Talk:Anatomical Record 13 (1918)

From Embryology



EDITORIAL HOARD Irvino Hardesty Warren H. Lewis

Tulano University Johns Hopkins University

Clarence M. Jackson Charles F. W. McClure

Uolvorslty of Minnesota Princeton University

Thomas G. Lee William S. Miller

University of Minnesota University of Wisconsin

Frbderic T. Lewis Florence R. Sabin

Harvard Untvecslty Johns Hopkins University

Georoe L. Streeter

University of Michigan

G. Carl Hdber, Managing Kditor

1330 Hill Street. Ann Arbor. .Michigan





V. 13

Cop. ^

M »


No. 1 JUNE

II. 10. JoKDAN. Studies on striped niusolo .structure. III. 'I'lie comparative histology'

of ciirdiiic and skeletal muscle of scorpion. Twenty-one figures 1

Margaret Hekii Lkwis. fhe effect of certain vital stains upon the development of

the eggs of cerebratulus lacteus, Kchinorachnius ])armii and Lophius piscatorius. . . 21

Ai>nERT M. Reese. The blood of alligator Mississippiensis. Kight figures 37

S. B. Grant. A persistent superior vena cava sinistra in the cat transmitting coronary

blood. One figure 45

Hollo E. McCottkk. The vomero-nasal apparatus in Chrysemys punctata and Kana

catesbiana. Seventeen figures 51

No. 2 JULY

IIouEK Ci. KisHEU. Histological structure of the retractor penis muscle of the dog.

Four figures (^two plates) 69

\Vak(j Nakahara. Preliminary note on the nuclear division in the adipose cells of

insects. Eleven figures 81

J. T. Patterson and C. G. Hartman. A polyembryonic blastocyst in the opossum.

(Jne text figure and two plates 87

|{. M. Stkonc;. Some observations on the origin of melanin pigment in feather germs

from the Plymouth Hock and Brown Leghorn fowls. Six figures 97

(lEORGE W. Corner. Maturation of the ovum in swine 109

Lloyd K. Reynolds. Hyperphalangism accompanied by supernumerary epiphyses

and muscular deficiencies. Twelve figures 113

Ivan E, VVallin. An inexpensive projection apparatus. Three figures 127


Ralph D. Lillie. Variations of the canalis hypoglossi 131

.Mary T. Harman. .\ case of superfetation in the cat. Three figures (two plates) . 145

Laboratory .\pparatus. I. .V simple electric thermo-regulator. H. S. Bxtrt. One

figure. II. \n automatic air pressure valve. III. .\ laboratory timing clock.

One figure. .VoonsTUs'F. Scharfk 159

Marsh Pitzman. Some suggested changes in nomenclature 165

Charles Eiioene Johnson. On the development of the liver in the genus Citellus.

Four figures 166

Newton Miller and James S. Godfrey. A note on the anastomosis of arteries and

veins in a cat. One figure 177


.\i.iiert M.' Reese. The anatomy of a two-headed lamb. Eight figures 179

Fi.dHKNi'H R. Sauin. Preliminary note on the differentiation of ungioblasty and the

method bv which they produce blood-vessels, blooil-plasma and red blood-cells as

in the livuig chick ".




Miiilim on llir iiminm«ry (cIhikI. III. A coiiipiirison of the developing II in iii:ili- iind foinnlp iiltiino nits from the lute fetal stages to ten

«. -. vrii ligtin'M 205

II I ltM.<-.M vMi.l I. Kw/. Vn iiHjimliilile slide liiisket. Five, figures . 227

Nn. ,-) (XTOHKH

J A I iiiifiiiul study of the effect of stress and strain on bone de VI irns 233

... AM) JtAYMiiM) Pkaki.. Sex studies. IX. Interstitial cells in the

• organs of the rhickcn. Six figures 253

i.i.-ii-ii. li. .\\it.\ . rhiigocytoHi.s liv ostoorlasts. Four figures 269

S K Wiluamh AMI |{ \\ Haiiii. The anatomy of a double pig (Synrephalus thoracopagus I . Seven figiircn 273

KiMiKiui K. Oi.ivKii Mi-iiil>r:ini-.s nf the right iliac region.. . . . 2S1

('«KHo\ (iiixArtni: »M) lloHMiii llii,i. IlKfsToN. Detailed study of a monster with

■Taniorarhischisi.i ami nlhrr anomalies. Six figures 289

Iv*\ K. .V teaching model of a 10 mm. pig embryo. Three figures. 295

KdwamI) t'AKKiiLt, Dav. Two conveniences for the laboratory. Three figures.. .. . 299 MAmin W. Lton, Jk. An hereditary case of congenital absence of one kidney. One

figure " 303


G. ("aki. HuBKR. On the morphology of the renal tubules of vertebrates. Twenty-two figures 305

C Caki. llrnr.K and Arnold H. Egoerth. On the morphogenesis of the papilla foliata of the rabbit. Six figures 341

\V«hh»:n H. Lcwi.s and Maroaret R. Lewis. The duration of the various phases of mitosis in the mesenchymp cells of tissue cultures 3.i0

I'aii. B. Katon. The cfM-liac axis. One figure 369

F. IV>TAi.r-. IVsarrollodo los mil.sculos oculares recto cxterno y recto interne, on el

rmbritin humnno. Diei figuras ". 375

John a. KiTTKi-xoN. The poatnatjil growth of the kidney of the albino rat, with observations on an adult human kidney 385

II. S. f;t"r>Ei.i.. An anomalous case of blood formation in the connective tissue of the

sciatic nerve in man. Six figures 409

I*. K. Unkback. A scheme for drawings in a course in embryology. Three figures. . . . 419


.\|»m T. IIahwan. .Vnother case of gynandromorphism. One figure 12.')

J .M. .VtNimt (iAi.vAo. The finer structure of the I'iliary ganglion of Ophidians 437

A. or. I.KMiw ToHKi'-H. On the ciirlrlaginoiis tissue of the heart of Ophidia 443

J MoHKiHA MA ItcH'iiA. Slaiiiingof adult cartilage by Lundwall's methods 447

J«ur... |{ ('aim l)n the ilevflopmenl of the lymphatics in the heart of the embryo pig 451 KiTii It^Nh I III the relation of the licad chorda to the pharyngeal epithelium in the IMKpnilirvo: » contrilmtioii to the development of the bursa pliaryngca .iiid the toii.silla ' * .'••1. Ti-n figures 465

• iiAi.i. I)»:TWii.rR. «ln the use of Nile blue sulphate in embrvonic tissue >. iM..|ii iiitniiofi . 493

sTrDn:s ox sthiim:!) muscle sTHrcTURK



Department of Anatomy, I'nivcrsity of Virginia


I. ixTHonrrTiox

Tho present investiiiatioii tleiils witli the striped muscle of the Floruhi scorpion.- iuchicling tlie ('tail'), prealuloininal, ceplialo-thoracie. leg. and heart muscle of both the adult and the new-born. In the latter material occur unportant developniciital stages, which tlir()\v valual)le light uptm the more involved adult structure.

As was to be expected the strijied nuiscle of scorpion is ver\' similar to that of limulus, and accordingly confonns more closely to the vertebrate tyjje of strijx^d muscle than to the arthro])(Kl type as excmplilied, for example, by the sea-spider, .\iiopoldact j'lus lentus, already describetl, and by certain insects.

Aj)art from being, as far as I am aware, the first detailed microscopic stmly of this muscle, the chief \alue of this investigation inhei"es in the information it yields, by reason of the greater

' The two earlier studies of this series eoiieeriied the musc-le of the king crab (I'roc. .\iii. .\ssoc. .\iiat.. .Viiat. Hec, Jan., I'.llll; eoniph'te pafH-r in a forlheoiuiiiK piililieation l)y the Carnegie Institution of Washingtonl. anil that of the seaspider (.\nat. Hec, vol. 10, pp. 493-508, 1916). The pertinent literature is fully reviewed in these papers, and a l>ililiographic list ineludeil, and will not here l>c repeated in detail.

' This tnaterial was very kindly furnished alive liy Dr. William I'atten, of Dartmouth College. It was variously fixed in the strong rhrom-aeeto-osmic solution of I'lcinniiiig, and in the alcolinl-nitric-arid solution of /immerinann, and stained liy the lleideidiain iron-hemaloxylin method, in eertain eases followed hy Van (lieson's stain.



'2 II. K. joiin.vx

il<;irn<>>s in irs|H'ct of soinc importunt details in this tissue, n-nanliiiK the minute stmetiue of slripe«l muscle in general. TIM'S*" »iet nils conceni es|H'cially : (a) the nat ui-e of the sarcolemma ; (h) the n'lation of tin- telophragma to the sarcolemma and to the nuclear wall: (e) the mode of muscle giowth and development, hoth with n«s|x>ct of the multii)lication of the nuclei, and the iiicrejiso in the numher of the myofibrils; (d) the decisive evidenc*' in contravention of the interpretation of striped muscle structuH' in ti-rms of 'muscle cells' (Ajiatliy; Baldwin) andextrar4>lluhir myolihrillae. in analogy with connective tissue; and (e) the relation l)<'tw«><>n the nuiscle fibrils and the tendon fibrils. The cardiac muscle of scorpion and of Limulus conforms with vertebrate heart muscle in its sjTicytial structure, in the more or less n'Rular railial arrangement of the myofibril bundles (lamellae) and in the definite character of the relatively less complex cross striping. I'he skeletal nuiscle also of these fonns agrees with vortobnite skeletal muscle with regard to the major stripes, and in the ab.^'iice (or exti-eme temiity) of the mesophragma and the ucces.sory disk (N-disk) of l^ngelmann and Rollet, conspicuous in certain insect muscles. However, the cardiac and skeletal muscles of these forms are, in respect of minute structure, somewhat mor«' clos<'ly similar than is the case in vertebrates generally, possil)ly the ivsult of relatively clo.ser functional similarity. Whether the above-named similarities between arachnitl and vertebrates' muscles have genetic significance, as would s<>em probable on the basis of many other observations on other structures eiuunerateil by Patten,' or whether they indicate only functional eciuivalence remains an o]kmi (luestion.


n. Skeletal muscle of adult

The .skeletal muscle fibers are long multinucleated elements tap«>rinK towards their extremities. They vary greatly in diameter in difTen*nt r<'gions of tin- body. Those of the post • Pntti-n. W. Thr Kvoluticiii nf ilif \'crtcl>ratca and tlirir Kin. Ulukiston, I'hilailrtiihin. IUt2.


ahdonion arc very stout and of irregular form in eross-soction (fig. 1 a and !>); elsewliciv in tlic Ixxiy tlic HIxts aro of much smaller girth, more unifonn in size and of an a])i)roxiinateIy circular outline in cross-section. In resjjcct of minute structure the skeletal stri])ed muscle Hljers of the .sevend regions of the body are essentially identical. Hence the description may confine itself largely to the coarser fibers of the post-abdomen.

In ligure 1 a and b are illustrated two larger and more irregular fibers, in transvei-se section. 'Hw vesicular nuclei are located centrally within an iiregular core of finely granular, verj-^ delicately reticular, saicoplasm. Peripheral to this is a wide border in which the myofil)rils are arranged in the fonn of lamellae radially dispo.setl. Many of these lamellae show a peripheral split of gi'eater or less depth, giving to these lamellae a Y oi' \' shape. Certain lamellae show a double ])erii)heral sjjlitting. This condition most probably is a gi'owth phenomenon, in the number of the lamellae, ami the size of the fiber, being attained by a radial longitudinal fission of parent lamellae, as is the case in growing skeletal fibers of certain teleo.-<t fi.'^hes, e.g., the rain-bow trout. Certain iil)ers show a grouping of the lamellae into Imndles, suggestive of K()lliker's columns (Cohnheim's areas, in cross-.'^ection ; fig. 1 b) of mammalian skel<>tal mu.scle. The irregular form of the fiber in cross-section exjilains the variation in wiilth of fibers as seen in longitudinal sections.

The fibers are invested by a distinct sarcolenuna, which varies in difTer(Mit fib«>rs between a relatively coarse memltrane and one relatively tlelicate. Closely atUierent to the sarcolenuna superficially is a more or less delicate and wide-meshed fii>rillar connective tissue, with smaller more elongate and generally more chromatic nuclei. .Vfter Van Cieson's stain the connective tissue (iK>rimysium) is coloretl red, while the sarcolenuna remains unstained. After prolonged staining in a solution concentrated with acitl fuchsin, both the telopliniginata i ground membranes; z-meml)ranes) and the sarcolenuna stain pink or red.

In longitudinal sections the nuclei ai'e seen to be elongate elements with irregular sinuous or serrated contour (figs. 2 and 3). Such sections, jjroperly stained with iron-hematoxylin, show a

4 H. K. JOltU.VN

(li.stiiict cnKss-st nation <»ss«Mitially like lliut of vortolirato voluntan- striiM^d inuscli>. Tlio st nations of the relaxod fiber include the linliter .l-«lisk. tlie darker C^-disk, and the Kranular telophnifnna. 'rh«'re is not the slightest indication of an additional niesophragina ( M-nienil)rane) and tlic accessory disk of ICngelinann and Hollet (^N-lineJ. The granular teloiihragnia stains n-latively intensely. The granules represent swellings at the |M»ints when' the inyotihrils are attached to the telophragma. Centrally the telophragma passes through the sarcoplasmic iion-tilirillar con> and becomes intimately attached to the tips of the nuclear s«>rrations. \\'ithiM the core the telophragma ap|)ears non-gr;iiiular and stains less intensely (fig. 2). Peripherally the telophragma makes a close union with the sarroleninia (Rg. li). This extra-fihrillar portion of the telophragmata is also non-granular and less deeply .staining. The sarcoI«>mnia is usually festooned hetwcn^n successive telophragmata, |)roital)iy a fixation artifact. The serrated character of the nuclear wall may he in jmrt similarly an artifact.

Figure :\. at the left, shows a iM'culiar condition freciuently met with. Here the sjircolenunas of two adjacent fibers have become blended, and the common membrane has been thrown into a zig-ziig fonn, the apices with their attached telophragmata altenuiting in the apposed fibers. Such result is made ])ossible by n'a.^»n of a local ])aucity or practical absence of inter-fil)er connective tis.sue allowing thus of a fusion of the adjacent sarcolenunas. The process of such fusion can be pictured by reference* to illustration figure 4, which represents the tajiering end of a lilwr and part of an adjacent fiber. Here the connective tissue l)etwe<>n the two filn^rs is relatively abundant and closely wiherent to the highly festooned sarcolemmas. If we imagine the conne<-tive ti.s.>iue n>moved, ;ind tiie adjacent san'olemmas thus l>rought together .so that the crest of a festoon on one side fits into the trough W'tween successive festoons in tiie ai)posed win-olemma. then by a blending of the two .sarcolenunas, and under the opposed lateral tractions of the attached telophragniata, the conditijui shown in figure ^ would become realizinl. Certain steps in .such a process, esfMH-ially the one where a doulile


.sarcolciiiina witliout iutervoning conuectivo tissue occurs, can actually be traced in the sections. Similar conditions have been described in inamnia'ian heart muscle (Heidenhain, 'Plasma undZclIe'j.

b. Skeletal ruuscle of (he new-hnni

In the no\v-l)orn scorpion not all of the muscles ;iri' <'(|ually well developed. Those (jf the cephalo-thorax, abdcMiicn and legs seem in general functionally mature; though the oval, regularly contoured, and more chromatic character of the nuclei (figs. 5 and (i) indicate a lesser degree of progressive differentiation. (Vrtain fibers in the region of the head are in the early stages of transformation of myoblasts into multinucleated fibers.

Figine ."> illustrates a lightly-stained mature fiber of this stage, with a festooned sarcolemma closely adherent to which is a fibrillar connective tissue. Figure 6 is more deeply stained and thus shows a more conspicuous Q-disk and the granular tt^lojihragmata. Tliis fiber is moreover distorted, apparently l)y a force causing compression along the longitudinal axis. The folding or bending thus produced caused an alternate bulging on ojiposite sides, and the liend and the bulge are always along the levels of the telophragmata. This i)henomenon demonstrates a firm connection between myofibrils and the telol)hragmata. It demonstrates also the absence of a mesophragma of a structure similar to that of the telophragma.

In figure 7 is illusti'ated a mid-phase of contraction. Here a new disk, the H-disk (of Hensen) becomes conspicuous. The ll-disk is an incident of contraction. Figure S illustrates the completely contracted condition, in which ilarker contraction bands alternate with lighter disks of apjiroximatc^ly eiiual or somewhat gieater longitudinal width.

Figinv 9 represents a contracted hl)er ujion which is sui^erposed terminally a traction stress. I'nder .<uch conditions a flexible mesophragma (M-membrane) if present, but suppi)s*xlly ordinarily invisible in vertebrate strij>ed nuiscle because of its alleged tenuity, should be brought into view l)y reason of a tliicUening after releas(> of lat(>ral tension. Such was the case,


jiH pn'vioiisly (l<'scrili<Ml. in the leu imisclc of the soa-spidor. Hut ill 111*' scorpion striiH'<l inusch- no iiu'soplirugina becomes'rniliU'.

In linun- 10 is shown the region of transition from a small RhfT of the jaw to its connectinK tendon. The myofibrils and tendon fibrils an- in direct continuity, the latter apparently n'pn*s«>ntinK modification products of the former. The tendon stains m<m> deeply than the muscle in iron-hematoxylin preparations, and faintly red after \'an (liesons stain. The cros.s striiM>s fade out at the level of transition from the muscle to the tendon.

In the head region certain Hbers can be seen representing early stap-s in the metamorjjhosis of a myoblast into a. long multinucleatiHl slriiK'd fiber. Figure 1 1 illustrates a myoblast with four miclei, the next to the lo\v<>nnost one in process of amitotic division. The lower extremity shows also a transition to the developing tendon. The sarcolenmia is relatively delicate. The myofibrils are relatively comjiuct. and give to the cell a rt'latively dwi^^r color. No cross striations have yet appeared. Such first terminally and apjKar progressively towards the center. The first striations are the telophragmata; the Q-disks apjK'ar only some time later. Tlic nuclei imiltii)!y by direct division. Coincident with the nuclear multiplication and the apjK'arance of striations, the fiber elongates. Diu-ing this process {il)ers occur in which as many as twenty nuclei may l)e count<>d in close series in a relatively short extent of the central sarcoplasmic core.

Figure 12 shows two atljacent young fibers from the anterior n-gion of the cephalo-thorax. The jieripheral lamellae of myofibrils <H'cupy less than a third of the radius. A few lam<'llae may he seen in pn)cess of radial longitudinal splitting, by which event the lamelhu* are increa-sed in number and the fiber grows in diameter.

c. Cardiac muscle of the adult

The heart tul)e consists of a single layer of muscle fibers ftralM'culae), invested ix-ripherally liy a layer of connective


tissue (pericardium), l^oth the scorpion and the Liinulus heart hick iin eiulothcHal lining. The myocardial fillers are for the most part arranged in an ol)li(iuc circiihir direction. The original fil)crs lia\c anastomost'd in the adult heart so a.s to form a continuous, hrandiing, spirally disposed, syncytial muscular membrane, 'llie fibers of this meml)rane vary in the degree of compactness and the diameter of their cross-section. In a medial longitudinal section of the heart tube groups of from three to four or more, compacter alternate with a similar numl)(>r of looser cros.s-sections. In the transverse direction the more compact portions of the fiber alternate with the loose portions due to the spiral arrangement of the trabeculae. .\nd in paratangential section, the compacter portions may be seen to pass by gradual stages into the looser portions of the opposite side. It appears probable that the compacter portions represent the extremities of the original myoblasts. According to this interpretation the derivatives of an original myoblast would extend for about two and one-half turns around the adult heart.

Figure 13 illustrates a transverse section through a more compact portion of a myocardial fiber. The midei are located centrally in a coarsely granular sarcoplasni. The delicate sarcolemma bounds a peripheral layer of similar construction. Between the peiii)heral and the central granular sarcoplasmic areas, the plates of myofibrils are radially disposed, and grouped into smaller bundles which anastomose internally, thus giving to the heart nmsculature a double-syncytial character, as in Limulus. A few of the lamellae are split peripherally. At the lower pole of the ilrawing, the point where the liber curves from a lateral to a ventral jmsition, the lamellae are cut very obliquely.

The heart muscle fiber in longitudinal section is practically identical with the skeletal nniscle fiber (figs. 14 anil 1.")). This very close similaritj- between heart and skeletal muscle filler was striking also in the of Limulus. In figure lli is shown a bundle of fibrils in extreme contraction. The change in .shape of the nucleus, and the finely serrateil character of its wall, as compar(>d with conditions in figure 14, demonstrate a very firm iMiion IxMween the telophragmata and the nuclear membi-ane.


d. Caniiac muscle of the new-born

The hoart tube of tlic new horn scorpion is very like tlial of the lulult, t'xrept that tlic syncytium is less compact, tlic fihcrs an" smaller and more rcgvilar in cross-section, the nuclei are utoutor. have a more regular contour, and are somewhat more chromatic, indicating a lesser degree of differentiation; and the inyoliliril lamellae are relatively few in number and more regularly disposed radially without any segregation into smaller bundles (fig. 17). A few lamellae are undergoing a longitudinal fission.

.\s indicated by the form of the nuclei, different portions of the myocanliuni are at different stages of differentiation. In the younger portions, the nuclei are actually multiplying by direct division (lig. IS); in other portions, practically adult conditions obtain (fig. 19). Certain portions have suffered contraction, imposed extraneously pos.sibly by the coagulation effect of the fixing fluid; in such portions the nuclei are strikingly modifietl (fig. 20), in that they are shortened longitudinally, caused to widen transversely, and the peripheral serrations are greatly accentuated. .Mso the telophragmata are less widely spaced in .such regions. The conditions shown in figure 20 can be readily conceived as derived from figure 11), where also four serrations with attached telophragmata are present, by compression exerted at right angles to the long axis of the fiber. Figure 21 illustrates a similar condition, and shows besides a stage in the direct division of the larger imcleus; and a mass of nucleated connective ti.ssue is shown at the right. The plane of tiudear division pa.vses between successive .serrations, which brings it to pass that the daughter nuclei retain their connection with the telophragmata.

m. Disri'ssioN .\nd conclusions

One f)f the most striking facts regarding the scorpion muscle, as also regarding the Linnihis muscle which it very closely resembles, is the es.sential identity in microscopic structure between the cardiac and the skeletal muscle. The axial location of the nuclei, the radial lamellar arrangement of the myo


fibrils, and the relative simplicity of the striations all inriicate a relatively low degree of diflerentiatioii. This circumstance is probably a reflection of a relatively low functional ref|uirernent and a coincident approximate similarity of action. Muscular movement is in both instances leisurely, and the meagre reciuirements are presumal)ly satisfactorily met bj- a relatively simple structural condition.

Furthermore, the minute structure of the scorpion muscle resembles closely vertebrate striped muscle (especially cardiac muscle; note figs. 1 to 3). ,\s a subdivision" of the arthrofxxls the muscle of the arachnids might be expected to resemble closely that of insects. But there is not the slightest indication of the mesophragnia and the accessory disk of Engelmann and Rollet (N-disk), (•()nsi)icuous in certain insect muscles. The fact that the mesophragnia is not discernible in a fil)cr like the one shown in figure 9, where the contracted fiber has been stretched terminally, leads me to conclude that a mcsophragma is actually absent in this muscle; and that this membrane is characteristic only of very highly difTerentiated fibers. Moreover if a mesophragma occured, of the same nature as the telophragma as is claimed by Heidenhain, we should expect, under the conditions represented in figure ti, a double series of foldings, that is, at the levels of the alleged mesophagmata as well as at those of the telophragmata. Whether the similarity here also, that is. between the muscle structure of certain arachnids (scorpion; Linmlus) and vertebrates (as is more probably the case as between cardiac and skeletal muscle of these forms) signifies likewise only a functional eciuivalence. or whether it has any phylogcnetic significance, —or po.s.sibly both — is a (juestion whose further consideration lies outside the scope of this investition. All thai need be said by way of interpretation is that the-^ie additional facts harmonize with Patten's theory of the arachnid origin of \'ertel)rates.' Moreover, lioth tlie ■<ct)rpion

  • Among the large body of evidence presented in support of this theory Patten

noted that "The iiriiohnids resemble the vertebrates in more general ways, as in the minute structure of curtilage, muscle, nerves, digestive and sexual organs,' (p. XVIII, Historical Sketch); but he nowhere illustratesor describes in detail the microscopic structure of the arachnid striped muscle.


niul Limuliis mytK-artlia arc .syncytial in stiuc-turt", very like thr iiiyocartliuni of vortphratos; but the scorpion cardiac musculature is only one filxT in thickness, while that of Linuilus is many fihers thick: and hoth imisculatures practically lark lon^iludinally disposed iihers.

'I'hc observations al)ove recorded repardinij the relation of the telophrapmata to the nucl(>ar wall, when taken in connection with sin ilar observations in Linuilus nuuscle and that of cat and mouse,' wouKl seem to definitely dispose of any further profitable attempt to interpret striped muscle in terms of 'muscle cells.' and extra-cellular niyofilirillae. The close union between the telophragmata and the myohbrils, sarcolemma, and serrations of the nuclear membrane, above described, demonstrate that the fiber is the cellular unit. There can remain no longer any (|uestion of a smaller definitive cellular element. -Moreover, the history of the development of the myoblast disproves .such interpretration. The myoblast is a stout fusiform cell, envelope<l by a membrane which becomes the definite sarcolemma and .supplied with a nucleus which divides re]ieatedly by amitosis. The cell becomes measurably elongated, showing first more anil more conspicuously the telophragmata and then the (^-disks appearing in a medial jirogression, the simple myoblasts thus becoming transformed into the definite multinucleated striatetl muscle fiber.

.\nother interesting similarity between this muscle and that (tf vertebrates concerns the manner of growth. This is by radial longitudinal splitting of the myofibril-lamellae, a process .substantially identical with that occurring in embryonic fibers of certain teleost fishes, e.g., the rain-bow trout.

This radial disposition of the myofiliril-lamellae resembles also the condition in the adult heart muscle of vertebrates. liut in the latter the fibrillar constitution of the lamellae is readily di.s<'ernible in cross-sect irms of the fibers. In the case of the scorjnon and Linuilus skeletal fibers the finer fibrillar condition

' Jordnn. H. K. Tlir mirrosropic structure- of niammnliun ciirdiar muscle with aperiBl rcfrrcncp to Ro-callcd muscle cells. Annt. Roc, vol. 8. 1914.


of the lamellae is much less conspicuous and occasionally quite impossible to recognize in transverse sections: in the cardiac muscle of these forms the myofil)ri!s are less compactly aRgregated into lamellae, and in consequence are easier to distinguish. The condition in the post-abdominal muscles approaches that described for certain insect wing-muscle, e.g., Libella, in which, according to E. Holmgren*^ the radial lamella cannot be further analyzed into constituent hbrillae. Hut when one observes the .scorpion muscle very carefully, especially in thin longitudinal section under the oil-inunersion lens, one may .satisfy himself that the lamellae here actually do consist of fine myofibrillae, some of which appear to cross between adjacent lamellae. Moreover when the coarser myofibrils are carefully followed for some distance, they may be seen to resohe ir)to still finer hbrillae, to the limit of visibility. The internal fibrillar (lamellar) constitution of this muscle fiber thus appears to be of a syncytial character, and the evidetice with respect to the successively finer resolution of the lamellae into more and more delicate myofil)rillae offers support to Heidenhain's protomere hypothesis of histologic structure as applied to striped muscle, namely, that the ultimate vital units are metamicro.scopic fibrils. '

This material shows also very clearly that the tendon is a derivative of the myofibrils. Myofibrils and tendon fibrils are in direct continuity: the sarcolemma apparently blends with the more peri])lu'ral tendon fibrils.

A feature of Limulus heart muscle which brought still closer the similarity between its structure and vertebrate cardiac muscle was a simple type of intercalated disk. Such di.sks could not be detecteil in the scorpion heart. This does not necessarily mean that they were not actually present somewhere, nor that they could never appear in any heart of any age or any condition of functional strain, ."^uch a statement could only be made with approximate certainty if a complete heart tube 'of an old specimen were carefully searched, which has not been done. Even

« Sec llciclcnliaili. M. •l'l:i.«m.i uiul 7a-\U\' p. .iSI, 1!I|I. • lli'idfiiliiiiii. .M. 'Plasma and Zcllc," p. 5S2.


III mliilt 1-iiiuilus such disks an- coinparatively rare. If my hy|HillM>sis. which is supported by a number of suggestive obserVHtidiis, and is in harmony with the actual histologic facts, is rorn-ct. namely, that tiicsc disks represent irreversible contracliitn bands, possibly the result of a local impairment of function, then one would perhaps hardly expect such disks ordinarily ill Ihe scorjiion heart: but they might possibly occur sparsely in aged specimens.

The blending of the sarcolemmas of two adjacent fibers, and the ensuing alternation of the tips of the resulting zig-zag membrane, and of the attached telophragmata calls for discu.ssion. It is interesting to note that essentially identical structures occur in vertebrate and human heart muscle." The interpretation which at once suggests itself regarding this alternation of structure between the apposed .surfaces, is that the adjacent fibers have been drawn slightly in opposite directions by the tensions exerted during de\elopment or function. Rut Ileidenhain calls attention to the fact regarding vertebrate cardiac muscle, in o|)position to this interpretation, that the numerical relationship of the serrations and the attached telophragmata on the two surfaces of the common sarcolemma is always as n to n + I; and he is led to the explanation of this j)heiiomcnon as the result of a spiral distortion of a fiber about a longitudinal axis.

The sim|)ler conditions in the scorpion skeletal muscle lend support to Hcidenhain's hypothesis. In the case illu.strated in part in figure 3 the condition.s are as follows: the two fibers are from the peripheral region of a post-abdominal segment : the fiber at the left comes to a point above, the fiber at the right terminates similarly below; above and below near the termini of the pointed ends the telophragmata of the two fibers are at the same level; between j)oints, the telophragmata of the adjacent fibers alternate in such a manner as to leave fifteen sarcomeres on the left and only fourteen on the right. The pointed termini of the two fibers at the end of this field may be accounted for by the a.ssumjition of a spiral twisting about a

• Hridpiihitin. .M. 'Plaatna iind Zcllc,' pp.641 to 544; 615 to 617.


common axis, producing thus in a longitudinal section the asyninu'trical alignment of the opposite sarcomeres, according to the conception of Heideiihain. It should again he noted that there is with respect of this phenomenon also a very close similarity between scorpion skeletal muscle and niamn)alian heart muscle.

One of the most puzzling and interesting features of the developing scorpion nniscle. both skeletal and cardiac, concerns the multiplication of the nuclei hy the direct method of division. Not a single mitotic figure api)ears in the muscle tissue of this form; but abundant instances of amitotic nuclear division may be seen. The plane of division is alwaj's between successive points of attachment of the telophragmata, so that the daughter nuclei remain susj)ended by these membranes. But in this connection it should I)e noted that in the earlier stages of metamorphosis of the myoblast into the muscle fibers, when nuclear multiplication is most active, telophragmata are not present, at least not in such complete form as to be recognizable under the ordinary magnification. Such dexice would .^^eem to aid, by rea.M)n of the absence of possible restraining membranes, the intensive and raf)id multiplication of the nuclei, necessitated by the metabolic requirements of the 3'oung, rapidlj* differentiating muscle fiber.

This material would seem to offer a favorable opportunity for the study of the underlying causes of amitotic nuclear division as opposed to karyokinetic <livision. This may constitute the basis of a future investigation, .~^in(•e the technic here employed failed to disclose a centrosome or archoplasmic .substance, it must remain uncertain whether such is actually lacking or whether it simply was not visibly preserved. In consetiuence no final argument can be made on the basis of this material for or against the hypothesis that amifosis is the consequence of some factor which puts a restraint upon the activity of the centrosome substance in a growing tissue. However, in the certain absence of jKithologic factors here, it n\ight with some reason be a.s.sumed that the possible restraining agent might be lack of relatively adeciuate luitriment, caused by the rapidity of growth


mill (litTtTiTitiatiDii of tin- myoblast, ami tlius intorforing with rfi\tr«»«iinal activity. However tliis may l>e, it .seenis oloar, in view of the doiihlo function of the nudcu.s, namely, as an organ in control of constru<-tive tnptal)oli.<m and of specific heredity, that the nniuiremenls of growth and differentiation are am|)ly met by the increa.><e and distribution of nuclear materials by din«ct division. The cause of amitosis in this muscle may inhen' in an economic adaptation to essential needs, rather than in an indirect restraining influence upon the activities of the karyokinet ic ntechanism.

The regular reticular system of telophragniata and sarcolenuna is so conspicuous and robust that one feels inclined at first sight to interpret the whole mechanism as a connective tis.sue derivative. And the absence — whether apparent or real is uncertain of the telophragmata in the and in early stages of transformation into the muscle fiber, gives further stn-ngth to this inclination; as also the fact of the very close union between connective tissue and sarcolemma. Hut two chief facts negative any .such interpretation: (1) the intimate union between telophragniata and the nuclear membrane; (2) the difference in staining reaction to \'an Gieson's solution of the connective tissues and the sarcolemma: and the similar lack of affinity of the sarcolemma and the telophragniata for the acitl fuchsin. The first point is in harmony with the derivation of the telophragmata from the cytoreticuluni of the myoblast, by |)roce,s.s of condcn.sation and growth and a rearrangement into n'ctiiinear me.shes. somewhat according to the suggestion first nuule by MacCallum." Similarly, the sarcolemma is the adult repn-sentative of the original cell meiiibraiic.

The decided festooning of the sarcolemma in this material niises the riue.stion of the cau.sal factor in the production of this condition. Huber'" suggests on the basis of his studies of rabliit voluntary striped mu.scles in teased prejiaration th:it the

• MBrCiillum, J, B. On the histoloiO' und hisluKcncsis of the heart muscle cell .Xnnl. Ani.. IW III IS'tT.

'* Hutior. f!. Carl. On the form nn<l iirninKcmpnt in fasciculi ofsfriatcd volun»ry muiirlr IiIhtb. Aniit. Iter., vol. II, p. IOC.



festooiiiiijr (il the sarcolemnia, desfribod l)y certain authors, may perhaps be brought in relation with the ends of myofibrils which do not extend the entire length or tlie muscle fiber.' It is impossibh- for me to speak witli certainty renardiiiK the h-ngth of the separate myoftbrils in the scorpion or the I.inmhis nuisde. One gels the very clear impression from a study of thin iotiKitudinal sections that the hirfje myofibrils are divisible into fine fibrillar components, and that the entire fibrillar constituent forms a loose meshwork or syncytium. On this basis one miRht perhai)s lefjitimately conclude that many of the myofibrils at least do not extend the entire leiiffth of a fiber. Hut the rather intimate interconnection of the entire fibrillar component of the fiber, coupled with the fact of the continuity of the terminal myofibrils and the sarcolemnia with the tendon fibrils, makes difficult the application of Huber"s sugp;estion in regard to the formation of the festoons of the sarcolennna. The explanation that most strongly suggests itself to me, as the result of my observations in the scorpion muscle, is that, due to the intimate imion between myofibrils and the telo])hragmata and between the l.'ilter and the sarcolemnia; and the difTerence between the contraction etTect of the fixing fluid u])on th<' myofibrils and the sarcolemnia, the latter is forced to bulge in order to adapt itself to the greater shortening of the successive sarcomeric portions of the involved myofibrils. The same jihenomenon would result in the case of functional contraction.


Tlir illu.slrnlioiis wiTc inaili- with tliu aid <if llic H, and I.. drawiriK i-ainorii. tlic del ails l)ciii(5 added frcr li:irid. Tlic tii:iKi>>'<<">t>i»> <>> "H ca.Mc.s is I.VX) diamplcrs, rcduci'd oni'-lliird in rcprddiirtiiin in liKurrs 1 to 1'.'. oiit'-fourlli in timircs 13 to '21. fnlcss ollicrwisc spiM-iKrd the Iccliiiic cniployrd was tixation l>y Zimnirnnanii's alfoliol-nitric-acid soliilion fullowcd liy tlir inin-lipinatoxylin and \*:in (lirsitn's stains.


Kig. 1 Triinsvcrse sections of two liirRc fihors from the poHt-ubdomon. The filx-rs viiry uri'ntly in sliiipo iinil dianietiT in cross-scrtion. The nurici iirc rontr:illy pliiccd in n very (Icliciilcly reticular siircopliisni free of myofibrils; they are rireulnror oval in eross-seetion. The fibers are elosely invested by a sareolemnia of varying delieaey. which does not readily take the Van (iieson's stain. Intimately adherent to llie sarcolemnia is a more superficial fibrillar connective tissue ('perimysium') which stains a more or less deep red in Van (iieson's mixture. The myofibrils are arranged in the form of radially disposed |)lntes, many of which lamellae are in process of longitudinal fission, probably a growth phenomenon. The cleavage is initiated peripherally so that many of the plates have the form of a V or V in cross-section. In b the myofibril-plates are aggregated into columns (the probable analogues of Kiilliker'seolumnsof mntnmalian striped muscle) siiggi'sting in cross-section Cohnheim's areas. The smaller cephalo-lhoracic and leg muscles are nearer the size of myofibril columns. The shape of their cross-section c.\plains the variations in width seen in longitudinal sections of the post-abdominal muscles.

Fig. 2 Portion of a longitudinal section of a post-abdominal muscle in the relaxed condition, including three nuclei. The sarcolemraa is not indicated. The fiber contains centrally a continuous non-fibrillar sarcoplasm; this sarcoplasmic core contains an extremely delicate cytoreticulum and occasional small masses of granules (myochondria). The nuclei have an irregular or spinous contour in longitudinal section, in pari possibly the result of fixation distortion. To the apices of the peripheral projections or ridges are attached the tclophragmata or ground (Zi mendjranes. Where the latter |)ass among the peripheral myofdirils they have a granular appearance. The gramiles stain more deeply and mark the point of attachment of the niyofibril to the telophragmaia (fig. 6). Uesides the telophragmata there can be distinguished in the relaxed myofibrils also the J-and Q-disks. Not the slightest indication of an H-disk or a mesophragma (.M-membrane) appears.

Fig. '.i Portion of a longitudinal section of a fiber including both surfaces, and the i)eriphery of an adjacent fiber, showing besides the above-enumerated features also the sarcolemnui. The latter is regularly festooned between successive telophragmata, to which it is intimately attached. Peripherally, beyond the myofibrils, the telophragma is non-granular. At the left where the two fibers are apposed, the intervening common sarcolenuna a.ssumes a zig-zag character, the telophragmata of the adjacent fibers alternating, and so drawing the crests of the 'festoons' into sharp points. It seems probable that in such cases the adjacent sarcolemmas become intimately blended through the medium of a sparse amount of intervening connective tissue (fig. 4), the close ap|)roximation (and fusion) and the alternation of the opposite telophragmata being the result of a spiral torsion about a common axis of the two adjacent fibers.

Fig. 4 Portion of a paratangential section through the tapering end of a filH>r, and the periphery of .an adjacent fiber. In this case the inter-fiber connective tissue is more abundant, and so ])revents a blending of the sarcolemmas. This illustration shows well also the non-granular character of the extra-librillar portion of the telophragmata. The condition shown in figure '.i coidd be derived from the one here illustrated by an interlocking of the festoons of the adjacent




% "'






f '"I « 


q >






^ 8







tiU-r-H. Ml tliHl llif crcHls of mir would lit into (lie troiiKlis lit'lvvccii successive >|i|Miiii(<- fcMtoniiH, ami a 8uliiir(|uriil lilriidiiiK of the siin-olcliinias.

