear as doubly refractive or anisotropic substances under the micropolarisoope.

In a recent puI)lieation, Sunclwall describes granular interstitial cells in the choroid plexus of the ox, which because of their morphological and staining characteristics he places in the category of mast cells and which he believes may be concerned in some type of secretion. But for his statements that he did not observe such cells in the choroid plexois of other animals — rabbits, guinea pig. dog, hmnan — I should be inclined to consider them as identical with the vitally staining granular cells which I found, without exception, in the choroid plexus of the rabbit after intravenous and intraperitoneal injections of trypan blue. I also observed them in the few dogs which I examined. Sundwall mentions variations in the number of these cells with different anunals. I have occasionally noticed slight difTercnces, the letwencd number coinciding with the diminished intensity of staining, especially in the study of the dental pulp; it was generally found in aged animals, but the observations are not num " Sumlwnll, Tho choroid plexus with speciul reference to interstitial granular eelli. Anttt. Rec. 1917, 12. 221.


erous enough to warrant the statement that age is responsible for the decrease in the number of the cells.

An assiunod function of whatever nature, inherent in the specifically staining granular cells of the connective tissue within the various organs of the body, gives a peculiar interest to this particular tissue; far from being merely a supporting stroma, this tissue would have all the signilicance of a vital organ. .Such a view is expressed by Renaut"' in the statement that the connective tissue is ' ' like the largest of the glands with an internal secretion which exists in the body of the vertebras, because it keeps the elements ready for inmicdiate action or susceptible to such action at any time, wherever blood-vessels course through the connective tissue." I am inclined to consider these cells as endothelial cells, being in the most intimate relationship to the Ipuphatic ajjparatus and plaj'ing a more important r61e in this system than that of merely lining the lymph-channels.


Concerning the results of these experinients the following observations are to be particularly eniphasized :

An apparently exceptional behavior towards vital stains is exhibited by the endothelial cells lining the perivascular spaces, or, if the existence of such cells be denied, by specific cells of the perivascular connective tissue within the brain and spinal cord.

Unlike such cells in the perivascular connective tissue in other organs and tissues, they do not habitually take up the vital stain, but do so only under the influence of stimuli from pathological conditions. This lack of affinity for \'ital stains may be due to a (liff(^ronce in specific functions inherent in these cells. DilT(>rences in the behavior of endothelial cells towards the stain are observed with constancy as follows:

1. With reference to blood-vessels, affinity for the vital stain is absent, in general, in the endotheUal cells of the inner lining

'• Reimut, "Les cellules connectives rhagiocrines." Arch, d'aniit. niicroacop., 1907, 9. 495.


of artorios, veins, and capillarios; prosont o) in the rapillarios and vonuK's of thf spleen, the capillaries of the bono-niarrow and the blood sinuses of the hemal glands, i.e., in tissues the source of blood and lymph elements, b) in the Kupffcr cells of the liver, i.e.. in cells of which it is not yet absolutely certain whether they belong to the capillar}' wall proper or to the perivascular lymphspaces.

2. With reference to hinph-channcls, affinity for the vital stain is absent in general in the endothelial cells lining the inner wall of lymph-vessels outside the organs; present in lymph-channels within the organs except the brain and spinal cord.

3. With reference to the central nervous system, affinity for the vital stain is absent in general, in the perivascular spaces within the brain and spinal cord; present in these conditionally and in focalized distribution in the presence of pathological stimuli, in general, within the membranes along channels conveying l}7nph or cerebrospinal fluid.


i;.\l'I,\N A TKiN 1)1 1 KilKKS

FiR. I Dura. (IntruvennuB injcctioim of trypiin l>Iui'; lilood-vcsscis lillcil with cariiiiiu- nfliiliiil Tracts of vitally stainrd rclls niiiniiiK aloiiu thi' lilooilvcssols and also in iiidc|>ciid(>iit directions. MaKniticatiim KKIdiani.

Kin. - I'ia-araclinoid. (Intravenous injections of lithium earniine; hloodvcssols filled \vill\ Herlin lilue gelatin.) Two complexes of vitally stained cells: more or less distinct lines of Krunular colls radiating from those centers. Magnification 100 dtam.

Fig. 3 Choroid plexus. (Intravenous injections of trypan blue; hlood-vossels filled with carmine gelatin.) The niuiiorous vitally stained interstitial cells are <|uitc everdy distrihiitecl; there is no distinct perivascular arrangement Magnification KM) diam.

Fig. 4 Dura. (Injections of trypan Mue into the cislerna magna Kogion of trnnsilion from the hraiii to the olfactory tract. Tracts of vitally stained granular cells closely following the hlood-ve.s.sels. There was a striking lack of vitally stained cells in the dura over the hemispheres. Magnification 1(X) diam.

Fig. 5 Intracerehral lilood-vossol. ( Injection of trypan hlue into the cisterna nuigna.i Perivascular vitally stained granular colls. Magnification iOO diam.

Fig. I) Spinal cord. (Intraspinal injecticui of trypan Idno.) ICnormous accumulation of 'Iviriicheh/.ellen' in a region near the central canal Tracts of slender granular cells converging towards the focus; they follow the course of the lilood-vessels and increase in size as they approach this center. Photomicrograph. Magnification 18.') diam.

Fig. 7 Choroid plexus. Stained with Sudan III and hematoxylin. I^xtracellidar droplets of secretion with a clear center and a lijioid memlirane. Fron\ an apparently nornuil ralil>it. Magnification 100 diam.

Fig. 8 Choroid plexus. From a ralibit injected with cholesterol. Olohular and ring-shaped extracellular droplets of secretion; stained with Sudan III ; the tissue is very faintly stained with hetnatoxylin. Photon icrograph. .Magnification '275 (liatn.

Fig. !) Pia-arachiioid. i Intraspinal injection of try p in Mue. i Diffuse staining (if cellular elements and all the fillers of the cmineclive tissue demonstrating the complex structure of this memlirane.


^ \

^ \ » .- \ \ , X

\ !

.'V,\ \ : \

V \

' i

» .

') /

t* / f



v.», C»

'^=^.r.s. 4V»

'i ^ ^

V a i




k i




^'■%^^Ji>* ♦

» » •

• 1 • * r '

f %•



V( TllllR H AllHrilAlT Of TIIIH |*AI'i:il |HfirC[) liV Tllk Itllil.lMiiK.M-lllc nKKVKh. JULY lU

A I)i:S('I!IIl"l()\ OF \ CASI': OF FALSK lli:i{.MAlMlliol)l TIS.M

II. K. JORDAN Dcparlment of Anatomy, University of Virginia


.My iiitcrc.'^t in tlii.s (■use wu.s tirsi arimscd Ix' tlic autopsy materials included, besides two abdominal testes, a bilateral pair 111' syninictriciilly placed smnll alxinminal l)otlies suK^e-stive of ii\arian icmnants. These ixxlies i)ro\-ed on microscopical exaniinatimi to 1)(> altered lynii)li-notles, such as are typical for certain inlianiniatory conditions. The nodes are chanvcterized by fil)rosis, hyalinization, and arteriosclerosis. The case is tli<>refore one of masculine pseudohermaphroditisinus, and is oonse(|U<'n11y of relatively lesser importance. However, since tlic inicrosco|)ical data in refjard to the testes have a bearin}? <in I lie alleiied causal relationship l)etw(>en the interstitial cells and the development and maintenance of the male secondary .sexual chnracters (Bouin et .\ncel'), as also on the hypothesis that ' heat ' is dejiendent upon an internal .secretion on the part of cells, a detailed descrijition of these organs seems warranted. In view of the wide ffcncrjili/at ions current . as comjjareil with the relatively meager body of definite data, concerning the function of the interstitial cells, it would seem that any additional relevant microscopical ol)servations should be carefully recordeii.

The subject was a patient of Dr. ('. S. I'nderhill, of Mc( iahe.\ s\ ille. \'a.. who first directed my attention The aiilo])s\- was performed on October .'^l by the coroner of Hockiiridge ('ounty. Dr. .1. .M. Hiedl(>r, of Ilai'i'isonbui'g, \ a. The

' Hoiiiii el .\iiitI. I'.MU. lici'hcri'ln'.s siir Wi .•iiuniliciiioii |>hy.iici|ci)(i(|iii> ilo la kIhiuIc intor.sliticllc lUi tcslii-uU- dcs iiijimtiiif^rcs. .lour. <ir plivsiol. ct do piith. Kon., T. 0, i>. lOIJ.


II. !■:. .IDKD.W

iiitrriial and rxtcriiul R«>iiitiilia were pn'sci\cil in 10 per ci-iit fi)riiialiii. I am ^;|•(•ally iii(l('l)t<'il to Unth <>( tlicsc gent l( mm for their interest in this case and fur kindly sending me wellfixed tissues and full ailtojjsy notes.

Dr. I'liderliill's notes ijielii<le the fnllowinjj pertinent data: 'I'he siihjoet was thirty years of as*', white, of p)od i)hysiop;nomy. with narrow forehead, hi^h eheek-l)ones, and oidy a thin growth of hair on top of head. He had supernumerary fingers and toes, the foiH' additional digits being accessory to eacii i>l' llie fiftli

Kin 1 I'liiit<iKra|>h of iiui.'triiliiic fiilsc hrrmaphroditc, witli |>olydiictylisin and tali|i08 varus ( Kiiidiics.s of Or. (' I) riidcrliill.)

nioinhers. Hoih fe<'t were chibbed (taHpes varus). The mammary ghmds were well developed, the penis was rudimentary, the .scrotum was vestigial, and the hands, hips, thighs, and legs were of female conformation. 'I'here wa.s much fat everywhere; all organs were covered and infiltrated with fat; the subcutaneous fat was 4 inches thick. Opposite the internal ring on each side was a having the 'feel of ovary or testicle." Two wnde.sceiide<l te.sticles were removed from a .sort of to the internal ring, from which the finger could i)e pushed down into a 'pouch of skin where the scrotum should have been."


On subsequent incjuiry tlio following additional facts were ascertained: The subject was 'very feeble-minded,' left-handed, unmarried, with no tendency to 'venery.' His parents were 'about third cousins.' His mother had a severe case of heat prostration at the age of 15 to 17; she was in poor health when married at 22. She has been married twice; her one child, a daughter, by her first hust):ind is 'alive and well.' She had three chiklrcn by her second husband; one died in infancy, and a feeble-minded daughter, sister of the subject, is still living; 'she is able to work about the house.' The parents deny any abnormalities; they do not have supernumerary digits.

The specimen of the external genitalia shows only a .short penis, of 2.5 cm. length. The internal, or liidden, portion of the penis has a length f)f 9.5 cm. Taken as a whole, this organ is therefore about two-thirds average length, and not actually, as apparently on external view, 'rudimentary.' The scrotal region is marked by a small area of pigmented, loose, greatly wrinkled skin.

The bodies originally suspected of being ovaries were enca.sed in an envelope of fat, about 6 mm. thick. The larger measured 3.5 X 2 X 0.5 cm. ; the smaller, 2 x 1.25 x 0.5. The general shape was flattened reni^)rm. The central core of the larger body, apparently of connective-tissue structure, was triangular in transverse section, measuring 8 mm. from base to apex and 5 mm. along the base; the latter was indente<l by a hilus.

The central cores onlj' of both of bodies were embedded in paraffin, sectioned at 5 microns, and stained with iron-hematoxylin followed by van Gieson's stain. The preservation was superb. The bodies are unmistakably lymph-nodes, in which tile reticular tissue has increased enormously. .Approximately half the volume of these nodes consists of dense hyalinized reticulum. .\t first sight this appears homogeneous, but closer inspection reveals a fine fibrillar constitution. The larger denser areas occur in the cortex, apparently at the centers of the original nodules, and along the medullary cords. The reticulum of the denser masses shades o(T into the reticulum of the medullary siiuises. The lymphocytes are well preserved and appear normal.


The tosticlcs have a flattniod oval form, roundly triangular in transection. Thoy differ Rreatly in size: the larger measures 5 X 2.5 X 1 cm.; the smaller, 2 x 1.5 x 1 cm. Microscopical examination shows that they ar(> atrophic, similar in appearance to undescended testes, but with the usual degenerative phenomena greatly accentuated, the degree of abnormality corresponding with the size difference. The sections were stained with the iron-hematoxylin van Cieson's combination. The tunica albuginea is greatly thickened. The testes are characterized in general also b.y fibrosis and arteriosclerosis.

Seminiferous tubules are still distinguishable in certain regions of th«' larger testis, but their tunica propria is greatly thickened and converted into a fibrous hyaline wall. Throughout this wall nuclei of variable form, size, and staining capacity are sparsely scattered. The nuclei are relatively small; they have an oval, bilobetl or irregularly lobed form, suggestive of amitotic division. The larger oval nuclei are generally vesicular; the smaller and more irregular fonns are usually pyknotic. The perinuclear region, corresponding to the cj'toplasmic area, appears clear and homogeneous, the tissue as a whole thus resembling somewhat fibrocartilage. The surfaces of the wall have become sinuous or corrugated through contraction. The thickness of the wall varies as the diameter of the cross-sections, the smaller tubules generally having the more robust w-alls. In certain regions of the larger testis the tubules have collapsed and become solidified into cords of hyaline material.

In the smaller testis practically all of the tubules have become solidifietl. These denser, tubular, areas are separated by narrow areas of hyalinized fibrous tissue with few cells containing mostly pyknotic irregular nuclei, and an occasional small degenerating interstitial cell. All of the intratubular lining cells have disappeared in the smaller testis.

The p.'itent tubules of the larger testis contain a peripheral layer of irregularly c<ilumnar or polyhedral cells, one to three layers thick. It seems most probable that the lining epithelium wjus originally single-layere<l and that this became 'stratified' thnjugli a dislocation of certain cells following the shrinkage of


the tubules during the process of hyalinization of their walls. Certain of the smaller tubules are filled with cells arranged in the form of a reticulum. The reticulated appearance is due largely to a vacuolization of the cytoplasm. Certain cells contain a single large spherical vacuole, the nucleus having been pushed to one pol(>, giving these cells the appearance of fat cells.

The nuclei of the less degenerate cells are relatively large, oval, and vesicular, and contain a delicate reticulum; those of the more extensivelj' vacuolated and irregular cells appear collapsed and pyknotic. In the larger tubules an occasional cell may be found of spheroidal shape with large vesicular imcleus suggesting a spermogonium. Besides these very infrequent cells, the lining cells include no elements which resemble typical spermog<jnia or spermocytes; they are most probabh' largely remnants of Sertoli cells and possibly in very small part of prunordial germ cells. Between the tubules occurs a coarsened fii)rous connective tissue with numerous small, irregular, more generally p\knotic, nuclei. Relatively few scattered interstitial cells are discernible.

The chief point of interest in the case centers on these interstitial cells, considered in relation to the mixture of secondary sexual characters ami the absence of sexual desire in the subject. The interstitial cells were identified by comparison with those of a normal testis, the spermogenesis of which was previously described.- Compared with those of the normal testis, the interstitial cells of this hermaphrodite are very few, much smaller, approximately half the nonnal size, and the cytoplasmic content of lipoid granules, instead of being scattertnl. is grouped into one or several large clunips. The cells could be studied best in unstained preparations. Thanks to their brownish-j-ellow granules they could be easily detected. They are widely scattered; they ha\e a spheroitlal or polyhedral shape; certain have a large, oval, vesicular nucleus, but the majority contain only a small, frequently irregular and pyknotic, nucleus.

Jordan, H. E., 1014. The spermatoKcncsis of the mongoose; and a further comparative study of mammalian spcrmatof^enosis. with special reference to sex chromosoines. I'lil). 182, Curnonio Institution of Washington, p. 16.'J.


T1h« niicrospiipical evidence shows, then, that these testes contauiotl oriKinnlly l><>th the intra- and intertubular cellular elements characteristic of normal fetal male gonads, but that both liad sufTcnKl a gradual dcficnoration, probably in part patiiological, until only modified intratubular, chiefly sustentacular, cells and degenerate interstitial cells remained throughout the hyaiinizcil stroma. The persistence of even a few recognizable interstitial cells under the conditions which obtained in these testes, even in the more atrophic of the two, indicates a most remarkable vitality and suggests an important r61e on the part of these cells.

The case under consideration appears to be one of arrested male development. All things considered this pseudohermaphrodite is very decidedly more male than female. The female secondary sexual characters are superimposed on a male substructure. And it may be emphasized that the 'female' characters are not infantile; witness the well-developed mammary glands. Nor are they wholly dependent upon the considerable adiposis. For some reason certain female secondary sexual characters have gained formal expressicm in this essentially male individual. The underlying proximate cause is most probably the atrophy of the testes. The ultimate cause would seem to have been one conmion to the hernuiphroditism and the concomitant physical (supernumerary digits, club-foot) and functional (left -handedness; feeble-mindtHlnessj anomalies. In respect to this congeries of anomalies, presumably dependent upon a common germinal factor or disturljance. the is practically a duplicate of one previously described,* where a very similar pseudohermaphrodite was one of nine children, all of whom showed various anomalies, including polydactylism and gigantism, and adds another to a large and varia!)le group of pseudohermaphroditic malformations coincident with other minor anomalies.

Thus the 'rudimentary' penis, vestigial scrotum, and cryptorchism appear to be different aspects of a general anomalous condition. The case may be conceived to have developed as

' Jordan, H. E., 1912. Studies in human heredity. Bull. Philosophical Society, Univ. of Vir((inia, Scientific Series, vol. 1, no. 12, p. 293.


follows: The atrophy of the testes, and the suppression of the scrotum, supervened upon the inguinal retention of the testes; the development of certain female secondary sexual characters resulted from the atrophy. The effect of the original cryptorchism may have been modified by pathologic factors, for undescended testes commonly have an increased number of interstitial cells coincident with, and reciprocal to, a decrease in the number and activity of the intratubular .spermogenic cells. It would be of interest to know in this connection also the condition of the hypoi)h3'sis. but this organ was not removed at autopsy.

The well-attested lack of .sexual desire in connection with the paucity and abnormality of the interstitial cells is perhaps the most important phenomenon pertaining to this case. That the control of the male secondary sexual characters is not dependent upon the interstitial cells of the testis seems proved by this case. where a mixture of secondary sexual characters is maintained in the substantial absence of healthy interstitial cells. Such conclusion must be drawn also from the various castration and spaying experiments where certain secondary sexual characters, generalh' reciprocal, originate and persist in the absence of interstitial cells. The same conclusion is indicated also by conditions in certain turtles which lack secondary sexual characters and where the male nevertheless possesses abundant interstitial cells during the spring, for a period probably corresponding with the breeding season.

The hypothesis that 'heat' is dependent upon the integrity of the interstitial cells seems to rest upon a tinner basis of observational data. The case describetl by \Miitehead,^ of a stallion with a third abdominal testis, seems conclusive on this point. The whole of data relating to infertile hybrids (e.g., mules) and cryptorchid individuals supports the .same conclusion. For here heat remains in the absence of potencj', correlattMl with a numerical increase of interstitial cells and an ab ' Whitehead. 1{. II , lilOS. .\ |HTuIi:ir case of cryptorchism. and its hearing upon tln' prohliMM of Ihi- fuiictiun of the interstitial cells of the testes. Anat. Rec, vol. 2. p 177.



sence of complete spennogenic activity. The hermaphrodite here descriho«i adds similar e\'idonce, though of negative character. Here heat was absent, a condition correlated with few interstitial cells and these of degenerate character.

The sufficiency of this latter hypothesis, as well as the now more generally disearded earlier one of Bouin and Ancel, is controvert etl by Boring and Pearl' on the basis of their results from a microscopic study of the testes of chickens, in individuals of which of over six months they claim interstitial cells are whollj' lacking though of 'full sexual norraalitj- both in respect of primary and secondarj' characters' (p. 205), These results, however, are at variance with those obtained bj' the recent careful work of Reeves,* who reports interstitial cells in the testes of chickens at all ages up to eighteen months.

The coincidence of left-handedness in this case with various anomalies (polydactylism, rudimentary and atrophic genitalia) supplies additional interest. In my studies of hereditary lefthanilednes" I have been led to adopfthe hypothesis that the condition follows a fetal structural variation in the cerebral bloodsupply, of a nature to effect a better nutrition of the right cerebral hemisphere in contrast to the usual condition where the left hemisphere is better supplied. An examination of the fetal vascular system suggests that the arrangement of the common carotids, iimominate and subclavian arteries is such as to favor the po.ssibility of a larger blood supply to the left cerebral hemisphere and the right arm, thus presumably determining a bias towartls right-handedness. If this morphologic explanation of right-handedness and left-handedness is correct, then it is significant that also left-handedness should occur in this case where various other anomalies are associated with hermaphroditism.

In view of these data the conclusion is suggested that a condition of pseudohennaphroditism is of the same category as the

• Boring, Alice M., and Pesirl, Raymond. 1917. Sex Studies. IX. Interstitial cells in the reproductive organs of the chicken. Anat. Rec, vol. 13, p. 253.

Reeves. T, P., 191.5. On the presence of interstitial cells in the chickens' teates. Anat Ric , vol. 9. p. 3.S3.

'Jordan, H E, 1914. Hereditary left-handedness, with a note on twinning. Journal of Genetics, vol. 4, p. 68.


common anonialies of other systems. The accompanying functional (listurl^ances in this case, expressed in loss of sex potency and suhsidcnre of 'desire,' resulted from the induced atrophy foIl(»sving the inguinal retention and proljablj' later disease of the testes. Whether subsidence or accentuation of heat follows cryptorf'hisiii seems to depend upon whether the resulting testicular atrophy involves also the interstitial glandular cells or only the seminal epithelium.

AtTTHon n ADATRArr nr tmipi PAreR irhitkd



H. E. JORDAN Department of Anatomy, University of Virginia

The primary object of this investigation is to determine whether blood-platelets occur in lymph. The material employed consists of stained smear preparations of lymph from the thoracic duct of the dog, treated according to Wright's method for blood smears. The desirability for further studies on the structure of Ij'mph arises by reason of the discrepancies in the results reported by recent investigators. Schafer ('12) makes the unqualified statement that "both lymph and chyle contain thrombocytes" (p. 394). Davis and Carlson ('09) in their study of the cell-content of lymph, make no mention of platelets, permitting the inference that they did not see these elements in lymph. Vinci and Chistoni ('10), on the basis of careful microscopic studies of the lymph, of dog, cat, and rabbit, conclude that lymph lacks platelets (p. 209). Howell ('14) comes to the conclusion that "platelets do not occur in lymph from the thoracic duct of the dog."

The following general facts would lead one to expect to find platelets in lymph: 1. Platelets play an important part in the clotting of blood "by liberating an additional supply of thromboplastic substance and of prothrombin " (Howell, '14). Lymph, in common with blood, contains fibrinogen and thrombin, and under similarly favorable conditions clots promptly. Since thrombus formation in blood is apparently lurgcly dcpoiulent upon the presence of platelets, the latter would be exin-cted to occur also in lymph, which has a similar physical and chemical constitution and clots like blood under similar conditions. 2. With the possible exception of platelets, lymph contains the same cellular elements, in vastly ditTerent proportions, as blood. If any bar 37


ritT l)ot\v(»on the hlood-vusruhir ;iiul tlu- lymph-vascular systems can bo passed by all the other cellular elements of the blood, even the erytliropiastids, it seems reasonable to suppose that platelets also would not be excluded from lymph. 3. The origin of blood-plat elels from segmenting pseudopods, and from fragmenting larger areas of cytoplasm, of a certain type of megakaryocyte in the red bone-marrow is well established nVright, '10; Bunting, '09; Downey, '13; Jordan, '18). .Since the platelets pass readily into blood-capillaries of the marrow, no plausible explanation presents itself as to why they are excluded from the IjTnphatics of the marrow.

The recent work of Howell ('14) is the most thoroughgoing that has yet been devoted to the question of platelets in lymph. It is requirefl, therefore, that this work be first carefully analyzed, with a v\ow to determining whether any possible contingency remains which has not been fully met in this search for the possibly very elusive platelets in lymph. Such a contingency would serve as the only excuse for any addition to an otherwise apparently conclusive research.

Howell's investigation of platelets in lymph is part of a larger research relating to the mechanism of coagulation in blood and l>anph. He obtainofl the lymph of his experiments from dogs anesthetized with morphia and ether, by placing a vaselined cannula into the thoracic duct. The rate of flow was variable, but averaged about 0..') cc. per minute. The lymph from dogs starved for forty-eight hours clotted in from ten to twenty minutes. The more milky lymph from fed animals clotted in from thirty to sixty minutes in the case of earlier specimens; Ij^mph collected after an hour or two only clotted after from one to tliree hours. Howell searched for platelets both in the fresh and in oxalate<l Ij'mph of these experiments. He examined also the corpuscular sediment obtained by centrifugalization of oxalated lymi)li, but neither in the deposit from a first nor in that of a second centrifugalization could he discover platelets, though the same Tnethod applied to blood-plasma revealed them in great abundance. Howell C14) concludes that "platelets do not constitute a normal element in the lymph."


Acoordinp; to Howoll, clottinp; involves the cr)6poration of four constituents: 1, fibrinogen, in solution in plasma and lymph; 2, prothrombin, liberated by platelets and by lymphocytes; 3, antithrombin, present in lymph and in bl()()d-i)lasina in substantially equal amounts, and 4, thromljoplastin, liberated by platelets and lymphocytes and tissue cells in general, and operating to neutralize the antithrombin. Pure lymph uncontaminated with tissu(> juices is said to coagulate more or less imperfectly compared with manmialian blood, l)Ut it coagulated promptly and (irmly if tissue extract is added to it. Howell explains the inferior clotting of lymph as due to the comparative paucity of thromboplastic material, the result of the absence of the less stable platelets and the consequent restriction of its source to the relatively more stable lymphocytes. This explanation is in accord with the earlier cimclusion of ^'inei and Chistnni f'lO) that coagulation can occur in the complete absence of platelets. These authors concluded, moreover, that "the principal morphological blood element intimatelj' related to the phenomenon of coagulation is the white corpuscle" (p. 212).

In view of the facts that general considerations relating to the comparative microscopic structure of IjTuph and blood, and to the mechanism of clot formation as fornmlated by Howell, seemed to favor the probability that lymph contains platelets as expressly claimed bj' Schiifer ('12), it occurred to me that failure on the part of Howell, and of Vinci and Chistoni, to discover t"he alleged elements might be due to some unfavorable factor in their technical i^rocedure with reference to platelets. It is well known that the leucocytes in the blood-vessels tend to collect peripherally. .Vlso the extreme adhesiveness of the platelets and leucocytes in general is obvious. It seemed possible that the slow flow of the lymph (0.5 cc. per minute, Howell) might permit the elimination of platelets from specimens collected by caiinula from the thoracic duct, by reason of their adhesion to the wall of the duct. A final link in the chain of apparently complete evidence respecting the absence of platelets in lymph as given by Howell, seems to demand an ex:miination of lymph in the form of specially preparetl smears made directly from the thoracic duct.


Such hypothotical doinaiul fiirnishod tho roasnn for the present investiRntioii and suggested the special technic employed.

This investigation includes lymph smears from two different dogs prepared at an interval of several weeks. Control specimens of blood were made from these same animals. The technical procedure, for assistance in which I am greatly indebted to Dr. J. A. Waddel, professor of Pharmacology, at the University of \'irginia, was as follows: The first dog was quickly killed with ether. The thoracic and abdominal cavities were immediately opened, and a segment of the thoracic duct tied off and excised. The segment was thoroughly rinsed in physiologic salt solution at body temperature, to cleanse it of any possibly adherent blood. \ cut was promptly made into this .segment of thoracic duct and the lymph content allowed to flow onto slides. The lymph had a clear, watery appearance. Subsequently the segment was cut along the entire length and slides were smeared from the inner surface of the duct. The slides were at once treated according to Wright's staining technic. Control blood smears, stained according to the same method at the same time, showed abundant platelets. But the lymph smears apparently lacked platelets completely.

I was at first surprised to find erythroplastids among the cellular content of the lymph. I suspected that their presence was the result of the death struggle, they having possibly been sucked back from the subclavian vein. However, if their presence was tf) be accounted for on this basis, typical platelets should occur for the same reason. 1 fiiul that Howell, and Davis and Carlson, also record the presence of a certain small number of red corpuscles in lymph. In the second experiment almost immediate death was caused by administration of chloroform by mouth, with the idea of obviating any possible reverse flow from the subclavian vein at death. But the preparations from this specimen also showed the presence of a few red corpuscles. Accordingly, we may conclude, I believe, that erythroplastids are a normal constituent of lymph. Nor did any of this second series of slides reveal typical platelets.


The ((uestioii arises as to how the presence of the erythroplastids in lymph is to be interpreted. Davis and Carlson ('09) only regard the presence of crythroplastids as due to blood admixture (p. 16). In accord with the numophyletic theory of bloodcell origin, such red corpuscles might represent differentiations of certain lymphocytes, serving as hemoblasts, in their slow passage through the lymphatic tree. However, if this conclusion were correct, one should probablj* expect to find also erythroblast stages; but such could not be detected in my slides. It seems perhaps more reasonable to conclude that the few erythroplastids present in lymph have their origin in the red marrow drained by certain perivascular lymphatic terminals of the system. This conculsion, however, requires that a consistent answer be given to the question as to whv platelets, which theoretically could enter the system in the same way, cannot be found in lymph from the thoracic duct. It seems altogether probable that platelets do actually enter the marrow lymphatics in small numbers. The lymph of the thoracic duct contains elements generally liberated by platelets, namely, prothrombin and the thromboplastin (Howell), but in relatively smaller amounts than found in blood-plasma. Platelets are notoriously unstable structures. In their relativeh* slow passage (compared with the blood circulation) through the lymphatics to the thoracic duct, the relatively small number that may have entered the system may have entirely disintegrated, meanwhile contriliuting the relatively small amount of prothrombin and thromboplastin characteristic of lymph; to which the lymphocytes under favorable conditions may contribute a certain additional amount, as suggested by Howell. Considering the origin of the hulk of the lymph from the plasma via the tissue spaces, and the continual disintegration of blood-platelets, the presence of these elements (thrombin and thromboplastin) in lymph can be accounted for at least in part on the basis of this relationship between blood plasma and l\'mph. Similarly with respect to the thromboplastin of the ti.ssue cells.

Starling ('15) inclines to regard blood-platelets as precipitation products in plasma, following contact with foreign bodies or the lowering of its temjjerature from 37°C. to 18° (p. 837). He


baws his opinion on llie observations, following Huckmaster: 1. When a film of blood, held in a platinum loop and kept at body tenip<'raturo, is examined under the microscope, no platelets can l>e seen: on cooling, platelets make their appearance. 2. AMien noncoagulablf plasma, e.g., peptone plasma and oxalate plasma, is kept for twenty-four hours at 0°C., after removal of all formed elements by centrifugalization, a precipitate forms 'indistinguishal)le from blood-platelets.'

A\'ith regard to the first observation it may be justly urged that the fact that the platelets are not microscopically visible in the jilatinum-loop preparations at body teniperatur(> is not conclusive proof that such are not actually present. The re^ fractive index of the platelet protoplasm at body temperature may be so close to that of blood-plasma as to render the platelets practically invisible; on cooling, the refractive index of the platelets may change to a point where they become visible against the background of the fluid plasma.