FilC. ■*> I'ortion of ii loiiKiluiliiial scclioti of a filicr from the ccplialo-tliorax of Ihc iH"W-lHirii. This lilicr is ill tlu- rcla.vii coiulition. Note the oval, refjularly I'ontiiiireil, a ml more cliroiiiatic character of the nucleus in these youiiuer fillers. At the left is shown the connective tissue, with one nucleus, closely ailherent to the fistooncil luircolemimi. The connective tissue stains red in \un Ciieson's solution: the sarcolenuna remains unstained. When the staining; ))rocess is greatly proIonKi'd the telopliraKXiata as well as the sarcolemiiia arc colored red.

I'iK. ti i'ortion of a loiiKitudinal section of a relaxed fiber from the same region, <lis(or(ed liy a coiiii)ression-8( ress aetini; at risht angles to the long axis of the filMT. The line of transverse hending is always at the level of the telophragma, thus demonstrating the close attachment of the myotihrils to the ground-mendirane, and thi' aiisence of an additional niemlirane Iniesophragma) of the nature of the telophragma.

Kig. 7 I'ortion of a longitudinal section of a fiher from the .same region at a mid-phase of contraction. .\n additional II-<lisk has now made its appearance.

Fig. N I'ortion <if a longitudinal .section of a contracted fiher from the same region, showing the contraction hands alternating with lighter disks of approximately twice the longitudinal width. The positions of the telophragmata are indicated l>y the contraclion-liaiids. At the right is shown the relatively coarser sarcolemma.

Kig. !> I'ortion of a longitudinal section of a contracted fihcr from the jawregion. I'pon the contracted condition is superposed a longitudinal traction. This circumstance should l>ring into view the mesophragmata if actually present (as is the cas<> in certain in.sect muscles); but not the slightest indication of such a membrane api>ears.

Fig. 10 Longitudinal section of a small fiber from the head region at the exlri'tiiity where it is attached to the chitinous cxoskeleton by a short tendon. The fiU-rs of the latter are directly continuous witli the myofibrils; they stain ~<>mewhal more tleeply in ordinary preparations, and pink after Van Gieson's ?<tain.

Fig. II Longitudinal section of a large multinucleated myoblast from the head region. The myofibrils are more compactly arranged and hence the (iU-r to stain relatively more deeply. \o cross-striations have yet made their ippearaiice. Such appear first towards the extremities in slightly older myoi'hists. and progress centrally. M the lower extremity the tendinous connection is diffcretitiating from the myoblast fibrils. The definitive striped muscle fiber uriws by process of elongat ion of a single myoblast. Such fibers become multinurlented by a process of nuclear amitosis. In slightly older stages as many as 20 nuclei have lieen counted in .scries in the same, relatively short, sarcoplasmcore.

Fig. 1'2 Transvers<- .sections of two adjacent young muscle fibers from the hcnd region of this same s|H'cimen. The myolibril-platt's are radially disposed, of relatively short width, and only a few are in process of longitudinal fi.ssi(m.

Fig. 13 Transverse section of a smaller traU'cula ( fiber i of the cardiac syncytium from a paramedial longitudinal s<'ction of the heart tube. (Flemming's fluid ;iron-hcnintoxylin stain I. The heart -tubi' wall is one trabccuhi in thickness.



^v //,







16 ,'





The miclci arc ccnl rally IdcalctI in a granular, very dolicati-ly reticular, sjircopiasin, froc of myotihrlls. The myofibrils are pcripluTally arrangfil in radial InmoUno, which increase in nunil>pr liy InnRitudinal fission. The lamellae are Kalhereil into smaller l)iin«lles, or siil)-tralieciilae. which nnastomos*- freely and thus form with the main traheeulao or fihers a <loul)le syncytium, as in the Limulus heart. The 'fibers' are invested by a delicate sarcolemina, In-tween which and the pericentral myofihril-lamellae is a frequently relatively extensive, coarsely Kranular. non-filirillar sareoplasm. .\t the lower Imrder of this section, at which point the 'fiher' turns from the lateral to the ventral wall of the heart IuIh«, the

20 II. K. J OKI) AN

liiiiiflliif iirr cut very iil>li(|Ui'ly. Thin (ilnT is cut »l n piiiiil wliorc the iiitcriiiil synrytiiiiii in iiiiiri* coinimrt, pnilmlily nciir llu'fxtrrinity of the i>ri|;iiiiil myiiMiist.

Kig. 14 Portion of n loiiKitniliiinl Brrlioii of a ciiniiiK- tnilx-riila (filirrl from tlip niinu- lioarf. nntcriorly. Tin- intoriiiil sym-ytiiil structure i» horo Honiowlint loowr. (Itlu'r portioiiM iirr still more loosely nrr»nn<'<l. This filipr is in the n-laxfd i-oiulilion. Note tli<- close similarity, an essential iilentity, as also in the ease of I.imuliis, lietwi^en cardiac alitl skeletal muscle I fi(js. 2 and ^), especially from the standpoint of sliajM' and character of these nuclei, and the relation of the telnphraKinata to the luicle.'ir wall and the sarcolemma.

I'ii;. I.i Portion of a lon)(itndinal .section of n rnrdiac lrali<'cul,'i in the relaxed rondition, from the same heart. Compare with skeletal fillers, (ignres 3 and 6.

Fi(t. 16 Portion of a longitudinal section of a contracted .smaller cardiac tralH'cula. from the same liearl. Note the close series of contraction hands, and compare with skeletal muscle, figures. The shape of the nucleus and the serrated character of its wall, can lie ri'adily intiTpreted in terms of the condition illustrated for the relaxed fiber (liu- ' ' ' ^ which condition can lie conceived to pass into that of this illustration liy .a process of a coincident longitudinal short eiiinf; and a transv<-rse widening of the .sarcomeres (inokommata) and the nucleus.

Kin- 1" Transverse section of five adjacent earrliae trabeculae ('fihers'l from a loiiKitudinal .section of the heart tulie of the new-horn. To the left is shown a portion of the peripheral connective tissue ('pericardium'). The traheculae are oval in cross-section with a large central non-fihrillar, hut granular nnd very delicately reticular, sarcoplasmic area, in which the oval nucleus is generally locale<l. The nuclei show the regular contour and the relatively more chromatic condition, characteristic of the earlier stapes of development. The myofiliril-l.'imellae are peripherally arranged in a radial manner. Only a few of the lamellae are untU'rgoiiig a loiigit\idinal radial splitting at this stage. The filler is invested liy a delicate s.arcolemma. immediately internal to which is a narrow granular zone of non-fihrillar sarcoplasm.

Fig. IS Portion of a longitudinal .section of a cardiac fiber from the same heart tube, showing the amitotic multiplication of the nuclei, and the intimate continuity between the telo|)hragmata and the sarcolemma peripherally, and the nuclear wall centrally.

Pig. ID Portion of a longitudinal .section of a more differentiated cardiac fiber of the same heart, in the relaxed condition. The mn-lear wall is thrown into spinous processes or ridges ti. which are attached the telophragmata.

Kig. 'in Tiber from the same heart ami in the same ri'laxed condition, but niechani<'ally shortened iprobably thnnigh the contracting influence of the fixing fluidi, causing thus a longitudinal shortening and transverse widening of the nin-leiiH, and a coincident accentuation of the peri])heral serrations, but without a detachment of the supporting telophragmata.

Fig. 21 Similar fiber, fnmi same heart, showing in addition a stage in the process of nuclear amitosis; and a peripheral mass of nucleated connective tissue lat right) closely adherent to the festooned .sarcolemma.


MARGARET RKED LEWIS From the Harpswell Laboratory, South Har/iiiretl, Me.

The stains which were used in the following experiments were those which have been frequently used in observations upon tissue cultures of chick embryos, namely Janus pjeen, neutral red, brilliant cresyli)lue 21), gentian violet, pjrol blue and others.

The object was to determine whether any new light as to the action of these stains could be obtained by a study of the effect of these stains ujion the fertilization of eggs and also upon the later development of embryos, especially as to whether the effects of these stains is in any manner harmful or whether these stains are in reality vital stains, that is whether it is possible to stain the granules of a cell without injury to either the cell or the developing embryo.


The cerebratulus eggs proved to be the most plentiful as well as the most favorable material for ob.^ervations upon the effect of the stains upon fertilization and development. The eggs lived and (l(n'(>lope(l normally in the laboratory and were fre(|uently kept in a healthy state until in the late pihdia stage or for time or four weeks.

These eggs have a so-called gelatinous membrane surrounding the characteristic egg membranes, which enclo.-^e the egg. The egg membranes have a tube-like protuberance at the end of which is an opening. The gelatinous membrane has been suppo.<ed to become dissohed off after the eggs have lieen in sea water for some (ime (Wilson), but this is not the case as can easily be demon 21



stnitod by flio uso of stains. The gelatinous inetnbranc Ix'conios inucli loss visible after the e^gs are placed in sea water, but when a few tlrops of .lanus preen are acliled to the sea water this membrane takes up the stilin and can be distinguished as a pinkish blue layer, which has become separated quite some distance from the enclosed egg. The membrane takes up the Janus green stain, but does not readily permit it to pass through, and thus protects the egg from the toxic effect of the Janus green so that unstained eggs may develoj) up to 4 or 8 cell stage within the deeply stained gelatinous membrane before they are killed b}- the stain. Dr. Robert Chambers found that the gelatinous memlirane can be removed without causing any injury to the eggs by l)assing the eggs in sea water througli a ilouble laj'or of ordinary laboratory gauze. The egg membranes can be removed by using a tine cambric handkerchief in place of the gauze. When the handkerchief is used care must be taken not to break the eggs, as fre(|uently happens when the eggs are handled roughly. Bj* thus removing the membranes naked eggs are obtained upon which the stains act more rapidly than ui)on protected by the egg membranes. Also these eggs which are thus unprotected by the membranes can be stained by Janus green, in which case the toxic effect of the Janus green kills the egg in the same manner as it kills the tissue culture cells. The develoj)ment of the naked eggs obtained in this manner was normal; the j)olar boilies were given off; and the segmentation planes came in as usual, except that the ciliated blastula. instead of being confined within the membranes until the ])iliiiia stage, became at once free swimming and developed into normal pilidia.

The eggs of cerebratulus are relatively large spheres full of various granules and globules. The unfertilized egg has a clearer space at one side where there were fewer granules. The unfertilized eggs were subjected to the action of these various stains. In no case throughout observations was the cytoplasm or nucleus stained. Occasionally the nucleolus was faintly stained. When the eggs were placed in a weak solution of Janus green numerous minute granules in the cytoplasm became blue; when


))lacp(l in JU'utral rod solution thoro appeared an eciuallj* Rrcal nuniljor of larger rod glohulos; when placed in a double stain of Janus green and neutral red in addition to the blue granules and red globules many unstained globules scarc-ely evident before l)ooamo quite prominent. These are proi)al>ly yolk gloljulos. Kggs stained from ten minutes to one hour indicated no localization of the cj'toplasm. Detailed observations were not made upon the unfertilized eggs.

The spermatozoon is larger than that of the sanddollar and the head is sickle shaped. The middle piece is in the form of a niuiid disc, as is so characteristic of this body for most Echinodorms as has been shown by Retzius, and to it is attached a rai)idly moving flagellum or tail. With Janus green, the whole mifldle piece immediately stains blue. In a few moments the blue staining material clumps into one or more ma.sses of granules, which later become more or less extruded from the spermatozo(')n due to the toxic effect of the Janus green. Retzius figures the middle piece of many invertebrate spermatozoa in the form of usually 4, sometimes 5 or granules. In some as for instance his drawing of Lucina, shows those granules more or less extruded to one side of the head of the spermatozoiin. This is probably due to a less rapid fixation of these particular spermatozoa. Whoi\ i)lacod in neutral red a small rounil graiuile just above the collar and on one side of the sickle shaped head becomes bright red. There was usually only one such granule, but in a few cases it appeared to lie double. Brilliant cresylbluo 2b stains a granule purple in the same position as that which is stained red with neutral rod and in all probability the same granule stains red with neutral red and purple with brilliant cresylblue 21). With gentian violet the middle piece tirst stains violet, but very shortly the whole spermatozoon becomes a stained diffuse violet, especially the head, the middle piece, and even the small granule above the latter.


KruixouAniNirs tahma (SANnnoi.i.AR)

riir ('UK is quite (liffcroiif in sliapc and has differont cf^n imMiibrancs fronj that of the eerel)ratuhis and yet the results in regard to the various stains are practically the same. The sperniutoz()<)n is smaller than that of the cerebr^itulus and the head inst<'ad of heinp sickle sluiped is sjiear shaped, hut the arrangement of the stainahle material is the same, i.e., there is similar shaped middle piece whicli stains with Janus green and above the middle piece on one side of the spear shaped head is a small granule which stains red with neutral red and purple with brilliant cresylblue 2b. Meves ('12) pictures the spermatozoon of Parechiiuis miliaris nuich the same as that of Echinorachnius. Altliough Meves finds that the mitochondria occujiy the middle piece, he does not ilescribe any graiuile on tlie side of the head where the neutral red granule is located in the spermatozoon of Echinorachnius and also in that of Orel)ratuIus. Meves describes only two types of granules in the egg of Purechinus, one the mitochondria, which correspond with the Janus green granules of the egg of Echinorachnius, and the other the yolk granule. The reason that Meves does not describe the neutral red granule either in the spermatozoitn or in the egg may be due to the fact that it is difficult to fix the neutral red granule.

The eggs of the cerebratulus and also of the sanddollar were studied after fertilization in various stages of development, such a,s S cell. Hi cell, etc. The granules which stained with the difTerent vital stains could be differentiated from one another in (lie diffen'iit typesof cells of the embry.os. The stained granules differ in shades of color and size. .VU granules are more deeply stained on the surface of the egg than in the center. The Janus green granules are very small and often are around the edge of the neutral red graimles and in such cases the neutral red granules are .sometimes slightly bluish.

I'nder the conditions in which observations were carried out, it was imp()ssil)le to weigh out the stains in order to determine the exact solution of tin- various stains used. .\ standard solution of each stain was made up in sea water, or in a few in distilled water, and from this standard solution three


solutions of oaoh stain wore made by adiliiiK duv or more drops of the standard solution to a given amount of fresh sea water. The three solutions were as follows: a weak solution or one whieh contained so little of the stain that the color could scarcely l>e detected when held over a white surface; a medium solution or one which had a slight color; and a strong solution or one which contained enough of the stain to give it a distinct color. Thi.s strong solution corresponds in color with a 1 lOO.OOO .solution of Janus green, a 1 100, (MM) solution of brilliant cresyiblue 2b, a l-oO,0(M) solution (»f neutral red. and a 1 .lO.OOO solution of gentian violet. By this method the amount of a given stain in the solutions of the same strength in the different experiments remained the same throughout the experiments, although the exact amount of stain in gram weight present in the solution was not determined.

Pyrol blue was not soluble in sea water and even when a few drops of a solution of the dye in distilled water was added to the sea water the stain was at once precipitated. This precipitate was not toxic as the eggs continued to develop normally, although th{\v remained unstained.

Numerous experiments to determine the effect of the stain upon the fertilization and later develo])ment were carried out with both the cerebratulus and the sanddoUar eggs and spermatozoa. These experiments were repeated many times and were as follows:

1. Spermatozoa were i)laced in strong, medium and weak solutions of the various stains and then after different periods of time a dniji of these stained spermatozoa was placed in fresh sea water together with fresh unfertilized eggs, either with or without the egg meml)ranes.

2. Fresh unfertilized eggs were placed in strong, medium and weak solutions of the stains and after different periods of time were taken out and placed in fresh sea water to which a drop of sjiermatozoa had been adiled.

'^. I'nfertilized eggs which had been tn-ated with the stain were placed in fresh ,sea water to which a drop of stained spermatozoa was added.


4. I'nfoitilizoil eggs ami sporinatozoa were placed in strong, medium and weak solutions of the stain and left in these solutions.

0. CentrifuRod eggs w(>r(' treated with the stain and later fertilized.

(). Fertilized eggs in \arious stages of development, such as the jiolar liodies forming. 1st division plan(\ '2-cel!, 4-coIl, S-cell, Ui-coll, 3'J-eell. t)4-eell, hlastula, gastrula. pilidia antl later free swimming pilidia, were placed in strong, medium and weak solutions of the .stain and left in them.

Controls were always made with either normal eggs or normal spermatozoa in place of the staineil ones and only the results of those experiments who.>;e controls developed normally were consiilered.

The results of the ten al)ove experiments were as follows:

1. In only a few cases did spermatozoa which had been placed in a solution of Janus green fertilize the eggs, while those which had been placed in weak solutions of neutral red or lirilliant cresylblue 21) for perioils of time varying from two minutes to fifteen minutes did fertilize the eggs even though the granule on one side of the heail of the spermatozocui was colored in the typical manner. SiX'rmatozoa placeil in strong solutions of the stains for long periods of time did not fertilize the eggs. The presence or absence of the egg membranes did not appear to effect the ability of the spermatozoa to enter the eggs.

2. ^Vhen the unfertilized eggs were placed in the .Janus green solution the egg membranes protected the egg from the stain and, even though the outside gelatinous membrane itself became deeply stained, the egg remaineil unstained for some time. When the egg which had Ix'en placed in weak solutions of .lanus green for short periods of time (5 to 15 minutes) were taken out and i)laced in fresh sea water together with spermatozoa, they became fertilized and developed as far as 4 or S cell stage and in one or two cases as far as the pilidia. When the egg membranes had been removed before the eggs were placeil in the .Janus green solution, the eggs were not fertilized and so did not develop. ^^'hen the eggs and the spermatozoa were placed in the


solution of Janus groon at the same time, the eggs oecasionally became fertilized and developed jiolar l)odies and sometimes (li\ ided once or twice before the Janus green killed them. The lirilliant cresyll)lue 2b and neutral red on the other han<l did not kill the eggs, although the egg membranes did not protect the eggs from the stain. When the eggs were placed in weak solutions of these stains for jieriods of time extending from five to fifteen minutes the eggs became faintly stained, but could be fertilized and frequently developed into normal free swimming l)ilidia. A\'hen the spermatozoa and eggs were placed in the weak solution of the stain at the same time the eggs became fertilized and later develdju'd into free swimming pilidia, although the pilidia were red when in neutral red and blue when in brilliant cresylblue 2b. When a strong solution of neutral red or brilliant cresylblue 2b was u.seil the eggs usually died.

3. \Mien the .spermatozoa and the eggs had each been treated with the stain before they were placed together the results were not so good, although occasionally the eggs stained with either neutral red or brilliant cresylblue 2b were fertilized by si)erniatozoa correspondingly stained and developed into pilidia. When Janus green was u.^^ed fertilization did not take place.

4. The results in this case were much the same as those of 2 and '.i except that only eggs left in the weak solutioas of the stain developed.

r>. The centrifuged eggs d('\('loped ]iracti<'ally the same as the normal eggs except that there was a much larger percentage which died. The centrifuged eggs did not show any localization of the stain.

(). The fertilized eggs became increasingly resistant to the toxic efTect of all the stains except .lauus green as development proceeded. AA hile the early stages of development up to the blast ula were easily killed by even the medium solutions of the stains, the later stages of development lived and contimied to develop in the medium solutions and frequently pilidia liv»>d for .some days in the strong solutions of the stains except Janus green and gentian violet. .lanus green, on the other hand, was more toxic for the pilidia than for the unfertilized eggs. All


staK«'!< of tlrvclopiiu'iit wliicli wcri' |)n)t('i't('il by tlio ckk incml)rancs lived for a short wliilc in the modiuni or in the weak solutions of the Janus preen, but the hiter stages of developni('i\t which were free swinuiiing, were rai)idly killed, even by the weakest solution of the stain. The free ends of the eilia ininiediately nmtted together into deeply stained blue clumps which shortly became torn away from the emliryo.

A few experiments serveil to demonstrate the fact that .Janus green cannot be considered a vital stain in the above sense, for not only did it usually prevent fertilization of the- eggs but also in every rase where the solution of the stain was of sufficient strength to color any of the granules of a cell of the embryo, not only was the particular cell killed, but also the embryo itself died. The same was true in most instances where this particular gentian violet was used. The gentian violet was not the same as that used in tissue cultures, but was a stain acquired through the kindness of Dr. Neil at the Harpswell laboratory and was in all ])robability a more toxic stain than the one previously used in tissue cultures in wliich growth of cells occurred when the medium contained a very small amount of gentian violet.

On the other hand brilliant cresylblue 2b and neutral red each proved to be true vital stains, for the embryo continued to develop and the cells to divide, although certain granules of the ceils were colored and the embryos were kept in a weak solution of the dye. In many exi)eriments the granules of the cells were .so deeply colored that the red (neutral red) or blue (brilliant cresylblue 2b) embryo which resulted could be seen with the naked eye.

While the toxic effect of the \arious stains is in a way in proportion to the concentration of the stain in the .solution, the depth of color with which the stain appears in the granule is only to a small extent dcjjendeiit upon the concentration of the stain in the .solution. When the piUdia are left for twenty-four hours or longer in a solution of either neutral red or brillant cresylblue 2b, which contains so little of the stain that the color can hardly be detected, the granules of the embryo take up the stain until it api)ears in the granules in so nmch greater concen


tratioii thiit the embryos can l)e seen as miiiuto rod or blue bodies in the solution.

One of the intcrestitiK results of these observations is the fart tliat the spermatozoon and the unfertilized egK proved to be the least resistant to the toxic effect of all the stains, while the older the embr\o develoj)ed the more resistant it became. However a sufficiently stronp; solution of any of the above stains, except pyrol blue, was toxic and caused the death, not only of the egg or young embryo, i)ut also of the later embryo.

Each stain seemed to have an affinity for a certain type of granule. In the egg and later stage of development .hmus green stained the small round granules, which are scattered abundantly throughout the cytoplasm. In the spermatozoon the Janus green in every case stained only the granules, which lie within the middle piece. In the fertilized egg and later stage of development botli the neutral nnl and the brilliant cresylblue stained definite large and medium sized granules and in the spernuitozoon only a very small granule .situated at one side of the head and just above the middle piece. .V double stain could be obtained by means of .lanus green and neutral retl or brilliant cresylblue 2b. Neutral red stained the large granules dilTerent .shades of red. pink or yellow, while brilliant cresylblue 2b stained the same type of granules various shades of lavender or purple and with this stain the large granules frequently became difT(>rentiated into a pink globule or vacuole which contained one or more puri)le granules. If an embryo was stained with neutral red and then later by brilliant cresylblue, the blue stain replaced the red in the granviles and so far as could be observed diil not stain any difYerent types of granules from those previously stained by the neutral red. Cientian violet stained diffu.sely both the granules which can be stained by .lanus gnvn and also those which can be stained by neutral H'd or brilliant cresylblue "Jl).

In the sjiermatozoon only two kinds of granules wea-e ob.served, thos«> of the mi<ldle piece which can be stained by means of .lanus gi-(H>n and the one above the middlepiece which can Im> stained bv means of neutral red or brilliant cresylblue 2b. In


the t'Rg thoro were observeil tlire*^ kinds of granules, those which cjin l)e stained by meatus of Janus preen, those which can lie stained hy means of neutral red or brilliant cresylblue 21) aJid also a large granule which reniainiHl unstaiiu^il and which may be a yolk granule. The behavior of the various sizeti gi-anules or globules which stain with the neutral red or the brilliant cresylblue 2b corresponds with that of the so-called vacuole described by lx>wis and Lewis in th(> cells of tissui^ cultures.

I.ul'llIlS nSCATOIiUS lANCLKK I'l.sil OK (lOOSl:! KISII)

A few obser\ations were made upon the eggs of Lopliius piscatorius, which were brought ijito the laboratory in the gastrula stage. Unfortutiately these eggs were not obtaiiied again during the sinnmer. so that observations could Jiot b<^ made upon yoiuiger stages, and these obser\ations upoji the older embryos covdd not be repeated. However, they are given below l>ecau.><e they suggest that certain extremeh' interostuig results may be obtained by the use of neutral red upon the developing eggs of this form and possibly other fish eggs.

A mass of the eggs of Lophius piscatorius was put into a weak solution of neutral red in sea water and the remainder of th«> eggs was kept iji normal .sea water. The water was chaitged once or twice a day and the eggs in the normal se.a water develo|)ed into normal fre«> swinunijig embryos, while in the neutral i-ed develojx'd marked abjiormalities. The evident of these was that the chromatophores did not develop in the normal number aixd what chromatophores did develop remained small more or less round cells with black granules. The chromatojjhores of the normal embryo develop into large cells with many ramifying processes full of black gi-anules which fonn a gray network over the great(M' part of the embryo. The neutral red embryo remained more or less transparent, while the nonnal (^mbryo b(>came dark gray and more or less opaiiue.

The yolk sac of Lojiliius is filled with a colorless transjmrent fluid in which is sus]K>nded a consjiicuous yell()W oil droplet and in the normal embryo it is not possible to di.stinguish the boundaries of the yolk sac. When the young embryos were


placed iji tlic luutral rod solution, certain numerous granules within the cpitlicliul cells became stained red, although the cytoplasm of the ceil remained unstained and very shortly after this the red color appean-d in the yolk and contijiued to U- taken up by the yolk until the yolk became many times pinker than the solution of neutral red in which the embryos were placed. This sphere of reil fluid is cojisiderably smaller than the membranes around it and the oil droplet did not take up the stain Imt remained a clear yellow. When such a pink stained yolk sac was punctured the epp; membranes collapsed and the deeply stained, transparent fluid nii)idly flowed out and quickly became .so diluted with the surrounding medium that it could no longer be distinguished from it. The clear, yellow oil droplet usually remain(>tl within the collapsed egg membranes.

The action by means of wliich the neutral red was deposited in the fluid within the yolk did not correspond with that of a simple (lifTusion, Init it apiiear(>d to be ii-ssociated entirely with th<> large neutral red grauuU's which seemed to take up the stain from the solution of the stain in the sea water which surroimded the egg and to give it up again to the fluid within the yolk sac.

In the embryos wliich had develoixnl in the neutral red solution the cells of the central nervous system contained so many red granules that the neural tube appeared pijik. This pink color of the central nervous system persisted throughout the various stages of de\elopment.

.•Vnother esjiecially interesting abnormality in the development of these eggs of Lophius in a solution of neutral red was that, although the heart and the chief blood ve.>*sels ileveloped, there was jio circulation of blooil cells withux the blood vessels and heart. The heart did not begin to beat quite as soon as did the heart of th(> normal eml)ryo, but eventually the heart contracted normally. No blootl cells were ever ob.-^erved to pjiss through the heart or the blood ve.ssels, although the embryos romained alive in the neutral red solution for several days after the circulation of blood cells had Ihmmi establi.><hed in the normal embrvos.


Tlio lM>lmvior of tlic (•hr<>iiuitoi)h(»res and thv lack of a cirrulatioii of blood cells sopin to indicate that the neutral red may have sonunvhat the same ('ffect upon the developinn fish einliryo as that of alcohol upon I'undulus e^K^ observed bj- Stockard, in which case the chroniato|)hores remained small round cells and no lilood cells were observed in circulation, although the heart contracted rhythmically.

l-'ischel ("Oil carried out lunnerous experiments with various stains upon the embryos of the frog and also those of the salamanth'r. Fischel found that Ris!narckl)r;uiii. neutralrot, neutralviolett, nilblauchlorhydrat and nilbluusulfat are all true vitals stains, i.e., that the embryo continued to live and develop although stained. Fischel describes the same tj'pe of neutral retl granule as those described above in the embryo of Lophius. From his observations he concludes that the neutral red granule is a living plasmatic element. Fischel does not describe any changes in the chromat<)i>hores or in the circulation of the blood in the amphibian embryos which were stained with neutral red. This may be due to the fact that the embryos were first stained and then transferred to clean water, while in these observations the embryos were permitted to develop in a weak solution of neutral red.

It was impossible to obtain these eggs again, but it is hoped that some future summer it may be jiossible to acquire enough of fish eggs to carry out a more elaborate series of exix?riments.

The study of the spermatozoa, the unfertilized eggs and the later stages of development led to the conclusion that not only the embryo, but also the unfertilized egg and the sjx'rmatozoon react in the same maiuier as the cells of tissue cultures of the chick embryo do to the stains, that is they contain granules which are stained blue with Janus green, probably the mitochondria, and others stained red by neutral red or purple with brilliant cre.sylblue 2b. The Janus green staiii is toxic and resulted in the death of the embryo just as it caused the death of the ti.ssue cultures, while the neutral red and brilliant cresylblue 2li are less toxic and the embryo may continue to live and develop, although the granules in' the cells are stained. These two large groups of


granvilos, tho Jiinus proon pranulo und tho nrntral rod (franiilft, may \h' furtlicr ditTcn-iitiatcd liv tlic uso of other staiit- Midi as p>Tol blue, l)rilliant cresylblue 2b or iiile blue. In addition to these two types of K'a""'*'^- the e^R and certain cells of the enibrj'o contain yolk rIoIjuIcs or fat glol)ules. Dr. Paul CI. Shipley, who has made a careful study of the neutral roil Rranule, has suggested the te^rm vacuoles of segregation for this body and has studied its behavior in detail. The results of Dr. Shipley's obsorvatiojis will l)e i)ut)lisli('([ iji the near future.

In the discussion of the plastosomes of the living cell (p. tiOS) Prof. .1. Duesberg states that a number of authors have claimed to have stained the jilastosomes in the living cell. The used dyes according to Professor Duesberg are neutral red and methyleneblue, whil(> dahiiaviolett appears to be the most satisfactory stain. Daliliavioiett wixs first u.><ed by von la Valet ta St. George and after him by Hemieguj' i'Otij ajid Fauro Fremiet ('10). Michaelis ('00) and Laguesse ('05) found Janus green very satisfactorj', wliile Renaut ('11 ) used methylviolett and C'iaccio Cll) used brilliant cresylblue. The graiuiles stained l)y these different ob.servers cjuoted by Professor Duesberg camiot be the same granules and therefore not in every the mitochondria for n<Hitral red d<K\s not stain the same granules as those staineil by Janus greeii except in the of the granules within certain, not all, of the vacuoles described by I^wis and I.ewis ('15) (p. 3S2). The rea.son for the discrepancy in the results of the abo\e writers is without doubt due to the condition of the tissue sujjjjosed to be livijig but which iji many wjls probably dying, as can be seen from the following facts.

The action of most of the socalled vital stains cau.^es tho death of the cell more or rapidly and a concentrated .solution of almost any of these stains will kill the cell. In where animals are given increasing doses of stain until the body tissues are stained, the cells when i)la((>d und< r the microscojH' show first one |)i<-ture ajid later as the cells begin to die the siime cells show (piite a different picture. This can easily be tlemonst rated by a study of a living embryo such as the ceivbratulus pilidia in which the death of tlie i)iiidia cause s brilliant cresylblue 2b to fade out

}4 MAK(;.\ltKT iu;ki) i.kwis

of tlic Iiirne uninuU's aiid at the same time to bo taken uj) l)y the niitochoiidria. This sjiine phoiinmenon has boon dosfribcd by I^'wis and I^wis for the action of iiile l)hu' upon tii<> grajuiles of the tissue culture cf'lls, in wliich ease the vacuoles of the living cell were stained i)ink l)y means of nile blue but when formaldehyde vapor was ])assed ujuler tlu^ cover slip the color faded out of the vacuoles and the mitochondria then became stained a blue color.

The results obtained with pyrol blue and possibly with other staiiLs may also lead to confusion because of the change iji color shown by certain of the granules (I^vi, '16). I have observed in tissue cultures that when th(> cells first grow out (12 to 24 hours) into a medium which contaijis jiyrol blue, lioth the many mitochonilria and the few neutral red graimles are stained blue. As the growth increases in age the neutral red granules become more and more dwply stained and also increase in uuiHlx^r, while the color fades out of the mitochondria, so that the cells of a seventy-two-hour growth contaui numerous very dark blue granules and many vnistained mitochondria.

rischel states that the nucleus becomes diiTusely staiixed when dying and in previous ob.servations upon the cells of tissue cultures, the germ cells of the grasshopper as well as in the observations above, it has been fovmd that the slightest trace of color in tlie nucleus is an indicatioji that the cell is ijijured. This does not hold however for the nucleolus, for the nucleolus maj- remain faintly stained with pyrol blue through several gcncratioas of cells and also the chromosomes may be faintly stained with certain stains during division.

Observations upon the living embryo and also upon the cells of tissue cultiu-es show, when the cell has not been injured by the staiji and the nucleus is not stauied, that the cytoplasm of the cell consists of a homogeneous ground substance which contains certain gianul«>s which may l)e stained by means of neutral red and certain others which may lie stainetl by meaas of Janus green, and in addition to these two tyjx-s of granules there may Im' pre-sent fat globules which do not stain with either neutral red or .lanus green.


()l)s<'r\iitit)iis li;i\(' shown tlic proscnf*^ of tlic .hirius Kn***!! KruiuiK' ami also tlu^ neutral rcil granule in the nerves, the red blood eells, the endothelial cells, the ectodenn, the endodenn, the suKtoth muscle eells, the striated nuisde eells and the connective tissue of tlie chick embryo and also in various embryonic tissues of tiie i)ifj;, th(^ (ish, the firasshoi)|)er. the hermit crab, the cerel)ratutus, the sanddollar and the sea urchin. Thc-e granules an* i)r('sent in tlie germ cells of the grasshopiMT, the cerebratulus and the sandilollar and in all probability they are jiresent in all embryonic animal cells.


fl) DcKsiiKKii. J. I'M- I'lastosonicri ".\pparato Kotirolare Iiitorrio unci Chromidialapparat." Er|;cb. dcr Anat. und KntwirkliinRResch.. lid. 20.

l2) FisciiKl.. .\. l'.M)l rntcrxuchiiiigi-n iibor vitaU- riirlmiig. .Viiat. Ileftp, Bd. 16. Heft. I.

(3), Ci. 10l(> Dciiioii.strazionp dclla lint lira rondrinscimiru drgli oricanuli

(■('llulari I'liIoraMli col Moil pirrolo in cellule coltivati 'in vitni.' Ueale .\ccadeinia doi Liiicei, vol. 25.

(4) I.KWis, .\t. It. .\\t) LKWl.-i, W. H. 191.1 Mitochondria and other ryto plasmic structures in tissue cultures. .Vni. Jour. .\naf., vol. 17.

(5) Lkwis, .M. K. and Robertson, W. R. 1916 The mitochondria and other

structures ob.served by the tissue culture method on the male K(^rni cells of t'hortliippus curtipennis Scucld. Biol. Bull., vol. 2.'), no. 2.

(6) Mevbs, Kn. 1912 \'erfolKung des sogenannten .Mittelstiickes ties Kchini dens periniuins in befruchteten Ki bis zum Knile der Furchengsteilung. Arch fiir .\Iikr. Anat.. Bd. SO.

(7) Hetzhs, (i. 190I Zur Kenntni.'j der S|K'rmicn der Kvertebrateii. Biol.

I'nters., Xeuc Kolge Bd. II.

(8) Retzics. (!. 1910 t'ber den Bau des Kies der Echinodcrmen in unlw fruchteten und Ijefruchtclen Zustand. Biol. Unters., Neue Folge Bd. 1.-..

(9) StockaUD, t". R. 191.") .\n experimental analysis of the origin and rela tionship of blood corpuscles and the lining cells of these vessels. Proceedings of the National .Vcadeiiiy of Sciences, vol. 1. p. 556. lit)) WiusD.N, I'. B. The habits and early development of Cerebratulus lacteiis iVerrill). The (^uarl. .lourii. of Mirr. Science, vol. I;{. p:irl I. new series.


AMIKKT iM. KKP:SE DeparlmenI of Zoology, West Virginia Univeritily



In this study of the blood of the alligator both fresh and stained prejiarations were useti. The blood was obtained from an animal kept in the aboratory, by making a small slit between the ventral abdominal scales; the wound ([uickly healed. This operation was reiM»atetl whenever a new preparation was needed.

By sealing the cover with a ring of oil, to prevent access of air, the fresh blood could be kejit in normal condition for many days, during which ix^riod liie amoelioid activity of the leucocytes could l)e studied.

In making stained preparations various fixing fluids were used, but the ordinary dried smears gave the liest results. Of the stains used the best results were obtained with hematoxylin and eosin and with Wright's stain.

THK l•;K^■ll^HOc^TKs

The red cells of the alligator's blood are, as is well known, of the usual elli])tical form seen in the lower vertebrates.

According to (iulliver (2) there is some variation in the cells of closely related species. He found (3), as will also be noted below, that the corinisdes of dried blood are appreciably smaller than those of fresh blood.

He says that in Crocodilus acutus and in an unknown s|x»cies from V< ra Cruz the length of tlic corpusclf is somewhat less than twice the breadth.

TIIC ANaTOSIIC-AL MECORD, vol.. 13, N(l. 1


Oil the other h:ii\(l, Maiull (4) wlio studied C'. lucius says the lonntli is two or three times the width. In only three cases of several dozen nieasurenu'iits made of the corpuscles of A. mississippiensis did the writer find the length as much as twice the width, and in no case did it aiijiroarh three times the width.

In ( ". acutus tlulliver gives tiie avera^jje leii{i;th of the corpuscles as 20.31+ micra; the average width as 10.93+ micra. In C. fissi|x^s the average length was 1".).S5+ micra; the average width was lO.T'.t + micra. .Vs will be seen below, these variations are not nearly so great as those seen among corpuscles from the same individual animal in the Florida .Vlligator, A. mississippiensis.

Milne-Kdwards (5) gives the following measinrments, presumal)ly averages; A. sderops, length 123. <S(). micra; width 13.33 micra; \. lucius length 20.83 micra; width 11.11 micra.

In A. missis.sippionsis a considerable number of measurements was made, both of fresh and of stained corpuscles. The thickness of the corpuscles was also measured in a number of cases; this can be ilone only with the fresh blood where corpuscles can occasionally l)e seen in jirofile.

The average length of the fresh corpuscles was found to be JO. 77 micra the average width 12.78 micra; and the average thickness 4.17 micra. .\s noted above, and as will be seen later, these measurements are apparently greater than those of .stained corjiusdes.

The longest corpuscle found in the fresh blood was 24 micra; the widest was 14.00 micra: the thickest was 5 micra; the shortest was 18.50 micra; the narrowest was 11 micra; and the thinnest was 3.60 micra.

.\mong the stained corpuscles the average length was 18.69 + micra; the average width was 10.8.5 micra, both measurements being noticeably less than in the fresh blood.

The longest stained corpuscle measured was 22.10 micra; the shortest stained corpuscle was 14. SO micra in length, only 0.8 micra more than the width of the widest fresh corpuscle. The widest stained corpuscle was 12.80 micra; the narrowest was 8.10 micra.


Among all the forpusclos. hoth frosh and stainod, tlu> one that showed the greatest dilTerenee Ix'tweeii the length and the width was one that wan 22.10 niicra long and 10.40 niiera wide.

On the stained slides, for obvious reasons, it was not possible to measure the thickness of the corpu.scles, but the length and width of the nucleus were measured in each case. The greatest variation in the length of the nuclei was from 3.S0 micra to ti miera; in width from 2. SO miera to 4.30 micra: the average length was 4.S.") micra, the average width was H.titi micra.

The appearance of the erythrocytes, as seen in the flat. Is shown in figures 1, la, \h, and Id-h, drawn, as were all of the figures, with a camera lucida under an oil immersion objective from a stained slide. The proHle view, figure 1 c, wis, of course, drawn from a slide of fresh blood. .\s would l>e expected from the measurements given aliove. the ellipse varies considerably in ditTerent corpuscles. In the proHle the central thickening is jilain but the nucleus could not be determined.

The cytoplasm, so far as could l>e determined, was homogeneous, though it was examined under a magnification of 23(M) diameters. The cytoplasm was so transparent that, on the ordinary stained slide, when one corpuscle lay over another the nucleus of the imder corpuscle showed witli the same apparent distinctness as that of the upper cell.

This .structureless condition is in contrast to that descriljod by Bryce (1) in I>epidosiren, where he figures a clear band lx»neath the membrane, a fine network throughout the, an<l one or more clear areas and vacuoles in the cytoplasm. Whether examined under a magnification of KMM) or of 2300 diameters, the cytoplasm in .\. mississippiensis apixvin-d the same. Possibly more R'fined methods of technic might have brought out some details of structiu-e in the cytoplasm.

The nuclei, as indicated by the measurements and as seen in figures 1, \(i. Ih, etc., vary both in size anil shajx*. though they are usually ellii)soidal. They stain easily and darkly but not homogeneously; sometimes small unstained or more lightly stained areas are scattereil fairly evenly throughout the lujcleus, as in figure I; sometim(>s larger and more irn^gular artvis are


so<Mi ill Viirious parts <if tlic iiudcus as in linurc \ii. Occasionally a iiucU'Us is located at one cml of tlu' cell instead of its usual central position, and soniotiines a nucleus is seen at each end of the coll, figure 1 g.

No instance of mitosis was seen on any slide exaniin(>d, l)Ut erythrocytes are occasionally found with two closely adjacent nuclei, figinc 1 <■. which would s('<>in to have just resultetl from an amitotic dixision. I'Ijiiut If seems an evid<>nt case of amitotic tlivision of the nucleus just before the completion of the process, ^^^lile the cells in which two nuclei are found are usually of large size and elongated form, but one case could be found in which there was indication of a division of the cell body. This cell is .shown in figure 1 / is j)ossibly an artifact, but it is difhciilt to .see how the nucleus could have been pulled apart by artificial means as is shown in tlie figure.

Quite infrequently erythrocytes of the form shown in figure 1 (/ ai"e seen. It would, at first, seem possible that these were halves of just-di\-ided cells, but if this were the case they should frequently be founds in pairs, while, as a matter of fact, it is but seldom that two of them are found in the same micro.scopic field. It is possible that they may be comparable to the spindle cells found in frog's blood, but it seems more likely that thej' are merely artifacts.


On a slide of fresh blf)od, mounted with a ring of oil to j^revent access of air, as noted al)o\e, the amoeboid activities of the same white corpuscle may be studied for several days, though after a few days the motion is so slow that it can only be determined by making a .series of drawings at intervals of several mimites, as must sometimes be done to demonstrate the changes in shajM^ in amoeba.

In the fresh blood it is diflicult to identify the various types of leucocytes that may be seen in the stained blood, not only because of the lack of stain but also because the pseudopodia are usually more or less withdrawn in the stained l)lood.

I'igure 2 shows one type of leucocyte as .seen in fresh blood. The general outline of the cell is circular and a number of small,


sliiirply-poiiitcd i)scu(l()j)n(liii project frimi its ix-riplicry. lifinn uiistuiiicd, tlic nucleus is indistinct or invisihle, hut one or more small vacuoles may be soon. The cell is filled with fine, unevenly distributed graiuiles which chjuiKe their appearance as the cell rh.iiiKes its sh:i])e. Sucli Ji c<'ll, while it chanfjes its shape but little, changes (juite rapidly — about as fast as the changes se<»n in an active leucocyte in frog's blood.

Another type of leucocj'te, seen in fresh blood, is shown in figure :{; it is coarsely granular and changes its shape quite rapidly and markedly.

In stained preparations of alligator's blood .several types of l<>ucocytes may he distinguished. Of these the most numerous is shown in figure 4; since it stains with hemato.xyliii rather than with eosin it miglit he thought to be an extruded nucleus from an erythrocj'te exce])t that it is several times the bulk of su<'h a nucleus. It is of fairly large size and is usually circular in outline though the sha]je is variable. It is possible that it is a cori)Uscle in which the nucleus is verj' large and tlH> cytoplasm is so reduced as to he invisible. This would seem possible from the fact that occasional cells are found with a very large nucleus and a very thin peripheral zone of protoplasm.

Of almost, if not (juite, as frecjuent occurrence as the cell just described is a smaller tyjw shown in figures rt. n n, 5 b, and 5 r. This cell may, ijerhaps, correspond to the lymphocyte in the human blood. It varies considerably in size and shape but contains an oval or circular nucleus and a small amount of cytoplasm which generally gives the cell a pointed or spindle form, as seen in figures 5 b and 5 c.

Another type of leucocyte that is fairly coimiion is shown in iigures (> and ti a. These forms might, perhaps, be called mononuclear leucocytes: they are large, some of them being larger than any of the other types. The cytoplasm is clear or very finely graiuilar, and stains, with eo.sin, a pale pink color. The nucleus is very large and of an oval or circular outline: it does not stain so darkly as the nucleus of the erythrocyte. The outline of the cell is usually circular or polygonal.

The most striking in appearance of all the leucocytes. an<l, possibly witii one excei)ti(>n, the least numerous, is the ty|)«« 


that may ho coniparod to tlio rosiiiophilc roll of niainnials. It is a largo, usually circular coll tiiat may at once bo rocogiiizod by its coarsoly-gratmlar cytoplasm that tako a strong pink color with oosin. The nucleus is usually round or oval, and generally lies cl()S(» to one side nf tlic coll. as shown in figure 7. Occasionally two nuclei in a single cell may 1)0 soon, figure 7 h, as tiiough li\ di\ision of the larger nucleus; and an occasional elongated nucleus, as seen in figure 7 a, would seem to indicate an impending amitotic division. No was .soon in which there was any indication of the division of the coll as a whole. In many of these eosinophile cells, especially in those in which the cytoplasm did not take the stain, there was seen a heavy outline, like a thick coll wall, possibly caused l)y a iiorijjhoral zone of denser protoplasm; this appearance was usually most marked on the side of the cell farthest from the nucleus. This type is the most imiform in size and shape of any of the leucocytes.

The least immerous typo of leucocyte, if indeed it be a distinct tj'pe, is shown in figures 8, 8 a. and 8 b. Among the tens of thousands of erythrocytes seen, on several difTorent preparations, but throe of this possible type of leucocyte were seen; it is this extreme rarity that raises the doubt as to their being a normal type of corpuscle. They are all of rather small size and not very irregular outline. The cytoplasm is clear or verj' finely granular. The nucleus, or nuclei — figure 8 h shows no less than eightare so dark as to be almost a solid black. Whether these are really a nomial element of the blood or are some artifact or other abnormality it is difKcult to determine.


(1) Bryce, T. II. 1904-5 Tho histology of the Mood of the larva of I^-pido siren parudoxu. Trans. Hoy. Soc. Kdiiil)., vol. 41, no. 9, pp. 201-311 and 435-69.

(2) GcLI.IVER, G. 1840 On tho l>lood corpuscles of the Crocodilia. Pro.

Zool. Soc. London, vol. 8, p. 131.

(3) IS42 On the blood corpuscles of the British Ophidians, Ucptilcs and other oviparous vertebrates, ibid, vol. 10, pp. 108-11.

(4) Mandl, 1S.39 Note sur les Rlobules sanguine du Protde et des Croco diliens. Ann. dcs He. Nat., 2 .^eric, T. 12. p. 289.

(5) Milne-Rdward.h. A. 1856 Note sur les dimensions de globules du sang chcz

quclquc vertcbres. Ann. des Sc. Nat, T. 5, pp. 165-7.




All i)f the figures \viT(> drtiwii with a canicrii liicida under the same mugoification, I'j oil immersion olijoelive uml no. S e(>iii|M>nsating; orular.

FIrh. I, 1 u, 1 6 Three views of erythrocytes seen in the flat ; figure 1 c is a normal erythroeyte seen in profile, drawn from a slide of unstained and living Motni Figure 1 (/ is one of the rather unronimon pointed erylhrocytca, which may be


Miiiiply an iirtifart. I'iKuri- I <' is ii ri'il im-II willi, apparently, a jiisl -divided nu<*lcii8. I'imirc I /i.x a cell in wliieh the niiflruH is in process of division. Kin'irc 1 (7 in n rod cell with a nuclens at each end. KiRiire 1 h represents the single ease that was found that seemed to show an erythrocyte in which the entire cell was in process of division.

Figs. - and '.i Uepresenl two types of leucocytes drawn from living blood while CNhiliiling amoel>oid niolinns; figure .'? represents a more active cell than figure 2. and one in which the granules are coarser.

Fig. 4 Heprescnia a type of doiihtful character, which may he simply a cell with an enormous nucleus and almost no cytoplasm; it is the most common of the leucocytes.

F'igs. 5, .") n, ,")/>, and ,5 c Represent a type of small leucocytes very alnindantly represented.

Figs, (i and 6 a Represent a type of large mononuclear cells that may possibly lie the same as the one shown in figure 4.

Figs. 7, 7 a. and 7 l> .Show three very charaelcristic cells that seem to correspond to the eosinophile cells of mammalian blood; they are very coarsely granular and generally slain strongly with Eosin.

Figs. .S. S n, and S b Represent a type of small and unusual cells that arc so s<>ldom seen as to make it seem doubtful that they are a normal constituent of blood.


S. B. GUANT Washington University


The anomaly of the vascular system stated in the title was foimd (hiriiiR a oourse in comparative anatomy in the Zoolop^Department of Washington University. At the suggestion of Dr. E. A. Baumgartner, a study of the literature of this form of variation was untlertaken, the results f)f which are Iiere given.

This anomaly was found in a young and a})parenlly nonnal eat. Upon ojiening the thoracic cavity, a long slender vein, of uniform diameter, was exposed, which reached from the left innominate vein to the coronar>' sinus, about 4 mm. distant from its atrial end. This was recognized as a left suix-rior \ena c&va.

The coronar>' veins were all of normal size ami (hstrihution, antl their oix>nings into the coronaiy sinus were nonnal. But, the coronary sinus ended blindly about o nun. short of the jwint at which it should have opened into the right atrium. At this end of the sinus it received the vena cordis media. No rr>mnant of a connection could ho found upon the l)lind sinus end or the riglit atrial wall. Upon tracing the anomalous vena cava upward it was found to ojien into the left innominate vein at the point where this vein joined with the right to form the nonnal superior vena cava (fig. 1). In the upiM>r third of its extent the left superior vena cava received the sujx^rior intercostal vein, which was coTuposed of two intercostal branches.

Ui)on examining the ulterior of the right atrium, no trace of an o]x>ning of the coronarj' sinus was seen. However, the wall


S. n. (iRANT

hrn* was thin. No Tlu'lK'sinii valvo was prrsent. The valve of the inferior vena cava, howcMT, was well (level<»|)e<l. A large ilistinct foramen ovale was present.

I have found two references besides those p^iven by McCotter CIO)— l>eCat, Beyerlein and Hutton — of anomalies similar to the one under consiiieration. According to Marshall ('50),

W. ^fien. D,

y'^non S

Fig. 1 The heart and vessels viewed from the left and dorsaliy. It shows the course of the coronary sinus and left superior vena cava. Enlarged one-half.

I^eCat ol)st>rved (17."^S) a coronary vein which emptied into the left subclavian vein in an eight days old child.

Hutton ('15) found a case in which the coronarj' sinus ended in a sunken pitted area against the wall of the right atrium. Within th(> riglit atrium there was a shallow circular deiiression c(jrrespt)nding to the coronary oix-ning. There was no Thebesian valve, but a considerable remnant of the left venous valve. Hutton suggested two explanations of the closure of the coronary' sinus. The partition between the sinus and atrium was either; first, a composite structure, the result of fusion between


tli<' coronury scfniK'nts of the ri^lit hikI left vonous valves; or, socomi, tho rosiilt of an unusually voluminous 'riwlx'sian valvp, which ha<l cvcnlually fusod with tho marRins of the ostium of the sinus.

l^oyorlein ('14) described a case in a fift<'<'n months old child in which the coronary sinus extended to the wall of the ri^ht atrium. I'pon ohsorvinp; the interior of the right atrium, all tlH> openings apix-arod normal, but, upon probing the crironary sinus it was found to b(^ closeii. The Theijcsian valve. nf)t iK'ing mentioned, was evidently normal. Beyerlein offered two possible ('xplanatif)ns of his case: viz.: inflammation of the endocardium and atresia, although he siiw no evidence of; or, a mechanical influence due to suction after the formation of the left innominate, plus pressure in the atrium from the right superior and inferior venae cavae.

(iruber CSo), descrilx'd, in a fifty years old man, a coronary sinus without any atrial opening. In the right atrium in the plac(> of the usual coronary afXTture, there was a small groove leading into a lilind i)ouch (i nun. deep, (iruber tentatively called the free membranous margin of the groove the Thebesian valve, .\nton Siding ('9lj) also observed a case in an adult male in which the coronary sinus ended blintlly lo mm. short of the site of its o])ening into the right atrium. Within the right atrium there was a narrow aperture, guardetl b^- a feebly developed Thebesian valve, opening into a blind sac 10 mm. long In all of these cases the coronary simis drained through a {M>rsisting left superior vena cava.

The fact that the left .superior vena cava, of the present case, opened into the left innominate so close to the right superior vena cava, instead of some distance from it, is oiusily explained when one reflects that in the cat, with its narrow thorax, the transverse branch coimecting the two superior venae cavae becomes shorter and larger as the embryo in size until the left supi-rior v<>na cava may ai)]H>ar to ojx^n into the right at an acute angle.

The closure of the normal ostium of the coronary sinus probably took place after the transverse branch betwe<Mi the two

48 S. B. CiHANT

siijH'rior vonao cavac had Ikmmi foniuHl. As Ihitton Clo) su^H«'sts. (his closun' may liavi- lu't-ii duo to an visually lar{;(' Theliosiau valve, which may have closed it too effectively when j)rossun> was exerted upon it from the inside hy the flow of blood into the atrium. The left superior vena cava being still open, the coronary blood would tind an easy path to the right atrium through it, the left iimominate, and the right superior vena cava. If this suiiposedly large 'I'hebesian valve fused with the wall and pennanently closed the normal opening of the coronary sinus, the portion of the latter, or left duct of Cuvder, between the wall of the right atrium and the proximal end of the coronary sinus may have ilegenerated. It has been suggested by Beyerlein ('14), that the suction in the left .superior v(>na cava, caused by the of blood in the left innominate vein would aid in closing the normal sinus oix^ning. That the Thebesian valve, which is foniied froni the caudal portion of the right valve of the sinus venosus, may have been abnormal and possibly concerned with the closure of the coronary opening, is indicated by its absence and also by the presence of a patent foramen ovale. If th«' left venous valve were abnonnally small, it may have resulted in the failure of the foramen ovale to close, and in such a case, the right venous valve would be likely to be unusually large, anil, consequently, the Thebesian valve as well.

Gruber's and Siding's cases are similar to the present, in that the coronary sinus ended before reaching the wall of the right atrium; but differ in that in my case there is no cul-de-sac or ojHMiing in the atrium at the site of the normal ostium of the sinus. Ilutton's and Beyerlein's cases differ from the writer's and the other two, in that the coronary sinus in these cases extends to the wall of the right atrium, where it is closed.

Hutton's second suggestion as to causes of the anomaly seems to suit the case here described, but the fusion of tlu^ Thebesian valve to the atrial wall in my case must have taken place so early in development that no traces of it are left. Coupled with this may have been the mechanical causes suggested by Beyerlein.



Hkykrlkin, K. I'.il I Die iM-rsistirreiule Venn oiiv.i mipcriorc Bin iatrn nia AbfltiHsrolir fiir(ln.sCoronarvcncnl)liit. Fninkfurtfr Zeitxeh. f. Path., Hd. 15.

CIrvukk, \V. IS85 Anittoiiiischc Notizcn. An-li. f. path. Aniil., Hd. 19.

HuTTO.N, \V. K. Ull.'i .\ii iin<>ii>!ilo\i8 roroimry sinus. Jour. .Vniit. I'liyn., vol. 4'J.

Keibel and M.\ll 1912 Muniiid of Humiin KnihrynloKv. I.ippincott. Philadelphia.

LeCat, 17.'{8 Hi.stoirc dc raeudemic dcs Scicndos. Pari.s, IS40. (Quoted from Marshall.)

Makshall, J. 1S50 On the development of the Rrcal anterior veins, etc. Phil. Trans. Royal Soc, London, vol. HI.

McCoTTER, U. K. 1916 Three of the persistence of the left superior vena cava. Anat. Kec, vol. 10.

Siding, A. 1890 Ueher den .Vhfluss dcs Sinus coronarius cordis gegen den rcehten Vorhof. .\nat. .\n7,.. Bd. 12.


ROLLO E. M COTTER Department of Anatomy, University of Michigan, Ann Arbor


In a previous communication the writer ('12) described the vomoro-nasal apparatus in the opossum and other mammals. It was sliown that the Noniero-nasal organ, the voiner()-nasal nerves and the accessory olfactory bulb are parts of a special olfactory jiicchanisin the sjx'cific function of which still nMuains doubtful. It is with the ide^i that a careful comparative study of this apparatus in the different animals may lead to a more definite understanding of its function that this study was undertaken.

The obser\'ations about to be reported were based in part on serial sections of the heads of turtles and frogs and in part on ilissections of jirepared s|x>cimens of the .siime six>cies. Wax ]ilate reconstructions of the olfactory apparatus were made to show, so far as possible, the form and comjiarative size of its compojient parts. Figures 7, S. 1.") and Ki represent drawings of these reconstructions.


Two views have been advanced as to the structure that should be designated the vomero-nasal organ in these forms. One giouji of observers believe that the vomero-na.sal organ exists iji a very simple cojulition, and that in some sixM-ies it forms a shallow fossa covered by neuro-epithelium situated on the metlial wall of the nasal cavity, while in others the neun>-epithelial

52 . Ill >I. 1,1 1 K. M( CO'ITKIt

an>a Ims <>xl<'iul('<l on to tlic anterior and latofal walls of the na>;il fossji. Another kiou]) of workers cluiin that the voineronnsnl organ is a rudimentary structure and consists of a small duct that extends from tlie surface of the septal mucosa caudalwanl in the sMhnnicosji and ends hlimlly.

AccordijiK to Seyilel's {'{Hi) conununication the nasal cavity m Chclonia may be suhdi\ idotl into a cranially situated pars olfactoria and a mow caudally situated pars respiratoria. The vomeronasal *)rgiui belongs to the last named subdivision. In testudo graeca one can obser\e the vomero-nasal organ occupying a sliallow fossa on the medial wall. The separation of its epithelium from the neuro-ei)it helium of the olfactory region is completed through a narrow intervening zojie of indifferent epitluv lium. The \'entral and lateral portions of the pars respiratoria exhibits jio iunu-o-e]iit helium. II<> states that in emys europaea the pars respiratoria is more comjilicated than iji te.studo. Here the neuro-ejiitheUiun that comprises the vomero-nasal organ is foimd in four fo.s.sae which occupy the medial, the two sitle walls ajid \\\v floor of the ]>ars respiratoria. It is separated from the neuro-i'pit helium of the pars olfactoria by a low ridge that is covere<l by inditTerent epithelium. iSeydel accepts the view that the neuro-(>pit helium iji emys has extended from the medial walls onto tlie floor and side walls of the pars respiratoria. For an explanation of this ^'ie\v he refers to the course of the ner\'e fibers of the pars respiratoria. The olfactory fibers extend from the medial wall downward ajid cm-xe lateralward IxMieath the floor and upward on the lati^ral wall where they subdivide mto branches. These relations have resulted from a condition where the neuro-epithelium oc(U|)ied a small an^a on the medial wall juul has extended to the anterior and lateral walls.

In one embryo of C'hrysemys punctata Seydel observed that the neuro-epithelium of the vomero-nasal organ occupied a small area on the medial wall while on the floor indifTcrent epithelium was foiuid. In another the neuro-epithelium had ext<'nded from the medial wall onto the floor of the pars respirat oria.


Miluilkovics researches on emys europaea has lead hini to differ us to the position, fonn and structure of the vomero-nasal ornan. He lieliexcs that the \()nier(>-n;is!il ornan is a rudiinentury structure in tliese forms. He descrihcs this organ a« a small, blind, tubular structure extending from the surface of the septal mucosa caudahvard in tiie submucosa. It re<'eivcs at it«  distal extromity the ducts of the medial nasal glands.

Zuckcrkaiidl (10) observed a sjx cimcn of em\'s europaea and corroborates Seydel's important anatomical observations. In regard to the structure that Mihalkovics has dcsigruited the vomero-nasul organ lie is of the opijiion that it is the duct of the medial nasal glands. Zuckerkandl further observed that the olfactory nerv'es arise from two areas of nasal nmcosa. A dorsal branch aris?s from the mucosa of the j)ars olfactoria and a ventral branch receives tilamcjits from the vomero-nasal area. The dorsal and ventral branches unite to form a common olfactory nerve that jmsses through a large opening in the cranium togetlier with the nerve of the opposite side. .As these nerves ajjproach the olfactory bulb in their course caudahvard the nerve bundles become separated, the dorsal branch forms a large lateral bundle that distributes filaments to the ai)ex and tlie ventral surface of the olfactory bulb and extends tlorsiilwards over the medial and lateral surfaces. The \entral branch becomes the more slender medial ramus that sends filaments to the upper half of the inedial surface and to the dorsal surface of the olfactory bulb.

cnUYSKMVS pinc:tata

The naj^al fossa in chrysemys punctata consists of a principal nasjil chamber that conununicates anteriorly with a circular naris by means of a small cylindrical nasal pas.sjige and posteriorly with the choana through a larger posterior na-^il canal.

Hy referring to figure '■] it will be seen that the princijuil nasal chamber is oval iji transvers<> .section with the greatest diameter in the ]M'rpt>ndicular direction and its .shortest diameter in a horizontal plane. The anterior i'a.«al pas.sjige conununi TUK ANATOMICAI. HKCORD. VUL 13, No. 1


call's \vitl\ it oil the iintrrior wall al)<)Ut half tlic distance Ix'twocii the roof and tlie floor. 'I'hc posterior nasal eanal extends horizontally caudalward on a level with the floor. The otherwise smooth interior is interrui)ted hy many low ridges which course generally in an antero-posterior direction and suljdivide the cavity ijito numerous fossjie of varying sizes and depths.

H«'giiuiing just above the communication between the prineijMil i\iv.sjil chamber and the anterior nasjil jiassage are two ridges, one of which (>xtends caudalward over the medial wall, the other in the s;inie direction over the lateral wall ajid become less proi\otmced as they ajiproach the i)osterior wall. ridges which are covered l)y respiratory epithelium separate completely a large fossa in the roof of the principal nasal chamber. This fossa is covered by olfactory neino-epithelium and gives origin to the olfactory nerves, lielow the two ridges mentioned aljove is an extensive fossa that occupies the lower half of the anterior wall, the anterior jiortioji of the floor and adjacent portioii.s of the medial and lateral walls. This area which has a very irregular outlijie is covered by the ^omcro-nasal neuroepithelium and gives origin to the vomero-nasal nerves. By ref<^rring to figure 2, which is a transverse section through the anterior jiortion of the jirincipal nasal chamber, it will be seen that the vomero-nasal oigan occupies a suigle and extensive fossa situated on the lateral, medial and anterior walls of the nasjil chamber. By following this fossa caudalward in serial section it will be found that it becomes subdivided into four areas (fig. 3) by the apparent invasion from the cautlal direction of three low ridges capped by respiratory epithelium and thereby giving to the vomero-nasal organ the apix^arauce of occujiying four s<^j)arate foss;ie.

The olfactory neuro-epithelium (figs. 7 and 8) occupies the roof and adjacent j)ortions of the meilial, anterior and lateral walls of the nasal cliamber. It extends lowest on the anterior wall where it covers one-third the ilistance from roof to floor. From this jioijit the border grailually recedes donsjillj- until the caudal wall is reached. The vomero-nasal neuro-ei)ithelium


occupirs the lowor p<irtioji of tlio antrrior, modial and Iat4>ral walls ami (loor of tlic nasal cliainlKT. It ocfui)i(>s a siiigU- irn-RUiar fossa and is separated from tlic olfactory neuro-«pitholium by low rid)j;f's covorod 1)>' respiratory epitlicliuni. The reinairiinj? portion of the wall of th<^ nasal chaiuijer is covered tjy respiratory epithelium.

I'ii;. 1 A Iransvfisi' si-ctioii of tlio hoad of a turtle at aliout the middle of the anterior nasal canal. It shows the form of this portion of tiie nasal fos.ia. X 10.

Fig. 2 .V transverse section of the head of a turtle pa.ssiiig througli the anterior portion of the principal nasal chamber to show the form of the na.sal fossa and the [losilion and distribution of the voinero-na.sal mucosa. X 10.

The anterior na.sal pa.ssage extends nearly horizontally eaiidalwaril from the naris ;md communicates with the princijial nasal chamher about midway betwivn the roof and tloor. It is nearly cylindrical in outline and presents a low ridge that courses oblicjuely in a caudo-lateral direction from a metlit>cephalic ori{!;iir.



The i><)storior nasjil cjuuil ('Xt«')i(ls nearly horizdntnlly caudalward from the ])riiu'i|)al luisal (•lian\l)C'r. It is seinicircular in cross section. Attachcil to the anterior half of the lateral wall and to the cephalic portion of the roof is a crescentic valve-like fold that separates a dorsjvlly pla('<'d blind pouch that opens caiidahvard.

Fir. .1 A transvonto sprtion of the head of a turtle passing through the middle iif the prinripal tmsal chainlirr showing its form and size, the distriliution of the vomero-nasal and olfactory ncuro-fpit helium and peripheral course of the vomero-nasal ner\'e8. X 10.

The nerse fibers from that portion of the ncuro-epit helium of the vomero-na.sal orsan situated on the lateral wall collect into two limbs :ui anteri()r and a posterior. The former is the smaller and courses downward in the lateral wall of the nasal fossa. The latter is a broad flat band that passes medially lieneath the floor and is joined by the anterior limb.



Thoso c<)inl)iiicd Hlamunts form u broad shwt of imrvo fil>ers coursing mediallj' Ijeneath the floor to the iiiediul wall aiid at the same time receiving iidditioiuil fibers from the neuro-epitheiiun). In the medial wall the broad flat band of nerve filx^rs b«'come» sonuiwhat narrowetl and thickened and courses dorsally and somewhat caudally to the roof of the nasal fossa. Here it is joined by the bundle of olfactory nerves of the same side and the

Kij!. 1 A transverse seotion of the head of a turtle passing through the middle of the posterior nasal oanal. It shows the form of this passage and the relation of vomcro-nasal and olf.ntDrv iii-rves in tlic^ir cuiirsi' tlinmuh tin- ininiiim. X 10.

combijied filaments of vomcro-nasal and olfactory fibers of the opposite side. The combiiie<l filaments form a large round nerve bundle that courses caudahvard through a large o|x>ning in the cranium to the cranial cavity, .\lthough the right ami the left vomero-nasal and olfactory nerves course tlirougli the cranium together the nerve fil)ers of the lUtTerent bundles do not intcrmijigl(\ They lie contiguous to one another and may lx» separated from each other with little effort ivs will be s<H'n by ix'fcrriiig to figure 4.


As th»'y «!nter tho cniiiiHl rnvity tho vnmoro-nasal uml olfat'tory nones soimrato: the fornuT jmss dorsiil to tlic ulfactory nerves and to the ilorsnl surfuco of the olfactory l)ull) where the filaments spread out o\'er tlie \'oiner<>-iuisaI area.

Thi> ner\e libers of tlie olfactory neuro-epithelial area collect into numerous filaments that course tlorsahvard and converge to form a large oval bundle above the roof of the nasal fossa. The bujulle of olfactory nerves lie lateral to the vomero-nasal fibers in its course through the large opejiLng in tlie crauiuni.

P'ig. 5. A transverse sertion through tho anterior portion of the olfactory bull>s of the turtle to show the relation of olfactory and vomero-nasal nerves. X 20.

I poll reiidaiig the cranial cavity olfactory^ fibers separate from the vomero-nasal fibers and pass ventralward to be distributed to the olfactory area on the apex and ventral surface of the olfactory bulb.

The olfactory l)ulii is an ovoid mass extending horizontally forward from the forebrain. It is separated from the latter by a well definetl f)bli(iu<> groo\e that defines a verj* short olfactory peduncle. By referruig to figures .">, (>, 7, and 8 it will be seen that the surface of the bulb is sub-divided into two definite and separate arenas. An oval area occujiyijig the entire dorsal surface and upper half of the medial surface of the bulb to which the vomero-na.sal filaments are distributed is the vom«*ro-nasal



ana of the olfactory 1)1111). This arra is homologous to the ufcessory olfactory hull) of inainnials. Thcr olfactory an-u <)f the bulb is somcwluit inoro extensive. It occupies the apex, aii<l ventral surface ami extends 8onie distance upwaril on the



Set molnuiM*

Ki^. (■) .\ trnii.sversc section fhrough the niiddle portion of the olfactory bulb to shiiw the relation nf the vnnioro-nasnl ami olfactory areas. X 20.

lateral and medial surface. The filaments of the olfactory nerves are dislrit)Uted to this area. \n extension forward of the fore hrahx cortex separates the vomero-nasal and olfactory areas of the olfactorv bulb.

Viii- ~ -V nieilial view of a wax plate reconstruction of the olfactory apparatus of the turtle to show the origin cour.xc and distribution of the vomero-nasal and olfactory nerves. X 3.

Fin. .S A lateral view of a wax plate reconstruction of the olfactory appuratu."! of the turtle to show the oriisin, course and distribution of the voniero-n;is;i! and olfactory nerves. X 3.


111 transverse section (fig. (>) the olfactory bulb oxhil>its a large ventricle oval in form. One can (listinpuish tlie (liff«'rent concentric lay«>i-s that have Ixm-ii ilescrilxHl for tiiis portion of the brain in other forms. The ntMve fiber layer is incomplete and presents a dorsjvl .•ieRnient of vomero-nusal filxTs and a ventral st^gnient of fila olfactoria. The latter complet(^ly encircles the olfactory bulb in the lower mammals. The glomerular layer presents a dorsal and a \entral segment. The former is much thicker than the latter. Then follow the molecular, nerve cell, tile granular anil <']M'n(lyiiial layers in order.


The excellent ilescription of the nasal fossa in Rana by Kcker and ( laupp has been frequently consulted and, as far as possible, the same terminology has been used in this conununication. While he recognized tho origin of the olfactory nerves from the olfactory mucosa by two l)ranch<'s -a large dorsal and a small ventral —and that the vomero-na.sal nerves join the dorsal ramus, he faileil to determine the further course and termination of the vomero-na.sal nerves as a separate bundle. He states that the olfactory nerves on enterijig the cranium separate into two roots, an anterior distributed to the antero-ventral surface of the olfactory bulb, and a posterior root that is distribut(>d to the acces.><ory olfactory bulb.

Zuckerkandl ('10) was the first to recognize a separate vomeronasjil ajjparatus for the amjihibians. He describes the formation of the common olfactory nerves as of fibers formed by the union of the olfactory and vomero-nasal nerves but fails to state the relations of vomero-nasal and olfactory nerves in their pas.sage from the nasal cavity to the brain.


.Although the na.sal cavity of the frog has been very carefully descrilied by Kcker and (;au!)p and in text books on comparative anatomy it apjx-ars to me advisable, owing to the complexity of the arrangement of its subilivisions, to summarize briefly its more important features.



The nasul cavity of tho frog is situated in tin; ant<Tior part of the crauiuiii. It is very much flattmed dorso- vent rally and expands anteriorly anil laterally so that m a dorsal view it presents a semicircular outline. The anteriolateral curved

Fig. A triinsver8e section throURli tho head of the frog in the region of the external nasal opening. It shows the furni of the principal and middle nasal chambers and the distribution of th(j olfactory mucosa. X 2.5.

Fig 10 A transverse section through the head of a frog posterior to nuris. It shows the relation between the superior, middle and inferior nasal cavities and the distribution of the olfactory mucosa. X 2..5.

Fig. 11 .\ transverse section through the vomero-nasal organ of the frog showing the relation of the superior, middle and inferior na.sal cavities and the distribution of the vomero-nasul and olfactory mucosa. X '2.'i.

Fig. 12 .\ tnmsverse section tlirough the middle of the principal nasul cavity of the frog showing distribution of olfactory mucosa and thejrelation of vomeronasal and olfnctorv nerves. X 2. .5.

margins follow closely the curvature of the maxillae. It consists of two parts, the ri^iiht and the left nasal fossae. l'>ach fossa comnumicates with the exterior by an external luksal aiH>rtu«\ the naris, and with the oral cavity by an oval internal na.sal ajierture the choana.

(•.2 K. MiCOlTKK

IIk' nnsjil fossa is subdivided into several irn'Kiilar foninuinicatiiip cliainlicrs whicli lia\e lu'come separated to a greater or less extent dvirinn the pn)eess of development, by tlie ingrowth of septa anil ridges. It, therefore', presents for description a

Fi(j. 13 A transverso soction tlir«uf;li llic clioana of thr hoad of a froj;. It shnnx the ilistrilmtion of tlio olfactory murosa anil the relation of the vomeronasal and olfactory nerv-es. X -.5.

suix'rior or principal, a middle and an mferior nasal chamber, a lateral recess, the \omero-nasul organ, and two well defined narrow connecting channels, the infunilibulum and the isthmus.

The principal na.sal cluimber occupies a dorso-medial jiosition and comi)rises al)out time fourths of the fossa (figs. 15 and 10). By referring to figure it, it will be seen that a transverse section

FiK. 14 \ transverse section of the head of a frog passing through the posterior part of the !<up<'rior nasal cavity. It shows the distribution of the olfactory mucosa and the relation of the vomero-nasal and olfactory nerves. X 2..5.

of the cophalic portion of the superior cliambor has a circular outline and coninumicates with the exterior by means of the naris. .\f the lower part of the lateral wall can be seen the



plifu (crniijuilis whicli iiuirks off a dcn^p Kroo\(> which is the ix'Hijiniiig of th«' iiifuiidihuUini. In this rcj^ion the principal nasal cavity is covered for abtmt tliree-fourths of its circumference l>y olfactory neuro-ei)itheliuni.


pa*. IB


Fig. 15 A modial view of a wax plate reconstruction of the nasal fossa of the frog. It shows the origin and peripheral course of the vomero-nasal and olfactory nerves. .\l)Oiit three limes natural size.

V'tR. Ill .V dorsal view of a wa.\ plate reconstruction of the nasal fossa of a frog showing the origin and peri|)lieral course of ihc vomero-nasal and olfactory nerves. About three times natural sine.

Fig. 17 Kepresentsa dissection of theolfaclory apparatus of the frog showing the origin course and distribution of the vomero-nasal and olfactory nerves. Two and one-half times natural size.

Figure 10 r(>pix'sents a cross-section of the principal luisal chaiiilHTs ixisterior to the naris. It pre.>ients the liepinning of the ijifunihbuluni separated from the principal luusjd chamlMT by


the broadenctl plica tennlrmlis. It may be seen tliat the olffartory iicunM'pithi'liiiin covers tlie floor, the medial wall ami tlu' unat^T portion of the roof. The remaining one-thinl of the cireumfenMice is clothed hy respiratory The principal na.sjil chamber in liKure II present.s a n«'arly circular outline and communicates at its ventro-latcral margin with the middle nasal chamber by means of the infundibulum. The olfactory neurorpit helium covers about three-fourths of its circumference. The lateral wall is covered by simjile mucosa. In figure 12 the principal I'.a.-^jd chamlxr presents the form of an inverted shoe. The lateral limb of which comnumicates through the isthmas with th«> lateral recess. The olfactory eminence which extends dorsjilly from the floor aids materially iji givijig the peculiar form to this jwrtion of the cavity. It will be seen tliat the olfactor>- neuro-epithelium has become separated into two areas by a narrow intervening zone of ijidifferent mucosa. One area covering the olfactory eminence and a more extensive layer lining nearly all of the medial and lateral walls and the roof.

Figure \'A repn^s(>nts a section passing tlu-ough the choana. The principal nasitl chamber has an outline similar to that of figure 12. It conununicates directly with the choana. The olfactory eminence has increased ui height and breadth. The olfactory neunM'pit helium occupies two areas on the circumference of the chami)er. One caps the olfactory eminence, the other covers the upjjer part of the medial wall and about the medial two-thirds of the roof. The former is more extensive and the latter less extensive, than showji in figure 12. These two areas are separated by a broad intervening zone of respiratory epithelium.

Figure 14 re])r( sents a section posterior to the choana. The princijjal na.>*al chamber sIkjws marked reduction ui size and has a semihuuir outline. The olfact<jry eminence is very much flattened. The olfactory neuro-epithelium caps the olfactory (•minence and covers a |)ortion of tlu> medial wall and roof of the na.sal chamber; the n-maining circumference in this region is covered by respiratory epithelium.


The small iniddU' nasal rhambor is sitnatfii vfntro-lateral to the cephalic cxlrciiiity of the princiiKil chainhcr. It is much flatt<»ne(l ilorso-ven trail y and broad in a trans^'erse direction (figs. 15 and Ki). The naso-iachrymal duct communicates with its postero-lateral angle (fig. 10). Posteriorly it conununicatcs with the principal nasal chamber through the anterior part of the infundibulum (fig. 11), and more caudally with the inferior nas;il chainl)cr. In fact, the middle chamber appears to be merely an anterior sacculated expansion of the tear duct. It is lined by simple mucosa.

The inferior nasal chamber is an elongated, tran.sversel\' placed cavity lying ventral to the cephalic extremity of the j)rincipal nasid chamber. It is directly contijmous laterally with the lateral recess and medially with the vomero-nasal organ. The inferior nasal chamber as shown in figure 10 is an obliquely placod cavity, oval iji outline, and clothed by ordinary mucosii. In figure 1 1 this cavity is shown in direct conmiunication laterally with the lateral recess and metlially with the medial recess, the vomero-nasal organ. It is everywhere covered by simple mucosa.

The lateral recess is the direct latero-caudal extension and expansion of the inferior nasal chamber. It is oval in outline and follows the curvature of the maxilla. It communicates anteriorly with the inferior nasal chamber (fig. 11). medially with the iirincipal chamber through the i-sthmus (fig. 12), and more caudally with choana and oral cavity (figs. 13 and 14). It is lined by simple mucosa.

The vomero-nasal organ is a cup shaped structure that lies at the medial extremity of the inferior nasal chamber and communicates directly with it laterally. It is clothed by neuroepithelium (fig. 11). The infundibulum is a broad flattened chaimel which permits communication between the anterior portion of the principal nasal chamber and the middle and the inferior nasal chambers.

The isthmus is an obiitiuely placed .slit (fig. 12), broad in a sagittal plane that serves as a meiuis of communication between the i)rincipal na.sal chamber anil the lateral recess.


It will 1)0 soiMi from the fon-noijiR Jiiul by rcforriiiK to fipiires 9 to l(i iiiclusivo timt the i»<Miro-(>i)it helium is found in two sopanito and distinct ronions of thr nasal fossa.