In opposition to the deductions from the second set of observations, stand the results of the present investigation, namely, that in the smear preparations (essentially of lymph which was allowed to cool and thus suffered coagulation and precipitation of certain proteids) certain platelet-like bodies formetl which at first appeared indistinguishable from genuine blood-platelets* However, a careful study, as described in detail below, showed that these bodies are quite different structures, essentially compound precipitation products. Moreover, considering the chemical and physical similarity between blood-plasma and lymph, and in large part their conmion origin, it .seems strange that genuine platelets should precipitate only from the former and not from the latter. And in addition, the fact should again be empha.sized that the original results of Wright ('10) regarding the megakaryocyte origin of platelets have been confirmed by at least three subsetiuent independent workers (Bunting, Downey, Jordan) and his conclusions may therefore confidently be regarded as firmly established. Continued skepticism regarding the genesis of platelets as first described by ^\'right can only rest, it seems to me, on a failure to repeat in detail Wright's technic in the study of red marrow.


In view of Starling's interpretation of platelets as precipitation products, it becomes important to give in detail the steps by which I r«>:ichcd the conclusion, on the basis of my microscopic preparations, that the lymph from the thoracic duct of the dog lacks platelets completely. The slides were made a year ago and were at once carefully studied. Numerous structures were seen which sinudated platelets. The possibility seemed to present itiself tiiut these botlies were modified or disintegrating platelets. The investigation was temporarily abandoned with the idea that the material at hand did not permit of definite conchisions. I subsequently again studied the slides, and again arri\-ed at a point of uncertaintj'. A third, more intensive effort, has brought me to certain and definite conclusions: Platelets do not occur in lymph of the thoracic duct; the half-dozen bodies in my preixirations regarding which some douljt may remain do not materially alter the precision and generality of this conclusion; various and numerous precipitation products appear which simulate platelets, but which are clearly of different origin and of compound structure. This material accordingly jields additional ami significant evidence for the disproof of the conclusion that platelet-like bodies in peptone and oxalate plasmas are genuine platelets.

The stained smear preparations of lymph have a background of a homogeneous or ver\- finely granular vacuolat^'d substratum, ranging in color, depending upon a varying density, from a faint pinkish-blue to a light lilac color. Among the spherical vacuoles may occasionally be seen small spheroidal or oval areas of more condensed substratum. Throughout this substratum are scattered the lyinph cells. These are mucii more abundant in the specimens from the first dog, a difference depending, as Howell has shown, upon nutritive conditions, that is, whether the animal was st-arved or recently fetl. The small l>Tnphocytes predominate, forming at least 9() per cent of the entire cell-content. Large lymphocytes are also relatively abundant; similarly large mononuclear leukocytes with reiiiform lilac-colored nucleus, ami a wider shell of faintlj- pink cytoplasm. Occasional neutrophils, eosinophils, and erythroplastids also occur, probablj- of marrow


uiiil tissue-space origin. Occasional endothelial cells, detached from the duct wall in the process of making the smears, also occur. The large mononuclear leukocytes are in the vast majority of cases distorte<l or fragmented. cells are eviilently much more adhesive and fragile, both in respect to cytoplasm and nucleus, than the small lymphocytes. In the close vicinity of such broken leucocytes clumps of filirin-fibrils occur, suggesting that cells are more favorable regions for thrombus formation.

Scattered throughout the substratum are also innumerable spheroidal and bacillary lilac-colored metachromatic granules, most probably proteid in chemical nature. Such frequently collect in small groups and may thus simulate platelets. The resemblance is especially close where several such granules have collected within one of the minute oval coagula above described. It is only when constant reference is made to the platelets of the control blood-preparations that precise dififerences are established. In this work, whenever such a platelet-like body was encountered, reference was immediately made to the blood smears. The following are the essential criteria finally established for definitely distinguishing between such platelet-simulacra in lymph and genuine platelets of blood: The blood-platelets, like their sinmhicra in lymph, vary greatly in size; the size extremes of platelets are in the proportion of at least 1 to 10 volumes; but true platelets are very frequently collected into groups; their Ij-mph simulacra are generally scattered and never in large groups. The peripheral hyaline layer of blood-platelets is very pale blue, the central granules violet, after Wright's stain; in the lymph simulacra the correspcjnding parts are pinkish-blue or light lilac, with deep lilac or violet granules. Moreover, the granules in the case of the lymph structures are coarser and less regular (spheroidal and bacillary) than those of true platelets.

The inference ma\' perhaps be justly drawn on the basis of these comparative observations on Ijiuph and blood, with respect to platelets, that the so-called 'platelets' of peptone and oxalate plasmas, on which is founded the precipitate-explanation of platelets (Huckmaster; Starling), are homologous with the platelet-sinmlacra of lymph. The evidence seems conclusive that


platelets are preformed structural elements of blood — not precipitated products — and that they do not form a normal constituent of lynipli of the thoracic duct. This may mean only that the possiblj' small quota contributed to the peripheral portion of the lymphatic tree in relation to red bone-marrow, from which the relatively smaller amount of prothrombin and thromboplasfic substance of lymph may have a partial source, suffered disinte^riition during their relatively slow progress towards the thoracic duct, and accordingly do consequently not appear as structural elements among the blood-cell content of thoracicduct lymph.


Bdnting, C. II. 1909 Blood platelet and megakaryocyte reactions in the rabbit. Jour. Exp. Med., vol. II, p. 541.

Davis, B. F., and Carlson, A. J. 1909 Contributions to the physiologj- of lymph. IX. Notes on the leucocytes in the neck lymph, thoracic lymph, and Mood of normal dogs. .Am. Jour. Physiol., vol. 25, p. 173.

Downey, H. 1913 The origin of blood-platelets. Folia Haemotologica. Bd. 15, p. 25.

Howell, W. H. 1914 The coagulation of lymph. .\m. .lour. Physiol., vol. 35, p. 483.

JoKUA.N, 11. E. 1918 A contribution to the problems concerning the origin, structure, genetic relationships, and function of the giant cells of hemopoietic and osteolytic foci. .\m. Jour. Anat., vol. 24, p. 225.

Sfn.\FEH, E. A. 1912 Text-book of microscopic anatomy. (Quain's Anatomy, vol. 2, pt. I, p. 394. 1 Longmans, (irecn & Co.. London.

.SxAHLlNfi, E. II. 1915 Principles of human physiology. Lea & Febiger, Phila., p. 838.

ViNi'i, Ci., ET Chistoni, .\. 1910 Plaquettes et coagulation (Contribution k I'l'tude do la coagulation du sang). .\rch. ital. dc biol., vol. 53. p. 206.

Wrioiit, J. II. 1910 The histogenesis of bluod-platelcts. Jour. Morph, vol. 21, p. 263.



WILLIAM SNOW MILLER Anatomical Laboratory University of Wisconsin

In studying the distribution of the blood-vessels within the lung and their relation to the bronchial tree, there are certain disadvantages connected with wax corrosions, celloidin corrosions and the clearing method of Spalteholz which can be overcome by special injection masses used in connection with stereoroentgenograms.

Wax con-osions are very brittle and easily broken; moreover, they do not show clearly the relation of the blood-vessels and bronchi to any variation in the lobation of the lung.

Celloidin corrosions are less brittle than wax corrosions and, when once they are mounted in their proper preservative, they are more permanent; but they also fail to retain the relationship of blood-vessels and bronchi to variations in lobation.

Spaltoholz's method shows the lobation and its variations perfectly, but the blood-vessels and bronchi often fonn a confusing mass.

To offset these disadvantages, I have made use of the following method which shows variations in lobation, the distribution of the main arterial and venous trunks, and their relation to the bronchial tree.

The pulmonary artery is first injected with a starch mass in which vermilion gramiles are held in suspensibn; the puhnonary veins are next iujocteil with a starch niass in which ultramarineblue granules are held in suspension, ^\^len the injection masses have set the lungs are distended with air through a cannula and tube which is tightlj' tied in the trachea and a stereoroontgenogram taken. WTien this is viewed in the stereoscope it will be



fourul tliat the mass in the arti-ry gives a uniform dense shadow ; the mass in the vein gives a finely granular and less dense shadow ; the hroiu'hi distended with air have their walls well defined.

The use of starch masses for x-ray photographs is not new; but used in the manner I have indicated, by which a differential density is obtained, it is, so far as I know, new. The injection mass is |)repared l)v mixing cornstarch and the pigment with 70 per cent alcohol until the re<iuired consistancj' is obtained. If only the large vessels be desired, a mass that will flow with a pressure of 100 mm. of Hg. will give the desired results; but if the finer vessels are to be injected, the mass nmst be diluted with 70 per cent alcohol and the pressure increased.

No set rule can be given for the preparation of the injection, for I ha\e found that no two lots give just the same results. Each new purchase requires one or two experimental trials before the best results are obtained. This is especially true of ultnunarine blue, and in preparing an injection mass with this pigment a smaller f|uantity should be used than is the case with vermilion.




Department of Anatomy of lh<: University of Oregon Medical School, Portland,



The figure is from a sjnall portion of a class microscopical demonstration which the writer prepared to illustrate the bloodvascular supply of a rabbit's intestine. It also shows much more conclusively than microscopical sections how the absorption surface of the intestine is enomiously increased through the mucosa's being thrown up into numerous villi. When this preparation is examined with a binocular microscope the villi appear like numerous mountain peaks arising from the floor of the intestine.

This demonstration was prepared by first removing the blood from the intestine of a living rabbit by forcing a normal salt solution through the blood-vessels from the mesenteric vein. This was followed by an injection of chrome-yellow gelatin injecting mass, prepared after a fonnula given in a previous paper fProc. Wash. Acad. Sci., 1905), which easily passes through the most minute capillaries. This injection was followed by a second injection of a cannine injection mass (formula given in above-mentioned pajjor) thickened sufficiently with cornstarch to prevent its passing through the capillaries of the villi. When this injection was completed the various mesenteric vessels were iigatetl and a j)ortion of the intestine was removed and placed in cold water to solidify the injectictn masses. From whence it was transferred to a 10 per cent .solution of formalin to fix. Small pieces of the intestine were then dehydrated, cle.ared in cedar oil, and mounted on a slide with djvnunar so that their inner walls and villi faced the cover-glass.



The writer takes pleasure in expressing his acknowledgments to one of his students, Mr. R. G. Young, for his skill in reproducing a small portion of this preparation in a drawing for the accompanying illustration ffig. 1).

It will be seen from figure 1 that two moderately large submucosa vessels (S.A. and S.V.) sujiply the immediate neighborhood with numerous branches {s.a. and s.v.). These vessels are Ukewise confined to the submucosa, but supply each villus with an arterial and a venous branch {V.A. and V.V.), which follow up opposite sides of a villus to its apex, within the connective-tissue layer close to the epithelium. In this preparation the viUi arteries are fully as superficial as the corresponding veins. Each villus artery and vein sends off numerous branches to either side, which break up umnediatcly into a capillary network in the tunica propria close to the basement membrane. This capillary network is considerably finer at the apex of a villus than at its base. Sometimes a number of these capillaries form a continuous vessel (v.a.) which follows the long axis of a villus close to the basement membrane. It should not be confused with the main villus artery or vein.



5-^ s.v.


The figure is from a drawiup of a cleared luicniscopiciil preparation of a small portion of a rabbit's intestine, which had its arteries and capillaries filled with a yellow injecting mass and its veins with a red mass, looking at the inner surface. If any criticism is to be made of the drawing it is that the villi are a little too far apart. The arteries are colored red and the veins blue.

M.S'il., mucosa capillary network .i.e., branch of submucosa vein

S.A., submucosa artery I'. .4., Villus artery

s.a., branch of subnuicosa artery I'.a., small villus vessel

■S.V'., submucosa vein V. V., villus vein



hi; \\\ INC Ai'i'Ai; A rus

WILLIAM l'. \I.M;.\

Deparlmenl of Anatomy <>/ (Ac Cniversilij of Oregon Meilirnl Sclinnl, Purtlttinl,



This appjiratus, as sliow n hy tli(> fip;ure, consists of tho Bausch it Loinh ■■Siiiiiilc (Irawiiifz: (Miuipinciit" supported l>y a stami and a soeoiul stand lioldiiifj an ad.iusial)l(' dra\vin}i;-i)i)ard. The outfit serves every iiMiuii'cnicnt lor rapid and offifiont work, notwithstanding its total cost, cxclusiNc of the niici'os('o|)<>. was alxMit forty dollars.

'i'ho stand lor supi)ortiny: 15ausch iV Loinli's "Simple drawing ciiuipnicnl" has a wooden top (.1) '2'2 \ 14 inches, ahout 4 feet 3 inclies from tiie floor and is supported hy four legs constructed out of old gas pipe. 'I'hese were hent outward at the and strengthened Ki inches up by pieces. On top of the .stand will be .seen the Bausch it Lomb "Simph' (h-awing e(iuipment," which, including i)risni in front of th(> (\v(>)iiece for the microscope, is listed in catalogue at 83"J.o(l. As shown l)y tlie figure, the rheostat (R) is placed at the extreme left, next, in center, is a .small arc light and conden.ser (L), elevated 4 inches by a wooden stool. At the extreme right a microscope is clami)ed on the adjustai)le microscopical stage and bent at right angles so the prism (/•*) extends over the center of the drawing-board. It is an easy matter to ele\ate or lower the microscope by the adjustable micro.scopical .stand .so that the conilen.ser of the microscope (D is in liirect line with the light condenser (L.).

The framework for the adjustable drawing stand was constructed out of old gas jiipe. Its four legs w(>re bent outward at and aliout ti inch(>s from the floor were l)ound together


WII.I.IAM K. \i,i,i;.\

\t\ cross j)ioc('s of the saiin' iiiiitcrial. As will Ix' seen fmiii tlic (ifjiin'. uprJKlits '.\ feel Id inclics were screwed iiitn the re;ir ami (wo siile U'Rs und were cnniieclcd ;il liic tup hy cioss jjieces to ni\(' solidity. Tn proxidc ;iiii[)lc rnimi t'oi- drawing ami (■(Mitciiiiii the iiiianc oil the draw inn-l)oanl the two side lejjs and their uprights were placed 2 inches nearer- I lie hind le^ than the front lep. 'I'lie drawing hoard iD) is of 1 incli eedar, 2() x 21 inches. and is attached to <'ach of the three uprights by a metal daiui) (N), which allows for clainiiinfi; the drawinf^-hoard firmly at any tiesired distance frmn the eyepiece |)risiii iPw so that tli<> size of the imafje c;m be regulated from an inch in diameter to the total width of the draw iiifj-board. The total amount of moxcnient n|)ward or dowiiwai'd of the di'aw in^;-board is 'A feet 10 inches and the adjustable ]nicroscoi)ical stantl cm l)e raised about (i inches, j^ivin^ a total distance of al)out 4 feet 4 inches for projection of the imaKc.

If a greater diameter of the ima^e is de.sired, the adjustal)le ilrawinn stand and tlie prism (/') may lie discardetl and the ima^e can be |)rojected on a \(!rtical surface for or drawing a chart. I'or class demonstration the imago can ho projected on frround ghi'^'* to advantage.

For low magiiifieations it is best to remove the microscope condenser ((') i'lui de|)end entirely on the lantern condenser, or a third ct.ndenser placed before the mieroscop.' condenser will bring .'dxiut the s.imc results.

I'in- I i" friiiii 11 |ili(>ti>Kr:i|ili iif the iiiii'rd.scopic'il priijnctiii)! .iml drawitin apparntii.s. .4. staiiil fur liiiliiiiiK HaiHch ,V l.iinih's simple (lr!i« inn cmiipiiKMil , (', iiiirr<>srii|H- rotiilcnscr, It. :iiljii--litlil(' ili:i«iii)j-l"'ar(l . /,. small aic lidlil ami cimiIiioit: /'. prism iivrr i-vcpiiTc; If, rliiMislal; .V. i-laiiip for Imliliri); ailjiislalilr >lrauiiiK-l">ar<l.



AUTIIOltM MlMTHAl.'r urTlltl PAPftllt INHI^tH liV TlIK UlKt IDimil-lIK: nKRVirK, HKI^KUUCn Iti

A i)riM.i(\'i'i()\ ()i: liHAXCHixc OF Tin: xkiijai.



LaboratiiricH of the Daniel liniigh Inxtilulc iif Anatniiiy of tin- JcjTcrson Metlirnl

College, Philadelphia


111 ;t provinus article' concrrninpc the lesions pnxlufed hy (^ieetricity as ()l)serve(l aft<'r le}>;al electnteutioii. the writer called attention to a diiplicatiou (tf the neural canal in one of the cases studied (Case 111, C. (!.).

This structural peculiarity was first noted in examininjr sections at the level of the motor decussation, and upon careful study it was found at the lev<(l of the .sensor decussation also. The remaininfj; unstained sections also were examined, and that would fiive any assistance in the interjjretation of this condition were .stained and mounted. Of these the best and most characteristic were photographed. The accompanying illustrations represent only the canal area, and they have been so arranfi;ed that the top represents dorsal and the bottom ventral directions. Hitiht and left represent the san\e re.spective directions as far as can be determined.

Tlu^ order in which the illustrations ha\"e been placed .s<>ems to be the proper one according to the characteristics of the rest of each section. The first ei<iht illustrations represent this peculiarity at the motor decussation level and the last four illustrations represent the condition at the sensor decussation level. The order of the se(|Uence here may s(>em odd, perhai)s. but ;in exanunation of the superficial arcuate nucleus and fibers m.aterially assists in the arranjjement , as progressive enlargement of these structures is readily noted. Throughout all of the seclions the canal is much larger than s;>(>nis usual .'itid the direction of the long iixis varies at the two levels here described.

' Aiiiriioan of tho Mcilical .'^ricnces. Sciiti-iiil)rr. 1!UJ.




The first (Mght s<'ctitms \v(m-«' i)li()t()frr;ipli<'<l witli a Ki-mm. ()hjtH-tiv<« uiid SI 10 X ocular, inakiiiji; a luaj^iiilicutioii of al)oiit 12'> times. Ill the romainiiif; sections the canal was larger so that the 7.5 X ocular was used with the Iti-iiiiii. nhjective giving a nuiRiiification of about !K) tunes.

Figure 1. In this section the canal shows a |»'culiar Ixilge, but the ependyiiKi! liniiiy; is distincl and iinl)rnki'ii i linniiiliinil . though cilia are not dcinoiistrable. The long axis of tiu> canal is transversely jilaced and the iiiuuediate neuroglia is rather dense and devoid of nuclei. This seems to be the lowest section of the series.

Figure 2. In this section the canal is more extensive, laterally, than in figure 1, but less so dorso vent rally. Tlu' ependyiiial cells are clear and distinct, but cilia are absent. The neuroglia in the ijmn(Mliate area of the canal is dense and contains very few nuclei.

It is difficult to say which section comes next. If it is figure 3, then the canal has branched and the branches are at about their gT-eat<'st distance apart and are about to approach each other. If figure 4 comes fii-st, then the left canal (/) represents the direct contiimation of the canal of tiic spinal cmd wliil(> the right one (r) represents a blind branch.

P'igure 3. This is the section in which the double condition was first noticed. Here are two distinct canals, the larger of which measui-es about 112^ by (5.'^ and the .smaller 63^ liy 43^. They are about 224 microns apart. The ependymal cells are clear and distinct, but devoid of cilia. In the neuroglia imjuediately surrounding each canal there are many nuclei, whil<! in the central part of the area between them the neuroglia seems ([uite dense.

Kig. 1 level. FiK ■-' Kin -i I'iK I Vig 't

Thp ranal of tho oblonuata :it lower portion of the motor decussation

Tho cuiial of thi' oMonnatii iit n sliuhtly InKli'^r level. The (liiplicntion of the canal. L. left; U. rijsht. the two ennalH closer to each other, rhe curved canal. L, left limb; ){, rinlit linil). Fid. ft The canal at a hinher level (reverse picture) L, left limb; H, riRht limb; a, branch from left limb.



-A.- *






V ■■




FlKUro 1. In llii> section the twn canals arc rather lariie and closo logctluM'. TIk^ cpondyinul colls aic clcai', ilistinct. ami complete «>\(>i\\\\ hci<'. 'rii(> luiclcaf zon<' aroiiiul cacli canal is not so well inaikcd. If this section really precedes (ifiure 'i. then the caiuils are really diverginfi and one will end iilindly. The writer, however, believes that they are conxcriiinfj ;ind t'orni the ptnniliflr canal .seen in fifjure .5.

Iij>;ure .">. In this section the canal i- licnl ;ind the eix' is clear except at the toji of the left linii): here it shows a tendency to send cells into the filial tissu<> as thoiifiii an extension of the canal were indicated. This Ix-comes delinite in higher sections.

Figure (i. This illustration is a reverse picture, so that the parts have Ixmmi labeled as they should be and not as they ap])ear. At this level the left limb (/) shows a tendency to extend the canal dor.salh' (as this liml) points dne dorsally). i'lie right linib is unclianged. From the left limb, howev<>r, a branch (a) has de\'eloped and is quite cxtensi\-e. The ei)eudymal cells are dear and distinct throughout except in /. where an extension of the canal is forniing.

Figure 7. In this section the right limb of the canal is still unchanged. The branch («) is less distinct and it appears as though it will soon be closed i>f!. The left limb shows a welldeveloped extension of cells at h; this outgrowth is as extensive as the canal. These cells seem to r(>present a bud that will become hollow and constitute another branch.

Fig. 7 The canal near the upper part of the motor ileciissation level. L, left liiiil>; H. ri({hl limb; n. branch; h. cell cord extension from left limb.

FiH- s The canal at the upper part of the motor deciissalion level. I/, left limb; It, rijchl limb; «, branch tending to occlusion; h. accessory canal.

Fin. i) Till- oblonxatal canal at the lower pari of the sensor decussation level. I), dorsal; \', ventral; a, dorsal extension of ependymal cells.

Fig. 10 The canal at a slightly higher level. D, dorsal; V, ventral n, accessory canal; b, f-shaped group of connecting cells.

Fig. U The canal al a slightly higher level. I), dorsal; \'. veiitrul; «, accessory cnnal; h, f-shaped group of conned ing cells.

Fig. 12 The oblongata! canal at the upper part of the sensor decu.ssation level. D, dorsal; V, ventral; n, accessory canal.



^'•*«.^« .« BtfA. •


r '^'^ '

II '

^^KS^I ^^^B






l'"inun« S. Ill this section tlic ]\ix}\\ limb is as usual. The hraiicli a swms to lie closing. Tho loft limb (/) has afiain distinct bouii(lari(>s, but a part has bo(Mi soparafcd at I). This rc|)rcsciits the biidileil cells of tli(> ])rece(linji illustration, and fhrso have now formed a little canal: in all i)i(il)ability this is a little blind diverticulum. Wheth(>r the iiranch b is a i)liiid canal or joiiLs the main canal is imi)ossil)l(> (o saj', ;us no other .sections at higher levels show it at all.

The succeeding sections are all at the sensor decussation le\-el. and althoiitih the arrantrement may seem odd it coiifoi-ms to the internal structure of the (il)!on^ata at this ]e\'el.

Figure !). In tliis section tiie neural canal is ajjain single. It is much larger than in the preceding sections even though the magnification is only 90 times. Its long axis is dorsoventrally directed and is peculiarly bent. The ependymal cells are clear and distinct except at the dorsal tip of the canal wli(^r(» they seem to be fonning a dorsally directed bud.

Figiu-e 10. In this section the canal is .seen to be douljle. It seems as though the cells noted in the preceding section had formed (|uite a dorsal extension. These when hollowed out fonn another canal shown here at n. This canal is not coml)letely free from the main canal, as at !> there is the conn<>cting group of cells. The ependymal cells of the main arc distinct and clear except at the dor.sal area where the acces.sorv canal is connected.

Figure 11. This .section shows a main canal -till extensive; its cpend^nnal is distinct except at the dorsal area. The accessory canal at a is much smaller than in the preceding illustration and its epeiidyma is not distinct. It would seem .as though there was a tendencj' to occlusion. At /' tliere is a ["-shaped cord of cells that extends from the left side of the dorsal boundary of the main canal over and then down to the dorsal bouiuhuy of tlie accessory canal. These cells probablj^ represent some of the general mass of buckled cells from which the acces.sory canal is deriv(Ml.

Figure 1'-'. In this section the main canal is large and shows a tendency to branch at the lower right-hand corner. The


('I>rMi(lym;il cells ('\-<'ry\vh<'r(> arc clciir and ilistitict. The accfssdiy canal is larscr llian in the prcccdiiif^ sc'ctii)n and it is well sepamtcd from the main canal. This accessory curia! soems to he a hiind tlivcrticulum that arises at the lower part of the sensor ilecussation le\<>I and extends toward the mid-olivary level. It seeii\s to !)(> constricted and pn>l)al)ly does not reunite with tile main ( at all. Imi < nds either blindly or l»y becoming |)art of I lie liiiirili ventricle as it (the branch) extend^ to liinher levels.


()\ Tin; i'i{()ri:ss OF TMSAiMMiAirwcK or III).;


UlLULU V. .SMITH Thi- Dipfirliiirnt of Anatomy, Tulane Univerxitij <»/ Loiti.iiann


C'onsidoralilo ilifUculty lias hcfii ciicountcrod in tlio classificatioii <if (ishos and their ditToront genera, and especially in drawintl a distincf line between the (l.-uioids and Teleosts.

( i<>'!;enl)aur ('(ili) found in the ( ianoids aiul in a limited number ot Teleo.sts, interralated between the ventricle of the heart and the truncus arteriosus, two segments clearly distinguishable from one another. Of these segments, the caudal is calleil the conns arteriosus or bulbus cordis, while the cranial is called the l)ull)us arteriosus.

The cunus arteriosus is a well-marked muscular structure, furnislied with numerous valves, and is one of tlu; characteristics uf ih(> Elasmobranchs and (ianoids. Teleosts, with the excej)linn of a very limitiMl nujuber, have no conus.

It was onee thought that the rudimentary conus, or its total ab.sence. with l)ut one tier of valves, was characteristic of the Teleost heart. Stannius in liis researches ('54) found that Albula vulpes possessed two tiers ()f valves instc^ad of one. Likewise .S'liior ('()7) found the sanu^ true of Megaloi)s cyprinoides. Pterothri.ssus gissu, and Tarpon atlanticus. .Vmia calva has the shortest conus and a fewer number of valves than any other (Janoid. Some of lh(> Teleosts seem to have descended from a stem somewhat akin to Amia, notably the iierring group. .Sime of I hi fnmihcs of this group, for instance, Albula vulpes, Pterothrissus gi.ssu, .Magaloi)s cvi)rinoides. and Tarpon atlanticus have ahiKist as hirge a comis as .Vmia itself, but with two rows

' .\I:il«'ri!il ilsrd iliicfly finiii the I'nivcrsily and Ili'Ili'Viir lli>s|iil:il Mi'ilicjil Srliool oiilli'ctiiiii.

GG wii.m u t . sMiiii

i)f vnlvos instojid of thnH\ Other Hourly iclaiccl ffcncra luivo a distinct oomis with niily (Hic row of valves.

Hoas ("SO), in his diagram of tho Teleost hoart, shows the valves to arise from fibrous tissue. This is (Troneous, since it has .since been found that all conns valves of Teleosts have tlwir attached to nui.scle with some mu-scle fibers probably extending into the cusps. Boas' diagram for the Teleost repro•scnts the condition found in both Teleosts and Ganoids.

Favaro ('10) states that, from the point of view nf compaj-ative anatomy and embryologj', we are not obliged to recognize in the conus and bulbus arteriosus autonomous and distinct organs, such as are observed in single orders and species, but rather structures which in part correspond to one another. Thereby he contradicts Clegenbaur's statement that the conus and bulbus arteriosus are separate ami distinct .segments. However, when (me considers that Favaro had in mind only the part derived from vascular endothelium and ( iegenbaur that derived from splanchnic mesoderm, l)()th are correct. The intima and media of tlH> bulbus are derived from embryonic vascular endotheliujn and are homologous with the intuna of th(! coiuis and general vascular apparatus, (iegenbaur said that th(> conus difi'<'i-('<l the luillnis in thai the greater ])art of its thickness consists of heart nuiscle. This is (juite true.

Hcjyer ('00) advanced the theory that in Teleosts the conus had disappeared by intussusception into the ventricle.

I have examined serial sections of the hearts of the following fishes: Etnuiieus teres, ("oilia nasus, Clupea harengus, ("tengraulis my.sticetes, Engraulis mordax, Clupanodon coeruleus, Stolephorus compressus, Harengula macropthalma, Chanos chano.s, Opisthonema thrissa, Poniolol)Us i)seu<lnhareugus, Doro.soma cepcdiaiuun, Clupanodon lacepede, Xotopti^rus, Pantodon, MormjTUs caballus, Alepocephalus agassizii, Osteoglossuni bicin-hosujn, Mesopus pretiosus, Thpnallus signifer, Salmo irideus. ('luj)ea alosji, Oncorhynchus chouicha, Osmerus mordax, Balh\lagus benedict i, Nolemigonus chrysoleucus, I"]lojis .sjiiu-us, C'hrosomus erythrogaster, Chirocentrus dorab. Tarpon atlanticus, Hiodon tergisus, C'ampostoma anomalum, Albula


vulix's, I'tciois vdhiiis, I'lotosus anguillaris, Cantheriiies sjiiulw icliriisis, li:ili>t;ii)iis iiinliilut lis. Oslnicioii fornutuiii, ChiK"todmi uctifer, ( 'lijictodoii trifiiscijita, Hctcrotis iiiloticus, Opliioc'('i)liiilus strialiis, Ilippocaiiipus atterimus, SyiiKathus p<'l('aKi<"Us (iastrotokciis l)iaciiloalus, Aiiiia calva. PdIvimIkh spatliula. ami L(>pi( lost CMS plastnstdJIUIs.

'rii(> specimen of Ileterotis iiilotirus and OsteoKlossuiu hicirHiosuiii jn(>asiire(l 4(1 cin. in leiijitli. Tlie others ranged from () to 20 cm. The ilhistratioiis as here re|)ro(hiced arc not all of the sjune magnification. It was con.sidereil most convenient for comparison to employ un\' magnification necessary to distinctly show the conns, when present, and the attachment of its valves, "^rhe (ip;ur(>s are arranged in the order of what s(H>med to be the projiressiv<' disappearance of the conus arteriosus as found in specimens here studied.

I take this opportunity to thank Professor Senior, of The University and Hellevue Hospital Medical Collef^e, for the privilege of using material in the collection of this institution, and I sincerely ili.iiik Professor iJutli, of the University of Manila. Professor .Ionian, of T.eland Stanford University, and Mr. l!atlit)un, of the United Stales Xalional MuM-um. f()r their kindness in .sending me a numl)er of the sp(>cimens emnnerated.

1 tind tliat tlie hearts of Tarpon atlanticus, Osteoglossum l>icinlio.>um, ( )piiiocephalus striatus, Chaetodon aetifer. Coilia nasus. ("hanos clianos, .\li)ula vulpes. Notopterus, Plotosus anguiliari--. Dorosojiui cepedianum, and Ihierotis niloticus, all show a distinct muscular conus. Figure I illustrates such a conns in Dorosoma cepedianum and figun- 2 the similar condition in Ileterotis niloticus. It was deoned unnecessary to .submit illustrations of the remaining specimens mentioned since the coiuis in th(>m was of similar appearance and structure.

The hearts of Jialistapus undulatus. Pterois volaus. Stoh^ phorus c(inipr<'ssus, Pojnolobus jistnidoharengus, Elops sjiurus. ( lupanodon co(>rul(>u.s. Harengula macrothalma. (^i)isthon«>ma, Kngraulis mordax. .Mepocephalus aga.ssizii. and Utengraulis mysticetes show a rudimentary conus of elastic tissue including a small amount of .scattered cardiac muscle. Those are shown in figun>s M. 4. .>, (5, 7, and S. respect ivt>Iy.