Ihv olfactory niucosii consists of a very oxtonsivc and irregular area oi\ the wall of the principal nasal chamber. From an extensive area covering tiie anterior wall and tlie adjacent portions of the floor, medial wall and roof as shown in figures 9, 10, and 11, the olfactory mucosa extends caudahvard in two strips or zones separated by intervening zones of respiratory epithelium, figures 12, 13. and 14. The vejitral, caudal prolojigation cov<>rs the olfactory eminence and gives origin to the small ventral branch of the olfactory ner\'es. The dorsal, caudal prolongation covers a variable jxirtion of the medial juid lat«'ral walls and the roof of tlie principal nasal chamber. This portion of the olfactory mucosa together with the extension forward on to the anterior wall gives rise to the large dorsal brancli of the olfactory nerves.

The vomero-nasal nuicosa lijies the wall of the cui>shaped vomero-nasal organ and gives rise to the vomero-nasal nerves. .Ml the remaining portion of the nasal cavity is lined by respiratory ej)ithelium.

It will be seen from the foregouig description that the vomerona.sal apjmratus in the turtle eriuals in size and importance that of the ordinary olfactory mechanism. That th*^ olfactory bulb exhibits an olfactory and a Aoniero-nasal area which share about equally in its formation. In the frog, however, the vomeronasal ap|)aratus apparently performs a secondary roll in olfaction. It is very small comjwired to th(> olfactory mechanism. The accessor>' olfactory bull) is situated on the lateral surface of the hemispiiere caudal to the olfactory bulb and is only about one twenty-fifth the size of the olfactory Indb.

Hy referrijig to figures IJ, 13, and 14 it will be seen that the olfactory ner\'es collect into two sejiarate groups. Filaments collecting into nerve bundles on the dorso-medial wall of the suiM-rior chamber fonn the large ilor.sal ranuis of olfactorj' nerves and filaments collecting ijito ner\e bundles from the mucosa of the olfactory eminence form the small ventral branch


of tho olfactory iutvcs. Tlwsc Ijiiiiiclics coursf cuiul.ilwanl and join to form a single bundle at the caudal cxtroinity of the principal nasal chamber. From this point it courses to tho olfactory bulb where it becomes distributed over the antero- ventral surface.

The vomero-nasal nerves formed by filaments from the dorsal, ventral and medial walls of the vomero-na.sal organ form a single roujuled bundle that courses dorso-caudally in the medial wall of the principal nasal chaml)er. At the caudal extremity of this chamber it joins the bundle of olfactory nerves. Although there is a slight intermingling of the vomero-nasal ajid olfactory bundles the mriority of the fibers of the former can l)e followed in their spiral course caudalward where they wind Ix'neath the olfactory nerves to gaui the lateral side of this bundle and course over the lateral surface^ of the olfactory bulb to reach the acce.s.sory olfactory bull), which lies more caudally (fig. 17).

liti:r.\turk citkd

Ecker-Gaupp. Anatomic des Frosches.

MrCoTTEK, R. E. 1912 The connection of the vomcro-nasul ni-rvcs with the

accessory olfactory hull) in the opossum anil other mammals. .\nat.

Rec, vol. 6, p. 2(t!Ml.S. MlHALKovK's, v. 1S9S Na.senhiihle uiid Jacobsonsche Organ. .\nat. Hefte,

B(i. 2. Aht. 1. s. .-J-IOT. Seydki,, O. I.S9.5 (Ibor die Nasenhohle und das .Jucohaon'sche Orpin der

Amphihien. .\Iorph. .Jahrh., Bd. 2.3, S. 4o.'}-")4.i.

l.S9() Cher die Nu.senhohle und Jacobsonsche Organ dor Land und

Sumpf-schildkroten. Festschrift zum 70 Geburtstage von C. Clegenbaur,

B. 2., 1896. ZucKEKKANDL, E. 1910 Cber die Wechselbeziehung in der .Vusbildung dcs

.lucobsonschen Organs vmd des Riechlappens. .\nat. llefte. Hd. 41,

Abt. 1, S. 3-73.



HOMKR G. FISHER From the Anatomical Laboratory of the Johns l/upkins University


During the past few yearti, physiologists and chemists have been working with the retractor penis muscle of various animals, chiefly of the clog. The muscle has in every case been rcgarcieil by these workers as composed of non-striated fibers. The result* ()l)tained in these investigations have been discordant, (liiTcriii^ in some regards from the customary reactions, both physiological and chemical, of other smooth muscles. At the suggestion of Dr. Charles D. Snyder, this study of the histolog>' of the retractor penis muscle was undertaken, for it was felt that any theories of smooth muscle contraction, arising from a study of this retractor nuiscle, must be based upon an established histology.

The retractor penis nuisde in the dog is a cord-like structure, pale and translucent in its anterior portion but somewhat darker and more fleshy in its posterior fraction. It has it* origin by two separate bundles of fibers, one from each side of the sphincter ani nuiscle. These bundles pass ventrally and at a distance of about 1 cm. from the sphincter join in the median plane and run forward o\('r the ventral surface of the corpus spongiosum to the base of the glans where the fusetl bundles are insert eil into the corjnis spongiosum. In the medium-sized dog of 6 to 10 kilogiams, the nmscle has a length of about 50 nmi. while its diameter is about 3 nun. throughout its whole extent. The nuiscle is .^jurrounded by a dense fibrous sheath, continuous with the fascia covering the sphincter. On both sides of this are the bulbocavernosus muscles; these are attached partly into th(> sheath of the retractor penis muscle and partly into the



jvLT. igi7


fihnnis tissue of a incdian rai)li(> mi the ventral side of that muscle.

In tliis study the retractnr iicnis muscles of fi\'e adult dogs and of one puppy were examined. The nmscles were removed, fi.\pd in Houin's picro-formalin-acetic fluid and embedded in paraffin. The sections were stained in hematoxylin and eosin, iron lumatoxylin and Mallory's connective tissue stain. One m\iscle was cut serially and another was removed with the adjacent part of the bulbocavemosus intact; this block was sectioneil for the jnirpose of determining the r(>lations of the two nmscles. Representati\e sections were examined from different parts of the other retractor muscles.

A study of the sections showed in every instance that the retractor nuiscle is mixed, i.e., composed of both smooth and striated fibers. The anterior three-fifths of the muscle is Composed entirely of smooth fibers (fig. 1) while the posterior twofifths is made uji of fibers of both types (fig. 3). The mimber of striateil fibers in the jjosterior part is variable but their presence is constant. The proportion of striped muscle varied in the specimens studietl from one-third to one-half. The proportion of striated fibers was least in the muscle from the ]iuppy.

.\ characteristic field in the anterior j)art of the nuisde (figs. 1 and 2) shows large spindle-shaped smooth muscle cells varying in none of the essential features fnmi the usual text-book descriptions. The cells for the most part have their long axes parallel to the long axis of the muscle but there is some tendency to interlacing of the fibers and bundles of fibers (figs. 1). With Mallory's connective tis.sue stain the muscle plasma takes the characteristic red color. Sections treated witii this stain show the amount of connective tissue to be abundant. This tissue is partly of the white fibrous variety, but there is distributed throughout a relatively great number of elastic fibers. In certain of the sections (fig. 2) fibers have an undulating, wavy appearance. The nuclei of many of the smooth muscle cells in such a region tend to as,»!ume spiral fonris. Mcdill ('09) has described this phenomenon in smooth muscle from the walls of arteries and has reviewed the literature on the subject


particularly in regard to its rauso and signifirance. The spiral form may \)v duf, as is suggested, to tlie 'active or pa.s.<ive" contraction of the muscle cells, or it may be related to the shortening of the elastic tissue. While the cause of these spiral nuclei cannot be determined from the evidence at hand, it may be noted that in the tissues in which McCiill described the phenomenon, there is also associated with the muscle element a relatively large amount of elastic tissue. In the sections from the retractor muscle studied, the spiral form of the nuclei is most abundant where the elastic tissue is most undulating and presumably in the most shortened state.

The microscopic structure, then, of the anterior three-fifths of this nuiscle is (juitc characteristically that of any smooth nmsde, with, however, a considerable interlacing with yellow elastic tissue. In the posterior portion of the muscle, the histology is (luite different (figs. 1 and 3). Interspersed with the typical Inindlcs of smooth muscle cells are frequent small fasciculi of striated nmscle fibers. These fasciculi are often broken up with the individual fibers diverging somewhat from the original j)lane. In other i)laces, a few single anil isolated fibers are found in the midst of dense smooth muscle fascicuU. In gen(>ral, however, the grouping of the striated fibers is in small fasciculi, alternating with eciually small bundles of the non-striated variety. This rather general arrangement can be made out in a low-power photomicrograph of a characteristic field in the posterior two-fifths of the muscle (fig. 3). In this, it is seen that the general direction of tlie fasciculi is parallel to the long axis of the muscle.

Under higher magnification, the cross-striated fibers are found to possess the typical structure of such voluntary muscles from other parts of the body. The .striation (tig. 4) is very distinct and outspoken, very regular and at right angles to the long axis of the fiber. The alternation of light and dark bands is entirely similar to that of typical cross-striated muscle. The nuclei, oval and with only a small amount of chromatin, are found solely in the peripheral portions of the fibers. Between these fii)ers, a niininnnn amount of white fibrous connective tissue


will) hut few yellow elawtio filirils is foiiml. ^^'hen stained by Mallory's niothod, the cross-striations of these libers are beautifully brouuhl oul, if Kiunery's ('l(i) use of this -itain after Bouin'a fixation be followed. Inder this higher niaRnifieation, also, the smooth musele fibers are foiiiid to he ([uite similar to those of other organs (fig. 4).

In the posterior two-lifths of the retractor muscle, the projjortion of stripetl fibers to the unstriped is about equal, or the unstriped lihers may he somewhat in excess. This ]iroportioTi has been ealculateil from a study of the serial sections through the whole retractor muscle of an adult dog: it represents not an exact detennination of the amount of either tyjjc of muscle but rather a rough judgment of the jiroportion. In many single sections, as illustratetl by hgure 3, the relative amounts of striated and smooth muscle may be easily estimated. Apjiarently, the amoimt of striped nmscle in the one puppy stuilied was somewhat less than in the adult animals.

In the gross, it seemed possible that the striated element of the retractor muscle might be deriveil from direct extensions of the bulbocavemosus fibers which at their insertions attach partly to the sheath of this nmscle in its posterior portion. Serial sections, however, of a block made up of the two muscles removed intact failed to demonstrate any such origin; there was always a definite fibrous sheath interposed between the fibers of the retractor and those of the bulbocavemosus. While no definite continuous prolongation of the striated fibers from the adjacent structures was demonstrated, it is possible that they are derived from the sjiine anlage as the sjjhincter ani nmscle with which the retractor is so closely associated. The exact origin of the retractor jxTiis nmscle is rather indefinite, for the fillers arise gradually out of the sphincter.

The nerve supply to the retractor penis muscle was not (i( inonstratcd. It was suggested that if the nmscle was supplied by the sacral autonomics (as seems most likely), one might possibly he able to (ind nerve ganglia either in the muscle or in its sheath. Systematic search was made in serial sections for nervecells hut with negative resvilts as regards the demonstration of such collections of nerve-cells.


Thus it sof'iiis (juito iifrcssary to roiisidi r this n-tractor penis muscle us heiii(j; mixed, eomposcd of both smootli and striated fil)ers. Until 1915, however, this muscle, so far as it is possible to determine from the literature, was considered as a typical smooth nuiscle. This conception of its histology has led to its use by chemists as a i-elati\ely large mass of smooth nuiscle tissue, and determinations of the creatin, camosine and other" nitrogenous extracti\'cs in it have been made. The finding of a disprop-ortionatc amount of these chemical bodies in this nuiscle difTerentiated it chemically from other smooth muscles. The chemists. liii\\c\-ei-, haxc coiisideiid tliat tl:c findings harmonize with Botazzi's observation that tlie muscle is tlitTerent physiologically from intestinal or other smooth muscle, for it has a shorter latent ])eriod and sometimes presents two types of contraction, viz., the clonic and tonic. But it seems proper to suggest here that the hnding of a somewhat larger amount of creathi and other nitrogenous bodies in the retractor muscle was due to the presence of cross-striated fibers.

Retterer and his co-workers ('00-'1.5), in a series of articles dealing with the structure of muscle, has described the retractor muscle anatomically, both from a gross and microscopical standpoint, and has reviewed the literature on the subject. Chauveau was apparently the fii'st to examine the muscle histologically, and in 1S,")7 ileseribed it in the horse as composed of smooth muscle fibers (fibres musculaires ile la vie organi(iue). Comparative anatomists in subsernient jiublicatioiis either ditl not comment on the microscopic structure of the nuiscle or merely stated that it was involuntary. Retterer Clo) concluded that the nuiscle was striated but attributed the striations to the branching of the elastic tissue fibers ami not to the characteristic cro,<s-striations of voluntary muscle. These branches were described as leaving the longitudinal elastic fibers at right angles and encircling the nuiscle fibers or even passing into their substance, so as to give the appearance of heart muscle. Such ajijiearances have also been observed in liic present study, but the plienomenon of the cross branching ot eia>lic fibers was found to occur only in the smooth portion


iif tlic muscl(>. Il must lie cniphasizcd, liowcxcr, tli;il these lateral liiaiichin(;s of tlic elastic elements (»f tiie muscle are entirely (litTereiit from the striations of the striated fibers in the |>osterior ])art of the muscle where the cross-st nations oi)served are evidently due to the alternation of lipht and dark hands in the mynlihrils, as in typical cross-striated muscle tihers.

l^ota/./i ri.")). who made an extensive study of the retractor muscle from a ])hysioloKi( al sland])oint, divided the muscle into three j)arts an anterior or pn pucial end, a middle ])ortion, and a jiosterior or i)erineal end. He stated that the posterior fraction is darker than the other portions and suj^gested that it resembles striated muscle in its gross api)earance. On direct stinuilation he obtained from some s])ecimens two ty])es of contraction a clonic or twitch contiaction, and a tonic contraction (slow and gradual) while from other preparations he obtained only the tonic contraction. He writes (page 11) that "if the animal is very large the muscle is excessively long and tin II I take only the i)repucial and middle parts. If instead the animal is small, I isolate and take out also the extreme perineal end." Wliile he docs not state with which muscle ])re])arations he obtained the two types of contracti(jn, it is jiossible that thej' were obtained only from those preparations comprising the entire muscle. This would include the portion of the muscle containing the striated fibers which on direct stimulation should give a clonic type of contraction. On the other hand, the smooth muscle fibers might well lie expected to yield only the tonic tyjM' of contraction.

It seems, then, justifiable to conclude that the retractor i)eni.s muscle of the dog is mixed. In the anterior three-fifths the fibers are wholly smooth while in the posterior two-fifths the fibers are both smooth anil cross-striated. The divergent chemical and physiological ol)ser\ations b;i,sed on the assumption that this muscle was wholly smooth, have a possible explanation in the mixture of the two kinds of muscle fibers here described.


BIllMOdl! Al'll'l

BoTTAZZi, I'li.iriMi I'.il.") Kiccrclic sill M. Ucir.'iclor IVnia v hu iiliri prcpurati

Musroluri Lisci. TipoKrufiu dcllu U. Aceudcmia dpi Lineci, v. 2,

p. 43. Hicii.iA. (!. UND Constantino, A. liU'J HeitniRp zur Muski-lchemic. Zcit schrift fiir PhyaioloKisclu- Clipniic. Bd. 81, p. 120. KiN'UEKY, H. M. 1910 Some uses of .Mallory's connective tissue stain. Anat.

Roc, vol. 12, p. 2<.»1. McGiLL, Caiiolink UI09 The structure of smooth muscle in the restinf; and

in the contracted condition. .\m. Jour. .Vnat., vol. 0, p. 493. Ketterkh, K. et Lklikvre, a. 1909 Structure du tissu musculaire lisse.

Comptes Rend. Soc. do Biol., T. 61, p. 244.

191.5 Ivcs fibres musculaires des cordons retractor penis sont des fibre

cellules strics en travers. Comptes Rend. Soc. de Biol., T. 78, p. 136.

ET Neiville. H. 191.1 Des connexion et de la structure des cordons

musculoK'lastiques ou retracteurs du penis. Comptes Rend. Soc. de

Hii.l.. T. 7S, p. (iO.

iM.A'ii': 1


1 HctoudiPil pilot omicrograpli of typical field in anterior two-fifths of the retractor penis muscle of the dog. The smooth muscle fibers arc shown collected in bands and converging somewhat at the anterior termination of the muscle. KnlargcMienl . 240 diameters.

2 Hetouclicd photomicrograph, under higher iiiagnilication. of the rectangular area shown in figure 1. The fasciculi of smooth muscle fillers are shown separate<l by somewhat abundant connective tissue. The miclci of the nonstriated fibers exhibit a spiral form. Eidargement, 750 diameters.





1 V





^^ ^'^1






j Retouched photomieroRraph of a typical field in the posterior two-fifths

of the retractor i)onis nmsde of the dog. The alternation in the section of striated and non-striated fillers is well liroURht out and the e(iual proportion of the two types of fibers is indicated. Enlargement, ISO diameters.

4 Retouched photomicrograph of the squared area in figure 3. The typical cross striations of the cross-striated muscle fibers and the peripheral distribution of their nuclei is well lirought out. On both sides of these striated fibers are dense bundles of smooth muscle. Enlargement, 7.50 diameters.




I >

'1.4 TE J



WAUO NAKAllAKA From the Deparlmenl of Enlomology, Cornell Univerttity


As to the liiological significance of aniitosis, Fleniming's ('91) theory that "it represents either degeneration or an aberration, or perhaps in many cases is tributary to metabolism through the increase of nuclear surface" (Wilson, '00, p. 117) is generally regarded as rcjirespnting the truth. Cells which have divided amitotically and arc active in their metabolic processes may eventually degenerate and perish. This, however, is no direct proof of the first i)art (jf Flcnuning's statements here cited.

My study on the relation of nuclear divisions and metabolic activity in the adipose cells of various insects furnishes good evidence to show that amitosis docs not mean the approach of degeneration, or aix'rration at all. but this kind of nuclear division may be chiefly, if not entirely, to .secure the increase of the nuclear surface to meet the jihysiological necessity which is ilue to active metai)olic interchanges between nucleus and cytoplasm. This is the theory first suggested by Chun more than twentyfive years ago {'^)0), when he studied amitotic nuclear division in a giant cntodermic cell of the radial canal of Siphonophores (Flenuiiing. '91, 'i)'J; Wilson, '00), but was somewhat neglected l)y subse(juent writers. According to this theory, amitosis is primarily concerned with the vegetative function of individual cells, and so amitosis can no longer be regarded as one of the two es.sential methods of cell-mult ijjHcation. It may perhaps be consid(>red in association with such phenomenon as the ramification of the nucleus with its increasing activity, as in the case of silk-gland cells of many insects, ami the division of


S2 WAIU) \.\K.\H.\K.\

the cell-hody following that nf nucleus is a relatively subordinate phononionon.

In this i)ai)er I wish to ])()int out thiit in the case of the adipose cells of insects, amitotic nuclear di\'ision occurs preparatory to as well as sinuiltaneously with a certain metabolic activity of the cell, in which niudei take the r61e of essential importance. Fuller accounts of the relation of nuclear diNision and the metabolism of the cell in question, together with more extensive discussion on the general subject of aniitosis. \\nll be given in my further jjajjer.

The following observation done primarily on lar\ae of Pieris rapae, brings out the general feature of the changes observable in larval adipose cells during their activity. Changes similar to those described below, have been observed by many previous writers and also by myself in the case of various other insects.

In an adipose cell from a larva of the first stage the nucleus is round and shows no sign of division; the cytoplasmic area is small in most cells and contains only a few vacuoles. These vacuoles we interpret as indicating the places occupied l)y fatdroplets. This rather unspecialized condition of the cell changes in the following stage by areal expansion and more vacuolate appearance of the cell-body and by frequent occurrence of peculiarly sha])ed nuclei. Sonic few of these nuclei show nothing but their irregularity in shape, while most of the others apparently represent different stages in the process of amitosis.

I have shown in figures 1 to 7 sketches of nuclei, all representing possible stages of amitosis, and many of which one can find in every section. I have not been able to find any particular way in which these nuclei diAide, except that the division itself is effected by the constriction of the nucleus across its longitudinal axis, thus making the nucleus show a bilobed condition. More rarely, in the case of long, slender nucleus, constriction may take place at more than two different places in the nucleus, and in that case the latter shows multilobed appearance. Nucleoli and chromatin grainiles are apparently evenly distributed throughout the nucleus, and neither of them seems to behave unusually during the process.



In tlic tliird stage larva, one may notice the fact that some of the cells begin to show peculiar spherical granules in the cell-body. As pointed out by Herlese ('99), P6rez ('02, '10), Henneguy ('04), and others, these granules are of albuminous nature, and occur more abundantly in close proximity t(j nucleus, than in the periphery of the cell. They become very abundant at the fourth stage and at the last larval stage almost all of the adipose cells are seen to be hlltd with the granules. The cell

Fins. 1 to 7 Nuclei of ailipo.-ii- (.'ells, representing possible stages of amitosis. X 7.tO. Kigiire G represents predoniiii.'itinK type.

body becomes larger and larger with the advancing stages of the insect; ajiparently correlating \\'ith this, the occurrence of nuclear division is seen more and more freciuently, and finally, in old larvae, we find the condition as shown in figure S to be met with very commonly.

Sunmiarizing the facts, we may say that, in the adijiose cell, the luicleus continues to divide amitotically from early in the .second .stage, and the cell stores up albuminous gramiles in its cell-body, commencing the process late in the thinl stage. This



shows tliat the foll-miclci which have undorgoiip amitotic division, without lOKaril to wlicthcr tho cells rcinaiii nmhimiclcato or not, do not degenerate, and theceils proceed witii their active functional processes. This can he more strongly emphasized because we now see the fact that nuclei themselves take a direct part in the formation of alhumiiious granules, giving most conclusive evidence in support of the theory.

Fig. 8 A multinurleatc adipose cell from iin old liirva, contnining iillitiiiiinous granules in its cell-body. X 250.

It has been supposed bj^ some writers that albuminous granules are derived from the blood. HoUande's ('14) recent work. from the cliemical viewpoint, casts grave doubt on this assumption, and although lie has given no evidence, he has suggested that the granules may be of nuclear origin. It seems quite probable, from a cytological point of view, that such might really be the case, and especially so when we recall the fact that, in the of .silk gland cells of Pieris and Xeuronia at lea-st, the nucleoli, after eliminating their phosphorus, change into albuminous graiuiles and then extrude bodily out of the nucleus (.Nakahara, 17;.



Examining the nuclei of such cells as contain the granules, one can distinguish three different kinds of granules within them. One of the three kinds undouhtedly represents chromatin and another nucleoli, as can he judged from their appearances and staining reactions, characteristic of such elements. The third kind of granule is of nearly the same size as nucleoh but differs from tlio latter in that it shows stronger affinity for certain acid

• ' V \ - ' • • • • V

Figs. 9 to 11 Nuclei of adipose cells, showing the extrusion of acidophile granules. X 7.50.

and weaker for basis stains. This is exactly the sort of reaction shown by the cjioplasmic albuminous granules, and one may here a.'^suinc that these acidophile granules in the nucleus may be extruded into the cell-body, constituting the albuminous granules in question. As e\-idence in support of this theory, I have .shown in figures 9 to 11 some unmistakable cases of extrusion of the acidophile granules into the cell-body through the nuclear mcmbram-. These are not cases rarely obser\able but are those of very frequent occurrence.




Berlese, a. 1S90-00 Obsorvazioni Mil fononiciii ('he avvoDKniio (liiraiilc la ninfosi degli inaetti mctaholici. Utvista di Patologia VcKetalc, T. 8, pp. 1-444.

Flemmino, \V. 1891 I'cbcr TheiluiiK und Kornformen Im Leukocytoii und dorcn Attractionssphiirpn. .\roh. f. inikro.skop. Annt., Bd. 37. pp. 249-298.

1.S92 Zellc. EiitwicklunK und Stand dor Kcnntniaso iilxT Aniitose. Morkol 11. Bonnet's Kruchnisse der Aiiat. u. Entwick., Bd. 2, pp. 3782.

Henneoi'y, L. F. 1904 Lps inscctes. Paris.

HoLL.\NDB, A. Ch. 1914 Formations endoni'iics dcs cristBlloidcs allmniinoides et des urates des cellules adipcuses dcs chenilles dc Vanessa io et Vanessa urticae. Arch. Zool. Exp. Gener., T. 53, pp. 559-578.

>;.\K.Mi.\nA, W. 1917 On the physiology of the nucleoli as seen in the silkgland cells of certain insects. Journ. Morph. (in press).

P£iiEZ, Ch. 1902 Contribution d I'^tude dcs metamorphoses. Bull. Sci. de la France et de la Belgique. T. 37, pp. 195-427. 1910 Recherches histologiques sur la m^^tamorphose des niuscides, Calliphora erythri>cei)hala .Mg. .Vrch. de Zool. Exp. et Gener. T. 4, pp. 1-274.

Wilson, E. B. HKX) The cell in development and inheritance. 2nd ed. New York.

A POLYEMBRYONIC BLASTOCYST IN THE OPOSSUM J. T. pattp:r.son and c. g. hartman

Contribution from the Zoological Laboratory of the University of Texas, So. ISi



WTiile engaged in collecting material to complete the series of stages on the development of the opossum, Didelphys \-irginiana, we have recently come upon a polyembrA'onic blastocyst containing four enibr>'os. The history of the female from which this specimen was taken is as follows: The female (no. 300) came from the wild, and was received at the laboratory on January 20. It was anaesthetized and the left utenis removed at y.2() p.m. that evening. This uterus had six normal blastocysts, each measuring between 10 and 11 mm. in diameter and containing an embrj-o in the early somite stage. The incision was closed and the animal allowed to live vuitil the next day, when she was killed at 12.15 p.m. In addition to the polyembryonic blastocyst, the right uterus was found to contain one dead vesicle al)ove 7 mm. in diameter, and six normal blastocysts, each with an embryo. The average normal blastocyst from the right uterus measured about 14.9 mm. in diameter (fig. 7), while the polyembryonic specimen is not quite as large, measuring only about 12.88 mm. in diameter (fig. 4).

For several reasons this case is of especial interest. In the first place, it represents one of those rare cases in which a polyembryonic blastocyst has been discovered in a multiparous mammal. In the second place, it may have a bearing on the possibility of normal polyembrj'ony occurring in Marsupials, as reported by Bluntschli ('13) for the South American opos.sum, Didelphys marsupiales. Finally, as will be .shown later, the


arrannenicnt of the embryos suggests a close similarity to the paired coiulitioii tlial is normally found in tlie four embryos of the Armadillo, Tatusia novemcinrta.


In the normal blastocyst of this litter the eml)ryo has already sunk beneath the surface of the blastoilerm, and owin^ to the well-developed cervical flexure, the anterior portion of the embryo with its membranes ])rojects into the cavity of the blastocyst (fig. 8). In the surface view the posterior end of the embryo is dimly seen tlu-ough the membranes (fig. 7). The area vasculosa covers almost the entire upper half of the blastocyst. It is limited by a sinus terminalis, which forms a distinct marginal notch at tlie posterior margin where the vessels lead to the embryo.

In the polyembryonic blastocyst there is also a well-formed area vasculosa, with a less distinct notch on the lower side (figs. 4, 5). There is a bajing in of the sinus terminalis on the right side, suggesting a second notch, but no vessels pass from this to the embryos (fig. 6). With the exception of the united posterior enils of the two embryos lying al)ove in the figin-e, the embryos have only begim to sink down into the bla.stoderm (fig. o).

The four embryos are arranged ajiiiroximately in the form of a square, -with each embryo constituting one side (fig. 6). The area included within the square is somewhat higher than the general level of the embryos, giving to this central area the appearance of a vesicle. This is produced l)y the sinking of the embryos into the blastoderm. In sections the central area is seen to be structurally of the same character as the vascular area hnng just outside the square.

The exact nature of the paired arrangement of the embryos can best be descril)ed by reference to the sketch shown in text figure 1. For the sake of convenience in description, the embrj'os are designated by the Roman numerals 1-I\'.

Embryos I and II are united at their posterior ends, and constitute one pair. Embr\'os 1 1 1 and IV are similarly united, and



(•((iistitutc the <illu'r pair. All of ilie (Miibryos have the foreliiiil) hmls well started. The components of each pair are not equal in size, enihryn I being somewhat smaller than its twin, and ('ml)ryo III (Iccidedly smaller than IV.


In order to be able to give a brief description of each inthvidual embryo, it is necessary to have recourse to sections. The embryos were cut into a series of sections (12 n). Tin- plane of



I'itJ. 1 Outline sketch of the four embryos. Note tliat tliey are in two pairs, eniliryos I and II l)einn united at posterior ends, and likewise III and IV. The broken lines represent the planes of the three sections illustrated in figures 1 to 3. The numbers at the ends of these lines correspond to the sections in the series. X S.

section is .ipiiroximately transverse to the long axis of embryo II (fig. 1). There are .^Ot) sections in the .series, three of which (5)(), 242, '.V.W) are drawn in figures 1 to 8. Section 9() jia.sses through heart of embryo II, obliciue through embryo III, and transverse through the posterior enil of embryo 1\'. Section 242 passes through the 12fh pair of somites of embryo II. and the fore-limb buds of embryo 1\'. Section XM cuts the posterior end of embryo II, anil passes obliquely through embryo I.

Embryo I. This embryo is 4 nun. long, and has 13 to 14 pairs of somites. The neural tube is closed in the brain region, but is represented In- a Hat plate posteriorly. The heart is


jjrcscnt Init hinlily :il)ii()iin;il. Tlu' anterior half of the iiotochortl is formed, and the WOlfiian ihicts are well ileveloped.

Embryo II. This einhryo is iiraotically normal, and bilaterally synnnetrical. It is 5 mm. long, and has 20 pairs of somites. The neural folds, anterior to the level of the first jjair of somites, is open, hut a tyi)ical neural tube extenils from this point hack to the 13th jiair of snuiito. Posterior to this a eharacteristic open myelon extends to the end of the embryo. It has a welldevelopeil lu-art. notoelujrd, auditory vesicles, and NN'olttian ducts.

Embryo III. This embryo is 3.25 mm. long and has (i pairs of somites. The embryo is so hiffhly abnormal that it is difficult to make out all the details of structure in obliciue sections. However, the central nervous system is represented by a flat plate throughout the entire length of the embryo, except at extreme anterior end, where the folds are slightly elevateil. The heart is very rudimentary, but the Wolffian ducts are present. No notochord nor auditory vesicles could be distinguished in the sections.

Einhryo I\'. This embryo is 4..") nun. long, and has 15 pairs of poorly developed somites. The neural folds are open at the anterior end, but doseil at the level of the midbrain region. Posterior to this there is a flat neural plate, with a .slight indication of folding in the region of the limb buds (fig. 2, IV). A notochord is present for a short distance in the region of the heart. The WolfTian ilucts and auditory vesicles are both present.


At least three of the embryos are abnormal and would not have formed normal individuals. So far as one can tell from a study of the fourth embryo (II), it is normal, and prol)ably would have completed its develojiment. Regardless of whether or not any of the embryos would have reached full term, the is none the less inten-sting, because it shows that the blastocyst of the opos.sum can j)roduce more than a single embryo. There is no good reason to doubt that the blastocyst of this species may sometimes produce a normal jiolyt nibryonic group of indi


viduals. lint should swell ii fjroup coinc to f\ill Iciin, the iiidividuals could not he distiiinuislii'tl from tliosi' in the litter arising singly from tlic other blastocysts. This fact probably accounts for the extreme rareness of observed cases of polyembryony amonp; multiparous mammals, in contrast to the numerous cases on record for imiparous mammals.

In reference to the bearing of this case on the ])robability of polyembryony occurring in the South American species, we may call attention to the brief report of Bluntschli's Cl'.i) denumstration at the 27th meeting of the Anatoniische Gesellschaft, held at CJreifswald. This investigator demonstrated several embryos and young from the pouch, as well as one opened uterus. It is stated that the latter contained four embryos, each in its own amnion, i)ut all lying within a single chorionic cavity. The condition in this uterus was corroborated by analogous preparations of yovmger stages. Bluntschli con.siders the occurrence of jiolyembryony in Dideljihys marsupiales as probable, not only because of the conditions of the young in the uterus, but also because he had always encountered the young from the pouch of the same se.\. In the light of these facts it may be that the rare occurrence of multiple-embryo formation in D. viginiana has become a permanent phenomenon in the development of D. marsupiales.

As to the manner in which multiple embryos are formed in the blastocyst of the opossum, we cannot, from the study of a single, comjiaratively advanced stage, offer any definite explanation. It is, however, a significant fact that there are four embryos, and that the arrangement of these on the blastoderm suggests a certain similarity to the condition in the armadillo. .\s in the armadillo, the four embryos of the opossum are clearly in two pairs. Furthermore, one of the two embryos in each pair is larger than its twin, exactly as has been descrilied for an early stage of the armadillo (Patterson. '13).

i.rn.uAriKK cithd

Hi.rNTsnii.i. II. liii;< DciiKiiiMtratidii of cmhryoH of DulclplivH iiiursiipinles.

I'roc. of the Aimtomischcn (iosfllaolmft, ?7tli moftinR, (ircif.swiild,

May 10-13. 1913. Sii|>i>1<miiiiiI. Aunt. Anz.. vol. 44. p. •-•00. II.\RTMAN. C. Ci. 101(> Sludirs oil I lie (li'Vi'lopiiu'iil of tlio f)i)ossiiin Didi-lphys

virRiiii.'inii. Jour. Morpli.. vol. '11. p. 1 .s;j. Patterson, J. '1'. liU3 rolyciiitiryoiiic (li'vplopiiu-iit in Tiitu.siii iiuvomcincta.

Jour. Morpli.. vol. 24. ]>. .55i)-(>.S.3.



lto3 Sections %. 242. 331, respectively, of the series. The Roman numerals indicate sections through the four einliryos. X 37.








4 Dorsal view of tlic entire polypmbryonic blastocyst, showinp the four emliryos arranged in the form iif a .s(|\iare. X 2.95.

5 The area vaseiihisa with eniliryo.s. X 4.

6 The same, after staining aii<l cleariiiK. photographed by transmitted light. X 6.5.

7 Dorsal view of a monembryonie blastocyst from the same litter. The back of the embryo, lying in the center of the area va.sciilosa, is dimly seen through the membranes. X 2.95.

8 View of the inside of the same blastocyst, from which the lower wall has been cut away. It clearly shows the embryo projecting into the cavity of the blastocyst. X 2.95.

4, 5. 7 and 8 are each one-half of stereoscopic photographs. For all of these photographs we are indebted to our frii'iid Dr. Chester Ileuser of the Wistar Institute.





From the Anatiimical Laboratories of Vanderbitt University Medical .School



This paper is a consequence of work done under my direction by Katherine Knowlton. a candidate for a Master's decree at the University of Chicago during the year 1912-11)13. She was assigned comparative research on the pigmentation phenomena of feather germs from the Plymouth Rock and Brown Leghorn fowls. In the course of this work some new facts concerning the origin of pigment in feather germs were discovered. A preliminary statement concerning these results has been made (Strong, '15).

I ha\e recently restudied Mrs. Knowlton's material, anil I have prepared the illustrations which accompany this paper, as well as the manuscript.

The development of melanin pigment in feathers has been described in detail in a jM-eceding paper (Strong, '02). Others who have wTitten on the subject are mentioned in that publication. Since then, an important ))aper by Lloyd-Jones ('15) has appeared. This discusses inciilentall}' the origin of melanin l)igment in various varieties of domestic pigeons.

.Vs stated in pre\'ious publications, the feather is purely epidermal in origin. It is ditTerentiated from tissues which make up a cylinder surrouniling a central core of dense and \ery vascular mesenchyma, the dermal pulp (figs. 1 to (i). The meseiichyma is se])arated from the epidermal cylinder by a ba.<ement menii)rane (figs. 1, 2, 5 and li) which is well delined except




at the proximal end of the feather germ. The epidermal tissue includes an external sheath and a layer of stratified s(|uanious ei)itheliuni under the sheath, tiie inner sheath eells, neitlier of which become parts of the feather structure. Between the dermal pulp and the iiuier sheath c(>lls lies a broad zone of cells the intermediate cell region, (Hg. 1 ). This zone isdivided by radial extensions of the basement membrane (figs. 2 and 6) into sections called barb-vane ridges by Jones ('07), except at the proximal end of the feather germ. Each barb-vane ridge consists of so-called cylinder and intermediate cells. The cylinder

Basemeinl mernbrane; Dermal Tnelanopnorea ;'


eriTial melanophorea

/ /' ' iTiferior umDilicxi* / 'IriteTtnediate-cell region

'jinrier »lieatn-ce1l« 

Kijt. 1 Tartly diagrammatir longitudinal section of proximal end of foiitlicr germ. X 30. Melanophoros arc indicated l>y dots. Drawn with aid of camera liicida.

cells make up a layer, one cell deep (figs. 2. 4, and (i) which lies next to the basement membrane and its radial extensions. This layer has the form of a deep and narrow gutter or a letter U in cross section. The base of the U or gutter lies next to the dermal pulp except for the intervening basement membrane, and the other end is next to the inner sheath cells. The space inside the limbs of the U is filled by intermecUate cells (figs. 2, 4, and 6). .\11 of the feather structure except the loose pith in the quill is developed from these intermediate cells. Some of the intermediate cells are differentiated into melanophores which supply melanin granules to the differentiating feather cells.



For details conccriiing these remarkable pigmentation processes, the reader is referred to a previous i)ai)er (Strong, '02). There the histogenesis of the feather also is described in detail except in the case of the shaft and (luill whose development was worked out by Davies ('S*)j.

Sheatli Inner <keath-celU

Dasemenl membrine extension

lnterine(Jiste ceU nuclei

B^ooJ c«pill«rv-

Fi(?. 2 Siniill portion of truiisvi>rse section of feather germ showing an oblique section of a barli-vanc ridge and cylinder cell nuclei of adjacent l)arb vane ridges. X o(X). Back covert of Plymouth Rock fowl. I.«vel of section indicated by the line a in figure 1. This view was selected from a non-pigmented region. Drawn with aid of camera lucida.


The material used consisted of feather germs pulled from their follicles and placed in the fixing fluid inuuediately. The regions of the integument which sujiplied the feather germs were the back, breast, neck and wings. The scajnilars also were used. Two Plymouth-Rock males, two Tlymouth-Rock females, one Brown-Legliorn male, and one Brown-Leghorn female furnished the material. .\ majority of the feather germs came from the females as the males did not molt so extensively tluring the perioils when the material wa.s obtained.


The feather germs were mostly 1 to 2 cm. in length. In most cases they had just begun to expose the barbs at the distal end of the slu'ath. A few feathers in more advanced stages were also used.

(Jilson's fixing fluid was enijiloyed largely, the material remaining in it for three hours. Otherwise the technique was that described by Strong ('02).

The sections were cut in paraffin G to 10 microns thick. About forty of the sections at the proximal end of the feather germs remained regularly in the i)araflin ril)l)oii when 10 microns thick. Succeeding sections were more and more apt to drop out of the ribbon as more and more cornified material was reached.

III. Ki:srLT.s A. Origin of epidermal pigment

1. In intermediate cells. All of the germs studied were of feathers which contain large amounts of melanin jiiginent. Consequently, melanophores crowd the intermediate cell region in the preparations from the two fowls.

No e\'idence was obtained that these melanophores were not of epidermal origin, and I believe that they are specialized intermediate cells as I have stated before (Strong, '02).

Above the region where the melanophores ileveloj), the intermediate cells which enter into the feather structure have their cytoplasm characteristically jiacked with melanin granules. The pigmentation phenomena here are essentially like those described in previous publications.

According to Lloyil-Jones ('15) many ordinary intermediate cells may develop melanin granules autoctlionously. It is of course imjiossible to prove that this is not a frcciuent occurrence and I have indicated its po.ssibility (Strong, p. 170, '02). However, the mere fact that a cell containing melanin iiigment is not sometimes in contact with a chromatophore process in a microscopic preparation, does not prove that this pigment may not have come from a melanophore before the material was fixed



unless it ran ho shown that no inclanophorcs arc nrar onou^'h to have siii)pncil the pij^nicnt.

'Ihc niclanojihoro relationship is obvious at many points in a niicrosco])ic field. The evidence in my judgment points towards the orij^in of the melanin ])i{iment in intermediate cells from epidermal melanophores as the rule. This position is not inconsistent with the ol)servations made by Lloyd-Jones.

I'i)!. ',i riiiitiiinii'niKrapli .slion iiig a portion l4 nun. longi of u lon(;itU(lina| !<i-clion of feather (jiTni from near its proximal end. X I5(). Baekeoverl. I'lymouth Ilork fowl. The rejtion photographed was seleeted l>eeailse it had fewer melanophores than ii.siial. with eonse(iiient better outlines of individ\ial melanophores. Numerous melanophores whose hranehin^; processes connect with difTerentiating feather barhule cells may he seen. The lower end of the view corresponds with the line o in figure 1. and the area extends somewhat above the region indicated in figure I.

Ocular l();olij. I mm. Hausch and I.omli: picture reduced in publication.

2. In cylinder cells. A small luimber of cylinder cells were found ctmtaining clearly dcKnod melanin granules in their cytol)lasm. These cells occurred at the level in the feather germ where the meliinojihorrs were begimiinn to send out processes. .\ few of these cells were at the de())er end of the barb-vane ridges ne.xt to the pulj) except for the intervening basement membrane. Other cylinder cells containing melanin gi-anules were at the sides of the I ail -vai:e ridge, some distance from the

TUi: ASATOUICAL H>l<t>lu>, vol 13 Ni>.


pul|). Till' iiu'laiiiii K'w'i""'^ wen* always found on tlio outer sill*' of the luiflcus. i.e.. tlio side away from the dornial pulp. The Kraiuiles occurred in small tlense groups. It was not possible to determine their origin but we believe that they were formed in situ and did nf)t come from mclanophorcs.

.i". In itinrr slniith nils MclaniM firanuics occurred in some of the inner sheath cells of seven out of sixteen series of feather

SUatK Inner sneiln-celU

riOmenlec) oarDule oelU.;

Intermeaiate-cell nuclei Lpioermal mfclanopnoTes.

DerTTiii puip -fiipC?'

Fi({. 4 I'li<itoiiii<TiiKr:ipli of .small ixirtiini of transverse section of feather Kcrm. Neck covert. Brown IcKhorn fowl. X 'M). Level of line a in figure 1. Kpidermal melannphores may lie seen in their usual abundance, more or leas monopolizing the intermediate cell region. Intermediate cells differentiating into feather l>arl)ules are shown with their cytoplasm packed with melanin granules received from mclanophorcs. .\l the left, histological details have been ndih'il with a pen. a camera lucitla being employed. (>c. 10; obj. 1.9 mm. oil inuner.sion i Hausch A- Lombi. The picture was much reduced in publication. One barb-vane ridgi' appears cotnplctcly, another at the left is almost all in the picture, and a large portion of a third at the right is included.

germ sections, according to Mrs. Knowlton. In all cases the pigmented inner .sheath cells were found within the first millimeter from the inferior uml)ilicus, i.e., at the proximal end of the feather germ, but not in every section of this region. This pigmentation was irregular and sjjarse. The granules were of the same size, form, and color as those of the intermediate and cylinder cells.

The melanin granules were found usually on the outer or peripheral side of the nucleus, as in the case of pigmented cylinder


cells. As there wore no mclaiioplioros near crKHinh t<> have distributed niclaiiiii niaimli's to these cells, we eoiiehjded that they also must have a capacity for forming melanin granules. Such cells never accumulate many Kranules. ami they always retain tlieir normnl character as stratified s(|uaiiHiiis e])ithelium.

B. Origin of dermal pulp pigment

Melaiiopliores were found at tlie i)roximal end ot the dermal pulp near the inferior umbilicus of feather germs from both the Plymouth Rock and Brown Lcfrhorn fowls (figs. 1, 3, and 4). Dermal pulji pigment was not found in the feather germs of over ten species of birds previously studied by me, and 1 know of no other ol)sci\ations of it, Ral)l ('94) found no melanin pigment in the dermal pulp of nestling-down feather germs from the chick.

The dcinial iii(lan()])hores we observed occurred in the mesenchyma not far from the ba.sement membrane, as may be seen in figiUTs 1, ") and (>. Usually a good many were grouped near the shaft-forming epidei-niis. 'I'hat these bodies are inelano])liores is indicated by the following observations. 1) They are similar in shape and size to some ejiidcrmal melanophores. 2) The nucleus is often clearly seen, proving that they are cells, which would not be the case if they were artifacts. 3) bodies are made up of granules of the same size, shape and color as the epidermal melanin granules. 4) The arrangement of the granules is characteristic of a melanophore. The granules surround the nucleus, where they are packed densely. Naturally, some artifacts and foreign l)odies resembling melanophores superficially, have been found in the sections, l)ut they have always Iteen distinguished without serious difficulty. .V ([Uestion naturally arises as to whether these jiulp melanophores may not have been dragged in from the epidermal cylinder in cutting sections, .\gainst such an interpretation we have several arguments. 1) There are too many of the melanojihores to warrant concluding that their positions are accidental. 2) There is no evid(>nce of distortion or tearing of the surrounding tissue such as


U. M. .STK()N(i

would lio oxportod if tlicy had lu'cu moved thoro l)y the miorotouif knifo or l)y any other artificial eause. 3) These inelanophores lit in too naturally in the svuTounding tissue. 4) They are arranged with considerate uniformity around the periphery of the pulj) near the basement membrane. If their position had






Deepest lay^r ofepidermis

Kift. .5 Small portion of Iraiisvorsc section of feather Keriii near il.s proximal pikI. Breast covert of I'lymontli Kock fowl. X 700. Level iiulicated liy line 6 in figure 1. The dee|H'sl portion of the epidermal cylinder next to the hasenient menilirane is shown at the right. The elongated nuclei are separated by radiations which simulate cell boundaries. The basement membrane is poorly defined at this level. .\t the left the mesenchymal pulp is drawn partly diagranunatically. and its character is not indicated. Kour dermal nielanophorcs app<'ar.

been the eonsequenee of eutting, it would be expected that all of them would l)c on one side of the pulp.

We ol)taincd no e\i(leiicc as to what particular cells form these dermal pulp melai:oi)hores. The de\elo])nicntal stages resemble those of the epidermal melanophores, but the pulji melanophores do not become so large as the latter, and they do not form processes so profusely. In fact, the majority of the pulp n.elanophores apparently do not form processes at all. Son.e have small inconspicuous processes, and a few have larg( r i)rocesses. The largest melanophores observed had no processes.



It is charact eristic of tiic inilj) in('laiioj)horcs observed that such processes as occur are directed towards the ejHdermis either radially <ir in the line of a secant. These ])rocesses, as pre\iously slalcil, aic usually short, thick, and closely packed with melanin granules that never .■^eeiii to he distrihiited to other cells. The following talilc gives observations made by Mrs. Knowlton.


Fulp motano phorra with no


With nvlially dircctMl proceimem

With proce— ea

at hcnt ancln

to raditu

15() transverse. . .

00 LonKitudinul

pw cent

65 67

prr ctnt

16 13

per cent

19 20

The melanophores with longer processes lie near the basement membrane, and they are not frequent (figs. 2 and 3). Mrs. Knowlton found them in tliree feather germs Lying so near to the epidermis, it is important to know whether they penetrate the basement membrane or not. We have been unable to find any such penetration, even when the melanophore or its process was against the basement membrane or apparently in its barbvane extensions. These melanophores are located either near the apex of a barb-vane ridge or between two barb-vane ridges as just noted. In the former case the process follows the ba.sement membrane until it reaches a barb-vane extension of the latter, where it runs up between the two adjacent barb-vane ridges apparently splitting the basement membrane partition. In tlie other type (fig. (l). the body of the melanophore lies at the deeper end of the ba.sement memlirane i)artition, and its process runs straight out, i.e., radially, between the adjacent barb-vane ritlges. No such process was found extending more than threefourths the radial extent of the ailjacent ridges.

^^'e are unable to explain the location and behavior of these melanophores. They are so rare that we have considered them more or less abnormal. In the region where they occur, various features indicate activitv in a radial direction Here and more


imiximally, tlu> cells of the syncytial cijitlK*!!!!!!! arc undorKoinR coiisuh'ral)lc chaiiRes in position. Proxinuiliy, pale refraction lines simulating inconiplcto cell boundaries lie radially between the nuclei, which are narrow and arranged radially. In figure 5 such lines have l)een drawn for the deepest nuclei at the right. They have not i)een ol)served to form comi)lete cell boundaries

, Inter rnedi ate ce\n

Inter mediate cell layer \ "Cvlmoei^ceu layer


membrane extension


_ aaemer.t ■membrane

Fig. 6 Sninll portion of transverse section from feather germ of Plymouth Rook fowl. Near level n in (igiirc 1. X 700. Drawn with aid of camera liicida and reduced in pviVilieation. Oc. 10; oI>j. 1.0 mm. oil immersion, (Raiisch and Ix)ml>i. .\ Imwment membrane extension separating two barb-vane ridges is shown with a dermal melanophore process lying in its inner two thirds. The cell body of the melanophore is only partly in this section. Adjacent layer.s of cylinder cells and iiiiiriiHili:iti> cells ,iri- imlic-itnl by their nuclei.

often if ever, and they disappear at the level where the radially directed basement membrane extensions develop. The frccjuent occurrence of melanin pigment only on the outer side of nuclei, mentioned in this paper, is very likely another illustration of radially acting forces. It is conceivable that melanophore proce.«<ses tire carried radially by such forces or by the developing barb-vane extensions of the basement membrane.


In longitudinal srctions some liKht appears to he thrown on the signilicance of the dernial pulp nielanophores in general. In a few sections, I have found an essentially continuous series of these bodies extending to the inferior uinhilicus as seen in figure 1. It is my opinion that these nielanophores are homologous with those of the skin dermis. The dermal pulp being continuous with the skin dermis, such a continuance into the proximal end of the feather germ is not surprising.

C. Observations on the melanin granules

So far as I could determine, all of the granules were rod shaped as is usual in birds. Even when e\ndently fully developed, there appears to be considerable variation in size. The longest are slightly over 1 micron in length. Others may not be more than 0.5 micron long. They are about two to four times^their thickness in length. It is not easy to make exact measurements of such bodies because of their minute size and their arrangement at all possible angles with reference to the plane of the section. In the Plymouth Rock and in some Brown Leghorn feather germs, the granules were of the usual brown, almost black color.


1. ^lelanin pigment granules occur occasionally in the socalled cylinder and inner-sheath cells of feather germs from the common fowl.

2. Further evidence was obtained that the melanin pigment of feathers is epidermal in origin.

3. Melanoj)hores were found in the dermal pulp at the proximal end of feather germs from the common fowl. They are presumably homologous with the dermal mfelanophores of the skin. Some of these pulp nielanophores have processes which are usually relatively short, but they do not appear to distriliute pigment to other cells, and they have no part in the histogenesis of the feather or its pigment. A few of these dermal nielanophores were found in contact with the basement membrane but none had penetrated it.



Daviks. II. U. l>vS<.) Dii- KiitwirkliitiK <I<t Kcdor iind ihro HczioliiinKPn fu

aiulrrn InlcKUiii<-iitK<'l>>l<l>'»' M<>rp. Juhrli., Bd. 15, pp. rM) lAh. JuNEit. I., 'rill- ili'Vrli)pini-nt of iiratlinK feiithrrs. Liib<>rati>ry Hull., no. 13,

OImtIih CollfKc IS pp. I.i.iivD-.rnNKs. (). l!)l,5 Stiidirs nn iiiluTitaiirr in piRooiw. II. A microscopical

and chemical study of llic fcutlior pinmonts. Jour. ICxp. Zoo!., vol.

IS, no. 3. April, pp. 4.>{-»05. R.Kiii.. II. ISm I'dM-r die KntuicklunR dcs Pigmentos in dcr Diincnfeder des

llulinclicns. Ontral)!., f. Phy«'<>l • Bd. S, p. 250. Strosc. U. M. I902 The development of color in the definitive feather. Bidl.

Mils. Comp. Zool.. Harvard, vol. 40. no. 3. October, pp. 145-185.

1915 Further oliservations of the oriRiii of melanin pigments. Anat.

Ree., vol. 0, no. I, January, .Mistract o'J.



Frniii lite Aniilnmical Lnl)oratory, Universilij of California

Since the first description of maturation of the ovum of the rabl)it l>y \'an Beneden in 1875, the study of this remarkable phenonieiioii lias boon extended to certain other niannnals. with the result that practically all observers now agi'ee that two polar bodies are formed by every o\'um, the first of which is extruded from the egg while still in the ovary, but immeiliately before rupture of the follicle. Formation of the second jwlar body then proceeds to the formation of a division-spindle, but goes no farther until the egg is discharged and fertilized. .Vfter the entrance of the spermatozoon the .seconil polar body is completed and extrudcil. In the event that fertilization does not occur, the ovum degenerates in the state in which it was freed from the ovary.

Full and comprehensi\-c observations of the proc(>ss of maturation have been made in the rat (Sobotta and Hunkhard '10), guinea-pig (Rubaschkin '()o), opos,sum (Hartman 'IG), a bat, Vesperugo noctula (\'an der Stricht '08), the domestic cat (Longley '11), and the mouse. The studies of Sobotta ('95) and others, lead them to believe that in the last named species many eggs form but one polar body. The interest excited by this statement has caused the mouse to be more fully investigated than any other, with tlie result that the views ciuoted have been proven ineoiTeet. The reader is referred to the work of I>ong and Mark Cll) for details of the question together with the most comprehensive account of maturation yet presenteti in any nuimmal.

The final disproof of this supposed exceptional case raises the (|uestion as to whether the process may not be identical in all manunals. To answer this a wider .search will be necessary, for



the species named above comprise but three rodents, one marsupial, one carnivore, and one chiropter. There are also old or undetailed accounts or ])reliniinary notes ujion the ova of the dog (O. \ an der Stricht 'OS), tlie rabbit (Van Beneden 75, Heape '05), the niarsuiual Dasyurus viverrinus or Australian "native eat" (Hill '10). an insectivore. Tupaia javanica (Hubrecht '95), and of a lemuroid ape. Tarsius spectrum (Hubrecht '02), and thus of the twelve orders of mammals there are four in which the process of maturation is well known in one species or more, two in which it is very obscure, anil six in which ripening ova have never been seen.

Amonp the la.«t are the ungulates, in which the species are either rare and inacces.*<ible, or if conunon, are so large that the search for the ova is very difficult. For these reasons it will be of interest to describe a small series consisting of 15 ova from 7 sows (sus scropha domesticus). I am indebted to Mr. Ralston B. Brown, Sui)eriiiten(lent of the Oakland Meat and Packing Company of Oakland. California, for the opportunity to collect the material, and to Mr. A. E. Amsbaugh and Mr. Felix H. Hurni for a.ssistance in its preparation. The sows were selected and the tubal ova found by the method given in a previous paper (Corner and Amsbaugh '17). Ovarian ova wer6 studied in serial sections of the follicles (celloidin) or by cutting hardened follicles into sliees, locating the discus proligerus and sectioning it in paraffin.

1. Killed during oestrus, probably on the first day. Unruptured follicles. Three ovarian ova sectioned, two show germinative ve.sicles, the other shows the first polar body and the second polar spindle.

2. First or second day of oestrus. One follicle ruptured, its ovum not found in the tubes. An ovarian ovum shows the first polar body and the second spindle.

3. Second or third day of oestrus. All follicles ruptured. Four imfertilized tubal ova sectioned: all show the first polar body and the second .sj)indle.

4. History unknown. Five unfertilized tubal ova found, of which four (studied fresh) show one polar body extruded.


5. Probably socoml day of oostrus. Copulation obsen'ed 24 hours before killing. One tubal ovum found, covered with spermatozoa. In sections, the male and female pronuclei are in contact. Two polar bodies.

0. Second or third day of oostrus. Copulation observed 10 hours before killing. Two ova found in the tubes; both were covered by spermatozoa. One of them, studied fresh, showed two polar bodies. The other was sectioned, and showed the pronuclei in contact, but unfortunately the polar bodies were obscured by damage to the sections.

7. Tubes contained segmenting ova of 2, 4. and G blastomeres. One of the two-celled embryos (studied fresh) showed two polar bodies; a favorable view of the others could not be obtained.


The first ova of an ungulate mammal to be studied indicate that the sequence of maturation is the same in swine as in previously stuilictl forms of other orders, the first polar body being extrudetl and the second polar division proceeding as far as spindle formation before fertilization, the second polar body being cut off only after the entrance of the spermatozoon.


«  VAN Benkden, E. 1S7.3 La maturation de I'oeuf. la fecondation, Cetc.) chez le

Lapin. l$iill. do I'.Vcad. Royale de Bclpiqup. T. 40, no. I'J, p. fiSfi. CoUNER, G. W. AM) .Vmsuaui;!!, a. E. 1917 Oestrus and ovulation in swine.

Anat. Rec, vol. 12, p. 287. Hakt.max, v. O. 191C Studies in the development of the opossum, Didelphys

virjiiniana L. Jour. Morph., vol. 27, p. 1. Heape, W. 1905 Ovulation and deffcneration of ova in the rabbit. Proc. Royal

Soeiety, London, vol. 71), B, p. 2()0. Hli.l,, J. P. 1910 The early development of the marsupialia with especial reference to the native cat I Dasyurus viverrinusi. Quart. Jour. Micros.

Sci., vol. 5f), p. 1. HuBRECHT. .\. A. W. 1,S95 Die Phylogenose des .^mniros und die Bedeutung

des Trophoblasts. Verhandelingen der Kon. .\kad. van Wctenschap pcn. te .\ni8terdam. Tweedo .'^ectie, Dec!. 4, No. 5.

1902 Furrhung und Keimblattbildung bei Tarsius spectrum. Ibid.,

Dccl 8, No. 6.


Long, J. A. and Mahk, E. L. ISMl Tin- imitiirntion of tlif fn\i of llie nio\i8C.

Puliliriilion N<i. 142 of the CariicKii' Iiuslitiition of WaKhiiiKton. LoNfiLEY, W. H. mil Till- iiiiktiiratioii of tlii> ccfc nnd oviilution in the (lumestir

cut. Am. Jour. Aniil.. vol. V2. p. 13'J. RrBASciiKiN. \V. 10(15 I'i'Imt <lir Ki'ifuiiR.s — iiiul I)<>fru(-htiiii):i<|)rozes8c dcs

Mccrsrhweinclu'iK'icK. Anat. Ilcflc, Bd. 29, S. 507. SoDOTTA. J. 1895 Die HefriirhtunK und Furchung dcs Eic8 dcr Mans. Arcli. f.

Mikr. Anat., Hd. J.i. S. 15. SoBOTTA, J. AND BrHrHKAiiD. (i. L. 1910 Kcifung und Bofruchtiing dcs Eies

dcr Wcissi-ii Ilattr. Aiml. Hcftc. Bd. 42, S. 43.3. VA.N DEK i^THKiiT. O. 1!K)S La Structure <ic I'ocuf chcz la chioniic. C. R.

dc I'Assoc. dcs .Vnatoiiiists. Dixicmc Hounion. Marseille.

1910 La structure <le I'oeuf des inaniiniferes, (C'hauve-souris, Vcspc rugo noctulai. Troisieme partie. Memoire presentc a la classc des

isciences, Academic Hoyalc de Bel|{i(|ue, 2 ""^ Serie, Tome 2.



I.I.oMJ 1{. KKY'.\()LX).S From the Division of Aiialomy of the Stnnjitni Medical School


This interesting (IcvcloiJUK'ntal anomaly, was observed in the practise of Dr. Robert (1. Reynolds of Palo Alto in March of H)l(). The x-ray photographs acconipan>-ing this article were kindly taken by Dr. l^cyiioids, through whose courtesy I am enabled to report the case. .\s it usually happens in these instances the family history reveals related abnormalities. On the father's side of the family there have been se\cral cases of abnormal development of the skeleton of the upper extremities. The paternal grandfather is said to have had syndactyly of the index finger anil the thumb of the right hand, up to the interphalangeal joint of the thuml). The terminal phalanx of the thumb is said to have been flexed at right angles on the proximal phalanx so as to lie in the palm. The left hand and both feet were said to have been normal.

The father's oldest sister who died at the age of six is said to have had a deformity of the right hand similar to that of her father. Otherwis(> she was normal.

The father's second .Mster who died soon after l)irth is said to have had congenital absence of radius and ulna on the right arm and the father's older brother is said to have had similar deformities as the jiresent case.

The father's hands both show muscular defects. On the left hand the thenar eminence is absent, and the opponens and abductor jiollicis brevis apparently are lacking. The flexor brevis may be present. '1 here are two gi-oups of intcrphalangeal



\.\.i)\ I) K. RKYXOI.nS

aiul one of iiu'ta(';irpoi)lial;iiint';il -ulii. in tlic tliuiuli. ^ Ct Iho x-r:iy pliotoKraj)!! itifi. 1 i sliows that there arc only two plialanjrcs. The h'ft thuinl) is strai^lit and tapering; only al)out two-tliirds normal sizo and tho skin is smooth, shiny and atrophic. It is (1.4 cm. lonji and is rotated out a little. The mobility in the

l"i)r. 1 Left liaml i>f falhrr

second terminal phalan.x of the thnml) is somewhat limited. The width of the left hand in the line of the metacarpophalangeal articulations is cin. There is no wehhiiig of the fingers: the rest of the lingers of the left hand lieing apparently normal. The thumi) of the right hand is somewhat longer than that of the left. l>ut the thenar eminence is also absent; the regionbeing


marked liy a (Icciilcd llatlcniiin. Tlic rinlit thumb is 7.4 cm. long and is rotated in rather than out, as is the with the h'ft. It is tapcriiiK and atropine as in the other htind, pf) a small l)Ut ap])anMitly normal nail. The i)ulp of the thumb is (-(impressed from side to side and looks atrophied. 'I'liere are two prominent sulci opposite the interphalanj^eal joint of the thmnb and the width of the hand at the meta('ari)al|)lialanneal articulations i.s decidctlly greater than in the left hand. The other fingers are normal. The width of the right hand in the line of the metacari)alphalangeal articulations is 9.2 cm. or 0.0 em. more than the left. The medial surfaces of both thumbs, a little distal to the metacarpn|)lialanf?eal articulations are marked by an oval callosity about '2 cm. in diameter and O.o cn\. thick which forms a prominent pad. .Vs shown in itig. 2) the medial surface of the thumb is concave, the distal portion being bent medially, that is toward the ulna, the distal phalan.x being (Hrected medially at ai\ angle of about 10 degrees.

In the x-ray jihotographs itigs. 1 and 2) there is a pronounced mortising of the basal ends of some of the metacarpals of the thumb with the basal ends of the metacarpals of the index fingers. This condition seems to be the result of a general naiTowing of the wrist bones. In each hand the greater nmltangular is placed medially until its medial border is almost even with the medial border of the lesser multangular. Hence the position of these bones and the con.scMiuent jKisition of the respective metacarpals may jio.ssibly account to some extent for the lack of mobility of the metacarpals of the thumbs. The proximal phalanx in each thumb is markedly concave on its palmar surface and has a slightly bent shaft. This is more noticeable in figiu-e 2. These peculiarities in the thmnb may be ex|)lained in part by the fact that the lack of mobility of the metacarpals has made it nec(>ssary to rotate the j^roximal phalanx medially in order to use the thumb in conjimction with the other lingers. As a result of this rotation we see the phalanges from a lateral view rather than a dorsal and therefore the concave surface is the ventral and not the medial .surface of the phalanx. There doesn't .s(>cm to be anything to account for the three distinct



liciiiy tuborcules shown in outline in the x-ray plate on the ventral surface of the distal phalanx nf eaeh tlninil) (fips. 1 and 2). The distal ends of tlie i)lialanges seem large and irref^uiar. This is more noticeable in the rigiit (fig. 2) than in llie left

Fin- 2 Hinlit h.ilMl of f:illirr

(fig. 1) hand. l>ul this may he due to the fact that in figure 2 the hand was not held as flatly on the plate as in figure 1.

The son, the subject of this rej)ort is 13 years of age, ") feet, 8^ inches in height and weiglis \'M) jjounds. The ])hotograi)hs (figs. 3 and 4) show the marked deformity in the general appearance of the left hand and I figs. 5 and (i) the lesser deformity in the right. The I! \' fingers are developmenlally (|uite



normal for the iiiidtUc iingcr of the li-ft liaiui was injured in a recent accident. There is no webhing of the fingers and with the excei)tion of the second and fourth fingers the sulci are normal. On the palmar surface of the fourtli finger there are two sulci about 1 cm. apart opjxtsite the first intcrphalaiigeal joint. The index finger has a similar peculiarity. The thumb has the appearance of the little finger of either liand. It would indeed be im


Figs. 3 and t Left hand of son

possible to identify it by means of a photograph of these digits alone, in the of the left hand. The terminal phalanx and nail of the thumb are shaped exactly like the terminal phalanx and the nail of the little linger.

On the right hand the nail ui the thumb, although smaller, is shaped more nearly like the normal thumi) nail, and the terminal phalanx is also shorter and heavier than the terminal phalanx of the little finger. In each instance the thumb is 0.7

TIIK ANATOUIr.vL fllHXlKD, VOL. 13, SO. '.



("III. loiipcr than tin- little (inner and markedly coiicavo medially. The space l)i't\veen llie index and middle linnt'is of the left hand is (greater than normal and the index tinp;<'i' is ooncavo medially. This is jirohahly due mostly to the fact that the boy usually prasps objects between the iiulex and middle fingprs, rather than between the thumb ai\d ind(>x (infers, because of the abnormal tlovolopn)rnt and relation of the thumb. I Ix- thumb

l"i(»s. ."i :iml tl liiRht liiiiul of son

of tlie rifiht haml is also concaxc medially and small objects, such as a coin, when nrasj)ed by the ri>;lit hand ai'e ])lace(i between the dorsal surface of the ternunal phalanx of the thumb and the tips of the fingers

The thenar eminencfs are absent with ai)i)arently an absence of all the thenar nmscles (figs. 4 and (i). Even the adductors appear to be ab.sent and facts no doubt account for the jMculiarily in mobility. There is a lack of mobility of the metacarpals of the Ihumlis which is very noticeable because it limits


the motion f)f the tlninilis to tlio niovonionts possihlo in the othor {iiifijcrs.

The position of the thunil) in relation in the other fingers gives the forearm tlic apix'aiance of Iw'ing much longer than normal even though measurements show it not to l)e exceptionally long for a hoy of his height. The hand is markedly widened at the metacarpo])halangcal articulations. The left hand is 9 cm. wide. The width of the four fingers lieing S cm. The right lianil is [).'.i cm. wide and the width of the four hngers S.4 cm. A similar tlifferenee was present in the father's hands as stated above.

The jihotogiaph in figure 4 shows the smooth, glistening appearance of the skin in the ]ialm of the hand. This is characteristic of both hands although a slight movement during the taking of the picture prevented its showing in figure 0. The skin seems to be drawn tightly so that very few sulci apjiear.

In the x-ray photographs (figs. 7 and 10) it is shown that the thumb of the left hand has three, well-formed, phalanges which in large part accounts for its length. In this case the age is such that the epii)hyses show exceptionally well. There is nothing to indicate that the middle phalanx of the left thumb is rudimentary and will later fuse with the terminal jihalanx. Indeed the middle is longer than the terminal phalanx anil has a we'll-formed, normally-located epiphysis (fig. 7). Each of the other phalanges has a well-formed epiphysis and seems entirely separate from the middle phalanx. The metacarpal of the thumb on this, the left hand, also is of interest for its form and length seem inoic nearly like a true metacarpal than the corresi)on(ling bone in a normal hand. Besiiles, the epiphysis is at the distal instead of the proximal end of the bone, the same as the on the metacarpals of the other tiiigers (figs. 7 and 10). The metacari)al of the thumb is relatively long. The x-ray i)hotographs were taken with the hands lying directly on the plate and with the machine o\-er the .<ame part of each hand. Measuring the negative we find that the nietaeari)al of the thumb is 0.7 cm. loiifr while the metacarpal of the little finger is only (1 cm. long. The circumference of the first meta



(•ari)al al.>*o seems less than the circunifoionce of the correspondiiiH liDiie of the little fiiiiicr. Xoiinally the inctacariJal of the thunil) is sliortcr and tliickcr than any of llic other in('tacari)als. The x-ray ])hotogi'ai)h of tlic i-i^ht liaiid is in sonic i-cspccts more interestiiip; than lliat of the left. The extra phalanx of this tliuinb is not so well formed as that on the other thiimh or

Figs. 7 iiiul .s Left and ii^;lii li.-mds of .son

those of the other corresponding phalanfjes liut nevertheless as .shown in hunre '.• it is a separate bone with a distinct diajihysis and epiphysis, articular surfaces and base and head. Besides showing the .same peculiarities in regard to the extra i)halanx, the metacarpal of this thumb shows a double epiphysis (figs. S and 11).

In all of the cases of three-jointed thumbs reported, heretofore, the middle i)halanx is con>idered the extra phalanx.

iiYri;iu"iiAi,A.\(;isM 121

Pfitziior, '!)() who has done more work than anyone else on the (Icvclopiiicnt of tho oxtreinitios. has roportod several rases of tlircc-joiiitcd tiiiinilis, hut in all of these cases the three joints were only temporary for the extra, middle phalanx later fused with the end phalanx to form one bone. In no ease did Pfitzner find an extra jjhalanx that hail a distinct epiphysis, articular surface, head, and which in general had the shajje of a normally developed phalanx as in the present ease.

Kig- !' Thumb of right h:iti(l taken latorally

Rieder '00. rejiorted a family, most of whom had three-jointed thumbs. The father ami four of tlie children had an extra phalanx in the thumb. Kieder claims that his cases ditTcr from those reported by Ptitzner in that each end i)halatix did have a distinct epiphysis entirely separate from the middle lihalanx.

Dwijrht '07 in his report on the variations of the bones of the hanil and foot considers the jiresence of an extra i^halanx in the tliuiiil). lie says that in no cas(> reixirtcd diil the extra jihalaiix have an ei)iphysis in those of an a^je at which it could be expectctl to l)e present. He supports Phtzner's theory tliat the extra phalanx is later fuseil with the terminal ])halanx.

1 ■_'■_* LU)YI) R. REYNOLDS

Salzar 'OS, also r('pt)rtr(l a case of a tlirco-joiiitcd thuiiil) on oach hand. In each instance the second phalanx was short and ajjpaniitly ludinifntary. The so-callid ps(udo-e])i])hysis shown in hjtures S and 11 is distinct and has the cliaractcristics of a normal ejMphysis. It woulii be hard to consider one a secondary e|)ii)hysis as has been (h)ne in most of the cases heretofore reported and in this case it seems (juite evident that the first metacarpal had two true epiphyses. The presence of an extra ei>iphysis. however, has not caused any increased length as might be sui)posed. f<ir although this metacarpal is longer than the metacarpal of tiie fifth linger of the same hand, it is not as long as the metacarpal of the thumb on the left hand, which has only one epiphysis. I am leail to believe that both epiphysis are true epiphyses, by comparing the x-ray photographs taken nine months apart. In figure 11, taken in November of liUti, both epiphyses h;ive the appearance of true epiphyses as much as they did in the photogiaph taken in March of 191(i (fig, 8) or almost a year before. Hence although the supernumerary or cephalic epiphysis may fuse earlier than the basal or normal one it is clear that the former does not fall under what Thomson '06 .speaks of as "scale-like epiphyses on the head of the first metacarjjal which makes its apjiearance about eight or ten, and rapidly unites with the head. Since the first metacarpal in the left hand of this case, has a distally located epiphysis an interpretation of the latter in Thomson's sense would leave this metacarpal without an epiphysis.

Freund '05 also reporteil a case in which the midiUe phalanx had an epiphysis at the distal as well as at the proximal end. He compares his case to those reported by P(it/ner (HI, and arrived at the same conclusion regarding the spurious character of the.«^ epijjhyses Pfitzner his conclu.sion ujion the variation in ossification foimd in many which came under his observation. He thinks that in of apparent supernumerary epiphyses the oestodasts, instead of breaking down the epijihysral cartilage in a straight-line front, extend into it in the fmin of luoccssc s or projections and thus ])roduce an appearance which roughly simulatrs an epiphyseal line.

H Y I'E UPH A L.\.S( ; IS M


But as the photographs (figs. 8 and 1 1 ) show the breaking down of the cartilage in tliis case is being earricd on in a straight line front. The portion of cartilage between the diaphysis and the extra epiphysis is as even and regular as in any normally developing bone. Hence Pfitzner's explanation does not apply here and this epiphysis seems to be a true ami not a pseudoei)iphysis.

l-iKsi. Ill :iiul II Saliio Imiul^i nboiit one vcar hiUT


'l"lu' ini)rtisiii>r of llic l)j\sos of sonic of tlw metacarpals is fiuite proiioimccd. On I lie iiicdial side of I tic liasal ciiii of the thinl left nietacar))al. for example, tlierc is a dee]) notch for tlie articulation of the fourtli metacari)al (fig. lOj. There doesn't seem to be any explanation for the jMesencp of this unusually deep notch.

Scrutiny of figures 10 and 11 will show that the carjial hones in tlie right and left hand vary both as to shape and relative position. In (fig. 10) of the left liand the greater multangular is considerably larger and comiiletcly o\ershadows the le.sser multangular bone. In the riglit hand (fig. 11) the lesser multangular is larger than the greater nuiltangular but the greater multangular is more laterally placet!. Figure 10 also shows the tri(|uetral bone to have a rectangular shape. looks not luilike a short phalanx. The triiiuetral bone shown in figure 11 is triangular and considerably larger than that in figure 10.

The thumbs in this case es|)ecially the left seem to be like a little finger in most respects except their position on the hand. Their movements are limited to the normal movements of a finger; they have the same number of phalanges as the other fingers: the nails resemble the nails of other fingers rather than thumb nails and their metacarpals have flic characteristics of the metacarpals of fingers rather than those of normal thumbs. I df) not think that the extra phalanx in either case can be cont<idered as rudimentary of the perfect development of each individual phalanx, the presence of a di.stinct normallylocated epiphysis and the absence of any indication of fusion.

Professor Meyer to whom 1 am indebted for suggestions and assistance has offered a i)o>sible explanation for the occurrence of sueh a finger or thimib. If 1 understood him correctly he thinks that the ai)pareiit absence of the thenar muscles may indicate that the normal impulse to develojiment of the thumb was absent and that hence the phalanges and also the metacarpals of the thumbs, assumed the form and relations characteristic f)f the fingers rather than of the thumb thus accounting for the extra phalanges in the thumb, the epi|)hyses in the supernumerarj' phalanges and also for the superinunerary



Fig. 12 Normal left liiiiid of u boy of approximately the samp age, height and

weight for comparison

ill the nu'tacarpal of the rifiht thuinh as well as the distallyplaced, single e])iphysis in the metacarpal of the left thumb.

Although Professor Meyer's suggestions seem to account for the results found in this case, he feels that we can not he certain that the thenar nutsdcs are entirely absent. However, sis far as can be determined from external apjicarances and from the limited mobility of the thumb thiy must lie exceedingly rudimary indeed if present.

IJli I.I.OVl) K. ItKVNOI-ns

lUHMiiCK M'llV

Annadalk, Til. 1S<)5 The iniilformiitions. dispiisofi and injiiries of the finf^ers

nnil locH and their siirKieiil treatment. KdiiiliurRh. Hi.iFF. M. .1. l.SLt) riM-r das xoiteiianiite ()s nietanirpi piillicis mit 3 .\l)l)il diincen. .\rchiv fiir An.ntoniie iinil Pliy.-iUilouie. nwHiiiT, Till >M AS liHlT Clinical atlas of the hones of the hand and fool. Philadelphia and London. P'HEfND, L. MKI.") l"l)er IVeiKhiepiphy.sen. Zeitsehrift fiir .Morpholojiie iind

AnthrnpoloKie. vol. S. Kaltin. U. I'.HM Kin Fall von .Misl>ilduiiKen der olieren lOxtremitiit d\irc'h

I'lxjrznhl. .\rchiv fiir .Vnatoniie iind Physiologic Jo.M'illMSTHAL. G. I'ljcr Brachydaktylie iind Hyperphalan(ric. Virohow's

Arehiv. Bd. 151 Heft 3. Ki'MMKl.. W. ISSI.i Die Mi.ssliildiinKcii der Kxtreniitiiten diireh Defeckte,

Vervvai-hsunK und (Mierz;ihl. Hililioth. Med. K., Heft. 3. Leboitq, H. I.SOli De la hraehydaetylie et de I'liyperphalaiipie ehez I'lioinine.

Hidl. lie I'aeadeniie royale de medicine de Helfsicine .S-anee du. 30 Mai. Pfitzxf:r. \V. IS'X) Kpiphyseiil>ildiinpen. Mannigfaltigkeit dersclhon. Arehiv

fiir .Vnatomie iind Phy.siolopie. .Vnat. .Vlith.

1892 Beitriine zur Kenntniss des nienschliehen E.xtremitatenskelettes.

Schwalbe's MorpholoRische .\rl)eiten, Bd. 1.

Normalmasse der Phalangen o. Beitrag zur Kenntniss dcs mensch lichen Extremitatenskelettes. Sch\vallie"s Morphologische Arbeiten,

Bd. 2.

1S94 I'ber das Skelett der nienschliehen Hand. .*<ch\valbe's Mor pholoijische .\rbeiten. Bd. 4.

1S9.5 Kin Fall von beider.seitner Doppelbildungen tier 5 Zelie.

Schwalbe's .MorpholoRi.sche .Vrbi'iten, Bd. 5.

1897 Kin Fall von VerdopiM'hinn des ZeiRefingers. Schwalbe's Morphologische .\rbeiten, Bd. 7, Heft 2. HiEDKR, H. UKX) Eine Familie mit dreigliedrigen Daiiiuen. Zeil.schrift fiir

.Morphologic und .Vnthropologie. Salzkii, H. Zwei Fiille von dreigliedrigen Daiinicn. .Vnatomischer .\nzciger,

Bd. 14. .'^tliAKFK. .\. 1912 Zwei Falle von symmetrisclien Mi.ssbildiingen der Finger.

Ztschr. f. orth. Chir.. Bd. 20. TiioMaox. .\nTnfK liKXi Chapter on osteology in Textbook of .\naloniy,

Cunningham, London. Second revised e<lition. l'KiKi..\iA.\x, J. isi;2 Der .Mittelhandknoehen des Daumens, seine ICntwickc lungsgeschichte unci Bedeutung. (iottingen. WiMJLE, B. 1891 Thi- occurrence of an additional phalanx in the human

pollcx. Jour, of .\nat. and PhjsioL, vol. 20. \Vi:u:KEN, il. 1884 Die morphologi.sche Bedeutung des 1. Daiimengliedcs.