•^i rt*

^-f m^'

•1 — ■_




fri— " ■ ^




The figures arc all drawn to incliKlc only llu- cephalic ends of the hearts of the specimens named. The sections from which the drawings were made were selected from serial sections of tlie hearts and. as is evident, many of the sections ol)li<|nely through the cephalic ends. The figures rcpn'sent hearts of specimens as follows:

I Doro.iotna eepedianiim H Clupanodon coerlueus

'2 Hcterolis nilolicus 4 llariMiKula macropthalma


.") CtciiKiiiulis inystii'otcs Kii)craiilis iiiiirdax 7 Opisthonoiiia Ihrissii N .Mcpoccplialu.s aKassi/.ii !l Morniyrus cnballtis 10 Cliipoa harrnEiis

B A. bulhiis arteriosus

II KathylaKUs ln'iicdii-ti I'J SyiiKalhus pclcaKifUs \:i lliiiiliin tcTKiiOis II llippix-aiiipiis attcritntis I.) ('hirorcnlrus <|iiral> l(i raiitodoii


C A, conus arteriosus. V. ventricle 69


III llic iciiiaiiiiiiji iiiiiiilicr of 'I'doosts cxjimiiicil, I he (•(Hiiis scH'iMs to have (lisai)i)('arc(l. The Kioator muiiher of tlicin ai)i)(>ar to liav<' lost it l)y intussusception (telescoping) into the ventricle. In (igure 0. Moniiyrus caballus, the valves are seen to be attached Id iiitussusceptetl conus muscle, caudal to the cephalic (Mid of the he;uM, whih' on th(> other hand, in some that have lost their conus, one finds the conus \alves not to be drawn into the ventricle, as in figure 9 but to be attached to its most cephalic end, as in figure 10, Clupoa harengus. In the hearts of Bathylagus benedicti.Syngathus peleagicus, and Hiodon tergisus, figures 11. 12. and 13, resp(>ctiv(;ly, the cusps are solely bulbar in their attachment and show no evidence of being drawn into the ventricle nor any direct strui-tin'al rehition with it. In Hippocampus atterimus antl Ciiirocentrus dorab the vah^es are merely attached to the bulbus and to the aortic end of the \-entricle with no evidence of being drawn into the latter, while in the heart of Pantodon both cusps are attached to the caudal end of the bulbus, which, on the left side, is drawn into the ventricle.

I beliexe that Hoyer's statement that in Teleosts the conus has been lost by intussusception or recession into the ventricle applies to species of Teleosts, but, on the other hand, it seems to me that in those hearts one .should find the conus valves caudal to the af)rtic end of the heai-t, with their bases attached to the receded conus muscle. In the figiu'es showing no conus and no drawing in of the valves, and in those showing the valves to have only bulbar attachment with no intussusception, it appears to me that in a limited number of Teleosts the conus is not intussuscepted into the ventricle, but is taken up by the caudal elongation of the i)ull)us.

The pheiiomeiirtn of the disappearance of the conus arteriosus without iiitiissuscejjtion of the valve cusps into the ventricle, as nt)ted, may possibly be explained i)hysiologically. One may consider that, on ventricular conti'action when the blood is forced into the conus and bulbus. if during the succ(>e(ling contraction of the bulbus, the greater resistance offered to the passage of the blood through the l)ranchial ves.sels nece.ssitates a


greater pressure ami a hyixTlmpliy <<( tli<- i)iill»ii>. llim there would result a stroiiRer eaiulal reKurgitation of the blood against till' cusps of tlie eoiuis vah'es. Tliis caudal regurgitation might lead to a caudal compression and conseciuent obliteration of the eonus.

I desire to express jny aiijjreciation to Professors H. D. Senior and Irving Hardesty for their aid and >uggestions offered me in the preparation of this paper.


Rdas, J. v.. \' . IS.SO Cher doii C'oiiims arteriosus l)ci liutirinu.s unci bei andern

Kiicicliciifisfhen. Morph. .lalirlj.. Bil. (i. p. 527. KwAHo. (i. 1011 Cono <■ l)ull)o Artcroso negli anamni. Monitore Zoologico

Italiano. Anno. I'J. p. 121. Gegenb.wr, C. 1866 Zur vcrRl. .Vnatomic lies Hcrzcns. Jenaischc Zcit sciirift, Bd. 2, p. 365. HoYER, H. 1900 Bulletin international de I'acadetiiie des Sciences de Cracovie,

no. 7, p. 263. .I<ini>.\.\ .\.\D KvKH.\i.\.N Bulletin of the United States National Museum no. 47. MiKLLER, J. 1846 (ibcr don Bau unil die Orenzen der Cianoidcn. Berlin.

1846. Senior. II. D. 1907 The conus arteriosus in Tarpon atlanticus. Biological

Bulletin, vol. 12, no. 6.

1907 Teleosts with a conus arteriosus having more than one row of

valves. Anat. Rec. vol. 1, no. 4, pages 83-84.

1907 Note on the conus arteriosus of Megalops cyprinoides. Biological Bulletin, vol. 12. no. 6. Stannu'.s 1846 Bernierkuiigen liber das V'erhiiltnis der Ganoiden zu den Clu pciden, insbcsondere zu Butirinus. Rostock, 1846.




WAYNE JASON ATWELL Department of Anatomy, Univcrtity of Michigan


Rocont studies have shown that the epithelial portion of the hypophysis consists of three parts. It has been clearly demonstrated that these three parts are distinct both ontogenetically and histolofiirally. Besides the anterior lobe proper and the pars intermedia previouslj- recognized, Tilney ('13) has shown that a third epithelial part, the 'pars tuberalis' is to be distinguished in mammals and in birds. He gives a brief account of its development in the cat and in the chick. Woerdeman ('14) tre^its of early stages in the development of an homologous part, the 'lobulus bifurcatus,' in reptiles, birds, and mammals. Baumgartner ('16) traces the development of the 'pars tuberalis' in reptiles. Parker ('17) has described its ontogeny in the Marsupialia. In a recent paper ('18) the wTiter has given a detailed account of the development of the hj7)oph}-sis of the rabbit and has followed the differentiation of the three epithelial parts until the tijue of birth.

The most striking feature in the development of the pars tiiberalis is its paired origin. This is noted by all of the authors above enumerated. In the rabbit the pars tulieralis is discernible very early. From the thickened epitheliiun which lies in front of the early Rathke's pocket two thickened ridges are fonned. These are the anlagen which fuse anil form the 'pars tuijcralis' — a thin lamina surrounding the infundibular neck and spreading out under the tuber cinereum.

The pars tuberalis is in many fonns considerably more va.scular than is the pars intennedia. It is further characterizeil by the tubular or acinar arrangenxent of its cells. Tilney




states that tlio walls of these acini are composed of one or two layers of colls, while Parker and Atwell speak of them as being composed of one layer only.

A\'(H'rdeman ('14) draws an interesting homology between the pars tuberalis (which he terms the 'lobulus bifurcatus' following Bolk) and the inferior sacs of the Elasmobranch fishes, which, as is well known, have a paijed origin. Woerdonian lacked material showing the development of the hypophysis in the higher fishes and in the Amphibia. On this account the WTiter felt that such jm homology must be considered precarious until the ontogeny of the hypophysis had been stutlied for these remaining vertebrate classes (Atwell, '18).

The present study was undertaken for the j^urpose of ascertaining whether a homologuo of the pars tuberalis is to be recognized in the amphibian h3'pophysis. In answering this question it was hoped that light would be shed on the question as to whethir a lobe comparable to the pars tuberalis is constantly present in the hypophysis of all vertebrates.


The material for this study con.sists of some eighty series of .sections of lanae of Rana pipiens, which range ii\ length from 2 to 25 nun. This wa-s augmented by the preparation of series of hirval Rana clamitans obtained just preceding and during metamorphosis. For the hypophj'sis of the adult frog, specimens of R. pipiens and of R. catesbiana wore u.sed. For purposes of comparison se\eral series of larvae of a toad (Bufo americana) were also prepared.

A series of wax-plate reconstructions was made from typical hirval and adult stages to illustrate the morj^hogenesis of the hypophysis. The hypophysis of the adult frog was first studied in the gross and sketched at a low magnification, using the camera lucida. .\fter sectioning, the same structures were identified and stutlied under higher powers. Oraphic reconstructions were made and compared with the sketches obtained in the gross.


It is not tho purpose of this study to treat at length the early appc.aranco of tho hy|)ophysis. This has boon done for the Amphibia by (ioette ('7')), Orr ('89), Hallcr ('97), and Corning ('99). Of the largo niunbor of series of sections studied only certain t3'pical stages have been selected for description hero.

3-mm. larva of Raria pipicns. At this stage tho anlage of the hypophysis is already well formed. The ectoderm at the anterior end of the embryo is separated into two well-defined layers. It is from the inner of these that the hypoi^hysis arises. ■ In sagittal sections the hypophysial anlage shows us a wetlgeshapcd mass of cells lying between the neural tube and the wall of the foregut. The apex of the wedge is directed caudully. There is usually considerable separation between the two laj-ers of ectoderm at tho base of tho wedge. The appearance given is that an attempt had been made at evaginating to form an hypophysial pocket.

.V transverse .section from near the anterior end of the hypophysis is presented in figure 1. It is seen that this part of the anlage shows a distinct bilaterality. Reconstructions of the entire hypophysis at this stage show that this bilateralitj* is confined to tho anterior portion of the gland. Po.ssibilitios as to the significance of these appearances are discussed on a later page.

7-mm. larva of R. pipieun. \ sagittal section of a 7-mm. larva is shown in figure .3. The oral plate is intact. The hypophysis, hyp., extends caudally from its attachment to the inner layer of tho ectoderm. It does not reach to the caudal end of the thin-walled infundibulum, inf. .V reconstruction, from sagittal sections (fig. 5) shows that the hypophysis maintains its attacluuent to the ectoderm by a relatively broad band. Just caudal to tho stalk the hypopliysis is widened and .shows a p;ur of thinner lateral shelves. It is bolievotl that give to the more definite lateral lobes of later stages.

S-nnn. larva of R. pipien.'< (fig. ti). The hypophysis has recent Ij' its attachment to the ectoilerm. It lies flattened out under the floor of the infundibulum juul is now somewhat witler from side to side than nasocaudaliv. Its caudal end is now ahnost



«'Voii with tho caiKl.'il (>xtroiiuty of the infundihiilimi. Tho rather sudden breaking loose of the hj-popliysis from the eetoderm has resulted in a marked change in the shape of the gland. Before (fig. 5) it was much longer in a nasocaudal direction. Now (fig. 6)



Fig. 1 Transverse section of anterior end of hypophysis of a 3-inm. larva of Rana pipiens, showing bilaterality of hypophysi.s iinlage; hyp., hypophysis; b.w., briiin wall. X 100.

Fig. 2 Transverse section of hypophysis region of a 12-mm. larva of R. pipiens. l.l., lateral lobes; ant. I., anterior lobe proper; inf.. infundibulum. X 100.

it is spread out from side to side. This rearrangement is a very disturbing factor in any attempt to follow the lateral lobes. They cannot be distinguished with certainty at this stage.

12-mm. larva of R. pipicih<;. A transverse section from an embryo of this length is shown in figure 2, a siigittal section in



Fig. 3 Midsagittal section of head end of a 7-mni. larva of R. pipicns. n.c, notochord; inf., infundibulum; f.g., foregut; hyp., hypophysis; or. pi., oral plate. X 75.

Fig. 4 Midsagittal section of hypophysis region of a I2-nun. larva of R. pipicns. ani. I., anterior lobe proper; p. int., pars intermedia; inf., infundibulum; nc, notochord. X 150.



p. int.

. p. Int

Tif^. 5 Reconstruction of hypophysis of a 7-inm. larva of R. pipona, viewed from the dorsal surface, caudal end at the bottom, l.l., lateral lobe; cut., entoderm; eet., ectoderm; st., epithelial stalk. X 100.

Fig. 6 Reconstruction of the hypophysis of an 8-mm. larva of R. pipiens, viewed from the ventral surface; caudal end is below, hyp., hypophysis; b.w., brain wall. X 100.


figure 4, and a dorsal view of a wax model of the epithelial hypophysis ill fiRiire 7. It will he seen that the hypophysis has its eiiudal end in close proximity to the anterior end of the notochord. The latter, in earlier stages, touchctl the dorsocaudal wall of the infundibulum, causing it to be indented ( fig. '.i). The hypophysis is already differentiated into tiu"ee epithelial portions, one of which is paired. The anterior lobe proper is the thickened central portion of the gland. At its caudal end it is bounded by a bulging transverse ridge, the pars intermeilia. P'rom tlie sides a pair of thinner shelves or ledges (figs. 2 and 7) extend; these are the lateral lobes, the paired anlagen of the pars tuberalis.

18-mm. and 20-mm. larvae of R. pipiens. In these and all subsequent stages the morphological differentiation of the three epithelial portions is very distinct. In figures 8 and 9 models of these stages are \iewetl from the ventral surface, with the caudal end below. The anlagen of the pars tuberaUs are visible as a pair of buds (l.l.) located at each side of the anterior end of the hji^ophysis. At the caudal end of the gland a curving transverse ridge, marked off by a groove, is the pars intermedia. This ridge is considerably longer than the width of the remainder of the gland.

A thickening at the caudal end of the infundibulum, corresponding to the extent of the pars intermedia, is the beginning of the neural lobe. It lies just dorsal to the pars intermedia and so is not visible in a ventral view of the reconstruction.

Fig. 7 Ucconstruction of the epithelial hypophysis of a 12-inm. larva of R. pipiens, viewed from the dorsal surface, l.l., lateral lobe; p. int., pars intermedia; ant. I., anterior lobe proper. X 100.

Fig. 8 Reconstruction of the hypophysis of an l.S-inni. larva of R. pipiena, from ventral surface; caudal end below, ant. I., anterior lobe proper; p. int., pars intermedia; /. /., lateral lobe; b.w., brain wall. X 100.

Fig. 9 Model of hypophysis of a 20-mm. R. pipiens larva, from ventral surface; caudal end below. Abbreviations as for figure 8. X 100.

Fig. 10 Model of the hypophysis of a 22-mni. larva of R. pipiens, from the ventral surface. Caudal end below. .Vbbreviations as for figure 8. X 100.

Fig. 11 Model of the hypophysis of a 24-mm. larva of R. pipiens, from ventral aurface. Abbreviations as for figure 8. X 100.


Already the two show evidences of coming into intimate relation with ouch other.

iSS-mm. and 24-mm. larvae of R. pipiens. Ventral views of wax-plate reconstructions made from these embryos are shown in fipures 10 and 11. The lateral lobes, l.l., which are the anlagen of the pars tnberalis, are larger and are flattened out under the infundibular floor. The pars intermedia conforms more closely to the shape of the neural lobe (fig. 11).

Larvae of R. clamitans and Bufo americana at metamorphosis. Very similar relations to the last are sho^\^l by a larva of R. clamitans possessing very short hind legs. A model from this embryo is shown in figure 12. That the spreading out of the lateral lobes under the infundibular floor is due to active growth is indicated by the irregularity in outline of the lateral lobes at these stages (figs. 11 and 12). Other larvae of R. clamitans obtained at later stages of metamorphosis show the lateral lobes in the process of separation from the anterior lobe.

A reconstruction from a toad (Bufo americana) which had just completed metamorphosis is presented in figure 13. The lateral lobes, l.l., are seen to be joined. They are united to the anterior lobe proper by a single attenuated epithelial strand. The lateral lobes do not long remain united, however, for another toad of about the same age (not figured) shows one lobe entirely free, while the other is attached to the anterior lobe by a narrow strand.

From what has been observed in R. clamitans and B. americana it seems safe to assume that in general the lateral lobes become separated from the remainder of the epithelial hypophysis at the completion of metamorphosis or very soon thereafter.

A dulls of R. pipiens and R. cateshiana. In the adult frog the pars tuberalis, derived from the two lateral lobes, is seen as a pair of epithelial plaques lying close under the 'infundibular' floor some distance in front of the remainder of the gland (figs. 14 to 18). Each plaque is appro.ximately circular, but not infrequently shows a greater diameter from side to side. Each is 0.3 to 0.4 nun. in diameter and about one-fifth as thick. They



Fig. 12 Ventral view of a model of the hypophysis of a larva of R. clamitani with small hind legs (legs 6 to 7 nun. long), ani. I., anterior lobe proper; p. int., pars intermedia; l.l., lateral lobe; b.w., brahi wall. X 100.

Fig. 13 Ventral view of a model of the hypophysis of a toad (Bufo americsn*) which had just completed metamorphosis, n.l.. neural lobe; other abbreviations as for figure 12. X 100.

Fig. 14 Model of the hypophysis of an adult of U. pipiens, viewed from Iba ventral surface, p. f., pars tuberalis; other abbreviations as for figure 12. A-B, plane of section shown in figure 18. X 3o.


lie rioso to the brain wall, indenting but not invading it, figure IS. and are surround(><l I)v tli<' pin mater.

Hy a proper coinl)inatioii of dehydrating and clearing fluids the pars tuberalis may be seen in the gross, preferably with the aiti of a binoeular microscope or a low-power dissecting lens. I'sing a mixture of absolute alcohol, 00 parts, and xylol, 40 parts, it was possible to obtain the camera-Iucida sketch shown in figure 15, which reproduces the ventral surface of the hypophysis of an adult bullfrog (R. catesbiana). Similarly the drawing from an adult specimen of H. pipiens presented in figure 10. V was obtained. To make sure that the parts observed in the gross were the same as those seen in sections, these two brains were later cut into transverse series. A grajihic reconstruction of the caudal half of the so-called infundibulum and adjacent structures of one is shown in figure 16 B, and a single transverse section passing through the center of the pars tuberalis is presented in figure 17.

It will be seen that the two plaques composing the pars tuberalis lie under fairly tliick lateral portions of the 'infundibular' floor. Between these thickenings and caudal to them the floor is very thin, X, Fig. 15. \\']ien observed in the gross this thin floor is almost transparent and this fact, no doubt, has given rise to the rather misleading description of the infundibulum as a "bilobed structure, emarginate posteriori}' and divided by a median longitudinal groove" (Holmes, p. 294). Such relations are not evident from a ventral view of a reconstruction fFig. 14). It seems to me doubtful whether the thickened 'bilobed' portion is properly to be considered as a part of the true infundibulum. 1 would restrict the name infundibulum to that portion possessing the very thin floor, A", figure 15, and exhibiting a single medial thickening at its caudal end, Ih., figure 15 (see also fig. 16). Possibly the bilateral thickenings, from their more anterior position, are to be considered homologous with the tuber cinereum of the higher vertebrates.

Another feature of interest is the presence of small but wellmarked depressions, one above each half of the pars tuberalis, figure 17. Whether these depressions are to be considered



Fi(t. 15 Camera-lucitla drawing of the 'infundibulum' and hypophysis of an adult of n. catrsbiana, from the ventral surface. Caudal end is below, anl. I.. anterior lobe: p. int., para intermedia: pi, pars tuberalis; th., thickening of infundibular floor; A', thin portion of infundibidar floor. X 10.

Fig. 10 A Caniera-Uicida drawiiiu showing a portion of the ventral surface of the brain of an adult H. pipicns. Main body of hypophysis had been removed ; its approximate position is indicated in dotted outline, p.t., pars tuberalis: O.I., optic lobe; //, ///, 1', 17, VII, VIII, cranial nerves. X 10.

B Graphic reconstruction, from transverse sections, of portion adjacent in 'A'.

Fig. 17 Trailsvorse section of the frog's brain shown in figure 16 A. p.t., pars tuberalis; ///, oculomotor nerve; o./., optic lobe; inf., infundibulum. X 12.


homologous with the 'tuberal recesses' of higher vertebrates is uncertain.

The anterior lobe proper has become the most ventrocaudal portion of the gland. This h>be, then, does not deserv^e the name 'anterior' from its adult position in the frog, but from its very evi<lent relation to the homologous lobe possessed by higher vertebrates. The lobe is oval in outline, being shorter in its nasocaudal dimension.

The pars intermedia is a long transverse ridge with bulging lateral terminations. It conforms in extent to the neural lobe which is definitely constricted from the infundibulum donsally and at the sides. In handling, the hypophysis readily separates from the infundibulum, the break occurring at the narrow attachment of the neural lobe (fig. 18). In such a case the neural lobe, pars intermedia, and anterior lobe proper come loose as a unit. The pars tuberalis remains attached to the floor of the part interpreted as the tuber cinereum (fig. 10 A).

In sagittal sections the neural lobe is cut in a direction transverse to its greatest length (fig. 18). As Stendell ('14) has noted, it presents in this section the form of a right-angled triangle with its lying against the pars intermedia.

The three epithelial lobes are histologically distinct. The anterior lobe proper, which has been so named from its resemblance to the anterior lobe of higher vertebrates, is very vascular and consists of cords of cells which are chromophilic and granular. The pars intermedia and the pars tuberalis may both be spoken of as chromophobic portions of the gland. There are, however, well-defined differences between the two. The pars intermedia of the frog, in contrast to the same lobe in mammals, for example, is traversed by numerous blood-vessels. The pars tuberalis, on the other hand, and this is also in contrast to the mammalian gland, is not invatied by vessels at all. This may in part be accounted for by its small size in the frog. Many of the cells of the pars intermeilia (adult II. catesbiana) contain rounded hyalin bodies, considerably smaller than the nucleus. These seem to be more numerous in the neighborhood of the blood-vessels. Such liyalin bodies are not to be found in the pars tuberalis.

iivi'Driivsis OK thA anuka



Entoderm. My observations boar out the assertion <if( 'orning ('99) that entodenn does not enter into the development of the frog's hypophysis after the manner described by Kupffer ( '94) and Valenti ('95). There is indeed a mass of cells, epithelial in appearance, in relation with the anterior end of the notochord in larvae of \{. pii)iens having a length of 2.o to 3 mm. I am not able to find evidence that this mass becomes incorporated into the hypophysial anlage. That this eiiithelium-like mass is comparable with the 'protochordal plate' of other vertebrates, as suggesteti by Corning, ^Irs. Ciage, and more recently by Parker, seems to me not unlikely.



f. I. I ; , ml. I.

Fig. 18 Parntiiedian sagittal section of hyjiophysis of an adult frog (R. pipiens). Plane of section indicated by line A-B, figure 14. Nasal end is at the left, inf., infundibulurn; p.t., pars tuber.ilis; n.l., neural lobe; p. in*., pars intermedia; ant., I., anterior lobe proper, X 50.

The pam (uhcralis. \ two-fokl origin for the hypoj)hysis of Ajnbylstoma has been recordeil iiy Kingsley and Tiling. These observers believe that the bilaterality involves the entire gland and that the two parts soon fuse. The bilaterality observed in the frog's hypophj-sis seems, in certain stages at least, to be confined to the anterior portion of the anlage and does not extend to its caudal tip. I have never seen convincing evidence that the two parts come together and fus<> in the frog.

It is with some difficulty that the two parts may be traced to the bud-like lateral lobes which later form the pars tuberalis.



\\ln'n tho hypophysis anlagp breaks loose from its parent ecto<i«'rrii, there is a very noticeable rearranp;etnent in its shape, as has been notetl. This is a disturbing factor in any attempt to frac<> the history of the lat<>ral lobes. During the stages just preceding and following the separation, however, a somewhat thjjiner shelf is sin-n on each side of the anterior half of the luihige (Hgs. .'> and 7). Tlu>se I have supposed to be the lateral lol)es.

Hy the time the embryo has attained a length of IS mm. the lateral lobes are definite bud-like structvu'es at each side of the nas^d end of the gland (fig. 8). The further development of the pars tuberalis consists in a spreading out of the lateral lobes uinler the brain floor and their subsequent detachment from the anterior lobe proper. This latter occurs, in Rana clanutans and Bufo americana at least, during the latter part of metamorphosis or very .soon after its completion.

In the adult frog the pars tuberalis consists of a pair of epithelial plates, free from each other and from the remainder of the gland. The plates have been described by a few authors, but all have spoken of them as insignificant rudiments and none has considered them homologous with the lateral lobes seen during the development of the hypophysis of higher vertebrates.

The terminology for the various lobes of the hypophysis differs so much among writers that it is customary to gather into a table the names to be found in the literature. In the following table I have included only those authors who speak of three lobes of the .A.nuran hypophysis.

Terminology of the lobes'of the Anuran hypophysis





Caiipp ('89)

Para jiostcrior

ParH anterior

Partes lateralcs

U. Haller ('97)




drOsc (Saccus



Stendrll ('14)

Ilaiipt Liippcn

Zwisfhon Lappcn

Pars anterior or Pars chiaaiiiatica

AtwcU ('18)

.■Vnterior Inbn proi>cr

Pare intermedia

Pare tuberalis



To hoiiK.ldfiizo tlic lobos of the hypophysis throughout tlio vertebrate ehi-sses it is iiocossury to have clearly in iiiiiul both their dovelopmontal and their adult relations. I propose to eiuinierate some of the criteria by which they may be classified in th<^ lifiht of some of our recently acquiretl knowledge. It must be borne in mind that many of the points to be given are as yet supported by relatively few obser\-ations. However, it is hoped that their presentation in concise form may be founfl useful. The adult relations may first be considered.

.Size. In gener.-il the anterior lobe proper is the largest of the three lobes, the jJars intermedia the next smaller and the pars tuberalis the sjiiallest. That these relations may not hold constant in all \ertebrates is quite possible, but they are certainly true for the frog and for some mammals. Woerdeman ('14) believes that the pars tuberalis (his 'lobulus bifurcatus') becomes progressively less important as the vertebrate scale is ascended.

Form and position. The anterior lobe proper tends to maintain an approximately spherical or ovoid shape. It is not mouldeil to any marked degree b\' surrounding structures. It is not intimately attached to anj' tissue of nervous origin and in many forms does not e\en lie in contact with neural tissue. On this account Tilney ('13) has named this lobe the 'pars distalis."

The pars intermedia is always conformed to the shape and extent of the neural lobe. In reptiles, birds, and mammals the pars intermedia is a thin epithelial layer applied to the neural lobe and derived from the superodorsal wall of Rathke's pocket. It more or less completely surrounds the neural lobe, which in these forms has its longest dimension in the sagittal direction. Later in life the pars intenn<Hiia invad«'s the tissue of tlie neural lobe to a considerable extent. In the frog also the pars intermedia corresponds to the neural lobe in shape. Here the latter has its long axis extending from side to side. The par> intermedia has a very similar shape, with the exception that it is


rnundod and bulging where it protrudes beyond each side of the anterior lobe (figs. 14. 15, and 1()).

The pars tul)eraUs Hivewise conforms to the shape of that part of the brain wall upon which it lies. It is essentially thin and hunina-like. In nianunals it surrounds the infundibulum and spreads out under the tuber cinereum. Recon.structed and viewetl separately, it may be described as 'saucer-shaped' (Tilney, '13) with a perforation for the infundibular neck. In reptiles thin bands and zones may be formed about the middle of the anterior lobe in addition to the thin plate which is spread out under the brain floor (Bamngartner, '16). In the Anura the pars tuberalis does not extend dorsal or caudal to the attachment of the neural lobe. It consists merely of a pair of small rouniied plafiues, located nasalward from the remainder of the gland.

It is important to note the relations of the pars tuberalis to the membranes of the brain. It lies in the pia mater in close relation to the brain floor, but does not appear to invade the neural tissue after the manner so characteristic for the pars intermedia.

Histology. Much remains to be worked out in regard to the normal histology of the hypophysis. It is my purpose merely to emphasize the fact that the three lobes may be readily distinguished by their histological characters. Tilney ('13) was first in attempting to classify the differences in structure of the three parts. This he did for a number of manunals and for the domestic fowl. According to Tilney, tlu> three parts may be differentiated as to cell structure and vascularity. The pars tuberalis is more vascular than the pars intermedia and less vascular than the anterior lobe proper. Its cellular arrangement is that of cell masses with occasional small, relatively thickwalled acini. The cells are basophilic with rather scanty cytoplasm.

Th«' histological features of the three parts of the reptilian hypophysis an; thus saimnarized b)' Baumgartner ('16):

The pars intermedia has a laminar arrangement of columnar clearstaining cells. The part derived from the lateral lobes are arranged


in columns (or sometimes acini) of clear-stainine polyhodral cells. Tlif anterior lobe proper i.sfornu'd of columns or acini, witli clear-staining ami (larkly-staininR colls which may be acidophilic or b;i.-<ophilic. In Ken('r:il, the pars intermedia and the parts derived from the lateral buds m:iy bi- considered the chromophobic and the anterior lobe the chromophilic part.

Parker ('17) has called attention to the tubular or acinar structure of the pars tuberalis of the hypophysis in marsupials.

In a recent j)aper the writer has ("Uiphasized the differences in structure^ in the several parts of the rabbit's h>'pophysis at birth. Of considerable interest in distinpuishiiiR between the pans intermedia and the pars tuberalis are the peculiar spindleshaped and branching cells which have been observed in the former (Retzius, Hi^rring, Trautmann, Atwell, and others). They ha^e been called ependymal and neurogliar elements (iStendellj. It is important to note tiiat these cells have never been described for the pars tuberalis.

The frog forms an exception to some of the features found in the Amniota. In this form the pars intermedia is fairly vascular while the pars tuberalis is non-vascular. In the specim<>ns studied (which were obtained in October and November) the pars tuberalis does not show a tubular or acinar arrangement of its cells. The pars intermedia i.s clearly in contrast to the pars tuberalis on account of the numerous hyalin bodies to be found in the former, but never in the latter (R. catesbiana).

It will thus be seen that there is very good agreement among the several vertebrate classes, so far as studied, as regards the main features of structure of the three epithelial lobes.

Development. It is by careful ontogenetic studies that the distinction between the three epithelial lobes is most strikingly brought out. Thus the pars intermeilia, in those forms poss<^;sing a hollow hypophysis anlage. develops from the dorsal wall of Ratlike's pouch. In those forms having a solid hypophysis fundament, as the frog, the pars intermedia is <lerivcd from the dorsal or caudal tip of the soliil anlage. During development, in all forms, the pars intermedia becomes intimately united with the neural lobe.



Tho nnt«^rior lobe propor (iovolops from tho iiiaiu body of the t'pillu'lial anlag<>. In tlio frog tliis is the muklle ami anterior part of the kIuikI with the exception of the hiteral lobes. In vertebrates higher than the Amphibia the anterior lobe is formed mainly from tlie anterior or ventral wall of Rathke's pocket.

'i'lie pars tuberalis has a paired origin, being derived from the lateral lobes, lateral buds, or 'tuberal processes.' It had been lielieved that for niammaR the lateral lobes appear relatively late in development (Tiiney, ^Miller), but the writer ('18) has shown that thej^ are present very early in rabbit embryos. It seems likely that the 'bilateral origin' of the hypophysis observed in Ambylstoma by Kingsley and Thyng is an early appearance of the nnlagen for the pars tuberalis. This, at least, is my interpretation of the bilateral conditions abo\-e recorded for very young embryos of R. pipiens.

The definite lateral lobes are, during development, attached close to the epithelial stalk at the ventral or nasal end of the gland. They may fuse with one another and fonn a single epithelial plate closely applied to the floor of the third ventricle (most .Vmuiota). Thej- maj- maintain their attaclunent to the anterior lobe proper or they may become detached as cell masses (certain reptiles), or as two epithelial plaques (frog). In snakes and in some lizards it is even stated that the lateral lobes disappear entirely (Baumgartner, '16).


1. The hypophysis of the Anura consists of three epithelial lobes and a neural lobe. The lobes of epithelial origin are the anterior lobe proper, the pars intermedia, and the pars tuberalis.

2. Fn.m their development and their mature structure these lobes may be considered homologous with corresponding lobes of the hypophysis in all higher vertebrates.

3. The anterior lobe proper develops from the main central portion of the solid epithelial anlage. It comes to lie caudal and ventral to the other portions of the gland.

4. The pars intermedia develops from the caudal tip of the hypophysial anlage. It forms a long transverse structure conft)rming to the shape and extent of the neural lobe.