Prcisvertheilungs progr:inun der l'nivcr.-<ilat Halle.


IVAN K. WALLIN Departmetit <ij Aiiiitoiinj and liiologij, Marijuittc Cnivcrttity School of Medicine


Tlu ohJMt of this iijiparatus is the siil)stituti<iii of an Edison Mazda liimi) for the arc lain]! and the r( diiction in (Mist of a scrvircalilc apijaratns for drawinj;. The arc Ian )) as an illun inatinn unit in a i)rojcctior ap]>aratus has certain ohji ctionaMc features; the light is not always steady and it re(|uires aln'ost constant attention when in use. The perftction of the nitroRcn-fiiled electric Inill) has supjilied a lijihtinn unit which is only slightly less intense than the arc light.

A siir])lc and inexpensive jirojection apparatus would he valuable for students' use in the lahoratcry. T.he ajiparatus di scribed below, it seems to inc. will supi)ly a serviceable ai)|)aratus for class use at a very moderate cost.

This a])paratus is construct((l on the plan of the Edinger apjiaratus (fig. 1). It consists of an ui)right oak lieani (u) attached to an ordinary table, a microscope .support (;/;.. v.) which also supports a rod (.•-■./•.) carr>-ing n flcctor, lamp, condensors and diaiihragni.

The upright (u) is an oak bean', 2 by 4 by 42 inchis, containing a slot for the reeei)tion of the U'icroscoiie support. It is bolted to the sides of the tai)le and braced underneath to the top of the table.

The niicroscoiie sui)port (in.s.) is a steil i)ean'. The l>ack has a flange which fits into the slot of the U])right and serves to hohl it firm. The free end is constructed in the form of a clamp to fit around the limb of the n'icroseo]ie and is n'ade firm by means of a bolt (/>). The body of the supjiort has a slot (.•<) to receive the upright siipjxirt rod (s.r.). The n;icre..scope siipport beam is movable up and down on the wooden ujiright and can be trade secure at any iitint by ireans of the han<l nut (li.n.).

A different ty]>e of microscoix- supjiort is illustrated in figure 3. This t^^^e of su])port has the advantage that the iricroscojie may be attacheil or removed in an instant. The stage of the M'icrosco))e slides iiite) the sle'cve e)f the' I'-shapeel suppeirt. The- feieit allel pillar of the- micreisce)pe" are' turneel bae'k at right angles te) the l><)ely tube'. A sle)t in the' dia])hragm weiulil ace'eimnieulate' the' feieit. (It is an aelvantage te) have- a large diaphragm as it the> we)rking sjiace.)

The sui)peirt roel (.s.r.) and the .sujiport rings which heilel the le'nses, diaphragm, lamj) heuise and re'tlee'tor wire' taken freun an e)relinar>




hiliorsitdiy support or rinn stand. Tin- suijport riiijis in this type :ill have a coiiiniDn center. The siii)|)()rt rod is movable on the tnicro.icopc support heain so that lanip and eondensors may he properly centeri-d \\ it li the microscope.

The dia))luanm is a circular |)iece of wood which is held in jjhice by a su])port riiin. Curtains are attached to the diai)hrat;m and hanj: <lown over the edne of the tal)le. The curtains may be attached by means of hooks and eyes or nl<'v<' fasteners.

The reflector (r) and its housing are a jiart of an old style gas automobile lamp.

The linhtinji luiit is a 500 watt Edison Mazda steroopticon lan)j). The fdaments in this lamp are concentrated and form a compact unit. The lamj) house is made of sheet iron and is held in jilace by a ring support. It serves as a sui)i)ort for the lamp.

The condensor (r) consists of two 4-inch stereoi)t icon lenses. These can be obtained from stereo])ticon supi)ly houses. Thi' adjustieent of the lenses is illustrated in fiinu'c 2. The distance between the two lenses H and (' is arbitrary. The filament of the lamp (.1) should be at the point of focus of lens (fi). The focus of lens (D shoulil fall within the objective of the microscope (D). The focal length may be determined by holdinp, the lens to the sunlight in a smoky or tlusty atmos])here and mea.suring the distance from the point where the rays to the center of the lens.

The microscojie is an old tyiie from which the foot and iiiiiar have been removed. The stage has been reversed so that the sliile rests tm top of the stage. The fine adjustuK nt extension (cm.) c-onsists of a rod, a pi])e, a steel sprinp and a ca]). The ca)) is secured to the fine adjustment by means of a .set screw. The rod fits into the jiipe and is made secure for any length by a set screw.

The fine adjustment extension was made in the laboratory. The microscojie suiijxjrt was made by a blacksmith. If re(|uired in (|uantity the microscoiie supports could be cast in a foundry and obtaineil at a smaller cost.

The cost of the jiarts of this ajiparatus is as follows:

tprinlil iliard oak) $1 2."i

Micniscopc support, holts nnd nuts 3 50

Lenses for condensor 2 7.i

Keneclor .. 100

Ring supports and support rod \ ~r>

Lump ;{ IK)

Fine adjustment extension. ■")()

$14 35

Baiiseh and I.nml> Optical ('om])any liave recentl\' p(rf(i-ted a lamji house witli relleiti r and eondensors which might i>e adapteil to this a])|)aratus.

I am greatly indebted to Dr. 11. ('. Tracy for a.ssistanci' in the optical adjustnu'Ht of the ai)paratus.

AnTnoR*"* AiiHTKACT or mm papkh i^bdbd

BT TIU: hIKl rf)(iftAPniC BtHVK E AL'Ol^T 11.


HALPH D. LILLIE From the Division of Anatomy of the Stanford Medical School

My attention was attracted to the sul)ject of tho variations of tho hypoglossal canal by the discover)^ of a double hypoglossal canal in the dissecting room. Subsequent examination of about thirty skulls, about twenty of them Eiu-opean, revealed ten cases of complete division of the canal. Consultation of the ordinary reference books failed to give definite figm-es. So Dr. Meyer placed the collection of skiiUs in the anatomical museum and Indian skulls in the general nuiseum at my disposal.

Battels ('04) found the hjTioglossal canal doubled in 117 cases in 05 canal was double it was frequently dividetl into unequal parts so that the posterior canal was about twice the size of the anterior, or more rarely in the reverse ratio. Weigner believes that the tlivi.sion of the canalis hypoglossi is one ot the manifestations of an occi]iital vertebra. Graf v. Spec {'96) notes the division of the canal anil says that indications of division are almost constant in the adult but makes no mention of tho character of th(> indications. .Taboiilay and Lucy ('11) attribute the division of the canal to that of the nerve and Froriep ('11) showed



132 KM. I'll D. I. II. I. IK

that in ihr hoviiK' oinl)ry<) tho hy]>oglossal n('r\c corresponds to three segmental iier\-es. Schiifer aiul Syniiiintoii in Quain's Anatomy III, part '2. lOO'J. also regard the hjixjglossal nerve as representing tliree segmental nerves. Prentiss (10) working on pig emhrj'os, found eight ganglionic masses correspontling to four ganglia associated with the In'iioglossal nen'es as their dorsal root ganglia. These ga'nglia were in series between those of the \agus and of the first eerv-ical ner\'es. They are interpreted as representing four ln^^oglossal ner\-es. Five or six ventral roots of the In^ioglossal e.>dst, but according to Bremer ('08) the more anterior of these represent ventral roots of the vagus and glossopharjTigeal nerves. Martin ('91) working on cats, found live ganglia and five ventral roots and concluded all were hj^ioglossal. So it woukl seem that either tliree or four nerves exist developmentally and this could account for either three or four somites and their corresponding neural foramina and their corresponding vertebrae. Hence between the jugular foramen and the synchondroses petro-occipitalis and occipitosphenoidalis on one hand, and the atlas on the other, there must have existefl four or five occipital vertebrae with three or four neural foramina or hj'poglossal canals between them. Consequently we should expect to find traces of di\-ision of the hypoglossal canal into either three or four parts.

In fact we do find such divisions. As stated above Weigner noteil indications of cUvisions into three parts, also into two parts, the posterior of which was twice the size of the anterior. I also have seen a number fif such cases. Such occurrences could be interjjreted as jnutial or complete separation of the first hM'oglossal nerv^e while the two following remain together. I have seen indications of dixisiou into three equal parts two of which may be comjiletely separated. However, complete division into three i)arts has not b en obsen'ed.

McMurrich ('05) says "dujnng the cartilaginous stage of the skull the antcT-ior condyloid foramina are divided into three portions by two cartilaginous p.utitions which separate the three roots of the hj'poglossal nerve," and considers this as evidence of the existence of four fused vertebrae in the occipital bone.


These facts point toward tlirce original hyjioglossal canals. But a mniilx-r of oases wore ohscrvcd in which tlic canal was conij)lct('ly or inconii)lctoly divided into two ecjual parts and in which the anterior canal showed indications of further division into two equal parts. In the light of the work of Prentiss and Uremer I nnist look ujiou such instances as an indication of division of the canal into four equal parts. Of course, it may be in this case that it is really di\'ided into three and that the third h^-poglossal nerve is larger than the others. Poirier et C"liari)y report cases in which the canal was divided into four parts. I'rofessor Meyer suggested to me that some of the osseous processes observed may be secondary' ossification in the dural septa extending between the fasciculi of the roots instead of remnants of arches of occipital vertebrae. Such a ca.'^e was noted in No. 42, a dissecting-room specimen. In this skull the dried remains of a dural or connective tissue septum extended between two spurs on opposite sides of the foramen. Whether these spurs are to be looked upon as secondary ossifications of a dural septum or whether such septa are to be considered as unossified remnants of vertebral arches can not be decided in dried skulls.

It is also probable, indeed almost certain, that in some eases these di\'isions are significant not of diAision of the nen'e, but of vascular variation. Some of the variations recorded here are unquestionably due to aberrant vessels and not to diAision of the nerve. I refer here to those cases in which tortuous canals leave the main canal more or less obliquely to reenter in another place in similar fashion. On the other hand a large number of these variations are in all probabilitj' due to di\-ision of the nerve, for the nerve has been observed to pass through as two separate nerves, each in a canal of its own. In the more caudal parts of the body the metamerism of nerves and vertebrae iiolds strictly. Metamerism of the nerves exists in the head, showing best in the embrA'o. Indications of bony segmentation of the canal corresponding to the nervous segmentation are al.>^o found. Moreover an occipital vertebra may be partially separateVl. facts all point in the same direction.




SS Indian akitlh from Jersey County, Illinois

Character of hypoRloBHnI cunals




A sliRlit suiK-rior spur not

\ septum a few millimeters thick at inter

at cither opciitii):

nal opening. Tlir for.imina equal





Simple, smaller

Simple, larger


A slight superior, promincneo near tlie margin

A slight superior tubercle at the inner margin


Simple, n>iinJ

An inferior spine, a flattened internal opening and a slight indication above at the internal margin. Hound externally


A small a'ntero-siiperior spur

A small antero-superior spur



.\ rough superior internal margin


A heart shaped internal

An anterior foramen right through to a

opening with a promi

common external opening. A larger pos

nence sui>crior

terior foramen giving rise to a small anterior canal through to the common external opening and a large posterior blind pit. In the external opening are three small pits above and posterior to the two canals


A small antero-suix^rior notch

A small antero-superior notch


The internal margin is

A marked internal superior spine with an

rough superiorly

indication opposite on the lower margin. The anterior division is half the size of the posterior


A septum 3 mn>. wide by 1 mm. thick internally. The anterior canal is smaller than the posterior

A projection on the superior margin


A slight su|)erior spur

Two equal canals


Opposed spurs almve and below

Compre.s.scd dorso-vent rally


A broad septum within

.\ broad septum within the canal. The

canal. Anterior canal

anterior canal half the size of flic posterior

half the size of the






A slight superior tul>crcle

\ slight superior spur











■ Simple

A deep septum half the length of the canal. I'lqual division.



A slight double spur on the internal HUficrior margin



.\ slight spur from the roof near the internal orifice


A superficial internal sep

A rough su|K."rior margin internally. Tlie

tum. Small anterior,

medial side of the posterior condyloid

large posterior canal

canal opens into the skull posterior to canalis hypoglossi


A broad superficial inter

.\ slight superior spur. The anterior divi

nal septum. Anterior

sion half the posterior

canal smaller


Slight antero-supcrior spur on roof

Slight antero-suix-rior notch





Sim|)le, larger

Simple, smaller


Anterior superior spur. One to two division

Slight double superior spur


Antero-superior notch

Two slight superior elevations











A ridge below, a spur above









A marked superior spur





7'iro Indian skulls from La Conner, Washington

7645 7644

An internal septum. Anterior canal half the posterior


Simple Simple

Three Indian skulls from the Longueville graves, Plumas County, California


A superior spur

A narrow, sept urn





A broad superficial internal septum






UAI.l'lI 1). I.II.ME

(iBSKUVATIONS-ronlinurrf 16 Iniiian skulls from I'onct mound near MayficUl, California








A thin, narrow srptuni

Superior tongue-like process and

1 sIIkIk

in (he niid-pnrtion of

inferior (ulierclc

the oiiniil



Complete septum, .\nterior canal size of the posterior

half the


Spur on the superior internal niarfdn

Opposed spurs on internal .margin






A superior anterior roughening


Simple (young, first permanent molar)









Simple, flattened

Simple, flattened





A deep internal septum in inner third of canal






A thin, superficial internal septum. Anterior foramen half the size of the posterior



Both foramina round. Septum deep' in middle third of canal. Anterior canul larger

Flattened from top down.

Fire Indian skulls .

'rom near San Jose, California


Median .septum lu-ar in

.Median septum near internal foramen.

ternal foramen

Opposed spurs divide anterior canal into two


Antero-£U|K'rior spur and notch



Simple, round

Opposed spines. Anterior division half posterioV


Broad flat sepliun. Two internal foramina arc equal. External foramen single

Basioccipital of this side is destroyed



Strong sup«'ri<>r and .slight inferii>r spines on internal margin. Anterior division half poi<terii>r



Twelve whole skitlU or entire occipital bonea


9 10

11 12

ronipletely dividoil iiorve ill liDth cunnls

RoukIi superior internal margin

Slight median superior spur

(102.) Clear

Antero-superior notch

Medium superior spur. .\Iso superior spur in middle of anterior division


(13.) Strong septum anterior canal 2-3 times as large as posterior

An anterior-superior notch

Clear. (Occipital only)


.Slight median sui)erior elevation. Two small orifices just above internal opening

Clear. Only one hole in dura

Hough su|>erior internal margin

Clear. (Anterior part atlas assimilated)

Clear Clear Opposed median spurs


Septum in middle third of canal. Anterior canal is twice the size of the posterior

Anterio-siiperior notch. UJccipital only)

Clear internally. At external opening a small superior canal cut off



Twenty-two European skulls


IV. Clear

Antero-superior cleft and spur


V. Thin median septum

Thicker median septum in middle third.

in inner third

Spur in middle of superior margin. Small inferior spur and opposed superior ridge in middle of anterior canal


XIII. Large and clear, posterior condyloid canal opens laterallv

Large and clear


II. Clear

Median su[)erior tuljcrcle


XV. .\ntero-superior ir

Antero-superior and antero-medial spurs.

regular process 1 to 2

Small canal antero-superior to two spurs

nun. long


Thin septum in middle

Opposed elevations. Anterior part con

third of canal, .\nte

stricted and half the size of the posterior

rior two-thirds size of

posterior. ((> on medial

side of left mastoid;








XII. Oppo.m-d iiiitoro-«iiIH-rior itiul -iiiotiiiil spiirs which enclose n Hinitll caniil

Same as left but enclosed canal smaller


XIV. Clear


VII. .\htero-8iipcrinr nmri;inal elevalicin



III. Slight nntcro-su|wrior tuljcrcle



XVII. TriaiiKular. A fossa runs liack from posterior suix-rior corner into a small pit

Antcro-superior notch clear through


IX. Clear openin):. A small venous canal runs antero-medially from roof near posterior orifice to near the external orifice of canal



XVI. Clear



VIII. Clear



Heavy septum. Anterior canal small, almost slitlike. The slit points toward the condyle



V. Clear



III. Clear

A long superior spine which almost divide«  the main canal into a smaller anterior canal and a larger posterior. .\ wide opening in the postcro-latcral wall <)i>cning into the posterior condyloid canal.


II. Strong septum in inner half. Anterior canal slightly smaller



(04.) (Wcfxarate occipital) Clear

Median superior spur


BU. Opposed nieilian spurs, the su|M-ri(ir being stronger



Cll. Clear








IXK Al.lTf







n n



Si 8^


X a!


Illinois Indians <

California Indians \


Specimens from dissecting room I

European skulls

Total 1
















17 15 32

14 16 30

31 31 62

13 13 26




48.6 42.9 45.7

56.0 66.7 61.2

43.7 43.7 43 7

59 1 59 1 59 1

49.02 49.34 49 18

14 15 29

3 5 8

30 31 61








42 9

41 4

12.0 20.8 16.3


43 7

42 9

22.7 36 4 29.5

33.99 38.82 36.39


5 9







4 1. 5

26 18 44

11.4 14 3 12.9

32.0 12.5 22,4

14 1


13 4




16 99 11 84

14 43

35 35 70

25 24 49




22 22 44

153 152 305

Unpaired series. Fifty-nine left sides of skulls or of occipital bones from cadavers dissected in the dissecting room

1. .Viitorior superior spine and notch.

2. Anterior superior spine and notch.

3. Clear. .\ small canal from the lateral xvall of the canal opens behind the occipital condyl. Anterior to this are two other apparently blind small canals.

4. A superior tubercle in the middle of the internal margin and another midway between the first and the antero-medial internal margin.

5. Clear.

6. \ strong septum with a suiH'rior spur projecting inward. The anterior canal is smaller.

7. The superior wall of the canal is destroyed, the remainder appears clear.

8. A strong superior spur internally. The anterior division is half the posterior.

9. Clear.

10. .\ small au|M'rior elevation between anterior and middle thirds of the internal nuirgin.

11. Clear.

12. Large. Rough superior and aiiteru-mcdial margin.

13. (13L.) .\ strong complete septum clear through to the external opening. The anterior canal is half the posterior. .\ canal from the postero-lateral

141) UALl'Il I). l.n.I-IE

wall of thi- piiatorior riitial oimtis into the mcditil siilo of tho posterior roiulyloid canal. Anterior ciiniil has niarked anterior eonsi ruction. From anterior opening the anterior canal lias marked superior diverticulum which communicates with top of posterior canal. Latter has smaller anterior opening.

14. Clear.

15. I^arge, clear.

16. Posterior condyloid canal opens just posterior to posterior internal margin. The internal orifice had sliRht antero-supcrior notch. Opposed superior and inferior ridges near the external orifice divide the canal evenly.

1/. Antero-superior cleft.

18. Clear. One pit in sujiefior, another in postero-lateral wall of the canal.

10. Hough suj)crior margin. .\ small vascular canal goes laterally then for\vard from the posterior internal margin and reenters the canal at the external orifice.

20. Antero-superior cleft internally.

21. The superior wall is destroyed. Uest clear.

22. Strong median superior spine internally.

23. Small antero-superior cleft.

24. Clear.

25. Clear.

26. Clear.

27. (27L. ) A strong internal septum divides the canal equally. The internal orifice of the posterior canal is much larger. A wide postero-lateral opening into the posterior condyloid canal and a large superior diverticulum opening posteriorly and internally. The internal opening of the anterior canal has a sharp spine from the antem-niedial wall with a deep cleft below. A superior diverticuliun from the common external opening.

28. Clear.

29. Clear Slight antero-medial constriction.

jO. .\ septum in the outer third or half of the canal. The posterior canal is

half the anterior.

31. Antero-medial or -superior notch.

32. Large. .Vntero-medial notch. Posterior condyloid canal has ot>ening into postero-lateral wall.

33. Opposed spurs. Anterior division is half or third the size of the posterior.

34. Antero-suijerior spur.

35. (35L.) Internally an antero-superior spur has bridged the canal completely, cutting off a small antero-sujicrior canal which has a median superior spur that makes it I'-shaix^d.

3)>. Hough sujierior margin. Anterior constriction. Two small canals from postero-supcrior wall probably to the posterior condyloid canal.

37. Clear.

38. Clear.

39. Antero-suiK-rior notch. Canal from postero-supcrior internal margin probably to posterior condyloiil canal.

40. Clear.

41. Median and anterior superior notches.

42. Clear.


43. Clciir. •M. Clear.

45. Internal .siiprrior iiinreinal spur, l)ctwccii iniddle and postprior thirds.

46. Clear.

47. .Vnlero-sniM'rior spur.

48. .Aiitcro-iiu'dial tiit)ercle.

40. \ superior spine and an oppiwed inferior elevation. .Vlniost complete division, so that the posterior division is one and half times the size of (he anterior division.

o(l. .\iitero-superior internal noteh includes a small spine and opposed spine.s in the middle of the canal dividing it equally into postcro-superior and anteroinferior parts.

51. .\ntero-8uperior notch.

52. Opposed spurs. .Vnterior canal smaller.

53. Clear.

54 Median superior notch.

55. Strong median septimi.

56. Opposed spines — sujierior longer. Anterior canal smaller.

57. Clear.

58. Clear. Only one hole in dura mater.

59. Large. Strong slightly anterior septum in middle third of canal.

Fifty-nine right halves

1. .\nlero-sui>erior cleft.

2. Thin septum which is incomplete at the inner opening. .Vnterior canal «maller with further anterior notch.

3. Flattened from ahove down. .\n indicated spur in the middle of the su|)crior internal margin. .Vnterior half of orifice especially flattened. P^xternal orifice rounded.

4. Superior median elevation and an anterior superior spur with an opposed inferior elevation.

5. Narrow septum. .Vnterior canal smaller.

6. Clear.

7. Clear.

8. Hough superior margin. Klevation between anterior and middle thirds of superior internal margin.

9. .Vntero-superior notch. 10. Constricted anteriorly.

U. Strong septum with superior internal spur. Kqual canals.

12. Clear.

13. Snmll antero-superior cleft.

14. Opposed spurs which seiKirate the anterior third of the canal. In addition there is an antero-medial spur.

15. Clear.

16. Greatly flattened. Median spur within the canal on its roof.

17. Clear. Klatteneil.

14'J 1). i.llAAK

IS. (ISH.i I^irKt' nv!il iiiiuT iirilico. Al llir oiitor orifico ii thin strand of bono divide!* off a Hnmllcr posterior diviHiun. Postcro-liitornl to this is a pocket opening liack and inward with a round liolo 1 inin. in diameter in the middle which ofH'ns through into the externul orilice and a pit 'J mm. in diameter ahove, which opens anteriorly into a large superior diverticulum from the anterior canal.

10. Median suiierinr spine.

20. Clear. Large canal from posterior wall to posterior condyloid canal.

21. Clear.

22. Clear.

2;l. .\iitero-superior spur and notch.

2-1. Strong st'ptum. .Vnterior canal smaller. Common external opening has 8up«'rior diverticulum.

25. Slight spvir internally on upper third of antero-mediul margin.

26. Clear.

27. Superior spur on inner margin posterior division half the size of the anterior.

28. Clear.

29. .\ntero-8ui)crior spur and notch.

30. Clear. Foramen magnimi and condyle malformed.

31. Clear.

.32. Large internal opening clear. .\ canal starts off from the middle of the postero-sujx'rior wall o|x"ning separately at outer opening. .\n antcro-medial notch. Antero-supcrior diverticulum in middle of canal.

33. Clear.

34. Slight opposed elevations anteriorly on inner margin. Inferior elevation stronger.

.'tS. I.4irge. Clear.

35. Large. Opposed spurs between anterior and middle thirds.

37. Clear.

38. Large. Antero-superior spur.

39. Slight anterior superior eminence between anterior and middle thirds. 4fl. Clear.

41. Clear.

42. Opposed elevations and dried remains of a connective tissue septum. Anterior canal half posterior.

43. Slight superior spur lietween anterior and middle thirds of internal m.'irgin.

44. Merlinn and anterior superior spurs with slight opposed inferior ridges. A snial! canal from anterior superior corner returns by a curved course to superior wall near external orifice.

45. (IV.) .Vntero-superior notch.

46. Two su|H>rior internal spurs equally spaced with an anterior medial spur opposed to the anterior superior spur. .\ canal from posterior wall probably to posterior condyloid i-anal.

47. .Vn anterir)r constriction.

48. Snudl 1 to 1.5 mm. roiiiiil nolch on antero-medial internal margin which is set off by two spurs.

49. /Vntero-suiierior nolch.



50. .\Icili;in 8ii|MTi(ir lulnTcle.

51. (51R.I Strorin liDrizontal septum. SinallLT infi-rior, liirRt'r >iii|M.Tior ciinnl.

52. Larue, f'lcar.

53. Septuin in internal part of canal not quite to inner orifice. .Vntcrior cunul half the size of the posterior.

54. A complete thin, 2 mm. wide, superficial infernal septum. .Vntcrior canal slif^htly smaller.

55. Two equally spaced superior internal tulxTcles. .Vntero-infcrior spur opposes anterior one.

an. .\nteriorly constricted internal opening.

57. Clear.

58. Clear.

59. Clear

Althougli the uuniber of skulls from different gioujis or types was too small for a comprehensive Aaew, the above table indicates a variation in the frequency of division among the different racial tyi)es. This was also found to be true iiy Cartels ('04). Complete di\'ision shows itself in the Illinois Indians in only 12.9 per cent of the cases, while in the California Indians it was seen in 22.4 per cent. In the dissectmg room si)eeimens and in the Eurojjean skulls the incidence was 13.4 per cent and 11.4 per cent respectively.

Distinct differences between the right and left sides e.xist. Complete di^■ision is more frequent on the left side — 10.99 per cent as against 11.84 per cent — while indicated di^^sion is more common on the right — 38.82 per cent as compared with 33.99 per cent. The canal is clear in an equal numlier of cases on each side. Racial differences seem to l)e indicated here, too. In the Illinois IntUan skulls and in those from our dissecting room there is ])ractically no difference between the two sides, but in the Cahforuia Indians anil in the European skulls decided differences were found.

The canal is clear in an equal unmber of cases on both sides This agrees very well with Weigner's CM) who found the canal clear in 02.1 per cent of the skulls on the left and in .")9.2 per cent on the right. I find 49.02 per cent on the left and 49.34 per cent on the right. Weigner found indicated division in 19.4 per cent on the left ami in 18..") per cent on the right — praeti

144 KAM'M D. l.II.I.tK

(•ally in qcihuI nuinbors in 103 skulls. I (ind 3.^.99 per cent and 3S.S2 per cent for the left and right sides respectively. Complote division of the canal was present on the left iu 1.S.5 per rent and on the right iu 2'2.'.i p<>r cent of Weigner's specimens, hut in only 17.0 per cent oii tiic left and in 1 l.S per cent on the right side in my series.

Freciuently in of complete or indicated di\asions the two canals are unequal, one being about twice the size of the other. The anterior usually is the smaller. Indications of further division of the larger of the two canals occurs in a few cases. The two canals are often equal in size. In the latter case some few examples of further indicated di\'ision of the anterior canal were seen.

From the variations in form of the hyjjoglossal canal in adult skulls I am unable to decide whether there were three or four h>7)()glossal canals developmentally.

I take pleasure in thanking Professor Meyer for his assistance and suggestions.

LiTKK.vniU': citi:d

Bahtkus, I'. 1904 Ul>i'r U!is.sonuntpr.s<'hcidc am Srhiidcl I. Iiiterti .MoiLsi-lir. f. .Vnat. u. Physiol. Bd. 21.

Bremer, J. L. 190S .\herraiit roots and braiiclu'S of the abdiircnt and hypoglossal nerves. Jour. Conip. Xeur. Psychol., vol. 18.

Frokiep 1911 Poirier et C'harpy. Tr. d'anat. hum. Tome 1. Paris.

Jaboclay et LrcY 1911 Poirier ct Charpy. Tr. tl'anat. hum. Tome 1, Paris.

Martin 1891 Die EntwickehinR der neunten bis zwoiften Kopfnerven l>ei der Katze. .\nat. .\nz. Ud. (>.

McMl'RRlCH, J. P. 1905 The development of the human body. Phila.

Poirier et Charpv 1911 Trait*- d'.\natomie humaine. Tome 1, I'aris., C. W. 1910 The development of the h.vpoglossal K'tn^li" of P'K embryos. Jour. Com]). Xeur. Psychol., vol. 20.

ScHAFER ANi> Symington 190!t (^uain's .Vnat.. vol. 3, pi. 2. lllhed. London.

Graf vox Spek 1.S96 Bardeleben Ilandb. d. Anal. d. Menschen. Bd. 1, Abfh. 2.

WeiCiNER, K. 1911 Clier die .Vs.similalion des .Vila.-) und iiber die Variatiunen am Os occipitale beim .\Ionsehen. .\iiat. IJefte Bd. A'y.



MAHV T. IIAUMAN Kansas State Agricultural College'



There soems to be a lack of unity among authors as to what the phenonuMion of superfetation really is. Marshall ClO) defines superfetation as a condition in which "foetuses of different ages may be present in the same uterus."' He does not say that a second coition is necessarj' to produce this condition, but from his statement he imphes that it is. He says: "If oA-uIation takes place during pregnane}-, and if, owing to the occurrence of coition the o\a become fertilized, the phenomenon of superfetation maj' take place." At least, he makes it perfectly plain that the eggs belong to different periods of o\'ulation.

McMumch ('13), suggests that many cases which have been reported as superfetation are due to differences of nutrition, and are the result of the simultaneous fertilization of different ova. He says that it is not impossible for a second o^'^ml to be fertiUzed as the result of a coition at an appreciable inter\al of time after the first ovum has started upon its development. Howe\er, the difference between the normal emlirj'o and that due to superfetation is comparatively small, pro\'ided that the nutrition to both embryos are the same. He further states that for physiological reasons, the jiassage of t\w sjierniatozoiui to the egg is very soon impossible.

Longley ('10 and 'ID finds that coition is neces.>;ar\- to o\-ulatioii in the cat, and jiarticularly to maturation; therefore, in

' ('t)iitributioii from llic Zoulugiuul Laboriitorv, Kaiisiis ."^talc AKricultiiral C'dllcRC, No. 15.


14t> M AlC^ 1. 11 ARM A.N

or(l(>r to have suporfctation, it is necpssarv to have a second coition.

AiTowsniith {'34) gives an account of a number of cases of su])erfetation and suggests that it is impossible that they could be the result of the simultaneous fertilization of (HfTerent ova.

lionner (•')) classifies the reported cases of superfetation in man in tin-ee groujis: 1; His first grouj) is composed of those cases in which two mature children are Ijorn at the same time but bear marks of different jiarents. This grou]) of cases, of course, is due to ditTerent coitions, but not necessarily due to dilTerent jierioils of cn-iilation. In fact, they could not be due to very different periods of ovulation and be born at the same time in the same degree of development. 2) His second gi-oup includes twins liorn at difTerent times. He would saj' tliat these are the result of difTerences in development. 3) His third group, and the group which he would consider true cases of superfetation are those in which embryos are born at intervals too .small for a second conception. He cites a number of cases in the human, ranging in inters'als from two and one-half months to five and one-half months.

King ('13) distinguishes between what she considers superfetation and superfccundation. She defines superfecundation as the fertilization by successive matings of ova belonging to the same period of ovulation; and superfetation as the fertilization of ova of different periods of ovulations, followed by copulation occurring during pregnancy wliich leads to the simultaneous develojiment of two sets of ova. From King's definition, it is seen, that she would confine the term superfetation to those cases in which there has not only been a second coition but also a second period of ovulation.

Sumner ('!(») believes that superfetation may be the result of a .second coition, but not necessarily so. He thinks that the spermatozoa (in case of mice) may be retained in the uterine pa.ssage of the female for a length of time and still be functional, and that true casc-^ nf superfetation may be the result of a single coition.


Schultzf ('()(■>) (1()( s not say that suporfetation is iinpossihle; hut he explains all of tho cases reported in man up to that time on tiic l)asis of either the death or the retardation of the less advanc(tl fetus. Tiiis also seems to he the jxisition taken by Kuntz ('10). He states that the "two cases of apparent superfetation in cats which were examined hy the writer, afforded no evidence in favor of the occim-enee of superfetation."

Clodlewski ('14) believes that the process of suixrfetation is a physioiofricai impossibility.

It will be seen from the above that there is a pi-eat. difference of opinion as to what superfetation is; but all agree, at least, in one particular, that is, that embryos of different degrees of development are in the uterus at the same time. The case of the cat, which the writer is al)out to report, has also this characteristic in connnon with the other cases. There is a decided difference in tiic degi'ee of development in the embrj'os, and neither set shows lack of nourishment, if the blood supply may be taken as a criterion, or any dc gi'ce of decomposition.

The writer will use the word 'superfetation' to mean that condition in which the uterus contains embryos of different degrees of de^•e]opment. This condition may be the result of a second coition or a second concejjtion may have taken ])lac(' without a second coition.


During the dissection of the cats in the class work of general zoology, a marked difference in the size of the enlargeil ])laces of the uterus was noticed in one of them. This leil to an examination of the different embryos which revealed that three embryos, were developed near to term, and one apparently normal, was much smallc r and showed a much less degree of development. Nothing is known concerning the periods of heat or time of coition.

Figure 1 shows the gravid uterus about one-half natural size. The enlargements. A, R, C, and T). contain embryos but no trace of an ernbryo could b(> fouiul in enlarg(Mnent K. The blood ves.sels had been injected and the blood sujijily to tho uterus


148 MAUY T. llAltMW

is plainly shown. The supply to the smaller oiilarRoiiiPiits seems to he eciually as good as to the larper ones, if the size of the blood ves.sels may he taken as an iiulic-ation.

The right horn of the uterus contained one of the larger emhryos and the smaller one. The larger emhryo was next to the P'alloiMan tube wliile the smaller one was near the vagina. The left horn of the uterus had three enlargements, two of which contained embrj'os of the same degree of development as the larger one in the right horn. No trace of an embryo was found in the enlargement next to the vagina, although the fetal membranes extended into it and there also seemed to be an abundant blood supply.

The large embryos (fig. 2) are about DU nun. long, not including the tail, and from the external features appear to be near to term. The limbs are well formed and normal, ha\-ing joints, and on the ends of the digits are claws. The whole surface of the skin is covered with pits, althougli there is very little hair. Th(> nose and mouth resemble the nose and mouth of a newly born kitten and the external ears are very prominent. The tail is more than one-third the length of the remainder of the body. The body is closed ventrally. The one represented in the drawing differs from the other large embryos in that a short distance from the body the umbilical cord divides, and a part passes around either side of the body to connect with the placenta. The fetal membranes fill the entire enlargement of the uterus and fit very closely to its walls.

The snuill embryo which to all appearances shows no decompo.sition, is only 10 mm. in length. The umbilical cord occupies about one-sixth of the ventral surface. The limbs, fore and hind, are merely buds. The tail is about one-fifth the length of the remainder of the body. There are no indications of hair follicles. The mouth is in the process of formation. The mandibular ]) have met in the midventral line, hut they extend only about one-third as far as the maxillary processes. The lip gi-oove is shallow, in fact, there is merely the beginning of the separation of the lips and cheeks from the jaw proper. The oral pit is yet rectangular in shape. There is no indication


of eyelids; hut tho eyes are plainly visiMe from the outside as small dark splieres.

It seems that the smaller embryo is at a stage of not more than throe weeks' development, while the larger embryos represent seven or possibh' eight weeks' development.

Owing to the fact that the material was presented merely for the examination of thegross anatomy of the mother, the tissue is not favorable for microscopie study, but the conditions for the less athanceil fetus are the same as those for the larger ones. The fetal membranes of the one is in as good a state of preservation as those of tlie other. Both the anmion and the chorion of the smaller embryo has the appearance of being younger than they do in the case of tlie larger embryos, but they have no greater e\'idence of necrosis. The injection of the blood system of the mother l>efore jireservation has made jiossible a better study of tlie blood supjily to the embryos than could have been made otherwise. If the smaller embrjo should have been the result of an early death or retardation of development, this is not evidenced by any lack of blood supply.

Altliough, as was stated before, nothing is known concerning the periods of heat and time or times of coition, yet it seems that the condition of the fetal membranes of both the smaller and tlie larger embryos, the abuiulaiit blood supply to the smaller embryo as well as the lack of decomposition are e^^dences that it is not a result of early embryonic death, but that it is a result of a later conception.


The explanations of reported cases of superfetation are almost as numerous as the authors who have rei)orted them. There are those who do not believe that the supernumerary brood or accessory birth, if the writer may use that term, is (Schultze, '60) or can be (Clodlewski, '14) a result of concejition during pregnancy. As has been prexiously stated, Schultze (in the human) considers these cases as twins in which the one has been retained in the uterus for a longer jieriod than the other. He exjiluins the dilTerenee in the degiee of deveIoi)ment on tly? basis of lack


of f<Mi(l supply, tlunforc, (l('V('lnj)mcul is retarded. If this ho true of the cases reported in the human the gestation )>( riod lias been prolonged as niueh as five and one-half months. So far as the author is aware, this is a much longer period than i.s recorded for prolonged pregnancies. Furthermore, if this explanation should be applied to the cases of the mouse as rrporl(d l)y Sunnier ("Ki) the period of gestation would l)e almost three tinus the normal iKriod of gestation. .\t liast, tlie ])( riod from the birth of the last normal litter until the birth of the suptrninnc rary litter in one ease was as nmch as\en days, which, if adtled to the period of gestation for the normal litter which is twenty-two days, would be fifty-nine days. This lacks only one week of being three times the normal period of gestation. It might lie added that the male was removed from the nest !)efor(' the birtli of the lirst litter. The writer, in giving extremes, would not i)resume to say that such a phenomenon is impossible; but it seems that the e\ndence is scarcely enough to justify the belii f in such an extended prolongation of the time of gestation.

Another explanation of the cause of superfetation which has been freeiuently given is that the organs of reiiroductinn are abnormal, such as a Intid uterus, etc. AiTowsmitli and others have shown that in a number of cases, at least, this could not aceoimt for the phenomenon. In the case of the sheep as reported by the above mentioned author, an examination of the slaughtered animal revealed no abnormality. While this explanation may soem perfectly plausible, it does not always agree with fact, and coidd not be accepted as a universal cause.

Whether or not ovulation takes place during pngnancy, has been a subject for much discussion and investigation. It is the general belief, one is safe in saying, that ordinarily during pregnancy, in man, oAiilation but if the condition of corpora lutea may be a safe criterion (McMunich '\'-h this is not always true. Chinstophcr ('8(5) reports having found a cat far advanced in imgnancy with four mature flraafian follichs each containing a mature ovule. There is also some evidence that ovulation may occur during pregnancy in the horse.


According to Marshall CIO), McMurrich ('13), Bonnar ('65), Arrowsinith (':U), King ('13), Hcrzog ('98), and Sinnncr f'16), supcrfctation may he caused by the fertilization of an egg ovulated during pregnancy. Marshall, McMurrich, and King further state tliat the plienonienon is also ilue to a second coition. Sumner suggests that perhaps there is a periodicity of ovulation which may not be interrupted by gestation, and that the spermatozoa may be retained in the reproductive organs of the female for a consideral)le length of time, and yet remain functional. His data show that either this must be true, or that the perif)d of gestation is greatly prolonged, even more than doubled. In one the last ojiportunity for copulation was nineteen days before the birth of the first litter, and forty-seven days before the birth of the second litter. Since spermatozoa have been found alive and acti\-e in tlie female reproductive organs some time after coition, it seems as reasonal)le to i)eliev(> that these spermatozoa would function as that development would be prolonged to such an extent.