"). The pars tuboralis has its origin in the lateral lobes, which appe;ir very early. These two lateral lobes spread out under the brain floor and become detach(!d, forming two discrete, rounded plaques lying close under the brain floor in the pia mater.

6. The pars tuberalis is a constant structure, during development, in the hvfjophysis of amphibia and higher vertebrates. It is characterized by its paired origin (the two lobes appear relatively earh' and have their attachment near that of the epitheUal stalk) by its l;uuin;u- nature and by its adult location in the pia mater covering the tuber cinereum of the brain floor.


Atwell, W. J. 1918 The dovolopnifnt of the liypophysis cerebri of the rabbit (Lepus cuniculus L.). Am. Jour. .\ii!it., Sept. 1918.

Baumgartner, K. \. 1916 The development of the hypophysis in reptiles. Jour. Morph., vol. 28.

Corning, H. K. 1899 I'bcr cinigc Kntwickiungsvorgange am Kopfe der Anuren. Morph Jarb., Bd. 27, S. 173.

Gage, Sdsanna Phelps 190C The notochord of the head in human embryos of the third to the twelfth week and comparisons with other vertebrates. Science. N. S., vol. 24, p. 295.

Gaupp, E. 1899 .v. Ecker's und R. Wiedersheim's .\natomie des Frosches. Zwcite .Abtheilung.

GoETTE, .\. lS7.i Entwicklungsgcschirhto dcr Unke. Leipzig.

Haller, B. 1897 I'ntcrsuchungen iibcr die Uypophyse und die Infundibularorgane. Morph. Jahrb., Bd. 25.

Herring, P. T. 1908 The histological appearances of the mammalian pituitary body. Quar. Jour. E.\. Phy.siol., vol. 1, p. 121.

Holmes, S. J. 1906 The Biology of the Frog. Macmillan.

KiNosLEV, J. S., AND Thyng, F. W. 1904 The hypophysis in .\mblystonia. Tufts College Studies. Scientific Series, vol. 1, no. 8, p. 363.

KuPKKER, ('. V. 1894 Die Deutung des Ilirnanhnnges. Sitz. Ber. d. Gesellsch. f. Morph. u. Physiol, zu Miiiichen, S. 59.

LuNDBORG, Herman 1893 Die Entwicklung dcr Hypophj-sis und des Saccus vasculosus bei Knochenfischcn und .\mphibien. Zool. Jarb., .\nat. Abt., Bd. 7, S. 667.

Miller, M. M. 1916 A study of the hypophysis of the pig. Anat Kec . vol. 10, p. 226.

Orr, Henry 1889 Note on the development of the .\mphibians, chiefly concerning the central nervous syHtem: with additional observations on the liypophysis, mouth, and the appendages of the head. Quar. Jour. Micros. Sc., N. S., vol. 24, p. 275.


Parkkr, K. M. 1917 The development of the hypophysis cerebri, prooral (jut,

and related structures in the Marsupiuliu. Jour. Anat., vol. 51, part 3. Retiics, G. 1S94 Die Neuroglia der Neuro-Hj-pophyse dcr Siiugothicrc.

Biol Untcrsuch., N. F., Bd. 6, S. 21. Ste.vdell, Waltkh 1914 Die Hypophysis Cerebri, .\chter Theil in Oppel's

Ix-hrbuch dcr vergleichenden mikroscopischen Anatotnie der VVirbel thiore. Tll.NEY, Fredkrkk 1013 .\n analysis of the juxta-noural epithelial purtioii of

the hypophysis cerebi, with an enibryological and histological account

of an hitherto undescribed part of the organ. . Internal. Monatschr.

f. Anat. u. Physiol., Bd. 30. Traotmann, Alfred 1909 Anatomie und Histologie dcr Hypophysis cerebri

einiger .Siiuger. Arch, f. mikr. .\nat., Bd. 74, S. 311. Valenti, G. 189.5 Sullo sviluppo doll' ipofisi. Anat. Anz., Bd. 10, S. 538. WoERDEMAN, M. \V. 1914 Vcrglcichenden Ontogenie der Hypophysis. Arch.

f. mikr. Anat., Bd. 8G.



HORTOX R. CASPARIS Anatomical Laboratory of the Johns Hopkins Urtiversity, Baltimore


The significance of the omentum has been since the earHest times ;i matter of imirh theory and conjecture. \'arious protective and other functions have been attributetl to it, but these in the main have btH*n basetl on theory. Only recently has there been definite experimental work to prove the accuracy of some of these assumptions. Most recently the interest has been centered on what part the omentum might play in absorption from the peritoneal cavity. Previous work on peritoneal absorption has resulted in emphasis being placed on the diaphragm as the main pathway of absorption, Adler and Meltzer, Muscatello, and others holding that drainage was b}' way of lymphatics, while Heidenhain, Starling and Tubby, Cohnstein, Hamherger, and Dandy and Rowntree hiivc shown that much of the absorption takes place by way of the blood stream. However, asserticins had been made by some that the omentum aided in peritoneal absorption; but not until Rubin in 1911 showed that loss fluid was absorbed from the peritoneal cavities of animals whose omenta had been amputated than from the peritoneal cavities of normal controls, was there any experimental evidence to support assertions.

Knowing the facility with which material passes from the peritoneal cavity through tlie diaphragmatic lymphatics, the question arose as to whether or not the omental lymphatics might play a like rfile, but only a short siu*vey of the hteratiu-e was necessary to cause the matter to resolve itself into whether or not the omentum contained Ijinphatics. Ranvier found them in new-born kittens, but claimed that they disappear by



rotrofuossion or by somo (It'Rcucnitivo process by the oiui of the third month. Oth«Ts h;ivo in;iiiit:iincd that there are no lymphatics in the onientiun, and some of the earUer workers who made very ext«>nsive studios of the l>TupIiiifir system do not mention the omentujn at all as one of tiie Ijinphatic-contaiuing organs. And just recently Shipley and Cunningham, who have made a careful study of the omentum, proving definitely that active absorption from the peritoneal cavity takes place -by way of the blood stream of the omentum, were unable to demonstrate the presence of lymphatics. On the other hand, Klein, Norris. and others speak of lymphatics of the omentum, but do not give convincing proof of their existence, and it is this latter fact together with the fact that it is very difficult to adequately demonstrate them in a place where one would think they should be easily demonstrable that is responsible for the uncertainty which those fe<>l who have had occasion to think of or look up the matter.

In my first attempts von Rechliiighausen's silver method was used. Omental sjjroads were made, the surface was rinsed with distilled water, a one-fourth of 1 per cent solution of silver nitrate was applied, and after one or two minutes the silver solution was rinsed off with distilled water. The specunen was then iiiunersed in distilled water and exposed to the sunlight until it assumed a brownish tint. sih-er markings were also brought out by th<> arc light, but loss effectively. Ry this method I found that oulj- the peritoneal lining of the ojnentuni was silvered. Then I followed Klein's technique of penciling off the surface of the omental spread with a camel's-hair brush in order to remove the peritoneal covering before applying the silver nitrate. In this waj^ beautiful rich networks of Ijnnphatics were readily brought out in the centrum tendinum of the diaphragm- just such a picture as that diagrammatically sk<itchod by Klein for the omentum, but in no case were lyznphatics found in the omentum by this method except in the new-born kitten where some of the lymph-ve.ssels were present in the very thin meshes of the omentum. However, since the silver did nf)t with this teclini(}Uo penetrate the perivascular fat and the thicker parts of the omentum around the blood-vessels,


it (lid not exc'luclo tlio possibility of thero being l3^nphatics in those ureas. Attempts were then mjide to inject with rolor inass(>s, but with no success. The extremely fine and delicate structure of the omentum made this method unsatisfactory. It was then decided to return to the silver method and attempt to .silver the perivascular structures by forcing the solution through the blond streuJM under increased pressure. Inunediately after killing the animal a cannula was inserted hi the abdominal aorta, the aorta was then ligated below the cannula and above the coeliac axis. A nick was then made in the portal vein to insure a free outflow, and distilled water was passed through the omental vessels to wash them out. Inunediately following that a onefourth of 1 per cent silver nitrate solution was forced through the omental vessels under rather high pressure. One could readily determine the extent of injection by the immediate milky opacity assujned by the tissues which the silver invaded, and this was the guide in regard to the amount of pressure to use. The omentum was then removed, immersed in distilled water and exposed to the sunUght until the brownish tint appeared. The spt^imen was fixed in alcohol and was cleareil by the Spalteholz method, and it was found that the solution had pa.ssed through the walls of the arteries and had invaded and silvered all structures in the perivascular areas. By this method and with local forced injections lymphatics were foimd in the region of the larger l)lood-vessels of the omentum in the rabbit, cat, dog, and in man. The lymphatic endotheUum is typical and unmistakable and is easily distinguished from that of the veins and arteries by its different jjattern. It must be said here, though, that in no case were the Ijinph-vessels numerous. After finding the lymphatics by the above method, it seemed unn^isonable that one should not be able to inject them with color suspensions, but further attempts along this line were practically without success. Whether the fact that removing the omentum from the peritoneal cavity, exposuig it, and stopping its circulation which gives it a tendency to dry rapidly and contract somewhat, would cause the deUcate-wallcd l>inpliHtics to contract and make injection diflicult is only a possible explanation.


Attempts Jit injection did. howevor, bring out one or two inton^stiiiK facts. In the cat und iulult pig, by making shallow injections in the ventral stomach wall along the greater curvature just anterior to the hne of attachment of the omenttim, one was able to get loops of l\inphatics running down from 2 to 4 or 5 cm. into the omentum. From the position of these loop.s it would appear that the lymphatics had been dragged down from the stomach wall in the growth and development of the omentuju. There was no e\ideiice of drainage of lymphatics from the omentum proper into these loops. Also when one injected in the corresponding prepyloric region of the rabbit's stomach, the injection mass would pass from the stomach wall through a single Ijinpli channel which cour.sed down through the ventral leaf of the omentum to its lower splenic attachment. This vessel had no definite relation to any of the large bloodves.sels, but passed across one of the less vascular areas and could not be considered a true omental lymphatic.

Attempts at injection further brought out the peculiar pattern of the blood-A'es.sels of the omentum. In the normal omentimi capillaries are comparativeh' rare, but there is an unusually rich arterial and venous anastomosis.

Reverting to the main subject, at this point it was felt that I had progressed practically no further than others who had clauned positive results. Therefore, in the hope of producing absolutely definite and convincing evidence for the presence of hiiiphatics in the omeiitiuii. it was decided to run some absorption experiments, following the method used by Shipley and Cunningham in their peritoneal absorption experiments. Instead of decerebrating the animal, it was anaesthetized with luminol-sodium (a drug which produces an even quiet anaesthesia which can be made to la-st from one to six or eight hours by regulating the amounts, but from which the anunal never recovers), was then placed in a box in which physiological conditions were maintained, and the omentum was cjirefully drawn out of the animiars body through a midline incision and was kept immersed in a carmine solution. In the .several experiments various lengths of time ranging from one-half to three



hours wore allowed for al)sorpti()ii to take placo and then all the possible Innph glands to which drainage might occur (including the mediastinal glands) were removed, sectioned, and were examined for the carmine granules. Also each time a careful




Fig. 1 Specimen of the oiueiituiii of an adult rat in which silver nitrate (0.2."i per cent) was injected into the arteries to show comparison of the size and shape of the endothelial cells of an artery, vein, aixd lymphatic. The transverse lines on the artery and the vein indicate the smooth muscle. The entire width of the artery is not shown, as part of it was covered by the lymphatic.

search was made for an indication of the channels of absorption, but both here and in the case of the glands exjunined the results were negative. Then it occurred to me that possibly I had failed to find the glands or that the drainage might be direct. Additional experiments were made, IjTiiph was withdrawn


from the cysteriii chyli and sjiicars were made, with the result that a few granules were found each time. At other times l>nnph was drawn from the thoracic duct proper and granules wi're found, but in neither case were they numerous. Otlier exp<'riments wore made whicii were sbnihir to the above with the exception that for the absorption material a true solution was use<l (potassiiuu forrocyaiiiile anil iron aimnonium citrate). (Hands were removed as before and were fixed in hydrochloric acid formalin without the formation of Prussian blue, but when the hnuph was drawn from the cysterni chj'li and from the thoracic duct and was placed in hydrochloric acid fonnalin, Prussian blue was fonned.

These experiments then confirm the findings with the pressure injection of silver nitrate into the blood stream — they show deftnitelj- that Ij-mphatics are present in the omentum and furnish a channel for absorptions from the peritoneal cavity. Whether the lymphatics really play a very active role in this latter respect I am not prepared to say, for while absorption was rather meager in the above experunents it might be considerably more pronounced when the omentum is closed u]) in the peritoneal cavity where pressure conditions are necessarily different, or the meager absorption might have Ixx^n due to the small number of IjTiiphatics present. The experinients also show that drainage from the omental Ijanphatics is into the cy.stemi chyli and on through the thoracic duct and not by way of the posterior mediastinal glands as liyjiotliecated by Crouse in his extensive paper on the great omentum.

I wish to take this opportunity to thank Dr. F. II. Sabin, F*rofessor of Histology in the Johns Hopkins IMedical School, also Dr. R. S. Cunningham, Instructor in Anatomy in the same institution, for their helpful suggestions and criticisms and also for placuig e(|uijimcnt at my dispo,sal.


BIBLIOGRAPHY, I , ANU Meltzeb, S. J. 1890 Kxperiiiu'iital contribution to tlie study

of the path by which fluids arc carried from the peritoneal cavity

into the circulation. Jour. Exper. Med . 1. CoHNSTEl.v, W. 1895 Cbcr Resorption aus der I'critoncalhohle. Zentralbl.

f. Physiol., 9. Crocse, H. 1912 The omentum. Bulletin of the Kl Paso Med. Soc, April. Dandy, W. E., and Rowntree, L. G. 1914 Peritoneal and pleural absorption.

with reference to postural treatment. Annals of Surgery, 59. Hambcriikr, H. J. 1895 L'ber Resorption aus der Peritoncalhohle. Zentralbl.

f. Physiol., 9. Heidenhaix, R. 1891 Versuche und FraRen zur Lehre von der Lymphbildung.

Pfluger's Arch. f. d. Kcsammt. Physiol. 49. Klein, E. 1873 Handbook for the physiological laboratory, I. Muscatello, G. 1895 tjber den Bau und das Aufsaugungsvermogen des

Peritoniium. Arch. f. Path. Anat., 142. NoRRis, C. C. 1908 The omentum: its anatomy, histology and physiology in

health and disease. Univ. of Pcnna. Med. Bulletin, 21. Ranvier, L. 1896 Anatomic gc'nt'rale.— Aberration et ri'-gression des lymph atiques en voie de doveloppement. Acad^'mie des sciences. Comptes

rend us, 122. Rubin, I. C. 1911 The functions of the great omentum. Gynecology and

Obstetrics, 12. Shipley, P. G., and Ccnninoham, R. S. 1910 The histology of blood and lymphatic vessels during the passage of foreign fluids through their walls.

II. Studies on absorption from serous cavities. Anat. Rec, 11.

1916 The omentum as a factor in absorption from the peritoneal

cavity. I. Studies on absorption from serous cavities. Am. Jour.

of Physiol, 40. Starling, E. H., and Tdbby, A. H. 1894 On absorption from and secretion

into the serous cavities. Jour, of Physiol., 140.


AN r.xrsr.M, i^icii r i.rxc

MATTHEW MAUSllALL Anatomical Laboratories, The Schtiol of Medicine, Univernili/ of Pillnhiirgh

In the course of the regular dissection an unusual right lung was found recently. In brief, its interesting features wore:

1. A rudiiuenturv middle lobe so hidden in the interlobular fissure as t(j make the lung appear two-lobed from a sunple surface inspection, and

2. An unusu£ll distribution of the second lateral bronchus.

The inferior interlobular fissure did not reiich the diaphragmatic surface, but tunu^d around the ventral margin on to the mediastinal surface and thence dorsahvard and slightly upward to the hilus. The superior interlobular fissure separating the middle and superior lobes was not evident from the mediastinal surface; that is, there was no evidence of a middle 1<)Ik> from surface inspection of the mediastinal surface of the lung. On widely divaricating the inferior int(Tlobular fissure, the middle lobe could be seen near the mediastinal surface, Ij'ing in a position ventral to a frontal plane passhig through the hilus. It was shaped as a biconvex disc about 1 cm. thick, elongated in a dorsoventral direction, with its medial edge very close to and parallel with the mediastinal surface of the lung, and its lateral edge roughly parallel to the medial edge and about 4 cm. from the mediastinal surface of the lung. The dorsoventral length was about ."> cm. It was attached dorsomedially to the hilus, the remainder being covered by pleura except the ventral extremity, which was fused with the superior lobe. Thus it was incomplctc^ly separated from tlie upper lobe by the superior interloljular fissure.

The relations of the structures at the hilus were those of a nonnal right lung. When traced (h^'ply into the lung substance the relations of the eparterial bronchus and vessels of the superior



lulu- wen- finiiul to lu- iioriiuil. lln' sitoikI hiti-nil bronchus, the nonnal hrouchijil supply of the midillo lobe, was observed, however, to send a large branch to the upper lobe, taking a path through the fusion betwoon the middle and superior lobes, while it supplied oidy a small branch to the middle lobe. The main-stem branches proceeded into ihc inferior lobe in the usual manner.

The tendency of the middle lobe to fuse with the su])eriur lobe is well known. The diversion of the distribution of the larger part of the second lateral bronchus to the upper lobe in this case seems to suggest there is some parallelism between the extent of fusion of these two lobes anil the ilistribution of the sectmd lateral bronchus to the su'perior lobe.

This lung merits recording since: 1) it shows an extreme variation in the relations of the interlobular fissures of the right limg. and, 2) it presents a well-marked case in which an upper lobe of the right lung receives the major part of the bronchial distribution of the second lateral bronchus, and the middle lobe receives the minor part.

autbob's abstkact or thu papbb laat-cD bt





University of Minnesota, Department of Animal Biology


It has been cluiiued that IjTiiphocj'tes and leucocytes of the blood will not take up colloidal dyes in animals stained intra vitam. It is also generally believed that the reaction to colloidal dyes is specific and that it serves to distinguish between l\Tiiphoid wandering cells and fixed cells of the connective tissue and those which have migrated into the tissues from the blood.

At the New York meeting (1916) of the American Association of Anatomists the writer showed that blood-cells, both Ijinphocytes and polymorphonuclear leucocytes, will take up the dye and store it in the form of granules if the cells are isolated from the blood stream. It was shown that if the dye is injecteil into a blood-ve.ssel which has been isolated from the general circulatit)n by a double ligature and the vessel is left in situ for twenty-four hours, it will be found that the poh-morphonuclears have stored the dye in large ((uantities. The hiiiphocytes of the vessel have also taken up the dye. .\nother experiment described at that time consisted of the injection of the dye into the intermuscular connective tissue. Sections of the muscle, which wore removed on tlie following day, contained many poljnnorphonuclears with dye granules. These results were described in detail in a paper published in Vol. 12 of The Anatomical Record.

In that pajier observations were described which gave further support to the view of Evans and Schulemann, that the cellular

' Aided by u grant from tlip llcsearch Funds of tho University of .Minnesota.




reactions find llioir cxpljuiHtion in the colloidal nature of the dyes. Acc<»rdinK to these authors, the dye particles are taken into the cell by phagocytosis and concentrated in cytoplasmic vacuoles in smaller or larper masses which in the end-stages appear as more or less definite granules. The writer reviewed some of the literature on phagocytosis ^nd showed that the reactions of the cells which are primarily concerned in intra vitam 'staining' with the colloidal dyes are exactly what should be expected if the process is one of ingestion and storage rather than of true staining of preformed cj'loplasmic structures. Phagocytosis on the part of leucocytes is of rare occurrence within tlie general blood stream, but in the tissues and in blood sinuses, where the velocity of the current is greatly reduced, the same leucocytes may be very active phagocj'tes. It, was shown that in their beha\'ior towards the colloidal dyes the leucocytes follow this same general rule. The conclusion was reached that the reaction is no more specific than is general phagocj-losis, and that it does not serve as a means of distinguishing between lymphoid wandering cells of histcv genous origin and those which have wandered into the tissues from the blood.

These studies have been continued during the past year with results which add further support to the conclusions reached in the first paper.

In this later work it has been found that the studj^ of the very early reactions following single intravenous or subcutaneous injections is of verj- great importance. This phase of the work seems to have been neglected by nearly everyone who has experimented with these dyes. The bulk of the present paper is devoted to an account of these early reactions.

It has been found that the polymorphonuclear leucocytes are much more active phagoc.ytes for the dyes than was evident from the first series of experiments. In order to demonstrate polymorphonuclears in the bl<K)d which contained dye it was necessary to inject the dye into a segment of a vessel which had been isolated from the general circulation b}^ means of two ligatures. Since then it has be(>n found that it is not neco.s.sary to ligature the ve.s.sel before the injirtion is given if the blood is removed


from the vessel within from one-half to two hours after the injoetion, or if a sepment of a large vessel is tied off and fixed within this time limit.

Rabbits were injected with from 20 to 50 cc. of 5 per cent Trypanblau in distilled water or nonnal salt solution through the marginal ear vein and fresh blood smears examined, or a segment of the inferior vena cava removed within the favorable time limit. Spleen, lung, and mesenteric lymph node were also removed at the same time and fixed in Helly's Zenker-formol mixture. Sixteen rabbits were treated in this wa}' and, with two exceptions, polymorphonuclears containing dye granules were found either in the blood or organs, and usually in all of these locations.

The brief period of time between the begimiing of the injection and the removal of the blood and organs probably accounts for the fact that the dye granules in the poljnnorphonurlear leucocytes are usually smaller and frequent h- paler than is the case when the dye is injected between two ligatures and the vessel left in situ for from nineteen to twenty-four hours. In the sections, and especially in the fresh preparations, one frequently gets the impression that the dye has stained or has been concentrated on the surface of the special granules of the leucocytes. In the later stages, in the case of the doubly ligatured vessels, or when the injection has been continued after the death of the animal, the dye granules are much larger and more irrt^gular in shape than the special granules. Th«' appearance of the granules in the early stages, together witli the; fact that it has never been possible to stain the special granules in leucocytes containing dj'e granules, seems to indicate that the dye granule is built up about the leucocyte granule, the latter serving as a nucleus for the larger dye granule.

There is some variation in the results of these experiments which is difficult to explain. In some cases no dye granules could be seen in any of the leucocytes of the blood or sectioned bloodvessel, but examination of sections of the organs would show that the polj'morphonuclears of the capillaries or interstitial tissue of the lung, spleen pulp, or .sinusoids of the liver contained beautiful


dyo pranulos. In soino other c.isrs the l)hie Kiiinules were f<juiui in the blood, but not in theorgiuis. The usuiil condition, however, wjis to finfl blue pranules in the leucocj'tes of all of these locations.

All of those cells which Evans has grouped under the term 'macrophages' (histiocytes of A.schoff-Ki\'ono) do not take up the dye until later, when most of it has disappeared from the pol>inori)honuclears, which shows t hat the latter are more actively lihagocAtic for the dye than are the former.

\\'hen the blood or organs are examined later than two and onehalf hours after the injection it is difficult to find polymorphonuclears containing dye granules. \n explanation for this is not difficult to find. Immediately after the injection the leucocytes are floating in a .solution of the dye, the concentration of which is relatively high. Consequently the cytoplasm of the leucocytes is in contact with innumerable particles of the dye, and owing to its phagocytic properties is able to take in the dye particles. The dye being very diffusible, its concentration in the plasma is gradually being lowered by diffusion into the tissues.

Evans has shown that the colloidal dyes readily pass from a region of high concentration to one of lower concentration. This probablj' explains the diffusion from the plasma to the tissue fhiids, and also from the pohnnorphonuclears to the plasma, as 80<in as the concentration of the hit ter has been materially reduced. In the circulating blood the poljnnorphonuclears are in contact with the dye during a relatively short period of time, and hence the dye which they have taken up is not so firndy anchored as it is in the case of the doubly ligatured vessel from which diffusion takes place ver>' slowly, resulting in the storage of more dye and the fomiation of more resistant dye granules. In the latter experiment the dye granules of the leucocj'tes are very large and brilliantly colored after twenty-four hours, while the leucocytes of the circulating blood contain no dye granules at this time.

In the death of the animal occurs during the injection, the condition.s of the doubly ligatured vessel arc duplicated. The circulation is stopped and diffusion from the vessels is very much reduced. The leucocytes, of, continue to live for .some lime, and being in intimate contact wilii a high concentration of


the dye in the quiescent state they are able to phagocytose more of it than would be possible in the circulating blood stream.

The complete protocols of the two animals which died during the injection of Pyrrholblau may be of interest. Accidents of this sort rarely occur when Trj'panblau is u.sed, but with Pyrrholblau one must be prepared to lose some of the animals. In this ease the accidents served a good, for they gave further insight into the factors governing the ingestion of the dye.s.

Rabbit 27. 50 cc. of a fresh 1 per cent Pyn-holblau solution in iionnal siilt solution was injected into the ear vein and into both femoral veins. Total time for the injections: one hour and twenty minutes. Animal died one hour and ten minutes after the beginning of the injection. 20 cc. injected after death. Organs removed fifteen minutes postmortem.

Liver: Many polymorphonulears, including those in the larger ves.sels, contain dye granules. No dye in the stellate cells.

Vena cava (section) : Large and brilliant dye granules in nearly all of the pohniiorphonucle^irs.

Spleen: Ver>' hj^ieremic. Some free dye, but none in cells.

Mesenteric Ijmiph node: Negative.

Rabbit 29. 14 cc. of 1 per cent Pyrrholblau in normal salt solution. Animal died fifteen minutes after the beginning of the injection. Injection stopped at time of death. Organs removed two and one-half hours later.

Vena cava : The dye has settled on one side of the vessel. All of the polymorphonuclears which are surrounded by the dye contain largo and brilliant granul(\<.

Lung: Large numbers of ])olym(>rphonuclearsin the capillaries and interstitial tissue. .Ul of them are filled with very conspicuous dye granules.

Liver: Negative.

In the material from both of animals more of the jiolymorphonuclears contain dye and the granules are larger and of greater density than is usually the case when the animals are allowed to live until the organs are removed. In other words, the conditions are more like those obtained by i.solation of a segment of a vessel by double ligature. Reduced diffusibility


from (ho v<>s«'ls and tlio (|uic>scont state of tlie dye-plasma solution :uv probably the dftcrmininp; factors in both cases.

Although the later series of experiments have shown that absolute isolation from the blood stream is not necessarj' for ingestion of the dye on the part of the blood-cells, it still remains evident that a slowing down of the blood current favors phagocytosis. This probably explains the few cases of early stages in which dy«^ praimles were found in the polymorphonuclears of the lung and spleen, but not in those of the vena cava. However, it does not explain the condition in one animal in w^hich pol>7uorphonuclears containing dye were foimd in the vena cava and not in the organs. The number of polpiiorphonuclear leucocytes appearing in the sections of lung and spleen is extremely variable in the different animals. In this case the organs contained very few of these cells, and it is possible that if a greater number of sections had been .studied some pol3Tiiorphs with dye granules could have been found.

In the first paper of this series (Anat. Rec, vol. 12) the A^Titer reviewed some of the literature dealing with the subject of phagocytosis of Uving bacteria following intravenous injection. There is general agreement among the investigators of this subject that the organisms disappear from the circulation within a few minutes after the injection. The reticular cells of the spleen pulp and bone-marrow, and especially the stellate cells lining the sinusoids of the liver, are the most active agents in the removal of tlie organisms from the blood. Polymorphonuclear leucocytes may take up the organisms, but only in those locations in which the blood current is slowed down (sinusoids of liver and suprarenal body, etc.), or after the organisms have invaded the tissues. They are never able to phagocytose the bacteria in the general circulation in the larger vessels.

In order to gain some personal experience with living organisms the writer injected an emul.sion of staphylococci into the ear vein of two rabbits. The results agree with what has already been reported in the literature.

The first rabbit was killed fifteen minutes after the injection. Hlood smears and sections from the vena cava were negative.


Thr spleen contained cocci in its reticular cells, but the mesenteric lymph node was negative. In the lung cocci were found in macrophages and in polymorphonuclears. The organisms were more abundant in the liver than in any of the other organs. They were founil in stellate cells and polymorphonuclears of the sinusoids in the neighborhood of the larger bile ducts, i.e., in the peripheral portion of the hepatic lobules.

In the second rabbit the inferior vena cava and portal vein were tied off and the organs removed about thirtj'- minutes after the injection. Results were similar to those of the first rabbit. Blood smears and sections of the vessels were negative. In the liver the stellate cells and polymorjjhonuelears which contained cocci showed the same grouping as in the first animal, i.e., they were in the peripheral portion of the hepatic lobules in the immediate neighborhood of the interlobular vessels and bile ducts. There were not many cocci in the spleen, but a few reticular cells and ])olymorphonuclear leucocytes contained them. In the lung the polymorphonuclears were verj' numerous, and many of them contained cocci.

The results of the experiments of these two rabbits are quite .shnilar to those obtained with the intravenous injection of colloidal dyes. In the circulating blood of the larger vessels, however, an important difference is noted which requires explanation.

Study of the blood smears showed that the cocci are removed from the circulation within a very few minutes after the injection. This fact, together with the mechanical factor of the rapid circulation of the blood through tlie larg(>r ves.sels, probably accounts for the absence of phagocytosis in the general blood stream.

Colloidal dyes are not eliminated .S(j rapidly, and consequently the leucocytes are floating in the dye for a r(>latively long period of time (one-half to two and one-half hours), and they are coming in contact with innumerable particles of the dye during this period, and hence are able to phagoeytoseit. In regioiLs of slow circulation and in the tissues the leucocytes are able to phagocytose both the dye and the cocci. AMien escape of dye and cocci


is pnvriitod by tho double liguturing of a vossel, phaROcytosis on the part of Iho loucocytos is still more active. For the cocci this was checked by the injection of an emulsion of staphylococci into the doubly ligatured jupular vein of a living rabbit. The vein was left in situ for lo minutes and then removed and fixed in lit lly"s Huid. In sections of the vein the polymorphonuclear leucocytes were seen to contain many cocci.

Before leaving the .subect of experiments with intravenous injections of colloidal dyes it may be well to add a few selected protocols of animals kept alive until the organs and blood were removed.

Rabbit 2/f. 20 cc. of 1 per cent Trypanblau. Animal killed one-half hour later. Inferior vena cava tied ofif and fixed in Helly's fluid. Ljanph node and spleen fixed at the same time. Some polymorphonuclears of the vena cava contain dye granules, but none were found in the spleen or lymph node.

Rabhil So. 38 cc. of 5 per cent Trypanblau in normal salt. Organs removed thirty-five minutes after beginning of the injection.

Vena cava: Dye granules in the polymorphonuclears. One large mononuclear with dye granules.

Lung: \'ery few polymorphonuclears. They all contain pale blue granules.

Spleen: Negative.

Liver: A few stellate cells with dye granules.

Rnhbit 36. 58 cc. of .') per cent Tn,'panblau in normal salt. Organs removed one-half hour after the 1) ginning of injection.

Vena cava : Dye granules in pohmiorphonuclears. M iny of the granules are very large.

Lung: Rather pale dye granules in some of the polymorphonucl( ars.

Spleen and liv(r: Negative.

Liver: \'ery little dye in the organ. Some stellate cells have traces of it.


Rabbit 26. 40 cc. of 1 per cent Trypanblau. 20 cc. injected into each ear vein. Animal killed one hour after beginning of injection, thirty-five minutes after completion. Ligatured portal vein and spleen fixed in Helly. Dye granules in the pohiuorphonuclears of the vein, and the spleen contains a few polymorphs with dye granules.

Rabbit, 26. 22 cc. of 1 per cent Pyrrholblau in nonnal salt solution. Animal killed two hours after the completion of the injection.

Spleen: Dye granules in polymorphonuclears, but very little dye in the reticular cells.

T>iver: Dye in many of the stellate cells and in a few pol^^n<J^piionudears, but it is usually not in the form of definite granules.

Mesenteric lymph node: Very little dye in the node. A few polymorphs with dye granules. No dye in the reticular cells. A few histiocytes containing dye in the form of rf)unde(l masses of granules.

Lung: Great numbers of polymorphs with dye granules. A few small medium-sized l\^^^phocytes also contain dye granules, but there are no granules in the larger macrophages.

Vena cava: Some of the polymorphs have pale blue cytoplasm, but in none of them is the dye concentrated in the fonn of granules.

Rabbit 32. 50 cc. of 5 per cent Trypanblau in normal salt. Material removed two hours after the beginning of the injection.

Sjjleen: Dyt' granules in many cells of the reticulum and in many large macrophages.

Vena cava: Very few poljmiorphs in the sections, but most of them contain dy(> granules.

Lung: Not many polj-morphs, but they contain distinct dye granules.

Liver: Dye in stellate cells. No poljnnorphonuclears.

Mesenteric node: No dye granules in reticulum, but some reticular nuclei are l)lue, and some lymphocytes have blue nuclei or diffusely stained cytoplasm.

112 II \ I- DOWNEY

Rdbhil 33. 40 cc. o i)tr cent Tiypnnblau in norninl salt. Killi'd two hours mid twenty niinutos after the beginning of injection.

Spleen: Dye in the c<>Ils of the reticulum, but the total quantity of dye in the organ is .small.

V«'na cava: Dye granules in the polymorphonuclears.

Limg: Some free dye, but none of it in the polymorphonuclears or other cells.

hiver: Manj- dye granules in the stellate cells.

The abtn-e protocols are sufficient in number to give an adequate idea of the results obtained. It will be nof(>d that there is some variat ion in the re.sults even where the time elapsing between the beginning of the injection and the removal of the organs and blood was the same. Difference in the amount of dye injected does not seem to account for these variations.

When the experiments were first begun the dyes were given in aqueous solutions, but the hemolysis resulting seemed to be a disturbing factor. For this reason the dyes were suspended in n«)rmal sjilt .solution for the later experiments. This does not affect the storge of dye in the cells, but the amount of pigment in the organs seemed to be less when the salt solution was used.

At first it was found impossible* to demonstrate dye granules in the polymorphonuclears of blood smears, even though sections of the same veins from which the blood was obtained showed cli'arly that the poljniiorphonuclears contained dye Later it Was found that the technique of staimng {h.ssolved out the dye. WTien fresh preparations were examined without any stain the dye granules could be seen very clearly, especially when the preparation was examined by the light of a small Spencer lamp placed close to the mirror of the microscope. The heat from the lamp was sufficient to permit active ameboid motion. In these active cells t he blue granules could be followed as they circulated t hrough the cytoplasm and into the pseudopodia. i)repiirat ions can be preserved if they are fixed in heat or in the fumes of full-strength fonnahn and their nuclei can be stained with methyl green, which does not affect the dye granules. Mounting the sme.'irs dry, with just sufficient damar around the


edge of tho cover to support it, improves the preservation of the graiuilos.

As has already been stated, the granules in the poijTiiorphonuclears of fresh smears of the blood from these early stages are frequently very small. In such cases one gets the impression that they are the 'special' granules of the leucocytes which have taken up the dye. However, the very coarse and granules which are often seen, especially in the .sections of the vessels from these stages, are identical with the typical dye granules of the macrophages of the tissues, (iranules of tliis type are abundant in the polymorphonuclears of smears from blood or peritoneal fluid which have held the dye in suspension for a longer period of time. Such an experiment with peritoneal fluid will be described later.

Ill the earlier experiments, described in the first paper of this series, attempts to get dye granules in the polmorphonuclears of the subcutaneous tissue, following subcutaneous injections of the dyes, were unsucces.sful. However, good results were obtained with intermuscular injections, which pro\'ed that pohinorphonuelears th;i,t had migrated into the tissues could take up the dyes under certain conditions.

From the earlier experiments it was evident that a polymorphonuclear reaction always follows the subcutaneous injections. But, altliougli there had been no difficulty in getting the polymorphonuclears to migrate into the tissues, it had been impossible to get them to take up the dye, even though the tissue was saturated with it. The .second set of experiments was undertaken maiidy for the of clearing up this difficulty. That some progress was made will be shown in the following.

At first the later stages were studied, usually after repeated subcutaneous injections of 1 per cent Pyrrholblau. 5 per cent Trypanblau, or of lithium canuine, filtered and u.-^ualiy (fihited one-half with distilled water. The other two dyes were also alwaj's fihered before being injwted intravenously or subcutancou.sly. This is important in view of the fact that it has been claimed that any considerable difTerence in the size of the particles of the suspension will vitiate the results.


PolyiiKirphniiurloar leucooytos wrrp seen in many of these lir<pariHit)ii>, i)ut always without dye granules. The niacrophages or histiocytes (Evans. Kiyono) and fibroblasts seemed to have taken up all of the dye wliieh was not bound to the elastic and white fibers.

The early stages, taken within a few hours after a single injection, gave no better results. The tissue frequently contained jiolymorphonuclears, but they were without dye granules. One significant fact however, was noted in these earliest stages, and that was that the elastic fibers were stained ver>' brilliantly (Evans) and that the white fibers had also taken up a great deal of the dye. Various cells of the tissue had also stained difTusely, but no macrophages had deposited the dye in granular form. It was these observations which determined the continuation of the experiments.

The next st<>p was to get a series of continuous stages, beginning with one hour after injection and taken at definite intervals up to the time that the pohiiiorphonuclears disappear from the tis.sues and the dye becomes deposited in granular form in the macrophages and fibroblasts. It was thought deirable to get as

many of these stages as possible from a single animal and following a single subcutaneous injection, or several injections given at the same time.

The technique for this is very simple. A few cubic centimeters of the dye are injected into the subcutaneous tissue by means of a 'Record' .syringe. When it is desired to examine the tissue, a slit is made in the skin in the region of injection. With fine forceps a small bit of the softer part of the tissue is hfted up and cut out. This is then .spread out as thinly as possible on a coverglass bj' means of smooth, hne needles. Mopping up the excess fluid with filter-paper will greatly aid in this operation, as it will cause the tissue to adhere more firmly to the cover. Before the preparation has completely dried the cover is floated on Helly's Zenker-fornud mixture and Hxed for irom one-half to one hour. The cover and adhering tissue is then washed in running water for two liours. Tlie fis,sue is gradually dehydrated and iodine added when 7U or 8U per cent alcohol has been reached. Ziehl's


carlxil-fuchsiii is a pood stuin for thcsf preparations, as it docs not i-han^c tho color of the dyo Kranules.

The following stages have been studied: 2, 3, 4, 5, 6. 8, 10, 12, 14, 16, 18. 24, 30, 36, 48, 116, etc., hours up to several weeks. Not all of stages were ol)tained in any one rat. The twoto six-hour series was run in eaeh of tliree ruts, and tlie eight- to twenty-four-hour series in each of two rats. Different rats were u.sed for the later stages, and some of the earlier and more important stages were olitained at different times from different rats.

Study of the material from these series soon gave an explanation for the absence of dye granules in the poljnnorphonuclears of the previous experiinents. It was found that the subcutaneous injection of the dje was always followed by the migration of numerous polymorphonuclear leucocytes from the vessels, but that the time at which this took place was subject to coasidcrable variation in the different animals. In some eases it occurred as early as five or six hours after the injection, but in others these cells were not found until the twenty- f)r twenty-four-hour stage hatl been reached.

If the reaction occurred during the (>arlier j)eriods the polymorphonuclears contained no dye granules, and this is also true of those cells which were found in the tissue later than the twentyfour hour period. When the polymorphonuclears were numerous at the 12-, 16-, 20-. and 24-hour periods they generally containetl dye-granules, and it seemed to make little difference whether P\'rrholblau, TrA'panblau, or lithium carmine had been u.sed. In one of the most complete series inunerous polymorphs containing dye granules were found at the 12-, 16-, L8-, and 24-h()ur periods. In this .series it was quite evident that the number of cells containing granules was being rapidly reduced in the later stages. In the twentj'-four hour stage there were still many polymorphs in the tissue, but very few of them contained granules. After this time these cells were seen to degenerate rapidly, and in a few more hours they had completely di.siippeared from the tissue. By this time of the dye had been concentrated in the histiocytes (tig. 3).


Tlu- phfiioiiK'Hon se<'uis to be explained by facts which have aln-ady Iwi'ii partly doscribed by Evans and others. It is evident fn)ni tlie alcove that the time at which the poljniiorplionuclears appear is the most unportant factor. Tliis is related to the well-known fact that in the early stages the dye diffuses through the tissue, staining some of its elements from which it is later released. The .storage in the histiocytes and fibroblasts d«)es not take place until later, although the begimiing of this process was seen in several instances as early as five and six hours after the injection.

Figure 2 represents a portion of a field from a thrw-hour stage for wliich no eounterstain was used and figure 1 represents a fourteen-hour stage, stained with carbol-fuchsin. It is quite evident that all of the elements of the tissue have stained intensely with the Trjpanblau wliich was u.sed for the injection in both cases. The cells as well as the fibers have taken up the dye. Figure 2 shows the intense staining of tht elastic fibers, and figure 1 gives an idea of the immense (juantities of the dye which may be ab.sorbed by the hniiphoid wandering cells of the tissue.

If these two figures are compared with figure 3. which is from a celloidin section of the body wall of a fifte<'n-day stage, it is evident that the appe^xrance of the cells and fibers in the earlier stages is verj- chfferent from what it is in the late stage of fifteen days. In the latter the dye has been stored in granular form. and there has been a tremendous multipUcation of cells. The nuclei are fre«' from dye (they are counterstained with carbol fuchsih in this preparation) and the fibers are barely visible.

Considerableshiftingof the d\-enmst have taken place between the stages represented in figures 1 and 2 and the stage .shown in figure 3. Close study of the .series of preparations shows this to be th«' case. The beginning of the storage of the dj^e in granular form is fretjuently seen as early as five or six hours, but at this stage the number of cells containing dye granules is always very small. Their number increases progressively in the later stages, while the cells and fibers which were stained with it in the earlier stag»'s gradually grow paler. In other words, the dye is being liberated bv those elements which had absorbed it soon after


the inject ion, and ;it the siinic finic it is bfinn taken np rapidly and stored in the form of granules by fibroblasts and rajjidly multiplying 'macrophages.' While this process is taking place there must be a certain amount of free dye present in the tissue fluid, for if the polymorphonuclears appear at tliis time (twelve to twenty-four hours) they will also store it in the fonn of granules. If th(\v come later the dye has already been disposed of by the fibroblasts and macrophages, and there is none available for the polymorphoiuiclears. If they come too early they get none of it, because it is all bound to the fibers and cellular elements of the connective tissue (figs. 1 and 2).

During the time of storage of the dye in the 'macrophages' the number of Ij'iiiphoeytes in the tissue increases very rapidly. One can easily trace all intennediate stages between those which contain only a few dye granules and those without dye. A.schoff, Kiyono, Evans, and others have claimed that the reactions to colloidal dyes give final proof of the existence of two distinct and independent lines of Ij-niphoid cells. Those of lyiiiph-adenoid (hematogenous) origin presumably never store the dye, while those of tissue origin are marked by their great dye-storing ability.

This reasoning is no more logical when applied to the lymphoid wandering cells of th(> connective tissue than it is for the fibroblasts or polyinorphonuclejirs. For the fibroblast Iviyono admits this very fre<»ly. He cites Ribbert as finding that the number of dye granules in the fibroblasts depended on the amount of injected dye .solution. After from six to eight injections most of the fibrolasts contained granules, but there still remained a few non-granular fibroblasts. With direct subcutaneous injections they all contained granules, but the number of granules in the different cells was still extremely variable.

In conunenting on this, Kiyono concludes that the occurrence of the red gi-anules (lithium carmine) depends on the concentration of the dye in the tissue fluid. Kiyono himself found that in the omentum and loose subcutaneous tissue the number of granules hi the fibroblasts is extremelj' variable. In areas of inflammation he found that th(> fibroblasts which h:u\ l^ecome rounded


off coutuin inon- nuiiuTous ;ind coarser granules, which ho believes is related to the stronger nutritive stn-ani through these cells. His conclusion is, that the intensity of the carmine firanulation differs according to the functional condition of tiio fil)roblasts.

To the writer these two conclusions of Kiyono seem quite logical and well supported by facts. However, when Kiyono will not Jiilniit the application of the same reasoning to the h^nphocytes he is sunMy on \ery unsjife ground.

In the first paper of this series the writer showed that lymjihocytes of a doubly ligatured vessel may contain dye granules, and in this later stud}- of the subcutaneous tissue the siune fact is very evident. Tschaschin and Maximow have also come to the same conclusion, the former from a study of the subcutaneous tissue, and Maxiinow from a study of tissue cultures from lymph nodes and spleen. Why the lymphocytes of the blood stream and lymphoid organs should not take up the dyes under ordinary conditions was explained by the writer in his first paper.

Not all of the polymorphonuclears which arrive in the ti.ssue during the favorable time contain dye gianules. If we apply the reasoning of Aschoff- Kiyono to them we must conclude that there are two kinds of p<)l}^no^pllomK•lear special leucocytes, a conclusion wliich even these gentlemen would liardly venture.

The blood-ve.ssels of the subcutaneous tissue are very active in taking up the injected dyes, as is seen from figure 4, which is taken from the subcutaneous tissue of a rat injected with lithium carmine. This will be discus.sed in more detail later, but for the present it is of interest to note that the polymorphonuclears within ves.sels are frequently gorged with the dye. This is shown v(*r>' clearly in figure 4, where the polymorphonuclears are filled with large masses of the carmine. Exactly the same results wen' obtained with Trypanblau.

.\t the Minneapolismeetingf 1!)I7) of the .Vmerican .Association of Anatomists, .\ddison aiul Trorington gave a paperon theeffects of intraperitoneal injection of Trypanblau. They found numerous polymorphonuclears in the peritoneal fluid, but none of them contained dye, while many of the mononuclears did contain it.


III tlir (lisciissioii (if tliis piipcj- the writer siifificstcd tlinf flic l)()lyiii()ri)li(imi('lc>tirs prohubly would Ikivc tiikcii up the dye if it ii.ul l)Oon present in suflieient (|UJintities at the time of their arrival. 'I'he dye disapiiears very ([uickly from the peritoneal cavity. Mueh of it is absorbed direetly by the blood-vessels of the omentum (Shipley aii<l ( 'uiuiinKJiam) and the lymphatics of the central IcikIou of tlic diaphragm, and some is stored by the cells of the taehes laiteuses. clasmatoeytes. etc., of the omentum and by the mononuclear cells which are nonnall}' present in the ix'ritoneal fluid. Unless the poljnnorphonuclcars arrive very soon after the injection they will find very little free dye.

That J)<)l^^norphomIclears which migrate into the peritoneal cavity may take up Trypanblau if it is mail(> available for them is shown l)y the results of the writer's experiment with rats 44 ami 45. These rats wer(> injected intraperitoneally with 5 cc. of a ") per cent Trypanljhiu solution. Sixteen hours later .some of the fluid was removed by means of a fine pipette and examined. It contained numei'ous j)olymoi'phoniiclcai's. l)Ut they were without dye fii'anules. Another injection of 4 cc. of the same solution was then pveu, and fresh prejiarations examined and smears made six hours later. .Vs was to be expected, many of the polymorphonuclears now contained numerous dye granules which were identical with those of the monouclear cells. Many of the granules were suri)risingly large antl dense.

In order to make sure that the marked phagocytosis of the dye on th(> part of the polymorphonuclears was not due to the addition of salt to the .solution, rat 4.3 was injected with Trypanblau made up in distilled water in place of the phy.siological salt solution us((l in of the experiments. No differences could be (let(clcd lietween the material obtained from this rat and that obtiiined from rat 44 which was injected with Trypanblau maile up in sah snhitinn.

The results of this expi-riment set'in to indicate that, in this at least, phagocytosis does not depend .so much on the size of I lie particles which irritate the cell nuMnbrane (.\ddison and Thoriiigton) as it does on the availa'nlity of the material to bo phagocytoscd. It is tniethat thesizeof th;Mibject tob,' jihagocytoscd may



havo soincthinK li> do witli llii' rcacliDii in so tar as particular types of ri'Us an- iiivnlvcd, as is iiulicatrd l)y Mctclinikoff's classilicatinn of the Icunicytcs into ■iiiacroiiliafics' and 'niicmiiliaKos.' However, lliere are plenty of exceptions to tiie rule. TliusSciiott doscrihes I lie pliatocytosis by jiolyinori)honuclears of ontiro erythrocytes whidi were inject<'d into the peritoneal cavity, and Rowley describes a case of anemia in which all of tiie leucocj'tes of the l)lood were veiy active in the jjliagocytosis of red cells. Evidently pha}j;ocytosis depends on soniethinj; nion' tli;in the irritation of the cell membrane l)y particles of varyinji size.

in view of what has already been reporteil in the literature, it is strange that some investigators still believe tliat the colloidal dyes cannot l)e phagocylosed by leucocytes. Hibbert, who was one of the first to use lithium carmine, rejMjrted that leucocytes which enter atretic ovarian follicles contain carmine granules. Pari, a student of Ribbert's, found that by ligaturing the ductus choledochus he could get ])olyniori)honuclears in the heart blood which contained hthium carmine gnmules. In the same animals large Ijinjihocjies, large mononuclears, and transitional cells also contained cannine granules. Loele, working with Isaminblau, inj(>cted one mouse wliich had a dej)osit of pus on the surfaces of the i)leura, liver, and sjjleen due to pleuritis and peritonitis. Many of the polymorphonuclears contained blue granules. I.oclc's interpretation of this is interesting. Tie believes that the i)olyniorphoimelears do not (wdinarily stain with the colloidal dyes, because their granules are able to destroy (reduce) the dyes as rapidly as they come in contact withthem. Intheperiotonitis the leucocyte graiudes are .stained because they are .so injured b\- Iiarniful agents that they are no longer able to destroy the dye.

Cloldniann also ol)served dye granules in polymoriiliomiclears and large mononuclears of the blood after repeated injections of Trypanblau. However, he interprets these granules as pathologic foniiations, rather than as examjiles of true vital staining.

In experimental pneumonia Kline and Wint;'rnitz found dye granules in the involved jiorlioiis of the lungs which wen- cut off from the general circulation by plugs of fibrin in ihr capillaries.


For the lyiiii^hocyfos, Muxiniow scciiis to Ikivc (Icnioiistriifcil hcyoiid (|U(stion that they Jirc ublc to tiikc up the colloidal dyes !ift<T tlicy liavc undcrKoiic further differentiation. (loldnuiun and 'l'-(li:i<clnri >liould :i\«i lie li-lcd anions those wlio l)elieve thai l_\ni|)lioc\tes may phaficx-ylose theeulloidal d\'es, alt hough they I)oth believed that the phenomenon was one of ti'ue vital staining of preformed structures or of structures fonncd under the influence of the dyes.

While hunting through the tissues of different stages for polyinor])honu(lears containing dye granules, some interesting observations were made on the early reactions of the tissue to the dyes. Some of these observations do not agree with what has been previously reported in the literature. This is most of the previous work has been confineil to a study of the later stages, especially after the animals had been given repeated injections of the dyes.

It has l)een generally assumed that ditTuse staining of the cells, ('s])( cialix- of theii' nuclei, is alwa>'s an inilication of a pathologic condition of the cells or of their death. I'ari and others have shown that injured cells .stain diffusely with lithium carmine, l)Ut according to Pari dead cells do not stain at all with the carmine. These observations, together with those of (Joldman. Kiyoiio, Kvans. etc., all ajiply to the later stages.

If this rule holds good for the earlier stages, then we are forced to conclude that Trypanlilau is tpiite toxic for all of the cells of the connective tissue, for in the earlier .stages they all stain more or less intensely with the dye. Especially the nuclei show a great affinity for the dye soon aft(>r it is injected. Fibroblast nuclei are freciuently .stained when their e\toplam contains none of the dye i ligs. 1 and 2, FbL). The free wandering cells <if the tissue and the clasmatocytes usually ab.sorb much more of the dye, as is .shown especially well in the fourteen-hour stage from which figure 1 was drawn.

Th(> upper, large cell of figure 2 shows the beginning of a i)rocess which is nior*- marked in the later stages, esi)ecially in the fourtecn-hour stage of figure I. The miclei of the free cells are verj' dark from the large (luantities of dye which they have ab


sorlx'ii. 'riif pcriplicnil ixirtimi of the cytoplasm is also very (lark, l)Ut that part of it which iiiiincdiatcly surrounds the nucleus is <'oniparativcly free from dye. The picture resembles (ho so-call( (1 Hiincfcid-IIcnscii pictures obtained wh.-^n 12 per cent cane sugar is added to Anipliibian blood. Meves explains the resulting changes of the erythrocytes as follows. The sugar solution produces a firm precii)itation membrane at the surface of the erythocytes. while the hemoglobin in the interior of the cells remains fluiil. \\ith the death of the cell the nucleus swells, and in so iloing it absorbs th(> more (luid hi'moglobin, while the surface membrane and the marginal band prevent the collapse of the corpuscle.

The free cells of the connective tis.sue are e\'idently similarly afTected by the Trypanl>lau, for in general appearance they are ((uite similar to the Amphibian erythrocytes which have been treated with a sugar solution. In lUbS, the writer described somewhat similar ])icluri's wliicli i-csulted fi'oiii the Iri'atment of mast cells with a(|Ueous solutions. The soluble metachromatic sub.stance of the granules dil'fuscd into the nucleus just as does the hemoglobin in the case of the erythrocytes treated with sugar .solution.

The cells of figure 1 seem to have passed through a similar jirocess. The dye has jirobably coagulated th(> pei-ii)lural but enough of it has diffused into lli>" luicleus to cause the lattn- to sw( 11 ami to absorb the sui'rounding mor,' fluid portion of the cytoi)lasm, with the dye which it contained. This condition docs not long, for in the twenty-four-hour and later stages such cells are quite exceptional.

By no means all of the free cells are in this condition. Some of them contain v. ry little (ly>', and in some it has already begun to be deposited in the form of (li>linel dye granules as early as five and six hours after the injection, fn the late stages the numlx r of tin .se latter cells increases rapidly, while the number of the vacuolated cells with the dark nuclei decreases. It is (lifTicuIt to .say just what becomes of all of thes;>. If they all degenerate one would expect to find far more degencj'ating ells in the I)reparati<in than are actually present. .V few degenerating


cells are always prosont, even in the early stap^. but tlieir appearance is very different from that of the cells which contain so much of the dye. Their nuclei become pyknotic and stain intensely with the carbol-fuchsin which was used for a counterstain on most of the prcpanitions, while their cytoplasm contains little or none of the colloidal dye. Such a degenerating cell is shown in figure 1, deg. Its nucleus is the only structure in the field which is stained with the carbol-fuchsin.

Most of the fibroblast nuclei of these early stages also contain more or less of the Trypanblau, but it is quite evident that few of these cells degenerate. Intermediate forms between the dye cells of the early stages and the dye-granule cells of the later stages could not be found, so it would seem, that most of the dye diffuses out of these cells to be deposited later in granular form by other cells. It is quite possible that some of the dye may remain in the cytoplasm of these same cells to be worked over into granules later. However, all of it eventually leaves the nucleus, and much of it is probably set free at the same time that the dye is diffusing out of the fibers, for if the pohTUorphonuclears reach the tissue at the time when the nucleus and cytoplasm of the cells in C[uestion have begun to grow pale they are able to get enough free dye to be able to store it in granular form.

In these early stages the endothelial cells of many of the capillaries and smaller vessels are in the condition of the large, dark, vacuolated cells of figure 1. Later they lose every trace of the dye. The statement which has often been made, that endothelial cells of the vessels do not store colloidal dyes, applies, therefore, to the later stages only.

Although it can be shown that cells hke the ones illustrated in figure 1. with the nuclei filled with dye and dark peripheral cytoplasm, later give up the dye and recover, still one could hardly claim that they are normal anil healthy while in this condition. It is quite likely that they have been injured by the dye, and that the staining of their nuclei is due entirely to its toxic effects. The presence of luuiierous cells with df-nse pyknotic nuclei staineil deeply in carbol-fuchsin, cells which are clearly degenerating, is further evidence of the toxicity of the dye when it first


coiiH's in contact with the tissues. It is clear that some cells are killed, and that others are able to rid themselves of the dye, especially that which has been absorbed by their nuclei, and repain their foniier nonnal appearance.

Anions the cells which show the most injurious effects of the dyo are the smaller wandering cells of the connective tissue and the eosinophil leucocytes, which are always abundant in the normal subcutaneous tissue of the rat. The behavior of the latter is of special interest, for in them we have a case of true staining of preformed struct iu"es which is almost invariably followed by the death of the cell.

All through the earlier stages, and in some cases even in the very late stages, granular cells were found w'hose nuclei were in various stages of degeneration. The most common fonii of cells is illustrated in figure G. Its nucleus is very pyknotic and stained intensely with earbol-fuchsin. The cytoplasmic granules have stained with the Trypanblau used for the injection. In many other cells of this same tj'pe the nucleus is still smaller or it has broken up into several pieces. The staining reaction of the granules varies somewhat, although in the cells with pyknotic nuclei they are alwaj's quite dark, and they never take any of the earbol-fuchsin.

Numerous cells of this type had been observed before it became possible to classify them, although it was evident from the first ones observed that they did not belong to the dye-granule cells (histiocytes, macrophages) shown in figure 3. The regularity of size and shape of the granules, and the comparative freedom of the intergranular protoplasm from dye, together with the pyknotic imcleus .stained in earbol-fuchsin and the fre(}uent perinuclear .space (fig. 6), immediately placed these cells in a group by themselves. It was not until cells like the one of figure 5, and others grading from this to the type drawn in figure 0, were found that it be<'aiiie jiossible to classify tliis group.

The only granular leucocytes of the rat which have a ringshaped nucleus similar to the one of figure 5 are the eosinophils. Comparison of the eosinophils of untreated nonnal rat ti.ssue with the cells in question in the vitally stained animals .shows that the latter are eosinophil leucocytes.


ThcmicU'Usof tht'eosinophils never stains with Trypaiiblau. but their jaanules are often brilliantly stained, even before the nucleus shows any degenerative changes. As the granules take more of the dye and enlarge somewhat Cfig. 6) the nucleus degenerates rapidly, and finally becomes reduced to a small, dense, pyknotic sphere. This is the condition of most of the eosinophils encountered, although cells in more advanced stages of degeneration are by no means rare.

In the eosinophils we have an example of true vital staining of preformed granules with a colloidal dye. This must be clearly distinguished from phagocytosis aJid storage of the dye as it is illustrated in the cells of figure 3.

Since the cell of figure 6 is taken from an eight-hour stage, it is evident that the toxic action of the dye may cause verj- rapid degeneraton of the eosinophils. However, the cell of figure .5 was found in a preparation of a fourteen-hour stage, from which we must conclude that either the eosinophils of the injected area are not all killed at the same time or that others migrate into the region of greatest dye concentration from outlying regions of lower concentration. The large ninnber of degenerating eosinophils in some of the later stages would seem to favor the latter view.

That true vital staining of preformed structures may occur while other cells are phagocytosing the dye and storing it in granular fonn is by no means a new idea. Evans and Schulemann ('15) admit this very freely when they claim that the granules of the epitht^liimi of the plexus chorioideus, hypophysis, epithelial body, and adrenal body are vitally stained 'secretory droplets.' They compare this staining with that of the elastic fibers and cartilage matrix, for which it is not claimed that it is due to phagocyte wis.

The eflfect of the dye on the eosinoi)hils, the numerous degenerating cells of other types, the staining of the nuclei of the fibroblasts and hmiphoid cells, together with the frequent vacuolization of the cytoplasm about the nucleus of the latter in the early stages, seem to indicate clearly that the dye has a decidedly toxic action when it is introduced into the tissue. Th(> eo.sinophils, however, seem to be the onlv cells which are unable to recover


fmm tlu' «>ff("<'ts of it. Lutrr, when all of the dyt> has been stored in Rranuliir fonn in the fibroblasts and histiocytes it seems to bo conijiarativcly harmless, for degenerating cells are not so numerous, although they are never entirely absent from the preparations.

The contUtion of the blood-vessels of the subcutaneous tissue is of special interest in view of what has been n-ported for the blood-vessels of the omentum by Shipley and Cunningham. These authors proved exp.'rimentally that not only the capillariee, but also the larger veins and arteries of the omentum are verj- active in the absorption of solutions of foreign matter from the ptritoneal cavity.

The writer's experiments with sul)cutaneous injections of colloidal dyes show that the vessels of this tissue possess similar absorptive powers. M(»st of the material which was gathered within from ten to twenty hours after the injection contains snialhr and larger vessels which are filled with the dye. This conditfiu is of sueh fr-cpi^'nt oeeurrenee tliat it ean hardly bo due to accidental puncture of the vessels during the injection. The dye seems to pass through the endothehal cells, for in the earlier stages they are stained intensely with it. I^ater, as in the eighteen-hour stage illustrated in figure 4, the endothelial cells are free from dye, but there is a groat deal of free dye in the lumen of the vessel, and the polymorphonuclear leucocytes of the vessel eontjun dyv' granules. The mat(^rial was taken from a rat injected subeutan^ <»usly with litliium carmine. A mass of free cannine (Car.) is shown in the lumen of the vessel, and the polyninrphonulears (Pm.) contain large cannine granules. The endothi'lial c.-lls of the vessel are stained with the hemalum used as ;• eounterstain, but not with the carmine.

The vessels of nunu rous rats injected with Trypanblau show this sjuiK" condition, so it is evidently not of accidental occurrence. A nirchanism for the rapid distribution of the dye to all parts of the b)dy is. therefore, supplied by the blood-vessels. The exi.stence of >uch a mechanism seems necessary in order to account for the extreme rapidity with which some of these dyes are able lo reach the most remote parts of the body following single


intraporitonoal or subcutanoous injoctions. The part phiyctl by the blood-vosscls would !)»' surprising if the work of Shi pl«\v and Cunningham liad not aln'ady demonstrated their al)ility to absorb foreign fluids.


Poljinorphonuclears are ver>' active as phagocytes for Trj'panblau and lithium carmine, but in order to demonstrate this in the general eirculation the blood from the larger vessels must be examined within from one-half to two hours after the intravenous injection. Pohnnorph<inuelears in the lung, spleen, and liver removed within this time limit may also contain dye granules before the reticular cells of these organs have gathered up any of the dye.

Polymoj'phonuclears in the vessels and tissues of the organs will also phagocytose living bacteria injected intravenously, but those of the general circulation will phagocytose the bacteria only when the latter arc injected into a doubly ligatured vessel or after they have reached a vessel or sinus in which the velocity of the blood current is greatly reduced.

Subcutaneous injections of the dyes always cause the migration of numerf)us pol\inorplionuclears from the vessels, but the time at which this reaction takes place varies in different animals. If the leucocytes are numer<jus during the twelve- to twent3'-four-hour period they usually contain dye-granules.

Pol^^llorphonuclears of the peritoneal cavity will als(» take up Trypanblau and lithium carmine if the ilyes are injected after the leucoc^ies have reached the peritoneal fluid.