However, in the case of the cat described by the wTiter, another difficulty arises. The mature embryo is next to the vagina, and in the same horn of the uterus between it and the Fallopian tube, is an embryo almost at maturity. While one would scarcely be justified in sajnug that the passage of an egg from the ovary to this po.sition in the uterus under these conditions, is imjiossible, yet it hardly seems probal)le, for the fetal membranes are in very close contact with the uterine walls.

King suggests, in case of her rats, that perha])s the ovaries I'xuicliimed iuilependently, and that the first litter was the result of the ()\ul:ition of one ovarj' and the second litter the result of the ovulation of the other ovarj'. So far as the data given in her paper shows, this is merely a suggested jiossibility. No structures were examined to sui)i)ort the theory.

In the caseof the cat, this would be equallyasimprobableasthat the egg was fertilized by sjKrmatozoa which have lieen retained in the uterus of the mother for, as has been stated before, in the same horn of the uterus between the vagina and the small emhrvo was an embrvo which was near to term.


'I'lic writer would raise tlie (|ue.sti<>ii whctlur it would not lie as rrasoimhlp to beliovo that the egji ininht be retained in the uterus some time before it was fertiHzed and yet bo able to develop into an embryo, as that the si)erniatozoa eould function some time after being diseharged. The evidence seems to point to a good blood supjily to the smaller embryo, and the implantation seems as good as the more ailvanced eml)rj'os. If it is small on account of retardation of develojiment, some other cause for this arrested (levelo])ment must be sought. If the egg whicli gave rise to this embryo is of a much later ])('riod of oNiilation than the eggs which gave rise to the other embryos, there is the mechanical difficulty of the passage of the egg from the ovary to the position which the embn'o occupies in the uterus. This is tnic whether or not the egg was fertilized by a spermatozoon from a second coition, or from a spermatozoon which has been retained in the uterus for some time.

It seems reasonable to the writer that the different cases of superfetation may l)e the result of dift'erent causes and that it is hardly necessary to ascribe all cases to the same cause. C'onsitlering all the conditions the writer would suggest that the less developed fetus could be accounted for as a result of delayed fertilization.


1. The word 'superfetation' has been used to mean that condition in which the uterus contains embryos of different degrees of development. This condition may be the results of a second coition or a second conception may have taken ])lace without a second coition.

2. .Ml hough superfetation is rare and abnormal, many cases have been reported in man as well as in a number of other mammals, which do not seem to be satisfactorily explained in any other way than that a second conception has taken place.

3. Perhaps not all cases of superfetation may be attributed to the same cause.


4. In the caso of tho rat doscribod in this paper,' it seems as reasonalile to tliiiik of tlie less advaiieed eiiihryo as the result of delayed fertilization as to account for it on the ground of delayed development oi' a scconil coition.


Arrowsmith, R. 1S;W .S\iperfetafion (?) in the sheep. London Med. Gazette,

vol. 14. Bonnah. Okohgk Lindsay 1S0.5 .\ critical in(|uiry of superfctiition, with cases.

KdinbtirRh -Mi-d. .J<mr., vol. 10, p;irl 2. CliRlsToPHKK, W. .S. lS,S(i Uvuhilion during proRnancy. .\mer. Jour. Oli stetrics, vol. 19. Herzoo, Maximillia.n 1898 Supcrfetation In the liuniun race. Chicago .Med.

Recorder, vol. 15. Kino, Helk.n 1913 Some anomalies in the gestation of the all>ino rat

(Mus NorwpRiciis albinus). Biol. Bull., vol. 24. KiuKiiAM, \V. B. 1907 The maturation of the e(jf; of the white mouse and rat.

Biol. Bull., vol. 1.1

1910 Ovulations in nianiinals with special reference to the mouse and rat. Biol. Bull., vol. IS.

KiNTz, .Vlhkht 19l(i A note on supcrfetation. Interstate Med. Journ., vol.

23, no. 3. Lo.N'GLEY, VV. H. 1910 Factors which influence the maturation of the egg and

ovulation in the domestic cat. Science, X. S., vol. 31, no. 79.5.

1911 The maturation of the egg and ovulation in the domestic cat. Am. Jour. Anat., vol. 12, no. 2.

Marshall, Fra.ncis H. A. 1910 The physiology of reproduction. Longmans.

Cireon and Co., New York. McMurrich, J. Plavkair 1913 The development of the human body. P.

Blakiston's .Son and Co., Philadplphia. .SiMNKH, V. B. 1910 Notes on the supcrfetation and deferred fertilization

among mice. Biol. Bull., vol. 30. ScHi'LTZE, B. S. 18(56 I'cber Superfoccundation und Supcrfoetation. Jen aischc Zeitschrift fur Medicine und Xatur wissenschaft, Bd. 2.

'The author wishes to express her indebtedness to Mr. Wallace Park, .\ssislant in Zoology, for calling her attention to this case.



The prognaiit uterus of tlic cat, one-half natural size. EnlarRements A, C, and I) rontain embryos IH) mm. in length. Enlargement D contains the embryo shown in figure 3. Enlargement E does not contain an embryo.



MAItr T. llAltMAN





The embryo of enlarpeiiicnt D, fiKiire 1, two-thirds nntural size. The embryo is surrounded by the fetal membranes and a portion of the uterus. A part of the umbilical cord extends around one side of the body and a part of it extends around the other.

The eml)ryo of enlargement B. figure 1. enlarged 5 diameters. The embryo is surrounded by the fetal membranes and a portion of the uterus.






AL'TIKMt- *ir*TH*cr fiV TIIW PAPER OMUKD BY Till; hiiii.UMiKM'llir skhvkK AUGUST It.


H. S. BURR From the Anatomical Laboratory of llie Yale School of Medicine


The electric f liermo-refjiilalor liere (le.-<cril)efl is the result of several years (if sixiradic expeiiiiients. The object was to design an insiruiiieiit tiiat could I)e readily huili by any laboratory technician; that coulil be run on an ordinary liKhiiuK circuit without the use of a relay, .condensor or other apparatus of that sort; that would bo more positive and less subject tfi tlucluations due to jar than the tnorcury cohniin; and finally that would be more accurate than a bi-metallic renulator. It is Ix'lieved that this instrument fulfills many of these re(iuirenir>nts.

The principle is simple. It consists of a walking beam actuated by an (>ther taml)our. The expansion of the taml)our forces down the reach of the walkinj; beam so that the consec|uent raising of the other reach breaks the electric circuit. Finure 1 gives the details of its construction. The whole apparatus should be built of brass with the exception of the .stool pin which acts as the Iwaring of the walking beam, anil the platinmn-iridium contact points which are forced under pressure into the i-nds of the walking beam and adjiustinont screw. The dimensions given may be varieil to suit the convenience of the material at liand although the measurements given have lieen found to be the most satisfactory.

The most difficult pail of the construction is the ether tambour. While it is possible to make ii in a well equipped shop it will be found more satisfactory to purchase a thermostat manufactured by the Banner Incubator Company, Baraboo, Wisconsin or a Newiown thermostat manufacttired by the (!iant Incul)ator Company. .\ good mechanic, however, can spin a tambour i)rovided he has a sjieed lathe available. The method is as follows: In the end grain of a hard wood block a series of concentric giooves and ridges of semi-circular cros,s section are worked. Then by fastening a slit>ot of thin soft bras,s against the wood block, the whole being fastened to the face plate of the lathe, the soft can be worked into the grooves. Two plates are crimped in this manner, cut into circular disks slightly larger than the diameter of the tambour and soldered gas tight over a spreader pl.'ite.



H. s. mint

Tlic :is,s('iiil>l('ii inslniiucnt slmiild lie siipijoilcd iipsidc down in the fonslanl t(Mii|M'i;ituif Imx or inoni, thai is to say, it siiould lie fastened to tlio sidfwall with llio wooden liase up and the instrunienl lian^inn down so tlnit K'iivity will ke<'p tlie loan reach of the walivinfi Ix-ani pressed against the contact ailjustinent screw. The expansion of the ether tlien depr(>sses the shoU arm, raises the lonK arm and hreaks the circtiit. Two adjustments are provid(>d so tiiat any tcniiierature witiiin the ranp»' of the lieatinjr element may l>e maintained.

Almost an>- electric heating ai)paratus from simple batteries of carl)on lights to an electric ra(li;itor may he used with complete satisfaction. Where large sjiaces are to ho heated it will he found l)e.'<t to have sufficient heat on all the time to keep the temperature up to

\OooJg 7 Boae


\od,oror '^a\^^_^^^'^\{

witliin 5° of the rtiiuired temperature, the regulator being then cut in on a circuit with a sufficient heating element to readily raise the temperature the remaining .5°.

In installing this instrument the most satisfactory result will be obtained if the box or room is thoroughly insulated from the surrounding temiK-ratuif. For ordinary paratliiic work in the class room or research laboratory, a light, thick-walleil wooden box serves admirably. Such an oven with inside dimensions of 12 by 12 by 18 inches is especially adapted for parafline work since a temiK>ralure of .54° on a shelf placeil just above the heating element gives a temp<'rature on a shelf () inches higher that will flatten our paraffine .sections and dry them. A small room may be kept at a constant temperature provide11 iii.siil:tt('(| Willis :iiiy jrivcn space cun Ik- initintuintKl at a constant t^-nipoiatuii' with lcs,s than 0.5° of variation. Little work is rec|uir('<l to kt'c[) the ajjparatus in ^ood working order: it is only necessary at intervals to dean the platinum contacts.

The author is indeliled t.o Mr. \. K. ScharfT, Laboratory Technician of the .Vnatoinical Laboratory of tlie Vale Scliool of Medicine for help and many supne.stions in making tliis instrument.



Those who have had to dciJond upon the ordinary means for applj-ing air pressure for injectinK such iis n hand or water pump, recoKniz(^ the difficulty of accurately regulating the pressure or maintaining known constant i)re.ssure over any length of time. The e.^sential feature of an apparatus used in this lal)oratory for several years which accomplishes the above object is the automatic air pre.-ssure valve described below (fig. 2).

The valve consists of a piston working against a spring within a cylinder and connected with the lever of an air cock. The air from the reservoir, under high pressure, is led through the air cock to the rear of the cylinder ;ind forces the piston outward, compre.s.sing the spring, whicii ])ai-lially shuts off the supply of air through its connection with the lever of the cock.

The iiressure valve was made from a steam radiator air valve. The base is unscrewed from the bowl, the parts within removwi and a spring and piston in.serted into the bowl. Tlic spring is of Xo. 14 wire, coiled J inch in diameter and Ij inches in length. The piston is of leather, fa-stentnl to the pi.ston rod l)y a nut al)ove and below it. The piston rod is of i inch round brass, '.i\ inches in length, threaded for § inch at the end to which the pi.ston is attached and for 1] indues at the other end, which is fa.<tened to the connecting nxl. The vent of the air valve is enlargeil to allow the piston roil fri-e i)lay. A connecting rod of Yi ifl' round brass. 2] inches in lengtli, with a shar]> right angle bend 1 inch long at (>ach end, is fa.stened to the pislon rod by a link at one end and to the lever of the cock at the other. Both ends are burred to hold them in position. The link is made of ^ by f by 1 inch in size, b(>nt in the mifldle at an angle of 00°. A hole in one end admits the <-onnecting rod, allowing (rev sideways motion. Through a holi' in the other end, the |)iston roil is ]>as.s('d and clamped by a thumb nut on each side. Hy shifting these thumb sitcws along the piston rod, the distance between the piston and the lever of the air cock may be lengthened or shorteniHl, thereby changing the tension on the spring within the valv<> which thus decreases or increases the pressure.


The pi|H' convfviitK tin- :iir from tlic rosciNoir lo the r yliiitlcr is lc<l to 11 Y lulii'. Olio hniiicli of wliich loads through a low pressure reservoir (any larpe hottle, wliich acts as an air cushion on the mercury) to a l'-nianon»eter graduated to inillimetei-s and indicating the pre.ssurc beyond the valve. The other hraiich of the Y tuiie leads to the receptacle contaiiiinn the injectiiiu fluid. .\s the jiressure in the cylinder is rcduct'd liy the decrease in the amount of injecting fluid, the spring is released pulling on the lever of the air cock, opening it and a<lniitting a fresh supply of air. This maintains a balance between the air pressure in the eyjinder and a tciisinn on llie spring.

Ill .'. I.AIioKAlDlJV ll.\ll\<, CLOCK

.\U(ir.STi;s F SCHAUFF

The device ilescribed below is of gi-eat value in niea.suriiig the length of time that pieces of tis.sue or slides are to he left in the dilTerent reagent»s. It is the experience of every laboratory worker that either his entire time must be devoted to the staining of slides or great variations in the length of time in which they remain in the different reagents are almost sure to follow divided attention. The clock a?i described below has Invn used in this laboratory for over a year and it has been found that much time may be saved in .staining sections and great^er accuracy in their treatment obtained from its use.

The works of an ordinary clock are reniove<l from the case. With a number 28 drill make a hole through the face, just above the twelve o'clock mark and about J inch from the end of the minute hand. Through this hole a piece of insulated annunciator wire ti inches in length is passed. Strip the insulation for about | inch from the end of the wire and bend the bared poi-tion in the form of a flat looj) so that (he minute hand in pa.-ising will rub over it. The insulation not be removefl from that jiortion of the wire which passes through the liole in the clock face. \Vr;i|) the end of thi-; wire behind the clock face, around one of the cross stays of the works and leave the end free.

Drill a hole through the sliell of the clock through which the loose end of the anmmciator wire may be passed. If the clock used is an alarm clock, the clap|>er may be removed and the wire passed out through the clapjier slit. Place the works within the shell.

Fasten one wire of a 2 or 3 foot piece of double, flexible, electric hght cord to the end of the annunciator wire from which the insulation has been stripfwd f<ir the purpos<>. Wrap the union with tape. Fa.sten the stripiK'd end of the other wire to any convenient |)ortion of the clock. Fasten one wire from the other end of the flexible cord to a buzzer or bell, the other two to dry cells. Complete the circuit, interposing a simple switch so that the circuit may be broken at will.

For convenience in turning the minute hand a piece of bnuss tubing may Iw fa.stened to the thumb screw controlling it.


To uso tlio clock: wind it, turn tho niinutc hand to the (icsircd number of minutes hoforo the oven hour that it is dosirod to Icavo thr slides in the stain. Place slides in stain, (" the switch. When the time hits elapsed, liie minute hand connectinn the contact with the wire looj) will sound the i>uzz('r, which may he stoppe<l by openinft the switch. IJemove th(> slides from the stain and reset the clock for the next slatje. In the intervals Ix-twe^Mi the ac-tual proces.sos of setting the clock and chaniring the slides from one reagent to another, the mind is left free for other things.

By fa-stening a piece of sheet metal to the bottom of the shell of the clock so that it will project backward just behind the clock anrl fastening a strip of metal to the table so that the projection on the bottom of the clock may be slipped under it, the clock is rendered more stable.

TUC ANATtllllCAL BCCOltD, VOL. 13. NO. 3




In rharf/r of anatomical instruction. Dental Department of Washington University,

St, Louis

The clwiiigcs ill iioiiu'iii'laturi^ sugROslcd in tlic followiiij; notes are bused on their iilleped iinu'tical importance. The names ()l)j(H-ted to in my opinion tend to lie a source of confusion to tlie stu<lent and jiractitiuncr, wiiiic tiie sul)stitu1es arc olTered as a step towards clarity and simplicity in anatomical nomciK-lature. All the names olijectisl to liave IxH'n accepted in the H. X. A. revision. I hav(> not found tlu'se aliened sources of c()nfusion commenteil upon in our texts, reference hook:' or literature. Many of tlic-^e notes are of such a nature that they must ha%'e been considered liy others, while some may repres^'nt previously undetected enors. Hut any c|uestion of prioritj- is aside from my argument, which is tliat our reference books should at least point out thes(> soiu'ces of confusion. As a matter of convenience these notes will be arranged in the secpience of the R. X. .\.

Ostculoijia. Inst<'ad of os ilium, os isciiii.and os |)ubis the official names shoulil be pars ilium (o.ssis coxae), pars ischium (o.ssis coxae) and pais puljis (o.<sis coxae). The practice of speaking of part of a bone as a separate bone confuses students, so that eventually they are uncertain as to what constitutes a bone.

Introduce logic into the use of the terms tuberculum, tuberositas and tuber — that is use tuberculum for small projections, tuberositas for medium sized projections and tuber for large projections. The present haphazard arrangement, as for example tuberculum humeri as contra.-ited with tubeiositas unguicularis, is a severe tax on the memory without a compensating benefit.

The ofhcial term diploe might be dro|)ped entirely with advantage. Students on account of the special name are bound to fii'l that the spong3' layer of the cranium differs essentially from ordinary spongy bone.

Si/ndesmologio. It would be a great convenience to establL-^h an official name for a grou|) of ligaments and then designate jis pars the component subdivisions, as l"or example:

liganientum caraco-claviculare pars trapezoideum pars conoideum



Imu- practical imrposo ilic articiilalin lalo-cnnalis is llic worst utTciidcr, as no olHcial collcctivt> name is given. My siipKcstion would l)c:

liganiontuni coiiatcriilo fii)ialo (talo-oruralis) pars tihio-naviciilaro pars calcanco-liliialc pars lalo-liliialc anicrius pars 1al()-lil)ialc jioslcriiis liganipntuni collatcralc filmlarc (talo-cruralis) |)ars laio-lilmlari' aiitcrius pars taio-filiiilarc postoriiis pars calcauo-lil mlarc

MyoliHlid. Tiu' term septum intormuscularo is used when wiial is meant might he more siccurateiy piirased as seiitum intergi-egare (musculorum)— giegare, derived from grcx, a herd or group. As the suggestion is original no guarantee is ofTercd for the correctne.s.s of the Latin. Tiie term .«<'ptinn internmsculare might then he rcstM-veil for the minor iiartitions ix-tween tiie various nuiscles of one group.

The ()f1i<'ial term fa.-^cia superficialis might with great advantage lie (IropjM'd entirely, substituting the terms tela sulx-utanea adiposa or lianniculus adi|)(>sus ihrnugiiout . Sludeiils arc hound to he confu-^ed iiv tiie of the word fasci;i in conncclion witli a fatty layer, while its continuation in my oi)ini(in serves mi useful purpose, hut .simply iierpetuates the confusion of a past general ion of anatomists.

Stutlents have great diRicuUy in keeping clear the distinction hetween the ligaments, whicii hold down the tendons in the hancl and foot regions, and the ligaments of the neighhoring joints. Their protect liiat tiieir name .'^ound so nuich alike is in my opinion justified. I would therefore suggest that these hands he no longer called ligaments, hut retaining l)an<ls — retinacula. This is already i)arlly the custom, as in retinaculum nmi. peronaeorum suj). vel. inferius.

Spltnichiuilinjin. The (erm fundus vesicae to descrihe what would seem to he more logically named hasis vesicae might well in theory give rise to confasion. Students arc hound to feel that the term fundus vesicae descrihes the glohular upper free surface; of the hiadder. That even the higiily trained specialist jjractitioners have confused the term may he noted in an article hy Dr. Hransford Lewis, .lournal .\merican Medical .\ssociation, vol. (iO, pj). 17()-')-l7(it). I would therefore suggest that th(> term fundus vesicae lie (liscarde<l aii<l that hasis vesicae Ik- suhstituted.

The term ductus deferens is a great improvement over vas deferens, hut inasmuch :us this term might he applied to the duct of any gland, in Ix'half of accuracy I would nuich prefer a specific term, as either ductus testis or tluctus spermaticui?.

Anyiologia. The arteria epigjust rica inferior for obvious rca.sons shoulil he known as the a. rica (paries ahdominis), while the a. e|)igastrica su|)erior woulfl aulomati(;ally hecoine simjily the a.


epiKastricii. A loiin i-^tiililifilu'd inisnoincr, proltalily i-irolcssly luiiiicd b}' surgeons :iccor(liiin lo tlic coriLinon position of the body during abdoniiniil opi-i-jitions.

Tlie :u-tciia iliata coninumis should divide into external and internal subdivisions, (irantcd that this not an ideal nonienclature, it is in my opinion the l)eHt that ran he done and K'C'^'y superior to the pres<'nt eonfusiiiR aiTaiiKeinent.

Discontinue the use of the term plexus to deseriiic a nt'twork of smaller vessels, substituting either rete or some other suitable term. The use of i)lexus to describe both a network of ve.s.sels and a weaving of nerves is both tlu'oretically an<l practically wrong. The restri<'ting of the use of the word plexus to nerves would in my opinion help not onlv stuilents, but also scientists.

Neurnlogia. For the same reason as noted in the ve.s.-el discussion, discontinue the term ranuis aniU'Jtomaticus in connection with nerves, substituting ranuis or ramulus coiunumicans or some other suitable term.

Diminulives. The term glandula to describe macroscopic glands is unfortunate, as the dimiiuitive ending keeps suggesting to the student that the gland is of sub-macroscopic size. I would like to .-^ee the term glans introduced for the macroscopic glaiul.f, retaining glandula for the glanils scarcely or not visible to the unaided eye.

The word ductulus should in my opinion be substituted for ductus in connection with the terms ductus biliferi, ductus sublingualis, ductus jirostatici and ductus sudoriferus.

In conclusion let me as a non-specialist in anatomy again record my enlhusiasm for llie practical value of the R. X. A. revision. The authors of the B. X. \. tlid not put out their list as an ab.solutely fixed and peinianent thing, but specifically calletl for criticisms from those actually using it. I do not advocate that anyone use any of these suggested changes, but I desire to give my views publicity in order that they may receive consideration in any po.s>;ible new revision. In spite of the iuu|uestionable disadvantage of changes in nomenclature, any change which will yield an essential betterment must eventually come into its own. The knowledge that a genera list's point of view has often proven of interest and value to specialists encouniged me throughout this effort .


vr riiK DiiiLiouitApiiic MKitvicK Atrotmr II.


CHARLES EUGENE JOHNSON Department of Animal liiology, University of Minncxota


In a l)rief reference to the embryonic liver in the European ground squiiTel Citellus (Spermophilus) citellus, Voelker ('01) states that in this animal tho organ develops as two hollow outgi'owths of the ventral gut wall, and from these outgrowths solid cords of cells sprout outwards to all sides. Each of the two outpocketiugs opens separately into the foregut. This process, he remarks, does not conespond to the accounts that have been given for other mammals.

Voelker does not state whether or not these two outgrowths are paired structines, Init so far as may be judged from the figure accompanying his account — which, however, is intended for illustration of the pancreas rather than the liver and therefore does not present a favorable \new of the latter organ — they represent uni^aired outpocketings of the ventral gut wall, one situa,ted caudad to and in close association with the other. If this interpretation be correct, the anterior outgrowth probably represents the ]iars hejiatica and the posterior one the pars cystica.

In three other embryos of the same developmental stage as the specimen on which the description just given is based, he found the condition of the liver differing from that first mentioned to this extent, that the two primitive bile ducts are joined in a conunon diverticulum which is an outpocketing of the ventral gut wall. In a more advanced embryo (G mm., greatest length) he found the ductus choledochus entering the gut wall on the right side, this position having been brought ainnit by the rotation of the digestive tube.



.laiiosik I '95). ill his study of I In- tlcvclopmcnt of I lie pancreas in till' same spccit's, cxauiincil cnilnyos ranging in Icngtli troni (i nun. to 3r> nun. Ho relers to the condition of the liver only in the youngest of these embryos and here he finds that it is conipo.sed of two canals, one ilirected toward the left, the other toward the right.

From .lanosfk's meager descripti(Mi it is not jiossihle to know whether these two canals rejm'sent the d(>veloping liver in its entirety, that is, both hepatic anil cystic portions, or, as seems more proliable, the hepatic portion alone.

Although \'oelker gives no measurements for the youngest embiyo examined by him. it is clear from the ajiijcarance of the liver and the pancreas as shown in his illustration that it represents a considerably earlier stage than the one Janoslk describes for his () mm. eml)ryo; also, the condition described in Voelker's more advanced specimens, up to and including the nun. embrj'o, does not agree A\'ith the statement for the G mm. embryo in .lanosik's series. It is therefore doubtful that the two canals noted by .Tanosik correspond to the two hollow outgi-owths observed by \'oelker.

A study of a series of embrj'os of the American ground squirrels Citellus tridecemlineatus and C. franklinii, ranging from 1.5 mm. to 6 mm. in length, shows that in these members of the Genus at least the development of the liver is not essentially iliiTerent from that observed in other mammals. The results of this .study may be briefly summarized as follows.


The earliest stage in my series in which the hepatic anlage is present is a 1.5 mm. specimen. In this embryo the anlage is a thickened area in the ventral wall of the foregut, just in front of the yolk stalk, continuing laterally a short distance onto the .side walls of the digestive tube. The lateral extent of the anlage is greater than the antero-posterior extent, the area representing a transverse rather than a longitudinal thickening in the wall of the foregxit.


In oinhryos hctwoni l.o mm. and 2 mm. in length the hepatic thick( ninf^ Ix-conic.'^ a distinct outpmicliin^ of the gut wall. This outpouching is triangular in shape, boiiig broadest in its posterior i)ortion bordering the yolk stalk and from here l)econiing nanower toward its anterior end. It i.s relatively large and in these and slightly more advanced stages is decidedly asymmetrical \rith respect to right and left, caused by the indentation of its right antero-ventral wall by the ventricular loop of the heart. In surface \iew tlie pouch presents three lobe-like divisions, a right and a left posterior lateral lobe separated by the indented area, and an anterior median lolie Ijing in the angle formed by the \'itello-umbilical trunks as they converge to enter the sinus venosus.

^'ariations in the form of the hepatic pouch occui', depending apparently upon the variable mechanical effect of the developing vitelline veins and vitello-umljilical trunks, the septum transversuni and especially the heart.

In embrj'os about 2.5 mm. in length the hepatic anlage reaches its maximum development as a smooth-walled outpouching of the foregut wall. It is now a large, api)roximately synuuetrical, more or less spindle-shaped pouch situated transversely upon the ventral wall of the foregut with which it is comiected by a short hollow stalk. The enclosed ca\'ity conforms to the general external shape of the pouch but its size is much reduced by the increased thickening of the enclosing walls and in the extreme lateral portions of the diverticulum it has been entirely occluded.

Trabeculao niak(^ their appearance on the surface of the hepatic pouch in embryos but slightly older than the preceding and in specimens al)out 3.5 nun. long the walls of the diverticulum have in large part been transformed into such sprouts.

The hepatic outgrowth in embryos up to about 2.5 mm. or 3 mm. long represents the primarj' hepatic diverticulum or pars hepatica only. But now there ajjpears a thickening in the floor of the foregut caudal to but continuous with the primary outgrowth, and this constitutes the pars cystica. In embr>-os about 3 nun. or 3.5 nmi. long the pars cystica has become a thstinct evagination from the gut wall, occup>'ing the angle between the pars hepatica and the yolk stalk.


The thickciird area forniinp; the aiilaKc of tho pars cystica is from till" first continuous anteriorly with (he i)ars hcpatica and therefore is not a separate area hut rather a caudal extension of tliat tliickened portion of the tul)e which already has given rise to the pars l\epati(a. aiipearinf!; later than the i)ars hepatica for the reason that the part of the gut floor concerned witli its formation' is of later development than (he i)art in front of it which gives rise to the hepatic portion ])roper.

In emhiyos of 3.5 nun. to 4 mm. a further outpushiug of the pars cystica from the digestive tul)e has resulted in tlie formation of the ductus choledochus, which is a short broad tube into which the jxirs hepatica anti pars cystica now open by wide cliannels representing tlie hepatic and cystic ducts respectively, and through which these ducts communicate with the digestive tube. To the ductus choledochus as now constituted the pars hepatica contril)Utes the anterior and the smaller part of the lateral walls while the i)ars cystica contributes the greater ])art of the lateral and all of the posterior wall.

The distal portion of the jiars cystica is at this period distinctly marked off as the gall bladder.

In embrj'os about 5 mm. in length the ductus choledochus has been constricted into a tube diameter is about one-half that of the 4 mm. stage ; the gall bladder is about double its former size, is smooth-walleil, of globular form and has a sharply defined cavity the diameter of which is slightly less than the thickness of its wall.

At the junction of the ductus choledodius with ilu- hepatic and cystic ducts there is a bladder-like dilatation about twothirds the size of the gall bladder. This is a conspicuous structure in each of three 5 mm. sjiecimens in my series and appears to be a point of detail which does not occur in connection with the bile ducts in the second of the two species here concerned.


The early development of the liver in this species agrees in its main features with that of Citellus trideceniliueatus. The chief points of difference are in size and form. In Citellus


fraiikliiiii the oik.'vii is a imirh loss voluminous stnicturo and the pouch-like outf^rowtli, in contrast witli that of the first descrilM'd species, is from its first appearance quite symmetrical and more evenly exjianded, hcinR on account of its much smaller size modified only to a slight extent by encroachment of the heart and other neighboring structures.

No trace of a ventral pancreas was found in any of the stages of the two species examined. The entire absence of a ventralpancreas in the European ground squirrel has been reported by Voelker and Janosfk.

In a 6 mm. embryo the gall bladder and biliarj' ducts correspond clo.sely to the condition in the 5 mm. embryos of C. trideceinlineatus, except that the bladder-like dilatation associated with the bile tubes of that species is here lacking.


Brachet. a. 1S05 Rccherchcs sur le d(^vcloppement du diaphragmp ct du foio chpz Ip l.ipin. Joiirn. do l'.\nat. et do la Physiol., T. 31. 189(1 Itec'hcrchos sur le di'vclopppnient du pancreas et du foie (S^laciens. Reptiles, Mammif^res). Jourti. dc I'Anat. et de la Physiol., T. 32.

Bradley, O. Charnock 1908 .-V contribution to the morphology and development of the mammalian liver. Journ. of .Vnat. and Physiol., vol. 43, London.

Choronshitskv, B. 1909 Die EntsfchungderMilz, Leber, Gallenblase, Bauchspeiclieldriise und des Pfortadersystems bei den verschiedenen Abtheilungen der Wirbelthiero. Anat. Hefte. Bd. 13.

Debevre, a. 1010 Le foie, est-il d'originc endodemiique ou m<^sodennique. Bibliogr. anat., T. 19.

Fei.ix, W. 1S92 Zur Leber und Pankreasenfwicklung. .\rch. f. .\nat. u. Physiol. Anat. .\bth.

UA.M.MAK, ,1. -Vro. 1893 Einige Plattonmodelle zur Bcleuchtung dcr fruhercn embrj-onalen Leberentwicklung. Arch. f. Anat. u. Entw. 1S97 l'el)er einige Hauptziige der ersten embryonalen LcIktchtwicklung. Anat. .Vnz.. Hd. 13.

Hilton, W. A. 19t)l Early morphogenesis and hi.-itogenesis of the liver in Sus scrofa domestica. Studies of the Zool. Lab., I'niv. of Ncbr.. No. .53.

JanOsik, J. 1,89,5 Le pancreas et la rate. Bibliogr. .Vnal., T. 3.

Lewis, V. T. 1911 Die Kntwicklung der Leber. Ilandbuch der Entwicklungsgcscliichte <les Menschen. Kcibel u. Mall.

MiNOT, C. .S. I9I0 Laboratory text-book of embryology. Blakiston's Son and Co., Philadelphia.

Prentiss, C. W. 101.") Toxl-book of embryology. W. B, .Saunders Co., Philadelphia.


SwAK.N, A. IX'Mi Ht'clH-rolii'S sur lo (li'vflop|«'inriil dii fnii', ilii liilic digcHtir.

(le I'lirri^rr-oaviti'- du pi'-retdinc ct du iin'.sciitrri'. Jdiirn. dc I'Aiiiit.

ft delii I'hysiol., T. :t2. Thomi'son, p. 19()8 a note on the dcvelnpnicnl of tlif scpliim tninsvcrHUm

mid the liver. Jmirn. i)f Anal. Jind I'liywiol., vol. 42. London. V0KI.KKU, Dr. 1901 Beitnige zur Kntwickliin); dor Punkrcas hoi den Amniolcn.

Arch. f. Mikr. Anjit., Hd. oil.



1 Antcro-vcntral view of a reconstruction of the primary hepatic diverticulum of an embryo Citcllus franklinii 2 mm. loni;. X 150.

2 Ventral view of a reconstruction of tlic liver of an enibrj-o Citcllus franklinii 3 nun. long. X 150.

3 Ventral view of a reconstruction of the primary hepatic iliverticulum and a portion of the arcliojiteron of an embryo Citcllus tridcccnilinei-tus 2.5 mm. long. X 150.

4 Antcro-vcntral view of a reconstruction of the liver and a portion i<( tlic foregut of an embryo Citcllus tridccemlineatus 3.1 mm. long. X 150.


A. meil. /., anterior median lobe-like Lg., lungs

division L.p.l., H.p.l.. left and right po.sicrior

Fg., foregut lol)C-likc divisions

Gr., anterior constriction-groove be- P.ry.f.. pars cystica

tween hepatic diverticulum and P. hep., pars hepatica

foregut, yolk-stalk

Hg., hindgut H. A. Sanborn, artist


CHAnL»;H crOKNt: JnllNHnN







In the spring of 1916 a number of cats were brought, into the zoological laboratory for dissection. Among this number was one whieh showed a very unusual degree of anastomosing of arteries and veins.

A carmine starch mass was injeetec! into one of the carotids with the result that it i)a.ssed over into some of the veins, especially those of the hind legs an<l those of the posterior abdominal region. Since starch grains do not pass through the capillaries, it was e^-ident that direct union between arteries and veins 'were present.

Examination showed one large connection between the cauilal arteni' and common iliac vein which iiermitted the filling of the cava. In addition four others were found in the branches of the internal iliacs; seven in the branches of the femorals; three in the branches of the ilioluml>ars; two between the atirenolunibars; one lietween the intercostals which allowed the azygos to fill, and five between the branches of the right subclavian artery and vein. This is a total of twenty-three, besides there were evidences of others which we were unable to locate.

These connecting vessels were between 0.2 mm. and 0.5 mm. in diameter in all cases except that between the caudal arterv' and common iliac vein, in which case the vessel was more than a millimeter in diameter. The lengths varied between 1 nun. and 4 nun.

We examined the mother and two .sisters of the specimen but found no trace of anastomosing in any of tljem. Other members of the family are imder ob.servation and these with thtir offsjiring will be examined in due time. The specimen was alxmt nine months old when killed and ai)i)eare<l to be in good condition, with the exception that its hair never laid down smooth and sleek.




Fig. 1. Arteries are indicated in solid lines. Veins are shown in outline. The arrows indicate the anastomoses.


ALBKHT M. KEESE Weal Virginia University, Morganlowti, West Virginia


This lamb was liorn on Marcli 24, 101 (5, at Ringamon Creek, W. \'a., of a tluee-year-old Southdown ewe whicJi is still alive. It was sent to the \\Titer by Mr. C. O. Reed, taxidermist, of Fairnionl, ^\'. \':i. Accordiug to Mr. Reed the lamb fed with both mouths, and sccMucd perfectly well up to the evening of .June (), IDIC), Ijut was founil dead the next day. It was exliibited at various plaees under the name of 'Betty,' the posters announcing that "This Wonder Freak Feeds From Either Mouth, Hears From Four Ears, Antl Sees From Four Eyes." The animal, it will be seen, lived for about ten and one-half weeks.

When received the lamb had been skinned and all four legs had been eut off close to the body. In removing the skin so many of the nmscles of the neck hat! been removed or torn that it did not seem worth while to make a study of this feature of the anatomy. Some of the superficial blood vessels and glands were also injvu'ed. as will bo noted below, by the careless skinning. Since Mr. Ueeil especially retiuested the return to him of the skeleton, it was not possible to make a study of the nervous sj'stem.

Figure 1 is a ]ih()tograph of a ventro-lateral \iew of the anterior region of the animal. The heads are of about the .same .size, and are joined in the anterior cer\ical region. They are so close Idgctlicr that the adjacent ears (they were all. of course, removed with the skin) must have been clo.sely crowdeil together. With the exception, i)erhaps, of a slight loss of bilateral symmetry there is nothing unusual in the appearance of the thorax, unless it be a rather umisual dejith in a dorso-ventral direction.



sirrstiBiR, IVI7



A (lorso-latorul view of tlic aniinni shows a bifurcated ligaiiH'iitum iiucliar, the (li\isioii l;ikiiin placf in the aiilcrior tliinl of tlu" cervical rcniou; otherwise tiitre i> iio more uiuisiuil appearance than would be expected in sucii a monster.

Figure 2 shows the heart, and the ghinds of the neck and thorax, as seen in a vent ro-hit era! view, after the removal of the ventral and lateral jiortions of the wall of the thorax and the remains of the sujierficial muscles of the neck.


a. oiillinr of tlie llinrsix ln-forc its

rciiioviil aol, left aortic iircli aor, ri(tlit aortic arcl> ar, iirticiilar siirfaco for atlas al, atlas OM, uiiriolps ajc, axis

nj", pxtra l)oric of axis f<i, left precava ca', rislit procava rp, postcavn

cv', third cervical vertebra ri, seventh cervical vertebra <lai), dorsal aorta (//(/, left ductus Hotalli dbr, riKht ductus Hotalli f, external auditory meatus /. auricle-like lobe of luiiR fa, atlantal foramen fl. foramen transversariuni i7, left brachiocephalic artery I'r, right brachiocephalic artery jef, fused external jugulars jif, fused internal juKulars jcl, left external jugular veins jil, left internal jugular veins jrr, right external jugular veins jir, right internal jugular veins jf, fused jugidars /, various lobes of the lungs la, la', larynx of left and uf right heads Ig, Kinph gland

III-, left carotid artery of left head

Irr. right carotid artery of left head

III, lung

niv, niital valve of right heart

me', milal valve of left heart

0,0', abnonnal openings between cavities of the hearts

/),/)', parotid glands

/)/. left i>ulmonary artery

l>r, right pulmonary artery

pi7, left ])ulmonary vein

l>rr. right pulmonary vein

r. third rib

rlc, left conunon carotid of right head

rrc, right common carotid of right head

■icl, left supernumerary carotid

Kcr, right supernumerary carotid

si, left subclavian artery

sr, right subclavian artery

.v/H, sm', submaxillary glands

.1(7, left subclavian vein

sir. right subclavian vein

/, (', left and right thymus masses

la, la', trachae

tl. It', thymus of left neck

tr, rhymus of right neck

Ir', .second thoracic vertebra

(;/■ '.'/', thyroid glands

I', ventricles

va, ver'ebral artery

IT, vertebral vein

X. narrow interauricular septum

z, z', left and right azygos veins



Tho onorinous hoart, witli its two apicos, r and r', and it? pcricardiuni intact, is shown nearly filling the thoracic caxity; it will ho described hiter.

The Ihi/mus glaiul begins as two largo, white masses of lohulatcd material, /, /', at the anteinor border of the heart. These two masses are of a])i)7dximatcly the same size, the apparent

I'iK. 1 A veiilro-lalfral viow of tlio twci-lieadcd liimli ii> tlio condition in which it was received.

difference in size, as seen in fipjiire 2, beinp; due to the fact that tlie one on the animal's right side is seen in profile, .\fter extending cejilialad for about 5 cm. they out of the thorax (whose original anterior limits arc indicated by the cur\-ed, broken line, a) and lie on th(> ventral sm-face of the neck.