Phagocytosis of colloidal ilyes by jiolymorphonuchnir leucocytes depends entirely on the availabihty of the dye rather than upon any inherent physiological lUfference with respect to phagocytosis between these cells and the mononuclear macrophages of the tissues.

The three colloiilal dyes used in these experunents have a decideilly toxic jtction on the tissues when they are first introduced. .Vll types of cells in the connective tissue show diffuse .staining of cytoplasm and nucleus in the early stages. Many

128 11 AL DOWNEY

(t'lls iurv killed, but members of all groups of cells are able to eliiuinnte the dye and rt^cover, except mg the eosmophils. All of the »'<)sim)phils located in the region of the hijection are killed by Tryi)anl)iau, hut degenerative ehanges in their nuclei do not occur until after their granules have become brilliantly stained with the dj'e.

Bl<K)d- vessels of the subcutaneous tissue absorb large quantities of the dyes and probably carry them to other regions of the body. Endothelial cells are stained in the earlier stages, but not in the later stages when the vessels are full of dye. The polymori^honuclears within the vessels contain dye granules.

The experiments have brought additional evidence in favor of the view expres.sed in an earUer paper, that the intra vitam reaction to colloidal dyes is not specific, and that it does not serve to distinguish between blood-cells and tissue macrophages (histiocj^es of Aschoff-Kiyono, macrophages of Evans).



For titles not listed here see bibliography of the first paper of this series, Anat. Rec, vol. 12, no. 4.

Downey, H. 1913 The development of histogenous mast cells of adult guineapig and cat, and the structure of the histogenous mast cells of man.

Folia Haematol., Archiv, lid. 16.

1917 Reactions of blood- and tissue-cells to acid colloidal dyes under

experimental conditions. .\nat. Uec, vol. 12, no. 4. LoELE, W. 1912 Uber vitale Granulafarbung mit sauren Farbstoffen. Folia

Haematol., Archiv, Bd. 14. Meves, F. 1911 Gesammclte Studien an den roten Blutkdrperchen der .\mphi bien. Arch. f. Mikr. Anat., 1, Abteil., Bd. 77. P.\Ri, G. A. 1910 Uber die .Vnwendbarkeit vitaler Karmineinspritzungen fur

die pathologische .\natomie. Frankfurter Zeitschr. f. Pathologic,

Bd. 4. RiBBERT, H. 1914 Die .\bscheidung intravenos injizierten gelosten Karmins in

den Geweben. Zeitschr. f. Allgem. Physiol., Bd. 4. Rowley, M. W. 1908 A fatal anaemia with enormous numbers of circulating

phagocytes. Jour, of Exper. Med., vol. 10. Shipley and CcN.saNGHAM 1917 The histology of blood and lymphatic vessels

during the passage of foreign fluids through their walls. .\nat. Rec,

vol. 11.



1 Siihcutanoous tissue of rat. fourteen hours after subcutaneous injection of Trypanblau. The pyknotic nucleus of the deKeneratlnu cell, Deg., is the only structure stained with the carbol-fuchsin used as a counterstain. All of the other structures are stained with the Trypanblau. The large dark cells with the clear perinuclear space are of special interest. They are larp- lymphoid cells which have ab.S(>rbed great quantities of the colloidal dye. Fbl., fibroblast nuclei.

2 Subcutaneous tissue of rat. three hours after subcutaneous injection of Trypanblau. No counterstain. Shows staining of all of the fibers and cells of the tissue. A large lymphoid wandering cell which ha.s absorbed a groat deal of the dye is shown in the upper, left-hand portion of the figure. The nucleus cif this cell and of the fibroblast {Fhl.) stained very dark.


i:.\i'i:it!.Mi;N IS with colloiuai. dvics












® *





^ « 





He\«n A. Sanborn del




3 .\l>ili>minal wall of rat. ("clloidiii section. Six subcutanoous injections of Trypanlilaii nivcn at intervals of from two to three days. Animal killed fifteen days after first injection. Coiintcrstained witli carbol-fuchsin. Macrophages and Unn|ihocytes of various sizes loaded with dye granules. The fibroblasts (two in the center of the field) also contain numerous fine dye granules which are very different from of the macrophages.

i IJIood-vessel from subcutaneous li.ssue of rat, eighteen hours after subcutaneous injection of lithium carmine. Counterstained with hcnialum. Shows absorption of the dye from the tissue. Car., a mass of free carmine in the lumen of the vessel; I'm., polymorphonuclears containing large carmine granules; Ery., erythrocytes.

5 Eosinophil leucocyte from subcutaneous tissue of rat. eighteen hours after subcutaneous injection of Trypanblau. Counterstained with carbol-fuchsin. (iranules sl.'iined blue with the Trypanblau, nucleus stained red with the fuchsin.

ti Degenerating cosiMophil leucocyte with pyknotic nucleus from subcutaneous tissue of rat. eight hours after the subcutaneous injection of Trypanblau. Counterstained with carbol-fuchsin. CSranules blue, nucleus red.




I'l.ATI. 2





.* <» •

.. #

• 4 t






jlcUn M.6&nborn d«l




E. D. COXGDO^f Anatomical Department of Leland Stanford Junior University, California


The long-standing differences of opinion regarding the structure of adult heart nuisole have not as yet been brought any nearer to a reconciliation by the study of its simpler make-up in the embrj^o. In the various accounts which have appeared of the histogenesis of muscle in higher vertebrates, the mj'ofibril has been derived in turn from granules in rows or in groups of four, from rods and filaments, from nets combined with granules, and from a honeycomb structure either combined with or free from granules. No one view has ^ecei^•ed the adherence of even the majority of the more recent writers. The obser\-ations which follow relate to the earlier stages in the development of the heart muscle and were made upon chick embrj-os of the first four days of incubation. They suggest a reconciliation of the divergent views regarding the early contractile structure.

Photographs were emploj'ed to mea.sure certain elements of the sarcoplasm and also as a control to the conclusions based upon microscopical study. Several po.ssible errors need to be considered before the trustworthiness of this method of measurement can be admitted. Spherical aberration was avoided by using <inly a small central ar(>a in the microscopic field. Foreshortening was readily eliminated by measuring only those planes and lines all of whose granules were in sharp focus. It is not po.ssible, unfortunatel}', to do away completely with changes in the (liincnsioiis of the sarcoplasmic structures during preparation



f<ir thf slid«'. The usual Imlanoc was maintained, howovcr, botwtH'U tho fixiuR agents which contract and wliich swell the tissue. Shrinkage due to the passage of the tissue through the paraffins was reduced to a miiiiinuni by shortening the process to four minutes. The preparati<ins give but slight evidence of this defect.

The fixation was by Zenker's fluid or a combination of osmic acid, pota.ssium bichromate, acetic acid, and normal salt solution. Tiie .second fonnula was found the more serviceable of the two. The best results were obtained with the osmic acid only when its temperature was controlled during fixation. The staining was by iron haematoxylin. Its entire manipulation was carried on in 70 per cent alcohol, as water has an unfavorable action upon the mitochondria. It was impossible to obtain entirely .satisfactory fixation of the older heart because its bulk is so great as to prevent an exact control of the action of the osmic acid. The sarcoplasmic structure if well preserved is so dense that microscopic sections of 5 micra were found too opaque to give a clear view of the details. A thickness of 2| micra was the most satisfaotorj'.

During the developmental period ending with the fourth day a primitive .sarcoplasmic structure is present which gradually undergoes changes adapting it to its function of rhythmical jjulsation. The heart mu.scle in ten-day chick is made up entirely of this primitive type. It can be traced back to the earliest distinguishable heart rudiment. In optical section the .structure is made up of a series of granules staining with haematoxylin and connected by fine lines (figs. 1 and 2). It extends throughout the cj'toplasm. The areas included bj^ the lines are usually parallelograms approaching the form of a square. They are approximately 0.8 micron on a side. Triangles and various tj'pes of quadrilaterals also are seen.

The granules are shown well by an osmic bichromate fixation when followed by haematoxylin. The higher alcohols, oil of biTgamot, and xjlol partially dissolve them. Witliout cnlering into a di.scus.sion of the variations in definition of the term mitochondria, it is clear that it will be in agreement with the views of


I'lci. I riansvrrsc scitinii ul' priiiiilivc stn:ik ni i-liick i IckusciI al surface of spot ion). X (MX).

Fl(J. 2 Transvi'isc section of liearl of a len-soinile chick. X IKX).

Kip. 3 Growth in iilasiiia I'ulttirc of heart tissue from a chick of four days' inciiliation. X -70.

13S K. 1). <'ON(iDON

most of those occupied willi llie siil»jecl if (he term is jippHed to the pramiies.

The network wiiicli lias been deserilx'd could be pinduced as the optical section of a mesh in three tlimensions or of a system of jilanes completely enclosing protoplasmic bodies. The decision between these alternatives in the case of various tissues has long occupied the attention of those attempting to detennine the ultimate microseo|)ic structure of protoplasm. It is well known that the value of ob.servat ions upon the (incr details of fixed protoplasm as a means for determining the make-up of its living structure has been called seriously into (piestion by the di.scovery that various coagulation patterns can be produced in solutions physically similar to protoplasm by the use of difTerent conditions of fixation. It is of interest, then, that the hexahedra are fountl no matter whether osmic, formol, or coi-rosive sublimate fixing fluids are used. Also three observers lunc given good reasons for bs'lieving that it gi'adually changes during development into a structure which few would denj- as an actual element of the living adult muscle, namely, the myofibril. Evidently, then, we are not d(>aling with an artefact and the relative likelihoods of the net and the honeycomb being the reality in the sarcoplasm has a vital interest not only in r.'lation to the make-up of muscle, but also because of its bearing upon tlie ultimate structures of protoplasm.

Of the three observers who describe tiie network in optical section, MacCallum refers to the sarcoplasmic structure of the embrj-onic pig heart as a .Vt the same time he apparently believes that it contains flat membranes, since Ik^ speaks of .sarcoplasmic discs surrounded by the net. Briick also uses the word net in his discussion, ^'et he (Icfinitely states that then is a honeycomb structure. Wieman believes that in the chi<'k heart there is a net in three dimensions and applies to it the name cytoreticulum. It is n iufen-etl that they are too dclicati- to I),- visible except when jilaced nearly edgewise to til,' observi r. A ^ireat majority of t he spaces must be six-sided, since their cri)ss-section is so uniformly a (|uadrilateral figure. The structure, then, can well be described as hexaheilral.


A controversy n'gariling the distinctness in the .separation of the cell areas has been waged sine? the time when the cell theory was formulate<i. The d-cision in the adult heart depends on the interpivtation of the intercalated discs. These do not enter into the {[uesiinii for the eiiibryoiiic oi'-ian because they make tlieir a]ii)(arance late in development. It is generally agreed tiial wlun it is first distinguishable the h(>art rudiment, like the mesenchyme, is made up of cells s;'parate except for a few fine proc:ss3s. A iiuiiil)er of observers of heart development in birds and nianmials believe that flu'V soon fuse. In this group are Chiarugi I'STj. Hoycr ("01), Heideidiain (,'9*J), Kurkiewicz ("09), and Duesberg ('10). .V loss of identity is not admitted bj- MacC'allum ('97). Wieman ('07), and Schockaert ('09).

A comparison of sections of youngi-r ancl older heart ventricles indicate the possibility that there are .slight distinctions between the (1. gr. e of ell indcp. ndonce at three succe.s.sive ages. .Vt no time, hoWv'Vi r, was a cell membrane made out. In the ventricle of a tcn-.somite chick (here is a suggestion of cellular independence consisting of an arrangement of hrxahedral structure concentric to each nucl( us. \\hen a region is found, however, where the condition of tile planes can be well made out all the way from one imcleus to another, no break in tlidr contiiuiity is discovered or ol h( r modification to indicate a cell boundary. There is a lobatioii of the surfa<'e of the myocardium corresponding to each subjacent nucleus. The heart of .somewhat ohler embryos, including


the tliirtfH^n-somito staRo, contains small clefts in tlic interior nf tlic niyocanlinni partially iliviiiinK coll aroas(fi)i. 'Ji. Kurkicwicz ("OU) has (li'scril)t'(l thcni at Iciifith. l^vidcntly, then, the apparently coniplcto continuity of the sarcoplasni in tiic jjrcvious stage is tleceptive. The four-day heart no longer contains clefts. There are portions where the planes of the liexahedral struclure take a perfectly rectihnear course past several miclei (figs. 4 to 0). If our conehisions were to he l)ased on these localities alone there could be iiltlc doul)! of tlic complete fusion of tlie cells. Other regions have a more irregular ajipearance.

In a previous paper ('15), which describes the migration and growth of four-day chick ventricle in plasma cultures, it was found that the sarcoplasni extended out into tlie clot in anastomosing multinuclear columns whose only indication of cell Ixiundaries (^lig. 'A) are light constrictions of the columns. The complete cohesion of the tissue under conditions certainly speaks for a close union between its cellular elements.

.S'hockaert ('09) gives the only specific statement the writer has found that there are cell membranes in the embryonic heart. She publishes a photograpii of a section from a .seventeen-day ralil)it lieart which contains markings roughly simulating cell boundaries. The finer structure of the sareophisin is too completely lacking to make pos.sible an understanding of its original character. In Wieman's study of the chick lunut he does not dcscrii)e cell memiiranes, l)ut he has obtained mononuclear fusiform .siircoiilasmic bodi(>s by maceration. Thes(> caimot l)e exphiined away as coming from regions of the atria which are tardy in tlieir difTerentiation because some of tiiem are figured with a well-develoi)ed contractile structure. Kvidently, in spite of the indications of complete fusion .seen in the hex.diedral sy.stem .'ind of almost complete coalescence in plasma cultures, .^^ome kind of structural demarcation between cell areas is maintained in embryonic heart. It is in liarmony with this conclusion that although no cell boundaries are visible in the heart of the teii.somite chick, yet later a partial separation takes place by means of clefts.


? '» ^.



Figs. 4, 5, and 6 I'liotoftruphs at successive optical levels of ;i triilx-ciiln from heart of chick of fo\ir days' incubation.

142 E. D. CONtiDOX

Tin: sTurnriu-: of thi: i^hvthmically-contracting heart

All ('xaminatii)n was iiiudc of elovrn chick cmbrj'os containing from fifteen to scvontcon somites to learn the time of the beginning of i)ulsation. The heart was beating in four wliich were in the seventeen-somite stage, but in none of the younger chicks.

Begiiming witli tlie heart of the sixteen-somite embryo, series of parallel bars were found in the sarcoplasm which were believed to mark areas fixed in a condition of normal function. Their fonnation will be best understood after the sarcoplasmic structure has been described.

The comparison of the sarcoplasm of th(> non-beating heart with a later stage shows surprisingly little developmental change in structure. The volume of the hexahedral spaces is still the same. The granules have increased in size. In some regions the hexahedral structure has an irregular arrangement as in the younger heart. \l\ gradations occur between this condition and a nearly perfect alignment of the planes with intersections approaching a right angle. For the developmental jieriod with which we are concerned alignment is best seen in longitudinal .sections of the trabecular which make up the spong>- portion of the four-day ventricle. Where it is fountl the parallel bars usually also occur. They are associated with a uniform elongation of the hexahedral spaces in a common direction and result from a drawing together of the granules in the planes transverse to the direction of extension (fig. S). The optical section of the spaces in extension frequently have a length of 1.2 micron anil .•I breadth of 0.(5 micron.

Contraction pluises were made out less frequently than the opposite functional condition. It is possible to come upon regions, however, where the hexahedra are elongated transversely to the long axis of the trabeeulae. ' In describing the extension and contraction onl}' the two dimensions of the hexahedra parallel to the plane of t he cover-slip have been referred to. The spaces are so .•^rnall that the corresponding changes in the direction parallel to the optical axis were not satisfactorily determined. It is jjrobable that the greater irregularity of arrangement of hexa



♦^ >•




Fig. 7 DiuRniin of mitochondrial granules and outlines of hexahedrnl spaces of sarcoplasmic structure before the beginning of rhytlimic pulsation.

Fig. 8 Optical section of hexahedral structure in condition of fimctional elongation.

Fig. '.t Tr;insvcr.sp section of heart wall of a Ihirteen-somite chick, showing clefts between cell areas.

Fig. 10 Diagram of a hexahedral space with it^ membranes and granules.


hodra in \hv coinpaot wall than in the traboculac is due to a less uniform diroolion of contraction and extension, since the wall is not only a curved plane, l)Ut has trabeculae of the spongy niyocardium attaching to it in an irregular manner. The trabeculae, on the other hand, are cylinders attached only at the ends and so must contract in the direction of their long axis.

There is a certain amount of disorganization in the siircoplasmic structure common to the wall and the trabeculae. It is probably the result of agonal contractions that have disturbed the normal pattern at the time of death. \'arious methods were tried in order to fix the heart without the breaking up of the regular beat into irregular local contractions, but without success. A second abnormal feature that is to be seen in some regions of the heiirt is taken to be the effect of unsatisfactory killing and fixing. It consists of a massing together of parallel planes into long bands together with a partial disappearance of the transverse planes and the mitochondrial granules. It may be that the mitochondrial substance has become spread upon or through the planes as under these conditions they take a deep stain.

The consideration of the majority of earlier views regarding the origin of the myofibril can be brief, since they are plainly based upon preparations showing but incompletely the structure first seen by MacCallum and described in the preceding paragraphs. It is not necessary to consider separately the work on heart and skeletal muscle or to distinguish between the bird and the mammal, since variations in the character of the myofibrils in all of these instances is but slight. Wagener ('80), Mtodowska ('08), and Krukiewicz ('09j clahn that the myofibrils when first identified appear to a microscopic examination as structureless filaments. Ej'cleshymer f'04) came to a similar conclusion for the skeletal muscle of Necturus. Marceau (02) is not certain whether they are .segmented or not. Bardeen ('11) refers to them as having no definite .striation. Since the development of the mitoch<indrial concept many have become convinced that in one form or another it is the precursor of the nnofibril. Meves ('09), Ducsberg ('10), Asai ('14, '15), believe that the myofibrils can be traced back to the mitochondrial rods or


'chondriokonten,' and Torraca ('14) finds a similar oriKin duririK their rcKciifratidii. IJriick ('()()). flodlowski r02). and Rubasohkin CIO) derive the niyofihrii fnnii granular niitoehoiiih'ia. Luna ('13) finds that Ki'iHi'dc^ ajjpear first in skeletal muscle while in the heart the primitive eondition is an unsefjinented fibril. Altnian in 1894, before the name mitochondria had bet-n introduced, exfiressed a belief in the {^lanular origin of the myofibril.

It is not (lillicull to understand how many observers came to believe that granules, rods, or filaments jm'ceded the adult myofibril. Each of these structures can be found in the siircoplasm if the technique does not bring out the hexahedra and granules in their totality. .\s already indicat(>il. heavy short rods and structureless filaments api)ear when the hexahedral structure is not perfectly pr(>served. Tlie fine granules described by (lodlcwski ('00) were in many in.stances .strung along slender filaments. He saw in part both elements of the sarcoplasmic structure. Schlater ('()(•) distinguished even the optical .sections of entire liexahedra with their granules. He did not, however, obser\'e them to be united in a continuous structure throughout the protoplasm.

It has ahvatly been said that MacCallum ("97), Wieman ('07), and Briick ( '13) related either a structure of filaments or of planes to tile development of the myofibril. In Wieman's ('07) valualile account for the chick he finds mitochondrial graiudes con.stantly at the intersections. MacC'allum founil them less frequently. The figures of the two do not show the predominance of foursided optical .sections or as much regularity in the size and arrangement of tlie sjiaces as was found in the preparations used in the present stuilj-.

A question of greater significance upon which tlie present account is at variance witii their ol)servations is the numner of appearance of the myofibril. Both authors believe it takes origin within rows of the spaces as a residt of their subtUvision into still smaller compartments. Wieman fimls this process to b(> under way in the four-day chick. The measuremeiU of the spaces in the heart at this stage by the writer with the aid of photographs did not furnish any evidence that they fell into two groups in

14() K. 1). COXGDON

rcf«'roric»> Id tlicir siz<>. Hi' luis alsn cxprcsscil tlio Ix-lit-t tliut tlic font met ill' struct uiv at this tiiiu' docs not consist of fibrils, but that it is a slight inodificatinii of the primitive hcxahcdral spaces with their mitocliomhil firaiiules.

liriick i'i;ii (l('scril)e(i in the eiiilirymiie nuisch- cells (if the lainellibraneh Aiiodonta a primitive protoplasnii*- hoiieycoiiib which gives rise to Hl)rils by ihiekeiiing at th(> intersection of the phmes. Mitochondria is not only accunuilated at the intersection of jilanes as in the liigher forms, but is also present scattered along till' (le\-eloping Hbril. Hriiek's observations are all the more valuable as a coiidi-ination of the chief features ofthe accounts of .MacC'allum and Wieman, since one may conclude from his failure to mention their articles that he arrived at this view entirelj- independent of any suggestion from their work.


It has been siiid that rhythmic contraction is first to be seen in the seventeen- or possibly sixteen-soniite chick. It is in accord with the usual history of developmental processes that there siiould b(» a pi't'liminary contraction of a more simjile kind heading up tn it. If the claim had b;^('ii substantiateil that my()til)rils take their origin at a tleHnite period in development rather than by a gradual modihcation of a structure that has been in the heart from the first, ther>' would have been .some reason for anticipating that the contraction also would begin abruptly at the time of the appearance of the contractile structure. One writer jjlacil the liegimiing of pulsation at a time after the aj)pearance of the myofibril, another makes the two contemporaneous, and a third believes that the contracticm comes first. None of them a jjossible gradual assmnption of the contractile function. Wieman has pointed out that there is no longer a question regarding the time of appearance of a niyofibril, since it is a gradual ililTcrentiation of a struclure pn'seiit in the youngest heart cell. This coiisidci-ation suggested to MacCallum that the contractile function also may be very in its beginning.

.\ search was made for evidence of contraction in preparations fixed before the begiimiiig of rhythmic pulsation. It was found

p:mbryonic structure of avian- heart muscle 147

thiit ill flic licurt of the ti'ii-sdinitc chick siiuill ;ir.';is of licxaiicdnil structur.' often show cloiiKutioii. They uiv of too liriiilcd extent to bo oxphiinod as the result of stretching durinK th ■ handling of the tissue. It is not impossihle, then, that they are the expression of local niuscle contractions. If this !);> true it is still a (piestion whetlier the contraction occurred spontaneously or due to the stimulus of the fixing fluid.

Roux has de.-<cril)ed under the name 'faml)rosia' a widespread resj)onse of various typ;'s of embryonic cells to unfavonxble conditions which consist in their ilrawing away from each other. This can be seen in the early chick heart and is an undoubted expression of a protoplasmic contractility still iiio7\' primitive than ai)i)areiitly occurs in tlii' ten-somite heart.


The he.iil of I lie chicks younger than the sixteen- or seventeen-.somite stages, when rhythmical contraction begins, has a sarcoplasmic structure whose optical sx'tion is a net consisting of two systems of parallel lines intersecting to cut ofT .spaces approaching a square form. They iiuuisure in the fixed material about O.S micron on a side. The appi'rent net is probably produced by three .systems of membranes each pai'allel among themselves which inter.sect to form hexahedral compartments. At all intersections of three planes are .small uniform mitochondrial granules.

There is some slight indication in the arrangement of the planes of a division of the early myocarilium into mononuclear cell areas which corresiioiid to lobations on its exterior. For a period including the thirteeii-somite stage clefts appear j)artly separating cell areas. Yet at no time can cell walls be made out or any interruption of the planes cro.s.sing the region intermediate between two nuclei excej)t tlu^se occasional clefts. The myocardium of the four-day chick shows no clefts. It was oijserved in tissue cultures to migral<' out into the clot in I he form of anastomosing multimiclear columns which gave no eviilenc of breaking up into cells other than slight constrictions of th(> columns. In


spitr of this a|)pc;iniii('c of struct iim.l coiitiiiuity throuf^lioiit the siircoplasiu ill the four-diiy cliick, since Winnan has boon ablo to divide it into iiiononucloar liodics l)y iiiarcratioii and since temporary clefts ai)pear in tlic myocardial tissue it is to be concluded there is proi)al)ly some kind of demarcation into cell areas in the early myocjirdium.

The hexahedral structure of the i)uisatinfi heart throujrli the fourth day of incubation tiitVers from its earlier condition in the more rectilinear arrangement of its planes and in functional changes consisting in the elongation of the hexahedra over consiilerable areas in a eonmion ilirection. This brings the granules of the j)lanes transverse to the dinction of elongation closer together so that they give the appearance of bars. The opposite phase to contraction is apparently in part due to an elongation of the hexahedra in a direction transv(>rse to the original extension. The complete understaiuling of the functional changes in the form of the hexahedra was prevented by the inability to determine the extensions and elongations parallel to the optical axis of the microscope.

Several con.siderations suggest that the ili>thmical beat is precedetl by a less highly organized type of contraction. In the first jilace. the changes in the contractile structure through a period including the beginning pulsition and up to the fifth day are not very nuirked. Then also <ni;tl! arras in wliicli I lie liexahedral structure is elongated in a common din'ction can sometimes be found in the heart of a ten-somite chick. It nuiy be that these are the primitive local contractions brought about either spontane(msly or through the stimulus of the fixing fluid. The creeping apart of embryonic cells described by Koux and observed in the early heart is evidence for the existence of a VQTy primitive protopla.smic contractility.



Ai.TMAXN. 1{. 1S94 Dio Elemcntarorganismen und ihre BMichiingcn zu den

zi'llcii*. 2(i ed., Lcipzii;. Asm. T. I!)1I HeitriiKe ziir Histologic und Histogencsc dcr qucrgcstreifteii

Miiskrliitur dcr SiiUKcticrc. Arch, mik .\nut,. Bd. STi. Bardek.v, C K. 1900 The development of the iiiusculuturc of the body wall of

the pig, including its histogenesis and its relation to the myotomes

and to the skeletal and nervous apparatu.s. Johns Hop. Hospital, Rep.,

vol. 11. Bkxi>.\, C. Wcitcre Mittcilungcn iiher die Mitochondria. \'erh. physiol. Cles.

Berlin. JalirK. 1,S'.).S-1,S'.W. BrIVk. a. 191.'{ Cher die Muskelstruktur und ihre Entstehung, .sowie uber die

Verbindung der Muskein niit der .'^(■h:ile bci den Mu.scheln. Zool.

Anz., Bd. 42. Chi.micoi, Ci. 1887 Delia condizione anatoniische del cuarc al principio della

sua funzione c contribute all istogenesi delle cellule muscolari cardi ache. .\fti R. Accad. Fisiocritici. Siena. CoxoDOX, E. D. 1915 The identification of tissues in artificial cultures. .\nat.

Rec, vol. 9. DuESBERC!, J. 1910 Les chondriosomes des dellules embryonnaires du poulet,

et leur role dans la genesc des myofibrilles avec quelques observations

sur le devcloppement des fibrillcs musculaire striees. Arch. ZcUforsch.,

Bd. 4. Eycle.shy.mkr, .\. C. 1904 The cytoplasmic and nuclear changes in the striated

muscle cell of Xecturns. Am. Jour. Anat., vol. 3. CioDLKWsKi. E., Jun. 1902 Die Entwickelung des .Skelett- imd Herzmuskelge wcbes der Siiugetiere. Arch. mik. Anat.. Bd. 60. HfiDENHAi.N. M. 1899 Bcitriige zur Aufklarung des wahren Wesens dea faser forniigen Differenzicrungen. Anat. .\nz.. Bd. 16. HoYKR, H. June. 1901 Tber die Continuitat der contractilen Fibrillen im den

Herzmu.skelzellen. Anz. Akad. \ Krakau. Kt'HKiEWirz 1<X)9 Zur Kenntnis der Histogenese des Herzmuskels der Wir beltiere. Anz. Akad. Wiss. Krakau. LiNA. v.. 1913 .Sulla importanza dei condriosomi nella genesi dei miofibrille.

Arch. ZcUforsch, Bd. 9. MacC'alia'm. J. B. 1897 On the histology and histogenesis of the heart muscle

cell. Anat. Anz. Bd. 13. Marikat, v. 1903 Recherches sur la aatructure et le devcloppement compares

des fibres de Purkinje et drs fibres cardiaque.s. Bibl. .\nat., T. 10. Meves. F. 190<.) t^er Neubildung quergestreifter .Muskelfa.iern nach Beo barhtungen am Hidinorembryo. Anat. .\nz., Bd. M. Mtiiuowska, J. 1908 Zur Histogenese der Skeletmuskcln. Bull. Acad. Sc.

Cracovie. RriiAscHKiN, \V. 1910 ("hondriosomcn und DifferenzierunKsprozcise boi

.S'iugetierembryonen. .\nat. Helfte.. Bd. 41. Sfiii.ATKR, (i. 1906 Histologische rntcrsuchungcn uber da.i .Mu.xkolgewebe,

II. Die .Myufibrille de.« enibryonalen Hiihnerherzcns. .\rch. mik.

Anal.. Bd. 69.


ScHocKAEKT, A. 1900 Nouvelles recherchos conipariitivea siir le texture et le

devcloppement du myocnrdc chez les Vertebres. Arcli. Biol., T. 24. ToRBACA, L. 1014 II comportaniento del condriozomi nella rcgencrazionc dei

iiiiiKcoli striati. Arch. Zcllforsch., Bd. 12. Waoener, t!. H. 1S80 Uber die Kntstehung der Qiierstreifen auf den Muskein

mid die davon ahhiinRigen KrclieinunReii. .\rcli. .\nat. Physiol.,

Aniit. WiEMAN, H. L. 1907 The relation between the cytoreticulum and the fibril

bundles in the heart muscle cell of the chick. .\m. Jour. Anat., vol. 6.

AmiOK a AlMTItAlT !>»' Tllllt fAPEft I1I0I KD BT THE milLlonitAPIIIC AKRVICE HrpTCUUCR 16


ELIOT K. CLARK AND ELEANOR LINTON CLARK Anatomical Laboratory of the University of Missouri


The present study is part of a series dealing with the growth and reactive powers of the tissues and cells in the transparent tails of tadpoles toward extenuil stimuli. These experiments were begun with the especial oliject of studying the behavior and growth-regulating factors of lymphatic endotheUuin, but necessarily included the observation of blood vessels, leucocytes, and mesenchyiuc cells. One of the authors ('()9) had previously shown that the lymphatics react positively toward red blood cells extruded into the tissue, sending out new sprouts which grow toward them and actively engulf them. Also, the authors ('17) found that lymphatic capillaries will react positively' toward fatty subst.ances introduced into the subcutaneous tissue, growing toward the injected oil globules, and. in this with the aid of leucocytes, taking part in their ab.sorption. It is well known that lymphatics play an important jiart in the disposal of coal pigment, in anthracotic lung tissue, and in the absorption of foreign particles introduced into the peritoneal cavity and, for this reason, we thought it migiit prove profitable to test the response of lymphatics, in the transparent tails of tadpoles, toward injected gramiles of carbon and carmine.

The experiment of injecting small quantities of India ink into the tissue spaces of the tail lins of tadpoles was firstcarri(>d out in 191 3 in the course of other studies on growing lymphatics. 1 1 was found that th(> carbon granules were taken up i\v leucocyt<>s and by mesenchyme cells, but that lym|)hatics and Ijlood vessels



wvTv apparently iiulifforont to the iiroseiipe of foreign partieles of this nature in their vieinity.

Altlioiiuli the presence of cjirhon uninules in th(> tissue spaces did not prove to be a stimuius for the growth of lyinpliatics, it soenied advisable to record briefly the reaction of the various cells and tissues toward this substance and especially that of the niesenclmne cells, in view of the fact that the phagocytic powere of these cells has been disregarded or ilenied by most investigators of the subject. The region of studj* — the transparent tail of the tadpole — where each individual cell can be watched, recordetl, antl followed for days, or weeks if necessary-, and the method of chloretone anesthesia and observation in the upright chamber which permit of study in vivo under approxijnately nonnal conditions, both possess marked advantages over many of the regions and methcxls selected by others for the study of phagocj'tosis.

The tadpoles used for these experimejits were the larvae of Bombinator, Rana pipiens, and Hyla pickeringii. The method of introducing small amounts of foreign substances was the same as that described previou.sly for injections of paraffin oil (E. R. Clark, 'l(i) and fat (E. R. and E. L. Clark, '17). The injection materials — India ink, diluted one-half with tap water, and powdered cannine, suspended in distilled water — were placed in small vials, which were set in a of boiling water for half an hour. The tadpoles were anesthetized in chloretone, 1 to MM), and small amounts of the ink or caniiine were injected into both tail fins, through fine glass cannulae, under the binocular microscope. These su.spensions of granules were injected at ilifTerent points in the fins in some near the margin and, on a few occasions, they were injected directly into the lumen of a lymphatic capillary. The tadpoles were then transferred to the obser\-ation duunber, devised and described by one of the authors ('09, '12) and studied under the compoimd microscope in a 1 to 50()() solution of chloretone. Records of the site of injection, including all the cells and vessels in the neigliborhood, were MKule soon after the injection with the ai<l of the Ix'itz drawing eye-piece. The tadpoles were then returned to fresh water.


The injrc'ted repons were obsorvcil on succooding days, rofords wore made of the same cells and vessels and of the changes which occurred.

The results obtained from injecting carbon, in tlie fonn of India ink, and carmine granules proved to be identical, and a single description will suffice for them both.

Leucf)cytes were the first cells to respond to the presence of the carbon and camiine granules. About an hour after the injection some of these cells, cfmtauiing retl or black granules, could be seen near' the injection site. For several days after the injection, leucocytes continued to migrate toward the point of injection and to take up the injected granules. These cells became loaded with the carbon or cannine to such an extent, in some instances, that they resembled merely balls of pigment with no cellular matter visible, ^fany of the leucocytes, after taking up the ink or cannine, moved over to a near-by blood vessel and flattened out on its wall. In some cases such a pigmented leucocyte was obser\-ed to crawl through the vessel wall. But the removal of foreign particles by this method was evidently very slow, since many cells, loaded with the granules, remained close to the exterior wall of blood capillaries for long periods before entering the ves.sels. Other pigmented leucocytes did not migrate to blood vessels or Ijinphatics, but instead they wandered away through the tissue spaces of the tail fin. In the cjvse of cells which had taken up carbon granules, this process was difhcult to follow, owing to the fact that wandering cells containing black pigment are usually present in the tadpole's tail, under nonnal conditions, but it was easy to trace the wanderings of the leucocytes containing cannine, and these cells were found scattered througli the tail fin at long distances fr<im the site of injection.

Records made ten days and two weeks after injection, showed collections of leucocytes still present near the site of injection, all of them filled with black or red pigment and most of them adlu'rent to the walls of blood vessels. By this time the munber of leucocytes present in this region had diminished somewhat. iSiiice the phagocytosis of foreign j)articles by wandering cells


and leucocytes is such a. well-kiKJWii phenomenon, it seemed superfluous to inv«>stigate the cliaracter of these cells, in the tadpole's tail, and tlicir response toward the injected substances in anything but an incidental manner.

The response of the mesenchyme cells — and by this term we mean the stellate connective-tissue cells— was much slower than that of the leucoc>'tes and wandering cells, and for several hours after an injection no reaction was noted. On the day following the injection, an occasional granule of carbon or cannine c !uld be detect etl along the processes of some of these connective-tissue cells. From this time on the number of granules taken up by the cell processes of the mesenchyme cells increased and, after three or foiu* days, clumps of granules were visible in the cell bodies near the base of the processes as well as on the processes.

After ten days or two weeks the nmnljer of mesenchyme cells ct)ntaining red or black granules and the amount of foreign pigment present in these cells had both increased. This increase occurred at a tune when the number of mobile phagocytes had begun to diminish. The carbon or carmine persisted within the mesenchjnne cells, at the site of injection, for as long as the tadpoles remained under observation.

One of the authors ( '12) has made a careful study of the movement of these so-called 'fixed' connective-tissue cells. He found tliat their shape is continually changing: processes are withdrawn on one side and sent out on the other, and that by this means these cells are capable of a slow amoeboid locomotion. The sajne kind of movement was noted in the records made of these cells before and after taking up the foreign particles, but in compaiison with the behavior of tlu^ leucocytes this migration of the mesenchyme cells is quite negligible. The granules inside of leucocytes were conveyed to blood vessels and eventually transported by the blood stream to other portions of the bodj' or they were carried througli the tissue spaces of the tail toa relatively great distance from the site of injection. On the other hand, those cannine and carbon granules whose fate it was to be picked up by me.senchyme cells remained re'atively stationary, neai' the apot at which they had been originally introduced into the tail.


Figure 1 is ;i drawing made two weeks after an injection of a suspension of canuine gi-anules. The number of leucocytes in this region has diminished by this time and several wandering cells containing red pigment can be found scattered througli the fin at some distance from the point of injection. Those that remain for the most part have congregated near the wall of a near-by bloo(.l cai)illar3- and two of them are shown in the act of entering the vessel. Almost all of the mesenchyme cells of the region contain carmine granules. Figure 2 is a high-power drawingof such a mesenchvine cell, selected from a similar specuuen, which shows the arrangement of phagocytized carmine granules. As may be seen, these red granules have collected f)n the processes and in the cell body near the base of the processes. Figure 3 is a drawing of a number of mesenchjm^e cells made Hvq days after an injection of India ink into the subcutaneous tissue of the tail fin and .shows an identical response on the part of this type of cell toward the granules of carbon which have collected, in a suiiilar manner, on the branched processes and within the cell bodies near the base of the processes.

It is evident, therefore, that in a transparent region such as the tadpole's tail where each individual cell may be observed in tietail, in the living, and under practically nomaal conditions and where the different cell types may be distinguished with ease, that the mesenchyme cells are actiAC phagocj'tes of foreign particles, such as carbon and cannine, present in the tissue spaces. Their reaction toward these foreign substances is less rapid and intense than that of the leucocytes, but it is nevertheless definite alid probably important. The present observations gave the impression that gianules housed within the mesenchyme cells were more permanently located than those which had been engulfed by leucocytes, since that portion of the foreign pigment remaining at the point of injection two or tlu-ee weeks after it had been introduced is, for the most part, retained in the interior of connective-tissue cells.

We have made the statement that the lymphatics of the tadpole's tail showed no visible reaction towanl these foreign particles present in the tissue spaces. The sending out of new


sprouts wliicli p-«'\v tuwjinl tlic iiijcctod finuiiilcs ^thc churactoristic response of the Ij-mplmtics towjinl extruded blood cells and injccteil fat k1<>I'ii1<* — wjis never observed iix the case of the injections of carbon and cannine granules. H(wever, in repeating these exp(>rinu'nts during th(> past year, we found that the IjTnphiitic endothelium, although not stuuulatcd to grow toward these foreign substances when they were at a distance, would react toward them when they were in \erv close proximity. On a few occlusions, a small amount of India ink was injected directly into the lumen of a Ij'mphatic capillar}'. When such a tadpole was exiiniinetl on the following day, carbon gr-amdes were found to be enclosed within the endothelial cells of that h'mphatic which had been injected with ink. They were present as small black spots and as large clumps of granules within the areas surroimding the nuclei of the endothelial cells — the same region which stains with vital dyes (E. K. Clark, '09; Wislocki, 'IG).

A similar behavior on the part of blood-vessel endothelium was not noted.

That foreign particles, such as carbon, carmine, cinnabar, etc., introdueetl into the blood strean; or subcutaneous tissue are quickly phagocytized within the body has been knowTi for many years. Thus, Ponfiek ('Gi)) and Sicbel ('89) showed that such substiuices as cinnabar and indigo injected into the blood stream quickly disappeared from the circulation and collected in the liver, spleen, and bone marrow, where they were taken up bj' cells with large round nuclei, resembling leucocytes.

MetchnikoiT ("83), in his studies on iiiflaiiunation, described the phagocytosis of carmine granules and of red blood cells injected into Triton larvae and into the tails of tadpoles, by leucocyt«'s which have migrated from the blood stream antl by wamlering cells of the tissues. Later ('92), in describing further observations, he groups these two types of cells together as macropluiges or cells whose chief function is the phagocytosis of foreign {)arli(les and i)f cell debris.

Muscatello ('95) and MacCalhun ('03) found that carmine and carbon gramih^, injected into the peritoncjil cavity, were conveyetl to the lymphatics of the diaphragni by means of leuco


cytos which colloctotl in the poritonoal cavnty in largo niimbrrs and actively cnRulfod the foroign jjarticlos. These results diffennl from th(^ conclusiijii of Von Recklinghausen that such i)articles reached the lymph vessels in the free- state tlirougli large openings, or stoiuata, in the endothelial wall.

Recent studies on anthracosis of the lungs made bj- llaythorn ri3) and Klotz ('14) abandon the older view that the coal pigment is taken up by the epithelial cells of the alveoli and that it frofjuently reaches the h7iiph glands draining the lung in the fr(M' state, cutting through the walls of the h^nph vessels by means of the sharp comers on the granules. These authors find that all the coal pigment in anthracotic lung tissue is present inside of phagocytes, wliich, according to Haj'thom, are probably 'endothelial' leucocytes, which pick up the coal pigment in the alveoU, convey it to the IjTnphatics and l>Tnph nodes and also to the connective-tissue septa. They state that the pign\ent remains indefinitely intracellular.

The origin of these large mononuclear wandering cells and leucocytes is still undecided, but their importance as phagocj-tes of foreign particles and cell debris is unquestionetl.

The phagocytic power of the specialized endotheliiun in certain regions of the body has been emphasized by many authors. The endothelial cells of the liver and bone marrow and the reticul()-tMidt>thelial cells of the spleen and Ijinph glands have long been recognized as phagocytes of foreign material, and Evans ('15) has grouped this type of phagocyte with the large mononucloiir wandering cells and the clasmatocytes as macrophages. Mallory (14) even takes the extreme view that the whole group of mononuclear wandering cells and leucocytes, which are phagocytes of foreign particles and cell debris are derived from the endothehiun.

MacCallum ('03) found that the endothelial cells lining the l\^nph vessels of the diaphragm were packed with carbon granules aft(>r injections of India ink into the peritoneal cavity. The obs«'rvation of one of the authors (K. R. ("lark, '00), on the t;iking up of red blood cells from the tissue spaces, showed that the gi-owing lymphatics of amphibian larvae are phagocytic. In


tliis inst;iiice the foreipi suhstuiu-c, ui the form (jf reel blood eells. was taken into the liuuen of the lynii)h vessel. The observation of Wislocki ('16), tliat the colloidal dye trypan blue is stored in tlu" fonii of pianules in the iierinuclear areas of the hinphatic endothelium of tadpoles, rendei-s it hiphly probal)le that all the endotheliikl cells of the lymphatics in amphibian larvae possess the power of phagocytosis. This conclusion of Wislocki's is based on the work of Evans and 8chulemann ('14. '15) which apiH'ars to jirove that the storage of acid azo dyes, in the form of aggregations of dye granules, is evidence of phagocytosis on the part of cells which display this b(>ha\i()r. The present observations on the fate of carbon antl cannine granules introduced into the lumen of hinphatic capillaries demonstrates this phagocytic power of the Ijinphatc endotheliuju, for these granules were taken up by the endotheUal cells lining the l>inphatics and stored in tlie perinuclear areas.

The phagocj'tic power of the mesenchyme cells has not been so generally recognized in the literature. The ability of these cells to ingest foreigii material was not studied in detail by the earlier writers, while recent in\estigators frequently have doubted or denied its existence. That deposits of foreign pigment, such as coal jiigment in anthracosis, metallic silver in cases of argyria, etc., are j)r('sent in the connective tissues has long been known, but recent authorities claim that such material is stored solely in the resting wandering cells or in mononuclear leucocytes or in endothelial cells, such as the KuptTer cells of the liver, and not in the connective-tissue cells proper, or fibroblasts. MacCallujn ('03) is an exception to this statement, for in his description of the results <if injecting India ink into the peritoneal cavity, he mentions the i)resence of carbon gnuiules in true coiuiectivetissue cells.

Haythom (M.'i), in his studies on the histology of anthracosis, describes wanilering cells or endothelial leucocytes as the only cells which phagocytize the coal pigment. Carbon granules present in the connective-tissue septa he beUeves to have been conveyed tluTe by the mobile phagocytes and to remain pennanently within these cells. .Metclmikon' I's:^). in studying tlu


pffpots of injecting carmine giiinules into the tnins|)urent tails of Triton hir\ae and tadpoles, notes and figures these foreign granules inside of niesenchynie cells as well as in the wandering cells and leucocytes. However, in later studies ('92) he recedes from the opinion that the ronnective-t issue cells are phagocytic and states that mesenchyme cells which contain foreign pigment are cells which have ingested it while they were leucocytes and that leucocytes containing ingested pigment nuiy wanfler off antl become transfonned into cells which resemble mesenchyme cells. In a still later publication ('05) he denies the power of true mesenchjinc cells or fibroblasts to phagocytize foreign matter. Similarly, Buxton :md Torrey COO), in describing the disposal of carbon granules injected into the peritoneal cavity, note the rapid jiliagocytosis of these particles by the macrophages and consider that the carbon granules, wich are found several hours later in the connective tissue, have been conveyed there by macrophages, wliich then proceed to transfonn themselves into 'trailers.' However, the authors are not absolutely convinced that these 'trailers,' wliich are practically stationary and which store the granules indefinitely, are in\-ariably transformed macrophages and they state that these cells are frequently indistinguishable from ordinary connective-tissue cells.

Recently, Jones and Rous ('17) have denied the phagocytic power of niesenchpne cells. Their rather elaborate method of study consisted in making suspensions of individual cells by digesting with tryp.sin the clot of proliferating tissue cultures of embrj'onic chick material. The indivitlual cells were then freed from the tissue mass by filtering tlirough gauze, and new cultures were then made to which finelj^ ground carmine was added. .\ slight amount of phagocytosis occurretl, but mainly on the part of large cells which the authors believed to be endothelial hi nature. The authors concludetl that the connective■ tissue cells proper possess no power of phagocytosis."

' Since writing this article, the recent work of Leo Loeb and Fleisher ('\7) has come to our attention. These authors, in studying the behavior of connectivetissue cells in tissue cultures, question the decision of Jones and Hous that the large cells which phagocytize foreign particles, in their tissue cultures are endo


In tln' i»n'.s(»nt oxpi'iimciits, witli u roRion of study such as the t ranspHrcnt tail fin of Amphibian larvae and with the aid of chlorotone anaesthesia and the special observation chmnber, it was possilile to si'e and to keep records of the individual cells of the area at the time of the injections of India ink and of cannine pranules. and to follow these cells and to observe their subsequent beha\-ior with jierfect clearness and in tlieir normal environment. When the living cells are observed in this manner, it is perfectly clear that mesenchjnne cells possess the power of phagocj'tizing foreign particles of carbon and carmine and of retaining them indelinitely. It is very easy, by making records of individual mesenchjnne cells, to see that these same cells which were jiresent as connective-tissue cells at the time of injection later contain gninules of carbon or cannine and that they retain their characteristic identity as 'star-shajied' cells after taking up the pigment. -\lso it is clear that the wandering cells and leucocytes which ingest the foreign particles do not become transformed into connect ive-tis.sue cells, even in the cases where they wander off througli the tissue spaces, but that they persist as round or amoeboid cells, easily distinguishable from the branching connective-tissue cells.


The present observations show that three types of cell present in the transparent tails of tadpoles display the power of phagocytizing granules of carbon and cannine injected into the subcutaneous tissue. These arc: leucocj^tes (including wandering cells), stellate connective tissue-cells, and the endotlielial cells of the l>anphatics. The leucocytes actively migrate toward the sit«> of injection, while the me-senchjane cells and Ij-mpliatic endothelial cells apparently ingest only those particles which are in close proximity to them.

tlielial in nature. Locb and Flcishcr consider such cells to be fibroblasts which have increased in size during the process of regeneration and which are capable of ingesting foreign particles and cell d6bris.



Buxton and Toniiicy I'JOC Absorption from the peritoneal cavity. Jour. Med. Res., vol. 15, p. 3.

Clark, E. R. 1909 Observations of living growing lymphatics in the fail of the frog larva. Anat. Rec, vol. 3, no. 4, p. 183.

1912 Further observations on living growing l\^npllatics: their relation to the mesenchyme cells. Am. Jour. Anat., vol. 13, p. 351. 191G .\ study of the reaction of mesenchjone cells in the tadpole's tail toward injected oil globules. Anat. Rec. vol. 11, no. 1, p. 1.

Clark. E. R. and E. L. 1917 .\ .study of the reaction of lymphatic endothelium and leucocytes in the tadpole's tail toward injected fat. Am. Jour. Anat., vol. 21, no. 3, p. 421.

EvAN.s, H. M. 1915 The macrophages of mammals. Am. Jour. Physiol., vol. 37, p. 343.

Evans, H. M., and .S<hdlemann 1914 The action of vital stains belonging to the benzidine group. Science, X. S., vol. xxxix, no. 1004, p. 443. 1915 t*bcr Xatur und der durch saure FarbstofTe enstehenden Vitulfiirbungsgranula. Folia Haem. Arch., vol. 19, p. 207.

Haythorn, S. R. 1913 Some histological evidence of the disease importance of pulmonary anthracosis. Jour. Med. Res., vol. 29, p. 259.

Jones and Rous 1917 The phagocytic power of connective-tissue cells. Jour. Exp. Med., vol. 25, no. 1.

Klotz, O. 1914 Pulmonary anthracosis — a community disease. Am. Jour. Pub. Health, vol. 4, p. 887.

LoEB, Leo and Fleisher 1917 On the factors which determine the movement of tissues in culture media. Jour. Med. Res., vol. 37, no. 1, p. 75.

MacCallum, \V. G. 1903 The relation between the lymphatics and the connective tissue. Johns Hop. Hosp. Bull., vol. 14, p. 1. 1903 On the mechanism of absorption of granular materials from the peritoneum. Johns Hop. Hosp. Bull. vol. 14, p. 105.

Mallgry, F. B. 1914 Principles of pathological histology.

Metchnikoff, E. 1883. L'ntersuchungcn iibcr die mesodermalen Phagocyten einigcr Wirbelthicrc. Biol. Cent., vol. 3, p. 560. 1892 Lemons sur la pathologic comparfe de I'inHammation. Paris. 1905 Immunity in infective diseases. Cambr. Univ. Press. Trans, by Binnie.

MuscATEi.Lo 1895 Uber den Bau und das .\ufsaugungsvermdgen des Peritoniium. Virchow's Arch., vol. 142, p. 327.

PoNFicK, E. 1869 Studien iiber die Schicksale korniger FarbstofTe im Organisnuis. Virchow's Arch., vol. 48. p. 1.

SlEBEi.. \V. 1886 l*ber das Schicksal von Frcmdkorpern in der Blutbahn. Virchow's Arch., vol. 104, p. 514.

Slavjanskv 1869 Experimentclle Beifnige zur Pnetiraonokoninsifj, Virchow's Archiv., vol. 48, p. .326.

WlSLOCKi. (i. B. 1916 The staining of .\mphibian larvae with boniidine dye with especial reference to the behavior of the lymphatic endothelium. .\m. Jour. Physiol , vol. 42, no. 1, p. 124.



I'ig. 1 Cameru-lucida drawing of a region in the dorsal Kn of a tadpole, into wliirli a su8|)cnsioii of cacniino Rrannlos had been injected two weeks previously. The sketch shows that the red granules have been taken up by leucocytes and connective-tissue cells, a. Leucocyte, containing carmine, on the point of entering a blood Vessel. Enlargement = X 275.

Kig. 2 High-power drawing of a single mesenchyme cell, showing phagocytized carmine granules, distributed on the cell processes and inthe bod}' of the cell, near the base of the processes.

Fig. 3 Camera-lucida drawHng of a number of mesenchyme cells from a region of the tail fin into which India ink had been injected five days previously. Carbon granules have been taken up by the connective-tissue cells and are lodged on the processes of the cells and in the cell bodies near the base of the processes. Enlargement = X 300.

On The Time Of The Post-Natal Obliteration Of The Fetal Blood-Passages (Foramen Ovale, Ductus Arteriosus, Ductus Venosus)

Scammon RE. and Norris EH. On the time of the post-natal obliteration of the fetal blood-passages (foramen ovale, ductus arteriosus, ductus venosus)(1918) Anat. Rec. 15(4): 166-180.

Richard E. Scammon And Edgar H. Norris Institute of Anatomy, University Of Minnesota

Three Plates

It is gcnorally rocognized that two separate processes are involved in the occlusion of the fetal blood passages after birth. The first is the simple functional closure which takes place, in the great majority of cases, at or immediately following birth; the second is the permanent anatomic obliterati(jn which occurs at a later period. The mechanics and histology of the latter process are generally discussed in our larger treatises on obstetrics and pediatrics, and in our major an.atoTnical texts; but the thiie of postnatal obliteration is often unmentioned, although in a number the statement is made that the obliteration takes place in the first few days or, at most, weeks after birth.

The origin of the current concepts as expressed in these larger texts can be traced, we think, to the first statistical study on this subject, which was pubUshed ninety years ago by the French clinician Billard. This investigator (H)Ilected data on the obliteration of the ductus veno.sus, ductus arteriosus, and foramen ovale in a series of one hundred and twenty-eight children who died in the first eight days of life. He found instances of the obliteration of ail these pas.sjiges on the first tlay after birth. The foramen ovale and ductus arteriosus were closed in over fifty per cent of his cases on the eighth day while the ductus venosus was clo.sed in a still larger* immber. He therefore concluded that the obliteration of the fetal blood-passages proce<'ded very rapid!}- in th<' first few days of life- an opinion in accord with that held by a niimbor of wTiters in the eighteenth century. The n^sults of Billard's study were published in his 'Traits des maladies des enfants nouveau-n6s' in 1S2S. This work was extremely popular in its time; it passed through a number of French editions, wa.s translated into Italian and German and twice app<>jired in .Vmerican editions. In several publications of the middl.' of the last century Billard's figures are cited and his name is quoted in connection with them. In later works, howevei", the same opinion has been repeatedly expressed but its sourc(^ apparently has been forgotten.

A large amount of data concerning the chronology of the postnatal obliteration of the fetal blood-passages has accumulated since the time of Billard. This material is scattered through the periodical literature of legal medicine, obstetrics, and pediatrics and is also uieluded in a number of rather inaccessible brochures. In the following pages we have collected as much as possible of these scattered data and have arranged them in tabular and graphic fonii. In so doing we have confined oiu-selves to those records in which series of consecutive cases have been assembled. These records include ob.serA^ations on children who were born at term, and also a few cases of children who were prematiu-ely born. In most instances, however, investigators have failed to separate these two groups. A comparison of the records of the few known cases of prematurity with those of children known to be born at term shows no appreciable difference in the time of obliteration of the fetal blood-passages. We hav(> therefore combined them in our tables. In all cases the patency of the fetal passages was te.sted either by injection or by probing, excepting those of Faber ('09) which were examined micro-scopically.

The results made apparent by this combined series of observations are at variance with the current concepts on the subject as expressed in most of our larger texts, and also with the results of Billard. Billard's observations were confirmed by Bernutz ('65) who found the ductus arteriosus clo.sed in fourteen cases in a series of twenty-one children who died between the tenth and twentieth daj's, and in thirty-six out of thirty-eight children dying between the twentieth and sixtieth days. Since this time


no observer has substunfiatod these findings. .-ilthouRh a number of series niueli larger than those of Billard and Bernutz' have been collected. Thus Elsasser ('52), in a series of nearly three hundred observations upon children of the first month, found obliteration of the ductus arteriosus in about two per cent and of the foramen ovale in about three per cent of his cases; and Alvan^nga ('69) found practical!}' no instances of obliteration of the foramen ovale or ductus art^-riosus before sixty days. The findings of later obseivers (Alexeieff ('00), Theremin ('87-'95), KuchefT ('01) and others) agree essentially with those of Elsasser and Aharenga although they have noted some instances of earlier obliteration of the pas.-^ages.^


The compiled data on the obliteration of the foramen ovale are shown in extenso in table 1. Table 2 is a summary of these data giving bj^ periods the total number of observations and the number and per cent of cases obliterated. Graphic representations of these data are shown in figure 1 curve A. and in figure 2. In the summary and curve the data of Billard are omitted because his findings are so directly opposed to those of all other investigators that we conclude that either his method of investigation was defective or that his definition of obliteration was entirely different.

As will be seen in table 2. less than one per cent of the foramina are completely closed in the first week of life, and less than two and one-half per cent in the second week. In the latter part of the first month the figures indicate that the obUterative process takes place more rapidly as the opening is impervious in about one-eighth of all cases of this period. The rapiditj' of this process mcreases during the second month, for the interatrial

' The cases reported by Bernutz were not examined by him personally but were oollected iit his instance by the interne of a colleague in the Hospice des EnfantsTrouv<5s.

' Haberda ('96) studied the obliteration of the ductus venosus and ductus arteriosus in a considerable series of infants and children. As his data are not given in numerical form we are unable to include them in our summary. The general statements of this writer indicate that his findings were somewhat similar, as regards these two vessels, to our own.




Data on the obliteration of the Foramen ovale. Numerals enclosed in parentheses indicate number of obliterated cases

ornanTU AXD




< a






•4 O







o a
































Alexeieff. '00.



























Billard, '28...

118 (17)



British Anat.

Soc, '98...







286 (217)

Bitot, '37...





EUasser, '52.



63 (4)

62 (9)















Ogle, '57

62 (49)

Theremin '87













(10) 4

(13) 6


(18) 4

(47) 8

(51) 6





Theremin '95













291 (170)

































• 1 to 15 years t 7 to 10 months.



TABLE 2 Obliteration oj the Foramen ovale (SB58 cases)

Birth to 8 days...

8 to I.'} (lays.. . .

15 to 32 days

32 to 46 days

46 to 61 days

61 to 91 days

3 to 6 months.. 6 to 9 months. .

9 to 12 months. . 1 to 5 years . . . . 1 fo 15 years. . . . 5 to 10 years ....

10 to 20 years .... 15 years and over 20 years and over




HmiBBB or



















































comnumication is oblitoratcd in approxiinatelj^ one-.sixth of all cases in the first half of this poriod and in about one-fourth in the latter half. During the third month somewhat less than ten per cent of the cases are obliterated so that by the end of the first trimester the foramen or {do is finally closed in about onethird of all cases. After the end of the third month the process again goes on more slowly and the average of obliteration in the second trimester is about forty per cent, that in the third trimester about forty-five per cent and that in the fourth trimester about fifty-five per cent. Our figures for the period between one and five years show an average obliteration of fiftj-five per cent, which is five per cent less than that of the last trimester of the first year. This difference is due. no doubt, to the small number of cases tabulated in these two periods, and does not represent a real increase in the number of patent foramina. The figures of the .second five year interval indicate that twothirds of the foramina are clo.sed: but the small amount of data for the period between the first and tenth years makes it impossible to say with certainty just when this increase in obliteration


is bntuphf about. In the srcond drrado the percontago of open foramina is the same as in the period l)et\veen five and ten years and tlie foramen ovale is found to be obliterated in about seventytwo per cent of individuals of twenty years and over.

Curve A of hgure 1 expresses graph ieallj' the frequency of obliteration of the foramen ovale during different periods in the first year of life. This curve is readily di\'isible into three parts. Tlie first portion, which extends from birth to the middle of the

" ' >• Ol »3 W go 120 IBO JTO 360.1^

Fig. 1 Three curves representing the average percentages of obliterated fetal blood-passages at different periods in the first year of life. A, dotted line, foramen ovale, B, solid line, ductus arteriosus; C, broken line, ductus venosus. These curves are based upon the material summarized in tables 2, 4 and 5.

second week, is a short segment expressing the obliteration of a Uttle over two per cent of the It is followed by a longer segment rising abruptly and terminating in about the middle of the third month. Nearly half of the which are finally closed are obliterated in the period represented by this segment (sixtj- days), l^he third segment, extending from about the middle of the third month to the end of the J'ear, shows a very slow but continuous rise and expresses the obhteration of about ten per cent of the total number of cases.


Figure 2 is a curve illustrating the froquoncy of tho oblitfration of the foraiiu'n ovale throughout life. The details of the obliteration during the first year, which have just been described, are masked in this figure by the diminution of the time unit. Here again three periods can be recognized which correspond roughly to infancy, childhood, and adolescence and maturity.




^^^■-^ lS-80y«»rsTl.T







I.I 1 1 1 1

1 i to LSyur*

Fig. 2 .\ curve showing the average percentages of cases of oblit«rated foramen ovale at different periods throughout life. Based upon the material summarizcd in table 2.

During the first period the curve rises with extreme abruptness to a point at which fifty per cent of the cases are obliterated. The ri.-^e is continued but is much less rapid during the second period which extends from infancy well into childhood. In the third period, which extends to extreme old age, the curve rises very slowly to about seventj'-two per cent. In all probability this final percentage of obliteration is reached in early maturity although the character of our data does not permit a graphic representation of this point.




Tabic 3 shows tho compiled ihxta. upon the post-uutal obliteration of the ductus arteriosus, and table 4 summarizes these data by periods. The graphic presentation of this material is shown


Data on the frequency of post-nalal oblileralion of the Ductus arteriosus. NumeraU enclosed in parentheses indicate number of obliterated cases


Alverenga, '69 .

Bernutz, '65 — Billard, '28... Elaaaser, '52....

Faber, '12

Gerard, '00

Kucheff, '01...

Letourneau, '58 Theremin, '87..

Theremin '9.t Totals







12 67


429 (17)





























(6) 4 (2)

143 (18)

8 (1)




(8) 6 (3)

117 (13)

38 (36)







50 (26)




(23) 4 (1)

57 (27)






(14) 2 (2)


(55) 8 (8)

92 (70)











54 (54)

12 (12)

89 (84)









47 (33)

• 10 to 20 days.

•• 21 to 60 days.

t 2nd month (data for other periods not available).

■' to 30 days.



TABLE i Obliteration of the Ductus arteriosus (1095 cases)

Birth to 8 days 8 to 15 days. .. 15 to 22 days... 22 to 32 days... 32 to 46 days... 46 to 61 days... 61 to 91 days... 91 to 120 days.. 120 to 365 days.





148 143 117 75 57 92 63



1 3

16 13 28 27 70 52 84

ram csirr or


0.3 2.0 11.2 11.1 37.3 47.4 76.0 82.5 94.5

in figuro 1, curve B. As in the of (he foramen ovale, and for the same reasons, we have omitted Billard's data and also those of Bernutz from the summarized table and from the curve.

During the first week of life the percentage of obliteration of the ductus arterio.sus is even less than that of the foramen ovale (three-tenths of one per cent). In the second week the percentage rises to an average of two and in the third and fourth weeks to an average of a little over eleven. From this tune on to the end of the third month the process of obliteration is e.\tremely rapid; in the first part of the second month it averages over thirty-seven per cent, in the latter part over forty-seven, and in the third month seventy-six per cent. During the fourth month the average obliteration is eighty-two per cent. Thereafter the percentage increases quite slowly until the end of the year. The a\erage percentage of obliteration in the last three-quarters of the first year is nearly ninetj'-five.

The data a\ailable for the period after one year are small in amount and, with the exception of certain instances in Faber's series, all cases of this period were obliterated. It is quite possible that Faber's material included several specimens which, while containing remnants of the original lumen, were obliterated at other points. His method of examination might easily classify such cases iis patent. It is probable, therefore, that table 3 shows a much lower per cent of obliteration for this period than is actually the case.



Curve B of figure 1 is the graphic presentation of the data summarizes above. Like the curve for the foramen ovale already doserihod and repr(>s('ntrd in the .same figure, thi.s graph is reaciily divisible into three portion,s. The first segment extends from birth to the middle of the second week and shows a temunal obliteration of two per cent of the cases. The second .segment abruptly, cros,ses that of the foramen ovale, and teniiiimted at about .seventy-five per cent in the middle of the third month. The third segment, which extends from the middle of the third month to the end of the first year, shows a continuously decreasing rate of obliteration and terminates at nearly one hundred per cent. Probably all normal cases are closed shortly after the first year, although there are numerous records of individual cases of the anomalous persistence of the lumen of the ductus arteriosus in later life.'


The material compiled upon the obliteration of the ductus venosus is shown in table 5 and is represented graphically in curve A of figure 1.


Obliteration of the Ducltis venosus (,76i










Birth to 8 dara











27 64 78 36 20 41 19 12


8 to 15 dava


IS to 22 dayK


22 to 32 days


Second month


Third month


Fourth, tifth, and sixth months

100.0 100.0


• For the literature on the subject of persistent patency of the ductus arteriosus the reader is referred to the papers of Poynter ('16), Wells ('08) and Gerard ('00').


The procrss of oblilcnitidn is much more rapid in the ductus venosus than in the other fetal passiiges. During the first week the average is two and three-tenths per cent, in the second week it is eighteen per cent, in the third thirty-seven and one-half per cent, and in the last ten days of the first month about seventysix per cent: During the second month the percentage rises to nearly one hundred and thereafter all cases are obliterated.

The curve shown in figure I, while nuich more abrupt than that of the foramen ovale and th(> ductus arteriosus, is of the same general character and shows three segments. The first segment corresponds to the first few days after birth, and terminates between two and three per cent. The second segment rises with extreme abruptness to ninety-seven per cent in the middle of the second month. The third .segment, which is very short, rises gradually to a full hundred per cent by the end of this month.



In t)rder to study the actinty of the obliterative process in the various fetal blood-passages we have calculated from the graphs shown in figure 1 the average daily rate of obUteration for a series of periods in earlj' life. This was done by determining from the graphs the initial and terminal percentages of obliteration for each given period. The initial percentage was then subtracted from the terminal one and the figure thus obtained divided by the number of dajs in the period. For example, in the case of the ductus arteriosus the percentage of obliteration at the beginning of the second month was fift>--seven and at the close wjvs seventy-nine and one-half. Thus twenty-two and one-half per cent of all cases were obUterated in this period of thirtA' days and the average daily rate of obliteration was seventy-five hundredths per cent. Table (5 shows the results of these calculations for the three passages and the curves in figure 3 express them graphically.

It will be seen by the examination of these curves that they have certain characters in common. Each starts with a low rate




Approximate rate of daily and f

obliteration of the Ductus venosut, Ductus arteriosus, 'oramen ovale in early childhood










0.04 0.92 1.00 0.75 0.07 0.04


1 week to 1 month




3 to 6 months


6 to 12 months


1 to 5 vears


of obliteration, rises rapidly to a peak or niaxinium, and then declines to the base-line which represents the cessation of obliterative activity. In all cases the portion of the curve representing the decline in activity is less abrupt than is the initial rise.

Considering the curves individually, it will be noted that in the case of the ductus venosus both the initial and maximal rate of obUteration is much greater than that of the other two passages and that consequently the entire obliterative process is completed much sooner. The curve expressing the rate of obliteration of the ductus arteriosus is very similjir to that of the ductus venosus although the initial and maximal rates are lower and the apex of the curve falls at a later period. While the curve of the rate of obUteration of the foramen ovale shows the three conmion characters indicated in the preceding paragraph it is markedl}' different from the curves of the \cs.sol.s. It ri.scs less abruptly to a lower maximum rate of obliteration which is maintained with but little loss for a much longer period, so that the curve presents a plateau which is entirely absent from the curves of the vessels. Following this plateau, the curve falls at first rapidly and then gradually over a long interval to the base line.

The results of this study may be sunmiarized as follows:




lukrl mo.

2nd mo.

3rd mo.

15-6 mos. 6-12mo».


Fig. 3 Three curves showing the approximate average daily rate of obliteration of the fetal blood-passnges in early life. A, dotted line, foramen ovale, B, solid line, ductus arteriosus; C, broken line, ductus venosus. Based upon the data summarized in table 6.



The tinio of oblitoration of the three fetal blood-passages (the ductus venosiis, the ductus arteriosus, and the foramen ovale) is distinctly hittr than is cojnnionly assumed. The prnocss of oblitvniliim in each of these passages shows three fairly- distinct periods: an initial period with a low rate of obliteration, a middle period in which the rate of obliteration rises and the majority of cases iire closed, ami a tenninal period in which the rate of obliteration is again slower.

ObUteration is most rapid in the ductus venosus. Although slow in the first week, the process reaches its maxiniuni before the end of the first month and in the third month and thereafter all cases are closed.*

The ductus arteriosus closes more slowly. The obliterative proce-s-s, which is very slow during the first two weeks of life, does not n>ach its maxiiimm until the second month. Threefourths of all cases arc closed at the end of the first trimester and over ninety-five per cent by the end of the first j^ear.

The period of the obliterative process of the foramen ovale is a matter of years rather than months. Beginning ver)' slowly the process reaches its maximum activity near the close of the first nxmth and continues with a slightly diminishing rate during the remainder of the first trunester. At the end of this time approximately one-third of all cases are closed. During the second trimester the rate of obliteration declines rapidl}' and thereafter decreases verj- slowly for an indefinite period — probably until early maturity, although few cases are closed after childliood. At the end of the first year about one-half of all cases are closed, ill th«' second deccnnium about two-thirds, and in maturity about seventy-two per cent.

^ * It has been shown by the studies of Wertheimer ('86), Nitdtin COl), Fontan ClI) and others that the vein which sometimes occupies the center of the ligament of the ductus venoaus in older children and adults is a new vessel developed after the obliteration of the ductus venosus and is not derived from the remains of this trunk.



Aft or this pii,pc>r was in press we secured :i summiiry of onrhundn (1 ;ind eighty-*,'Veii obs Tvations by Parrot on the obliteration of the ductus jirteriosus. Parrot's findings are in general agroeint nt with thos.^ of other observers which we have sununarized above. Fnfortunatt'ly his cases under one year are grouped in such a way that wc? eaiuujt include them in our table 4, but if this were possible they would evidently affect our averages very little. Parrot found the ductus arteriosus patent in four in thirty-three of one yejvr and in one in fifty-four of two years. In seventeen crises of tliree y<>ars and over the the <luctus wa.s always obliterated.


Alexeieff 1900 The foramen ovale in tlie child. Diss. St. Petersburg. Alvarenga, p. F. DaC. 1869 Considi^rations et observations sur IV'poque de

I'occlusion du trou ovale et du canal arti^riel. Lisbon. Ber.ndtz 1865 [Quoted from Gc/rard, (1. ('00«).l BiLLARD, C. M. 1828 Traitd des maladies des enfants nouveau-n^s et k la

mamelle, fondi^ sur de nouvelles observations cliniques et d'anatomie

pathologique, faites & I'hopital des enfants trouvos de Paris. Paris. BizoT, J. 1837 Hecherches sur le coeur et le systtoe art(?riel chez I'homme.

Mem. .Soc. Med. d'Observation, T. I. EL6A8SER 1852 t'ber den Zustand der Fotuskreislaufwcge bei neugeborenen

Kindern. Zeitschr. f. Staatsarzneik., Bd. LXIV. Faber, \. 1912 Die anatomischen und physikalischen Verhiiltnisse des Ductus

Botalli. .Vrch. f. .\nat. u. Ent. Fawcett E. AND Blachford, J. V. 1900 The frequency of an opening between

the right and left auricles at the seat of the foetal foramen ovale.

Journ. .\nat. and Physiol., vol. XXXV. FoNTAN, C. 1911 Lc canal d'Arantius (i^tude anatomique). Th^sc, Lille. G£rard, G. 1900 Lc canal artrriel. Etude anatomique. Journ. de I'Anat.,


1900' De I'oblitt^ration du canal art^'riel (les th^-ories et les faits).

Journ. de I'Anat., T. XXXVL Uauerda. A. 1896 Die fotalen Krcislaufswege des Neugeborenen und ihre

Vcriinderungen nach der Geburt. Wien. Klobb 1859 (Amtl. Ber. XXXIII Versamml. deutsch. Xaturf. u. .\rzle zu

Bonn, 1857.)

• Depaul, Diet. Encyclo|x5d. d. So. M6d., 2. a6i., T. XIII.


KucHEFF, N. E. 1901 (The ductus Botftlli in cliildren.) Diss. St. Petersburg

(Quoti'd from summnrips in Jnhrcsl)cr. f. Annt. u. Entwickl., Bd. VII

(N.F.), and from Ciundohin: Die Besonderheiten des Ivindcsalters,

1912.) LETorRNTiAU 1858 Quelques observations sur le nouveau-n<5. ThJse, Paris. Parsons, F. G. and Keith, A. 1898 Seventh report of the committee of collective investi(?iilion of the .\nntomicaI Societj- of Great Britain and

Ireland, for the year 1896-97. Journ. of Anat. and Physiol., vol.

XXXII. NiDTiN, A. A. 1901 The duct of Arantius in childhood. Diss. St. Petere burif. Ogle, J. 1857 On certain cases in which the foramen ovale was still patent in

the adult. British Med. Journ. PoYSTEB, C. \\. M. 1916 .\rterial anomalies pertaining to the aortic arches

and the branches arising from them. University Studies, (U. of

Nebraska) vol. .XVI. Quincke, H. 1885 Ueber die Entstehung der Gelbsucht Neugeborener. .A.rch.

f. exp. Pathol, u Pharmakol., Bd. XIX. THfREMiN, E. 1887 Xote sur I'involution des voies foetales. Rev. d. Mai. d.

I'Enfance, T. V.

1895 £tudes sur les afTectioos congenitales du coeur. Paris., H. 1859 Ueber das Offenbleiben des Foramen ovale cordis bei

Erwachsenen. Vierteljalirschr. f. d. prakt. Heilk., Bd. LXII. Wells, H. G. 1908 Persistent patency of the ductus arteriosus. Amer. Journ.

Med. Sc. N.S., vol. CXXXVI. Wertheimer, E. 1886 Recherches sur la veine ombilcale. Journ. de I'Anat.,

T. XXII. Zahn, (Cited by .\bbot, M. E. in: Osier and McCrae, Modern Medicine, vol.

IV, 1908.)


iiii; i;Ai;i.\ AiTKAirwch; of iiii-; aslma.s oi' iiii-: I'AUs IN iiii: iivpophvsis of i'iif-:



Dr.pnrliiiriil •>) Aiiiilninii, Meilinil Sihiiiil nf the I'niverxili/ nf Miih!ijiin


[t is now well n'('()>riuz((l lli:it tlic cpithcliiil portion of tlip liyiJophysis consists of three ilistinct parts. The pars anterior propria is the principal cpitheHal lobe and constitutes the main l)iilk of the ulaml, the i);'.rs int(>nii( dia is a thin lavt'r. epithehal in iiatun". which becomes iiitiniately associated with the neural lobe. 'rh(» nicst recenth" rec(jgnizeil epithelial lobe is the pars tul)eralis — so named by Tilney ('13) on account of its close relation to the tuber cinereum. It extends forwaril from the junction of the pars intermedia anil the pars anterior propria, surrounding the infundil)ular stalk and spreading out for .some di.staiice under the brain floor.

The jnirs tubcralis has been .sometimes confu.sed with the pars intermedia- Lothringer ('86) and Herring ('08) — but recent studies have .shown conclusively that these two parts are different both in adult structure and in developmental historj-. Tilney, in summarizing the develo])ment of tlie pars tuberalis in the chick and the cat, states:

In atUlition to (In- histological ditTorcnccs between these two parts, llic ontogenesis of tlie organ as observed in the cat and the fowl still furtluT einpiiasizes the fact .that the pars tubcralis and the pars infmnlibularis (pars inlernuMlia) are morphologically distinct elements. Till- |)ars infiiiidibularis makes its appearance inniiediately after the aiiianc of the buccal portion of the liypopliysis is fornietl. The pars tubcralis ai'ises as a relatively late stniciiirc. It has its origin in two sccond.aiy ilivcrliciila or sprouts from the ImkIv of the pituitary sac. These sprouts, the tubcral processes, ultimately fuse with each other across the median line, displace the body of the pituitary sac vcntrad and thus secondarily a.ssume their juxta-iiein-al position.




Out" (if lis Al\V(>ll CIS) — ill ;i recent study of tlie tlevelopiiii'iit of tlie liypopliysis in the nil)l)if has obtainod somewhat (lifTereiit results. Wliile afireeiiifi with Tilney as to the distinctness of the pars tul)erahs and the pars intermedia, and also conHrminK the statement tliat the pars tulieralis is late in acquiring its adult relationship with the tuber cinereum, it has been found thai in the rabbit the anlagen of the pars tuberalis may 1)1' discerned \erv early. They were found to precede* the definite pars intermedia by a considerable period of time.

It was with the lio|)e of throwiiifi some light upon this point that the j)resent study was imdertaken. Accordingly we have been led to construct a luimber of wax-plate models of the epillielial hypophysis from chick emiiryos, l)eginning with .stages in whicii the tuberal processes might be recognized easily and then j)roceeding to succes.sively younger (>mbryos in an effort to determine the earliest appearance of th<' anlagcu.

Tlie literature relating to the lateral lobes and the pai's tuberalis in the hypophysis of the chick is not extensive.

Rossi ('{)()) speaks of a median and two lateral parts in the early hypophj'sis of the chick embryo, .\ccording to Ro.ssi the lateral lobes are secondary structures.

iM'onomo ('!)!)) observed a pair of "Seitenspro.ssen' in the hypophysis of the dove and of the domestic fowl. In dove embryos the sprouts appear b(>tween the fourth and seventh days of incubation. \o definite statement is made concerning the first ai)pearance of the sprouts in the chick.

Tilney ('13) first ob.served the 'tuberal processes' in a chick embryo of o days and 20 hours of incubation. From this stage the were traced to the formation of the pars tuberalis of the adult fowl, .\lthough a reconstruction was prepared from an embryo of fourda3's of incubation the tul)eral proces-ses were not seen in this stage. Embryos younger than four days were not studied.

Woerdeman ('14) notes that the lateral lobes are forming in a chick embryo of about 72 hours of de\-elopment. The thickened epithelium which lies in front of Hathke's pocket is constricted off from the mouth cavity by two lateral folds. Woerdeman con


siders that the lateral lobes so formed arise independently of Rathko's pocket.

Bruni ('15) observes the presence of two 'lobi laterali' in the chick at 82 hours of incubation. He also figures and describes the latrral lobes in older embryos but does not trace them into the formation of the pars luberalis.


We have prcpan-d wax-plate reconstructions of the epithelial portion of the hypophj'sis from cliick embryos of 48, 59, 67, 72, 96, 120 and 144 hours of incubation. A relatively high magnification was chosen for the construction of the models in order that all details of structure might be shown as accurately as possible. For all younger stages, including the 72 hour embryo, the magnification was 300 diameters. For the older embryos the magnification was reduced to 200 diameters.

Chick embryo, 4^ hours of incubation {21-2 pairs of primitive segments). Fig. 1. The hj^jophyseal pouch is well formed but opens widely into the mouth invagination. There is no indication of the lateral lobes. The anterior end of the fore-gut, which will later form Seessel's pouch, extends farther cranially than does the hj-pophyseal pouch. At this time the oral membrane is intact.

Chick embryo, 59 hours of incubation (30 pairs of segments). The hypophyseal pouch (Rathke's pocket) has deepened and now exhibits two lateral enlargements near its attachment to the oral epithelium. As later models show, these are theanlagen of the lat«'ral lobes from which the tuberal processes develop. .As may be seen from figure 2, Rathke's pocket is shghtly constricted just above the lateral lobes. The lateral lobes have the fomi of blunt ridges which protrude laterally and also somewhat nas^iUy. Their long axes lie parallel with the long axis of the entire h.vpophyseal pouch. This embryo shows one small perforation in the oral membrane.

Chick embryo, 67 hours of incubation. The lateral lobes are more prominent at this stage due to the fact that the hypophyseal pouch is beginning to be constricted somewhat from the





oral ojivity. The constriction of Rathkc's pocket dorsal to the lateral lobes is also more distinct than previously. Each lateral lobe contains a lumen communicating with the cavity of the main h}7iophyseal sac. Seessel's pouch is in contact with the dorsal wall of Hathke's pocket for a considerable area. This is the eeto-entodennal fusion which has been recorded by numerous observers.

Chick embryo, 72 hours of incubnlion, figure 3. The hypophysis anlage is closely appUed to the brain wall, causing the nasal surface of the pouch to be sharply concave. The lateral lobes are more prominent than in the preceding .stage. The lumen of the pouch extends well into each lateral lobe. One striking feature is the extensive degree of communication between the cavity of Seessel's pouch and the hypophyseal sac. The two open into each other almost to the summit of the ecto-entodemial fusion. This cau-ses the opening of the hypophyseal sac into the oro-phuryn.\ to be relatively larger than in pre\-ious stages. From an examination of embryos of this age alone the impression might be gained that the lateral lobes are being added to Rathke's pocket. A critical comparison of this and younger stages, however, indiates strongly that the lateral lobes of the chick do not arise independently of Rathke's pouch, but that they are formed from it. In this we support the observations of Rossi.

Chick cinhryo, 96 hours of incubaUon. The principal feature of interest in this stage is the beginning recession of Seessel's pouch and its separation from the hj'pophyseal .sac. The lateral lobes are more sharply marked off from the superior part of the hypophysis, but otherwise this stage does not exhibit any striking differences from the preceding. .

.\11 figures represent wux-pliite recon-structions of the epithelial hypophysis as viewed from in front and from the left side. S, Seessel's pouch, R, Rathke's pouch, /./., lateral lobes, t.p., tuberal processes, st., hypophyseal stalk.

Fig. 1. Hypophysis region from chick embryo of 21-2 pairs of primitive segments (end of second day of incubation). X 100.

Fig. 2. Hypophysis from chick embryo of 30 pairs of primitive segments (59 hours of incubation). X 100.

Fig. 3. Hypophysis from chick embryo of 72 hours of incubation. X 100.

Fig. 4. Hypophysis from chick embryo, o days (120 hours) of incubation. X 75.

Fig. 5. Hypophysis from chick embryo, 6 days (144 hours) of incubation. X 75.


Chick embryo. 5 days (ISD hoiirn) of incubation. By tliis lime a definite h yp< )pliysoal stalk has been fonned. It is hollow and affords a coniniunioation between the lumen of the hypophysis and the oral cavity. The lateral lobes have increased in size so that the transverse diameter of the gland, measured between the lateral extremities of the two lobes, is almost twice the transverse diameter of the superior part of Rathke's pocket. The lateral lobes ar<^ united by a prominent ridge around the inferior and nasal enil of the hj-jiophysis. This solid median protuberance doubtless corresponds to a vestigial '\'orraum' or 'corpus I«)buli bifiu-cati' of other vertebrates as described by Woerdeman. The lateral lobes are beginning to be solid, also. At this stage they sometimes contain lumina, which, however, no longer conmiunicate clearlj- with the main h3^pophyseal cavity.

Seessel's pouch is represented by a solid bud of epithelial cells just dorsal to the hypophyseal stalk (S, fig. 4). Curiously enough Economo labels this bud the remains of Rathke's pocket.

Chid: embryo. 6 days (144 hours) of incubation. The hypophyseal stalk is much elongated and has become solid. Near its connection with the oral epithelium may be seen the bud-like remains of Seessel's pouch. The superior, or distal, half of the hjT)ophysisis is bent dorsalh" and forms an angle of about ninety degrees with the inferior or proximal half of the gland. The constriction near the middle of the gland is pronounced. Distinct 'tuberal processes' have formed from the lateral lobes. Instead of projecting so much laterally-, thej' are now directed toward the brain wall. The tub(>ral proc(\sses are not located at the extreme nasal end of the gland but are seen to protrude from about the middle of the inferior half (fig. 5).


The lateral lobes, from which the tuberal processes arise, may be distinguished in a chick embryo having 30 pairs of primitive segjiients. From a careful study of stages preceding and following the rupture of the oral membrane it is evident that the lateral lobes are not fonned independently of Rathke's pocket


and later added to it, but are rather fonried secondarily from the nasjil wall of the early hypophyseal anlaK*'.

The lateral lobes, in all foniis studied, appear earlj- in devel«»pment. This would indicate that they and their derivative in higher vertebrates, the pars tuberahs, are of fundamental phylf>genetie importance. Thus gwat interest is attached to the broad homologies drawn by Woerdeman ('14).


Atwell, Wayne J. 1918 The development of the hypophysis cerebri of the rabbit (Lepus cuniculus L.). Amer. Jour. Anat., vol. 24, p. 271.

Brum, A. C. 1915 Sullo sviluppo del lobo ghiandolare dell' ipofisi negli Amnioti. Internat. Monutschr. f. Anat. u. Physiol., Bd. 31, S. 129.

EcoNOMO, C. J. 1899 Zur Entwicklung der Vogelhj-pophyse. Sitzber. d. kais. Akad. d. Wiss., math.-naturw. Classe, Bd. 108, Abth. 3, S. 381.

Herring, P. T. 1908 The histological appearances of the mammalian pituitary body. Quar. Jour. Ex. Physiol., vol. 1, p. 121.

LoTBRiNGER, S. 1886 Untersuchungcn an der hypophyse einiger Saugetbiere und des Menschen. Arch. f. mikr. Anat., Bd. 28, S. 257.

Rossi, U. 1896 Sui lobi latcrale della Ipsofisi. Monit. Zool. itul., 7, p. 240.

TiiJJET, Frederick 1913 An analysis of the ju.xtra-neural epithelial portion of the hypophysis cerebri, with an embryological and histological account of an hitherto undescribed part of the organ. Internat. Monatschr. f. Anat. u. Physiol.. Bd 30. S. 258.

Woerdeman, Martin W. 1914 Vergleichende Ontogenie der Hypophysis. Arch. f. mikr. Anat., Bd. 86, ,S. 108.




F. A. McJUNKIN Deparlmenl of Pathology, Marquette Univerrily School of Medicine


In an earlier report by the writer ('18) it was shown that the phagocytic mononuclear cells present in the peripheral blood arise by mitosis from the endothelium of the blood vessels. The method that was devised for this purpose consists of the intravenous injection of lampblack suspensions and is not, therefore, applicable to human tissues. The tis.sues of animals injected with carbon suspensions in which the endothelial leucocj-tes and cells are characterized bj' carbon particles ingested by phagocj-tosis are, however, well adapted for testing the action of various stains on these cells. It has been found that the staining method u.sed by Craham ('16) colors these leucocytes in a characteristic way in both animal and human tissue. Since paraffin or celloidin sections caiuiot be used, Graham emploj-ed frozen sections for his stain but owing to their thickness, they are not suitable for accurate cell identification. The purpose of this paper is to record a new method of tissue imbedding for obtaining thin sections to which the stain is applicable.

The .staining method of (iraliam depenils on the action of solutions of alphanaphthol on parts of the cytoplasm of cells. It was shown by O. Witt ('82) that a blue dye (indophenol blue) is formed by the oxidation in dilute alkaUne or acid solution of alphanaphthol and dimethyl-para-phenylenediamine. Winkler ('07) and others found that myeloblastic cells react in a characteristic way in tissues treated with these two substances and the


190 F. A. MrJUNKIN

grunulcs of the reacting rolls wore said to contain an oxidizing frrniont (oxydase or peroxidase) that oxidized (he two conii)ounds and caused them to unite with the production of a blue color in the neutrophilic and eosinophil granules. Later Loele ('14) found that the treatment of mjeloblastic cells with alphanaphthol solutions alone produced the same blue color in the granules. The part played bj- the cell granules has usually been regarded as an oxidizing one ('oxydase or peroxydase reaction). As applied by Loele the phenomenon consists of a pur])lish or bluish color in the cytoplasmic granules of the cell produced by treatment with old alphanaphthol solutions. If a dj'ostuff is produced by alphanaphthol alone, and this seems likely, an aromatic compound or compounds must be supplied by the leucocytic granules. If such is the case "indophenol reaction" is a better term than "oxydase reaction."

Leucocytic granule stain (Ciraham). Since hydrogen peroxide is added to the alphanaphthol solution to make it immediately activ(> and the swollen granules are heavily and permanently stained by treating the preparations with an aniline dye, the method of Graham is better than the other indophenol staining methods devised. To applj' this method, remove thin sections of formalin-fixed tissue attached to slides from the distilled water and stain in dilute Q-5) hematoxylin (Dclafield) for five minutes; wash them in distilled water, place in a saturated solution of lithium carbonate for five minutes, wash in distilled water for two minutes, and stain for ten minutes in 10 cc. alphanaphthol .solution to which 10 drops of 1 per cent pyronin (Griibler) have been added immediately before placing the preparations in it. Prevent evaporation by covering the dish. Wash the sections in distilled water, place in saturated aqueous solution of lithium carbonate for from five to ten minutes, wash in water and differentiate and dehydrate in 95 per cent alcohol for one-half minute. Complete the dehydration by inniiersing the slides in xylol and raising them above the surface of the liquid two or three times and blotting with smooth, soft filter paper. A flat oblong .staining, the .size and width of a slide, is used for the staining and differentiation of the slides. Mount in


coloplioniuni-xylol or acid-fr(M> balsam. Tho tinu' that tlio prpparations remain in tin* saturatod lithium carbonate aftor troatmont with alphanaphthol is important because this removes the excess of pyronin. With some tissue better results are obtained by staininp only five minutes in the alphanaphthol-pjTonin solution and differentiating for a shorter time in the lithium carbonate after this stain. If the one per cent pyronin solution is to be kept for some time, sufficient formalin should be added to make it 10 per cent formalin.

The alphanaphthol solution is made by dissolving 1 gm. alphanaphthol (Merck Reagent) in 100 cc. 40 per cent ethyl alcohol (made from absolute alcohol) at a temperature of 50°C. and adding 0.2 cc. hydrogen peroxide. Commercial hydrogen peroxide containing approximately 3 per cent hydrogen peroxide, as detennined by titration with decinomial potassium permanganate, should be used.

Soap method of imbedding. 200 grams transparent glycerine toilet soap are placed in a 5(X)-cc. Erlenmeyer or Florence flask, that contains 200 cc. distilled water. The soap must be so hard that it is brittle and cracks apart when cut with a knife, otherwise the soap solution will not be of the proper consistency. The flask is placed in the paraffin oven at 52°C. overnight in order to dis,solve the soap. Remove it from the oven and place for three hours in an incubator at 37.o°C. The contents should be a syrupy liquid and should solidify when a small amount is poured into a paper boat and allowed to stand at room temperature for one-half hour. If solidification does not take place, 20 grams of soap should be added to the flask, and the contents again melted in the paraffin oven. After the correct consistency has been obtained the soap solution is placed in 100-cc. wide-mouth bottles with cork or glass stoppers, with about 50 cc. to a bottle.

To imbed the tissue small pieces are taken from 10 per cent fonnalin and dropped into the melted soap contained in one of the bottles which is placed at 37.5°r. two to three hours and occasionally shaken. The liquid soap in the bottles usually become solid jifter remaining at 37.5°C. for a day or more. To melt the solidified soap, the bottles are placed in the paraffin oven for

192 F. A. McJUNKIN

an hour, the s(jap coolctl to A't'^C, the tissue added, and the Inittles rephic«'d in the incubator at 37.5°C. The soap solution with the tissue in it is eniplitd into a box of suitable size made from paper as in paraffin imIjediLing and the tissue arranged on the bott«.>m of the box with forceps. The box should be made from paraffined paper or the paper may be coated by pouring melted paraffin into it. At the end of about one hour the paper is removed, the soUd soap trinuned with a knife to thi; desired size about the tissue and the blocks attached to a heated metal disk just as paraffin blocks are attached. The blocks after about one hour are dropped into a siiturated solution of sodium chioriile in a pint Mason jar, and the jar placed hi the incubator at 37.5°C. overnight.

With forceps remove the block from the saturated salt solution, attach to rotarj- microtome and cut away the block until the tissue is reached. Carefullj' trim the block and allow to dry for from three to six hours until a ribbon 6 to 8 microns thick cuts perfectly. The di.sks maybe detached from the microtome and after the proper drj-ing again attached, so that the ribbon comes from the very surface of the block. The ribbon is placed in distilled water in a fiat dish more than 6 inches in diameter, and sections floated on shdes on which there is a thin coating of fixative made by adding 4 cc. of a very thick .syrupy celloidin to 16 cc. oil of cloves. The preparations after being pressed out and carefully blotted with filter-paper are placed in the paraffin oven for fifteen minutes, when they are removed, washed in 95 per cent alcohol for thirty seconds, and placed in distilled water where they remain less than five minutes. It)nization in the large volume of water in which the soap sections Jire first placed develops only a shght alkalinity, and in the thick soap solution ionization is practically absent. The sjiturated salt solution hardens the blocks since it prevents hydrolysis by mass action. If the tissue is quite fragile the ribbon may be placed in sjiturated salt .solution instead of distilled water.

Reaction of sectio7i.s obtained by the soap method of imbeddimj to the stain. The nuclei are blue and the granules of myeloblastic cells, eudothehal cells and endotlielial leucocytes arc red. (iraham


notrd fh<' red granuli-s in both ciidothclijil cells and ondothcliiil leucocytes and ixpluincd their presence there as the result of ingestion by phagocytosis of niyeloblastic ceils or the cytoplasmic granules of these cells. He does not make it clear whether the grannies were found in all endothelial cells and leucocytes.

In the thin .soap sections it is evident that many cells of endothelial origin contain the granules and that they are not present here as the result of an accidental phagocytic phenomenon is shown by the small size of the granules and their even distribution in the cell cjioplasm. The granules difTer from the myelobla.stic granules .seen in neutrophiles and eosinophiles in being f(>wer in number, snudler in size and more di.scretely distributed. In ti-ssues in which only mature ipol\Tnorphonuclear) niyeloblastic cells are present a casual glance is sufficient to distinguish the endothelial leucocytes containing the red granules since they are mononuclear. In ti.ssue containing myeloc.vtes, the heavier staining and greater number of granules of the neutrophilic and eosinophihc myelocytes separate them from the endothelial leucoc\-tes. The differentiation between myelocytes and endothelial leucocytes is well shown in sections of bone-maiTow

(fig. 2).

The characteristic action of the stain is best .seen in sections of the liver or other organ of an animal that has received intravenous injwtions of a lampblack suspen.sion according to a method devised by the wTiter ('18). Endothelial cells containing carbon and definiieiy lining the sinusoids have nnl granules scattered in their cytoplasm (fig. 1). Likewise carbon-containing endotheUal leucocytes in the vessels show the discrete red granules. The neutrophilic and eosinophilic leucocytes are more conspicuous than the cells of endoihdial origin owing to the greater number and larger size of their granules. There are a certain number of neutrophiles, eosinphiles, and endothelial leucocytes that do not take the stain; anil it is only in the endotheUal cells with a distinct amount of visible cytoplasm that the granules may be distinguished. The failure of some C(>lls to stain appears to be due to an error in technic but all attempts to correct this have failed.

194 F. A. McJUNKIN


Witt, O. 1882 J. Soc. Chem. Ind., p. 255.