.\ short distance cej)halad to th(> thorax the left thymus /, divides into two smaller, elongated ma.sses, //, //', which lie close on either side of the trachea, /</, and exteml forwards to the



I"i(5. 2 A vi'iilro-liitrriil virw of llio Iniiili after the removal of the ventral wall of the thorax and the 8ii|)erKeial muscles of the neek, to show the glands of the neek and the hear! in its perieardiiim.


angles of the jaws, where they lie, at least ou ttie right side, beneath the suljinaxillary glamls. These two anterior j)rolongations arc of the same character as the main, hut are separated from it by deeji furrows.

Tlie riglit thymus body, I', divides in the same way anterior to the thorax, into two masses, Ir, Ir', but the mass on the right, tr', was almost entirely torn away in removing the skin, so that its size could not be determined; in any but little of it would show in this view of the animal. The left mass, (r, is very large and extends forwards to the corresponding angle of the mandible, where it disappears beneath the submaxillary gland of that side.

Ljing lietween the thymus masses, ll and ir, are seen the fused external jugular veins of the adjacent sides of the two necks, jf.

The thyroid glands, ty, ty', are dark-colored, uiisymmetrical bodies, about 2 cm. in longest diameter, Ijing around the ventral wall of each trachea, ia, ta', just caudad to its larynx, la, la'. The two glands are of approximately the same size and shape, and do not have the two lobes and connecting isthmus characteristic of this gland. Whether this irregularly saddle-shaped mass represents a fusion of two lobes or merely one enlarged single lobe it is difficult to say. The greater part of each gland lies on the left side of its trachea.

The salivary glands. The submaxillary glands, sm, sm', are seen as lobulated masses, about 2.5 cm. in longest diameter, lying at the angle of each jaw. The gland of the left side of the riglit heatl was badly torn, so that its exact outline could not be determined, but it was apparently of about the same size and form as the other three.

The parotid glands, p, />', were all torn off in skinning except the right glantl of the right head, the ventral end of which is shown at p'. This gland is about 3.5 cm. in length and is of a crescentic shape. It extends from just cejihalad to the extt rnal auditory meatus to the angle of the jaw, where it i)artly overhes the submaxillary gland. On the left side of the left head a depression, /), in the muscles between the external auditory meatus, r. and the .submaxillary gland, .s>/(. indicates the former


position of the parotid gland of that sidr. So far as could be di'tonuined the parotids were all normal.

The suhlinyuah arc aiiparontly normal, as might be expected from the cluiracter'of the mandibles and tongues, and are not indicated in figun^ 2.

A large lymph gland, /</, is >ho\vii :it the base of the neck on the left side.

The exterior of the heart and the tnain blood vessels

Figure 3 represents a ventral \ie\v of the anterior end of the animal after removal of the ventral thoracic wall, the pericardial meinbrai\e, the su]it rficial muscles of the neck, and the thymus glands. The veins and the pulmonary arteries are sti]}pled; the other arteries are shown in outline.

The heart. At first glance, from this \'iew, the hearts seem quite distinct, thougli closely pressed together; just how closely they really are imited will be described later.

The two left ventricles are seen at v, v'; the two right ventricles, of which that of the right heart is much the smaller, are seen at r" and r'". A distinct groove in each heart indicates the line of di\-ision of the right and left ventricles. Covering the base of the right ventricle of the right heart, v'", and the origin of the right jHilmonary artery is a large fold of tissue, /, that has the ai)i)earance of an auricle but is, in reality, a small, median lobe of the lung, to be described later.

From each left ventricle arises a large aortic arch, nol, aor, that from the left heart Ijeing apparentlj' the main one; it is about 15 mm. in (hameter. This left arch bends around in the usual manner to the left and extends, ajiparently, as the single <lorsal aorta down the animal's back.

The right aortic arch, nor, also curves to the left in the usual manner l)ut empties, ajipareutly, into the left arch, dorsal to and between the left subchnian and brachio-cephalic arteries.

C'oimecting each aortic arch witli its corresponding jjulmonani' artc)->' is a ductus Hot alii. dt)l, dhr: that of the left heart is apparently completely closed, that of the right heart has a wi<le hnncii.



Kig. 3 A voiitro-lfitcral view of the liinil> after removal of the siiperfiiial muscles and the glands of the neck, to show the double heart and the chief Hood veasels.


T1h< hranchi's of the loft aortic arch will fii-st l)0 descril)od, tlicn thost' of the ri^lit arch. Tho first Itraiich to leave this arch, aft(r tho coronary which loaves just above the semilunar valvo, is tho largo iunoniinate or brachio-cephalic, il; it extends directly cophalad for about 3 cm. along tho ventral wall of tho left trachea, (a, and then divides into two equal branches, th(> common carotitls, lie. lir, which extend cephalad on either side of the trachea to disappear beneath (iii a ventral ^^ow•) the submaxillary glands of each side. A few branches are gi\-en otT tho common carotids to the thjinus and thjToitl glands.

A little more than a centimeter to the loft of tho origin of the brachio cephalic the left aortic arch gives off 'a vessel of about half the diameter of the brachiocephalic; this is tho loft subclavian, si. It extends cephalad for a coujjIo of centimeters, in a coui'se nearly parallel to the brachiocephalic, until it passes out of the thorax. In remo\ing the legs this vessel was of cut, so that it coukl not be traced further.

Just dorsal to the origin of the left subcla\nan a small vertebral artery, va, takes its origin and passes dorsad to the base of the neck.

About 3 cm. beyond tho origin of tho left subcla\-ian, from about the point where the left aortic arch straightens out to form the dor.sal aorta (as it appears in this \-iew), arises the right subcla\'ian arterj-, sr. From this curious point of origin it extends diagonally forward, dorsal to the anterior end of the heart, and leaves the anterior end of tho thoracic ca^^ty on the right side, opposite the point of emergence of the left subcla^•ian. Thus the right aortic arch gi\-es rise to no subchiAian arterj' at all.

The right aortic arch, aor, gi\es rise to but one branch, the riglit brachiocephalic, ir; this branch originates about half way between the base of the right heart and tho point of union of the two aortic arches mentioned above; it passes cephalad for a little over 2 cm. along the ventral side of the left trachea, giving off one or two branches to the thymus, and then divides into two large and two small branches. The larger branches, rlr and rrc, are tho kft and right common carotids of the right head; they


ext<^ii(l coplialad ou oithor side of tho traohoa, as did tho corrtspoiidiiiK arteries of tlie left head, aud disappear beliind the (•orresi)()iidin{4 submaxillary glands, giving off a few branches to the lliymus and thjToid glands. The two smaller branches of tlie brarliiorei)ha]ic have, for a better name, been called the left and inglit sui)ernumerary carotids, scl aud scr; they extend cephalad, about half way between the two tracheae, the left branch dividing at about the middle of its course into two equal parts. The riglit supernumerary carotid enters an intervertebral foramen close to the bases of the two skulls; the two branches of tlH> left supernumeraiy carotid enter two inter\ertebral foramina al)out 1 cm. caudad to the foramen into whicli the right branch entered.

The pulmonary arteries need little description. Each arises from its right ventricle in the usual manner and bends to the left to disajipear beneath ("dorsal to) the heart. They are of about the same calibre, though as seen in the figine the left pulmonary', pi, seems larger than the right, pr. Their further course will be described in connection with the dorsal \-iew of the heart. Each is connected with its corresponchng aortic arch by a ductus Botalli, as described above.

Tfie veins. The postca^•al vein, cp, is seen in figiu'e 3 emerging from beneath (dorsal to) the apex of the left ^•entricle of the left heart; it will l)e described when the dorsal aspect of the heart is considered.

The two precavae will now be described. The precava of the left heart, ca, as seen in figure 3, emerges from beneath the left auricle and passes cei)halad across the pulmonary and aortic arches into the neck to the left of the left trachea. .V coujile of centimeters after its emergence it receives a fairly large azygos, z, that curves to the left and passes caudad along the left side of the dorsal aorta and back bone, recei\nng the usual intercostal tributaries.

.lust cephalad to the point of entry of the azygos this precava receives two veins of iibout the same size as the azygos; one coming from the anterior wall of the thorax, it, may be the internal thoracic; the iiion- anterior one, svl, though cut when the


foreleg was removed, is aiiparently the hrachial or subclavaau. Aliout 1 cm. anterior to the last vein a small vein empties into the dorsal side of the jjreeava; it is the vertebral vein, re. A couple of centimeters cephalad to the vertebral the precava is formed by the union of the left external jupular, jel, and left internal jupular, jil. The former is the larger, and as it was nearly all removed in skinning the animal its probable course Ls indicated by dotted lines. The internal jugular follow closely the course of the eoiTesponding common carotid artery, Uc, described above, and the vagus nerve.

The precava of the right heart, ca', is of much larger diameter than that of the left heart. It receives just cephalad to the heart a large azygos, z', which is, in turn, made up of two branches, one from the riglit side of the vertebral colunui, the other from the mid-ventral line of that structure.

About 1.5 cm. cejihalad to the azygos the precava receives the rather small riglit brachial or subcla^•ian, sir, which had been cut, a short distance from its base, when its corresponchng leg was removed. No vessels corresponding to the internal thoracic and vertebral can be seen on this side.

A short distance cephalad to the subclavian the precava receives from the right the large external jugular, jer, most of which had been removed with the skin. Just cephalad to this vein is the right internal jugular, jir, a much smaller vessel that follows along the right side of the right trachea, along with the corresponding carotid arterj'.

Opposite to these jugulars the precava receives two or three veins, on its median side, from the thymus gland. Cephalad to the abo^•e veins the precava may be followed as a very large vein, h"ing between the two tracheae; this vein has been called the 'fused jugulars,' jf. About opposite the thyroid of the left neck it divides into two vessels, jef and jif, the ' fused external jugular' and the 'fused internal jugular,' the former diAides into four ves.sels, two passing beneath the left submaxillary gland of the right head, two pa.s.'iing, jMobably, to the skin and nmscles between the two heads, thougli this could not be accurately determined because of the removal of the skin and super


ficial iiiusclos of this rogiou. Tho fused internal jiiRiilar clividt-s into two vessels, one of whieh may l)e followed to the base of each skull, where it is lost, probably passing through a foramen into the skull.

Dorsal view of the heart

I'i^ire 4 represents a -dorsal view of the heart, with the stuiu])s or o])euings of all the blood vessels. The left ventricle of the right heart is seen at r', and above it is seen the edge of

Fin- -1 A (lors;il view of the double heart of the lamb, with tlic roots of the main blood vessels.

the right ventricle of that heart, »•'",• there is no external line of demarkation between them.

The left \('ntricle of the left heart is seen at r; the right ventricle of this heart docs not show in this \-iew.

The septa of the hearts are so abnormal that it is difficult to name the amides as either riglit or left. On the e.\treme left is the left auricle, au. of the left heart: this is distinctly a left auricle since it opens into the coiTcsj^ondiiig left ventricle, and icceives the large pulmonary vein, /)i7, from the left side of the lungs, but it is connected by a small though ilistinct opening


with the auricle on its rinht ; it is seen at nu in the ventral \'iew of the heart, fipiire '^.

In the median re^on of the eonibined hearts is a large, indefinite chamber, au', into which open not only the two right pulmonary' veins, pvr and pir', but also the right and left precavae, ca' and en, and the post cava, cp. The 'fusion auricle' is seen also in the ventral view of the heart, figiu-e 3, ait'. It is only partially separated, internally, from another large auricle, au", which, in turn, is connected with the right ventricles of both hearts anil with the left ventricle of the riglit heart. It is thus impossible to say whether these two auricles are right or left. The internal features will be described in more detail later. The veins entering the hearts will now be noted briefly. The left pulmonary vein, pvl. is seen as a large opening in the left auricle of the left heart. The two right pulmonary- veins enter the fusion auricle at pvr and pvr'; the posterior one being the larger.

Entering the fusion auricles, as above noted, is the single though partially di\-ided trunk of the postcava, cp. A septum di\ndes the trunk into two unequal parts which enter the fusion auricles at the same place.

The left precava, ca, extends diagonally across the dorsal surface of the auricles and empties into the fusion auricles just anterior to the opening of the postcava.

The right precava, ca', which is larger than the left, is seen as a prominent longitudinal swelling extending along the median region of the combined hearts to open into the right side of the fused auricles.

The arteries lea\ing the heart will need but little description in addition to what was given in connection with figure 3.

In the ventral view, as noted above, the left aortic arch, aol, appears to form the dorsal aorta, with the right arch, aor, emptying into its anterior border. In the present figure the right arch seems to be continue*! as the dorsal aorta, dao, with the left arch, aol, emptying into its posterior border.

From the right arch arises the large brachiocephalic arter>', ir, and from the left arch the left arter\- of the same name, il.


From the aorta at tho i)oinl of union of tlio two archos arises the left sulxlavian artory, s/, and from the aorta, caudad to this point, arises the curious right subclavian, sr, that passes straight across the anterior region of tlic heart to the riglit side.

Tlic h'ft puhnoiKiry artery, pi, is seen fni('rf»:inK from tlic ventral side of the heart l)ct\vecn tlie left preeava, nt, and the left aortic arch, ool: it divides info approximately equal parts that lead, as said above, to the left lobes of the lungs. The right pulmonaiy artery, pr, emerges from between the left acjrtic arch aol, and the right preeava, ca'; it di%ides into a small left and a larger right branch that lead to the middle and left lobes of the lungs.

The small left ligamentum Botalli, dbl, and a larger ductus Botalli, dbr, are seen leading from the left and right pulmonary arteries to the corresponding aortic arches.

Internal structure of the heart

The cavities of the two hearts are so abnormally connected with each other that it is surprising that "Betty" lived as long as she did.

The right ventricle of the right heart, figure o, v'", is not only connected with its i)ulmonary artery, />;•, but has a fab-ly large opening, o, into the left ventricle, v', of that heart; it is not directly connected with an auricle.

The left ventricle, r', of the right heart besides the opening just mentioned, ;ind its aortic outlet, o])ens by a wide aperature into the auricles au', aw"; this auricular-ventricular opening is guarded by a very well-developed set of mitral valves, we.

The right ventricle of the left heart, /'", besides its opening into its jn-oper jiulmonary artery, pi, is connecteil with the auricle <iti' by a wide aperalure with jioorlj* developed tricuspid valves.

The left \entricle, r, of the left heart has well tlevelopetl semilunar valves at its opening into the left aortic arch, aol, and equally well developed niitral valves, nir', betwi-en it and the auricle au.



It is in the auricles that the p-oatpst abnonnalitips oceur. 'riw l«ft auricle. <iii. of the left heart approaches more nearly the iKirnial than any of the others. Its apixMidane is shown at nil in hpn-e '.i. This auricle receives the left pulmonary vein. /*i7. ami oi)ens, as noted ahove, into its corresponding; left ventricle, r. It has also a small opening or foramen ovale, o', into the lar^e auricle to the rifiiht. an'.

The auricles an' and (/(/" are so little separated from each other i)v a small fold of skin, .r. that they form i)ractically one

Kip. a A .scnii<li;ipraiiim:ilir outline ot the (!oul>lo hoiirt !is soon from the dorsal side, to sliow. l>y nipiiiis of arrows, the coursn of the l>loo<l tliroiifdi the various auricles and ventricles.

\iirgi\ irriKular chamber that seem.s to represent throe auricles. Into the half of this fused auricle to the left of the septum open tlie following vessels and chambers; the left precava, ca; the two right pulmonary \'eins. /jrr, i)ir' : the auricle (in. tlirongh the foramen (/; the right ventricle, v". of the left heart; and the pcstcaval vein, cp.

Into the half of the fused auricle to the jight of the septum, x, which half forms a broad band diagonally the dorsal side of the hearts, opens the right precava, ca'.

As has been said, the left ventricle, r', of the right heart opens i)y a \\ide auricular-ventricular aperature into the common


(■liainl)cr of the flis<'<l miriclcs, riit', au" , below the ciluc of the soptuni X. Tlic fusotl auricles oi)eii widely into the iiurieular appendage, au' , seen on the ventral view of the heart, figure 3.

The respiralory system

The left and right lan-yngen, la, la', figure (>, and traeheae, la, la', have already been noted, and sinee they are distinct from each other, and show no iniusual features, they need not be describ(>d. The trachea gi'adually converge towards the lungs, into wliich tiiey enter by distinct but closely adjacent openings.

The lungs are not so nearly double as are the two hearts above descriljcd. The right lung, as a whole, is larger than the left.

The (Ha]ihragniatic lol)es, /, difTer l)ut little from those of a slice]). The apical lobes, /', are remarkable mainly for the long lobules that extend cephalad; that of the riglit side is the larger and is shown, in the figure, bent down and l)ehind the rest of the lung; that of the left side, shown extending cei)lialad between the two tracheae, is the lobule that is shown at / on the ventral view of the heart, figure 3.

The mediastinal and cardiac lobes are .so broken up into lobules that it is diflicult to differentiate them, but it seems likely that the lobes, /-, i', represent the two cardiac lobes, while the three small lobes, /', represent the mediastinal lobes divided into tiiree loi)ules.

In a dorsal nr ;i lateral view the lungs have almost the norn\al apiiearauce except for the cephalic prolongations of the apical lobes.

The tracheae eiitir the huigs at the anterior margin, and are, at this point, about l..") cm. apart.

.\t the l) of the right trachea th(> two l)ranches of the right pulmonary artery, fir, enter the lungs. Of tin se two branches the right is the larger and is distributed mainly to the right lobes of the lung; the left and smaller branch got s mainly to the central lobes of the lungs. At the base of the left trachea the



two bniiichcs of the left i)ulmoii!iry artery, pi, oiitor tlio lungs. Of these the h'ft is sUghtlv the hir^er: both of tliese hraiiches are iiistril)iite(l to the left lohes of tlie hin>!;s.

Ix^iving the hings from the region between the left and central lobes is the largest of the pulmonary veins, jirl, which enters the left auricle of the left heart (tig.;")). .Vbout 0.5 cm. to the right of this vein is another large vein. /;jt nowcr lino), which leaves

FiK- ti A photograph of the respiratory organs of the lamb as seen from the ventral side.

the middle lobes of the hmgs anil enters the left auricle of the right heart (fig. 5). .Vbout 1 cm. to the right of tliis vein, and somewhat cephalad to it, a somewhat smaller vein, pvr (upper line), leaves the left lohes of the hmgs and enters the left auricle of the right heart (fig. 5).

I/'aving the small anterior lobe, /', shown between the tracheae in figure (i, is a snuvll vein, /irl'. wliich enters the left auricle of the left heart along with the large vein, />rl. described above.


The digestive organs

Tho two heads being distinct the tongue and teeth are normal for each head: the salivary' glands have already been described. The two esophagi are separate antl norma! until they arrive within about 3 cm. of the stomach, where they unite with each other and empty, by a single opening, into that organ. The stomach is single and apparently (juite normal. The liver is also normal, as are the organs caudad to this region.

The skeleton

As noted above, the appendages had been removed before the animal was received, so that they could not be studied, but there is no reason to suppose that they were abnormal. The thorax is apparently normal, except, as noted above, for a possible unusual depth in a dorso-ventral <lirection.

A dorsal view of the two skulls, and of the spinal column as far caudad as the second thoracic vertebra, is shown in figure 7.

Except for a marked lateral twist in the right skull, and a slight twist in the left skull these two organs seem normal and need no further description. The twist in the right skull is so marked that the incisor teeth must have been almost useless.

The thoracic vertebrae, /(•', and the posterior five cervicals, cv', cv, seem from this view to be normal, except that they are rather wider, laterally, than normal.

The second cerAical, ax, bears no resemblance to the normal axis. In the dorsal \iew, figure 7, it exhibits two prominent ridges, separated by a deep groove; these ridges are fused in the median plane and project cephalad between the two atlases, at. This second vertebra articulates rather closely, in a more or less normal manner, with the third cervical, and presents a large, autero-laterally projecting process, ar, on each side, for articulation with one articular surface of each atlas. There is no sign of an otloutoid process.

The two atlases, at, are e.ssentially alike and seem nearly normal; that on the right side is slightly the larger. Two foramina are seen on each side of each atlas, the foraaieu transversarium,




//, and the atlantnl foramen, /a; the latter foramen on the mesial side of each atlas is a poiivc or notch in the antr-rior margin of the l)one. rather tlian a distinct hole.

A ventral \-ie\v of the cervneal and anterior thoracic regions of the vertebral column is shown in figure 8. In this \'iew the unusual width of the cervical \('rtchrae is shown in contrast to

Fig. 7 A dorsal view of the sliulls and the anterior vertebrae of the lamb.

the width of the thoracic vertebrae. The cervical vertebrae are also seen to be more or less closely and irregularly fused, so that it is ver>' difficult to distinguish the boundaries of the individual vertebrae, e.speciallj' of the last three.

The axis, ax, is seen in this view to consist of the two widely divergent articular processes, ar, which are not in contact with each other, and an irregularly rectangular bone, ax', lying in the median line and articulating with the median articular surfaces of the two atlases, al. This strange-looking bone is loosely attached to the articular processes, ar, but seems fairly closely united, in the anterior part of its mid-dorsal region, with


the ventral side of the two median processes of the axis, described in connection with the dorsal \iew. It is in the region of the posterior ])rocess of this curious bone, figure 8, ax', that the single spinal cord divides to pass to each brain. In this view the two atlases appear ([uite normal; one foramen transversarium is not visible in this view; the other foramina are very prominent.

KifC. 8 A ventral viow of the same vertebrae shown in dorsal view in the preceding figure.

Owinp; to the fact, as noted above, that Mr. Reed requested that the skulls be returned to him it was not possible to stud}' the nervous system.


C'.MiKY, Eben 1917 The anatomy of a double pig. Syncophalus thoracopagus, with especial consideration of the genetic significance of the circulatory apparatus. .Vnat. Hoc. vol. 12, no. 1. pp. 177-192.

CoNHow, .S.\K.\ B. 1017 .V six-legged rat. .\nat. Rec, vol. 12, no. 3, pp. 3(55-70.

Reksk. .\lheht .M. 1',)11 The anatomy of a double cat. .Vnat. Rec., vol. 5, no. 8, pp. aS3-9().

1914 The osteology of a double-headed calf. .Vmer. Nat., vol. 4.S, pp. 701-704.



FLORENCE R. SABIN Anatomical Laboratory, Johns Hopkins University

The (jucstion of the orif!;in of the vascular system can be solved by the method of studying the living blastoderm of the chick in hanging-drop preparations.

By watching chicks of the second day of incubation in these prei)arations it is possible to see all the processes bj" which blood-vessels and later blood-cells form. These observations can be made best on the area pellucida. Blood-vessels begin by the differentiation from mesodenn of a new tj-pe of cells, angiol)lasts or vaso-formal ivc cells. They differ from mesodenn in ha\-ing a much more granular cjiioplasm and in being more refractile. They differ also in their beha^^o^ and in their jiotentiahties. When a cell of the mesoderm divides, the daughter cells separate at least enough so that they can be recogruzed as distinct cells but angioblasts give daughter cells that remain together to form dense sjiicytial masses. These small masses soon jt)in similar masses by means of tiny processes of cytojjlasm put out from them exactly like the sprouts by which blood-vessels are known to grow. In this way angioblasts form a plexxis of dense masses of cells in sharp contrast to the more delicate plexus of mesotlerm wliich represents the early stages of the development of the coelom. The jilexus of angioblasts increases both by the division and the gi-owth of its cells anil by the constant addition of new angioblasts wliicii diffiM'cntiatc from the mesoderm.



\\'ithiii the jilexus of anpiohhists, vacuoles appear which represent a li(iuefactinn of the central ])art of the cyt()i)lasm to fonn bloo( The vacuoles begin aKainst the nuclei and may occur anywhere in the mass but are especially frequent under the edges. The vacuoles just under the margius leave the cells along the border of the mass to become an endothelial lining of a ea\ity. The comi>lete li(iuefaction of the central part of the mass into plasn\a takes from one to two hours and can be seen with great clearness in the living chick. The endothelial cells on the border of these cavities become less granular than the original angioblasts. Since the fluid is formed from the liquefying of the center there is no sign of distention of the cavities and no flattening of the cells along the border. There is a destruction both of the cytoplasm and of the nuclei of the mass to form the plasma.

This process of the liquefaction of the central part of the mass of angioblasts to form vessels takes place not only within the )>lexus but in masses of angioI)lasts which are still isolated and in this way small vesicles are formed which then join the main plexus by processes of c>ioplasm. I have seen such a tiny vesicle fonn from a single angiobhist pro\'ing that the lumen of a blood-vessel is intracellular.

There are two processes from which the formation of bloodvessels must be sharply distinguished in watching these living specimens. The first is the fonnation of the coelom. The mesoderm of the chick is originally a continuous sheet of cells as is shown in Lilly's figm-e 40, A and B, on page 70' and in Duval's figures 18-i to 188 on i)late 12. As the cliick grows larger this double sheet of cells forms a plexus with wide interspaces in the network where the mesoderm is entirelj' lacking lea\'ing nothing but ectoderm and endodenn. The coelom now begins as clefts full of fluid within the soUd bands, often at the nodes of the network of mesoderm. These spaces separate and, as it were, split the two layers of the mesodenn apart

' Lilly, F. U. lOaS The development of the chicle. Henry Holt and Comp.'iny. New York.

'Duval, .\I. 1S89 Atlas d'Embr>'ologie. G. Masson, Paris.


and Kradually flow together to iiuikc the ca\'ity of the coelom. The mesodenn on the Ijorders of these spaces gradually flattens into a mesothehmn. The process involves but httle destruction of tissue. At the stage of about six or seven somites a Hving specimen thus shows a double plexus over the area jiellucida, a dorsal plexus which is the developing coelom with large interspaces in the net and many tiny caxities representing the exo-coelom; and a ventral, more massive plexus of angioblasts with much more granular and more refractile c\-toplasm than the mesoderm. The granules of the angioblasts are strongly basophilic and stain intensely with haematox>'lin and with azur. The second structure from which developing vessels must be distinguished I shall call endodermal blebs or bhsters. They are collections of fluid beneath the endoderm, that is between endodenn and mesodenn, which are very frequent in the normal chick. They can be seen in any collection of mounted blastoderms. In the living chick they vary in appearance according to their size and shape. Their margins simulate endothelium to a marked degree. If they are distended and hence round, their margins will be thin, sharp and highly refractile; if they are flatter their walls will be still refractile but thicker. Often their nuclei seem to project into a ca\ity exactly like those of endothelial cells. They may be numerous, small and isolated or large and nmltiple reminding one of midtilocular cysts. .\s can be readily imagined from their structure they may change in shape rapidly, far more rapiiUy than true vessels change. They can be analyzed with the focusing screw both by following their margins over onto the endoderm and by noting their very superficial position. They occur under the ectoderm as well as under the endoderm but I think less freriuently. Xo one can follow the development of blood-vessels in the li\'ing chick without becoming thoroughly familiar with the appearances of these blisters. They have nothing to do with blood-vessels occurring both before the blood-vessels begin and afterward. They are however important physiological structures representing the method by which fluid is absorbed through tiie endoderm and ectoderm for the young chick.


Red blnod-corpu8clc8 can l)c seen to grow from the endothelial lininR of hloiMJ-vossi'ls. Tlioy may develop from little masses of the oriRinal anpiulilasts which l)ecomc i)artially separated by the liquefaction of cytoplasm around them. Such a mass of eells becomes a blood-island by ha\inK haemoglobin develop within the cells. The color of haemoglobin can be detected in the li\ing cells earlier than I have been able to fix a,nd stain it. Again an endothelial cell of a blood-vessel will divide so that one daughter cell jirnjects into the lumen. This cell becomes filled with bivsojihilic granules and develops haemoglobin. It is then a imicelluhir blood-island. It divndes and the mass is increased also by the addition of other cells which differentiate from the endothelium in the neighborhood and creep ahmg the wall to join the first cell. These cells .soon form a yellow sjnicytial mass projecting into the lumen of the vessel. \i this stage the islands have a smooth, sharji contour. .Vs they develop, cells begin to rountl up on their siufacc until the whole mass comes to look Uke a mulberry and then the red cells break free from the mass and float away in the blood-plasma.

These blood-islands I have seen develop in all the vessels of the area vasculosa, in the omphalo-mesenteric vein anil arteries and in the dorsal aorta. In the area pellucida they are most abundant in the i)l('xus of vessels just posterior to the area in which the omi)hal()-mesenteric arteries develop. In this area abnost any chick of fom-teen somites will show a plexus of angioblasts fuU of \acuoles and one of seventeen somites will show blood-islands. The circulation does not interfere with the development of the islands. If they clog the lumen the free cori)uscle8 either lodge behind them for a time or pass on thi'ough other channels in the jilexus.

One of the most interesting points al)Out the nui.sses of angioblasts and of the islands is that all of the cells in them divide at about the same time. Moreover all of the separate islands of a given area seem to divide at once. In these total preparations, all one can .see of the ju-ocess of mitosis is the nucleus in the stage of the metaphase and the actual tlivision of the cj^toplasm. When the syncytial mass is about to di\ide it becomes


intensely refractile and its cells become outlined; then one sees one nucleus after another pass into the nietaphase and finally all of the cells appear to be half as large as the mature cells. The island then becomes large by the giowth of all the cells to their original size. The blood-cells keep on dividing after they are free from the islands and after they have begim to circulate.

These studies have been made with chicks which were grown in the mixture of Locke's solution and chicken bouillon develoj)ed by Margaret Reed Lewis for tissue-cultures.' For the young blastoderms it is better to increase the amount of the sodiiun chloride in the solution to L0(3 per cent. A weaker solution lakes the haemoglobin and is also less favorable for all the other cells.

From these studies which are here reported briefly, certain jH-inciplos are established. Blood-\essels do not arise from the dilatation of tissue-spaces, but by the differentiation of angioblasts. They cannot be spoken of as arising from spaces because they develop within the bodies of these angioblasts; that is, they arise within cells, not between them. The processes by which they form are not in any way siimlar to the processes by which either the coelom or great systems of tissue-spaces like the cerebro-spinal spaces develop. The coelom forms by the splitting apart of two layers of mesoderm with Uttlc de.-^truction of tissue; the cerebro-spinal spaces form in a mass of typical mesenchjme with considerable destruction of tissue. Angioblasts differentiate tlu'oughout the wall of the yolk sac and in the embryo as well. I have seen the tlorsal aorta tlifferentiate in situ in the li\-ing chick even to some extent that part within the head. Angioblasts produce blood-plasma, endothelium and red blood-cells. .Vngioblasts and later endotheUal cells give rise to red blood-cells by develojiing haemoglol)in.

The term blood-island has been used since the time of the early embryologists, notably Wolff and Pander, for the masses of vaso-formative cells which can be seen in the lU'ca opaque of the chick even before the first somite. I jiropose however to

•Lewis. M. 1{. :in<l \\. II. lOU The growth of embryonic chick tissues in artificial media, agar and liouiiion. Johns Hopkins Hospital Bulletin, 2°J.


rt'strict the tenii hlood-islaiul to thoso masses of colls which can 1)0 shown to produco ha(>iiH)|a;lol)in and to boconio froo red blood-colls. Those masses are attached to the wall and hence are not strictly speaking islands. The more primitive masses which do boffin as isolated masses or islands of colls hut nmst lu"st i)roduce enilot helium and plasma, 1 shall call angioblasts. The retl blood-cells tlevelop after some plasma has been formed. .Ml of the blood-cells of the chick of the second day of incubation can be seen to have haemoglobin in the living chick and hence they cannot be consiilered as forerunners of white bloodcells.






J. A. MYERS Institute of Anatomy, University of Minnesota, Minneapolis


Schickele ('99) while studying the development, arrangement, and variation of thr mammary glands in rats noticed that ordinarily macroscopic examinations do not reveal nijiplcs in males.

In his work on the recognition of sex through external characters in the young rat, .Jackson ('12) called attention to the fact that the manunarj- gland nipples become ver}' conspicuous in female rats of about two weeks (post-natal life) while in male rats of the same age there is no external indication of nipples. The difference between the two sexes in this respect is so marked that he could reatlily distinguish the sexes at that age by the use of this single external character.

Steinach ('12) made successful reciprocal transplantations of testes and ovaries in young male and female guinea pigs. The males in which the ovaries grew gradually assumed the appearance of females even to such an extent that some of them developed functional mammary glands. The same experiments were applied to rats, antl from these experiments Steinach reached the following conclusion: "Das Rattenmiinchen ist flir diesen Punkt des Untersuchen deshalb ungeeignet, well bei ihm diese .\nlagen nicht einmal rudimentar ausgepriigt, bzw.makroskopisch nicht .sichtbar sind."'


•_>(»( i J. A. MYEUS

My studips (Myers, '1(5), which wore confinod cntirch- to the mammary glands of the female albino rat, led later to a careful. macroscoi)ic search for nii)|)les in males of corresponding ages. No trace of nipples was found in any male individual.

The above ol>servations made by four different workers are somewhat contradictory to the usual view that (he mammary glands of male animals are parallel in their development with of females until the stage of puberty, at which time the glands of the male stop growing and later atrojihy, while those of the female undergo a very rapiil development. In view of such an apparent discrepancy the present work was undertaken. .\ brief abstract of the results has already been published (Myers, '17 a).


The fetu.scs for this work were collected, fixed, sectioned and stained in the same manner as the female fetuses described in an earlier paper (Myers, '17 b). Cleared preparations were made according to the method described by Lane-ClajTion and Starling ('0(5).

In making macroscopic examinations it was verj' desirable to have the hair removed from the skin. At first a sharp razor was u.sed to .'ihave the abdominal wall. This method was later discarded, however, and the hair was removed from the entire body with sodium .sulphide as suggested by Frank and Unger ('11).

Some of the litters used for this study were slightly under normal weight. Therefore in each case it became necessary to compare males anil females of the same litter. In this manner it was possible to compare individuals of the same age and of approximately the same gross body weight.

In all, 70 individuals were examined, makmg a total of ap])n)ximately 840 mammary glands. About one-half of the entire number of individuals were fetuses. The remaining 35 were fairly evenly distributed among the described post-natal stages ratiging from liirth to ti-n weeks of age.



Fiflern (ind sixteen days. At fifteen days anfl nine hours the sex glauils of the fetuses are in the indifferent stage. In the fetuses of sixteen days and twelv'e hours the sex glands are so differentiated that with considerable difliculty the sexes can be distinguished. In the last mentioned stage there is no apparent difference between the developing nianunary glands of the two sexes. Therefore the description of the mammary glands of both of the above stages given in a pre\'ious paper (Myers, '17 b) is appUcable to either sex.

Eighteen days. In male fetuses of eighteen days and nine hours fresh preparations show the epidermis over each developing mammary gland to be somewhat lighter than the adjacent epiilennis. The ei)id('rmis is in most cases elevated so as to form a slight eminence. In one instance, however, a very shallow mammary pit was observed over a single gland.

In females of this age, it will be recalled, the area over the developnig mannnary gland appears lighter than the adjacent region, but instead of an eminence there is usuall}' a distinct mammary pit present over the developing mammary gland.

The stratum gcrminativuni and basement membrane of the epidermis in the region surrounding the developing primary duct, become continuous with the duct and form the peripheral layer of cells and basement membrane of the duct respectively. The primary duct passes through the corium into the tela subcutanea, as was found to be the case in the female. At this stage the primary duct somewhat resembles a large developing hair (fig 1). .Vt its free end is seen a considerable exj^unsion which may give ofT one or two small buds. The mesenchyma forms a thick condensed layer around the duct.

The primary fhiet of the female in size and general form is similar to the duct of the male; however, it does not show such a marked expansion at its free end. Its attached end also lies deeper from the surface of the epidermis, owing to the fact that it springs from the bottom of the developing mammary pit. The layer of condensed mesenchyma surrounding the primary duct is thinner than in the male.


J. A. M YE Its


f - ■ ^

Fi(t. 1 Driiwn from a section tliroU(?li tlio left third thorarir developing mumniHry uland of a male albino rat fetus of einhlecn days and nine hours. X 300. Zenker's fixation; hncniati).\ylin-oosin stain. Drawn with the aid of a camera lueida. Comparing this figure with figure 2 of an earlier paper (Myers, '17 b) for the female, note the difference in appearance of the primary ducts <p.d.). The condensed mesenehyma (cm.) immediately suriounding the primar>' duct is thicker in the male sex. There is less indication of a developing mammary pit {in. p.) in the male gland.


Nineteen chiys. When studied with a dissecting microscope, most of the niaiumary glands of nineteen day and six hour fetuses are not visible on the surface of the integument. An occasional gland, however, is representeil, as in the preceding stage, l)y an area of epidennis slightly lighter in apj)earance in fresh preparations than the adjacent epidermis.

The glands of the females of this stage are very conspicuous from the surface. In the female, the gland areas are very light in appearance and a well developed mammary pit is located in the center of each area.

The primary duct (of the male) at this stage has lost its resemblance to a developing hair. The attachment of the duct to the epidermis, however, is not unlike that of a developing hair. As the primary duct approaches the deep surface of the epidermis, its basement membrane and perij^heral layer of cells are seen to be continuous with the same structures of the skin. The attacluuent of the duct is, therefore, very near the surface. As in females of the same age, the primary duct passes into the tela subcutanea where it turns at right angles and courses parallel to the surface of the integument. Considerable variation is exhibited in the branching of the ducts. In some glands tertiary ducts are terminal while in others the primary duct is the terminal duct at this age.

The lumina develop in the same manner as reported for the female (Myers, '17 b). In general it may be said, however, that the huuina of the male mammary ducts at this stage are apparently somewhat further developed than those of females.

The stratum germinativum surrounding the attached end of the primary duct is somewhat thickened thus forming a slight ridge around the attachments of the duct. This ridge suggests the epithelial hood anlage, which is distinctly in exidence in females at this time.

The developing connective tissue of the male gland corresponds in structure and arrangement with that of the female.

Twenty days. At twenty days and six hours there is neither eminence nor mammary pit over the male manunarj* gland area in fresh specimens. The epidermis of the area is no longer



Fin. 2 Interniil view of a wax model reeonst meted from the left first inguinal gland of a male albino rat fetus of twenty days and six hours. X 50. When compared with figure 12 of a previous paper (Myers, '17 b) note the absence of an epithelial ingrowth' (cp. in.) and the presence of a short branch (s.) arising from the primary duct near its attached end in the male. The structures listed below are approximately the same in both sexes, e.6., end bud; p.rf., primary duct; s.d., secondary duct;, tertiary duct.


Fig. 3 Drawing of a section through the left alxlominal gland of a male newborn albino rat. X 300. Zenker's fixation; haematoxylin-cosin stain. Note the entire absence of nipple, mammary pit, and epithelial hood. Other parts similar to those of the female, r., irregularly arranged connective tissue cells; c.t., connective tissue forming sheath around duct; cp., epidermis; p.d., primary duct.


lighter than the adjacent epitlermis. There is therefore no external indiration of a dpvolojHnK iiiamnuii v Rland in tho male rat at this stage. As previously slmwn (Myers, '17 hj, the glands in the female are now quite conspicuous. They are distinctlyvisible with the naked eye. In seotions, the female gland shows a distinct nianimary pit with the nipple anlage at the bottom of the pit.

The primary' duct in the male near its attachment to the epidermis is often seen to give off a short solid branch (fig. 2), which was not observed on the female duct. In some instances there appear more di\isions of the milk ducts than have been observed in females of the same age. The second inguinal gland has failed to develop as rapidly as the others. It therefore appears somewhat rudimentary.

The lumina correspond to those described in the female, except that they are shghtly larger in the male.

Hair folHcles are developing in the immediate neighborhood of the attachment of the primary duct. This is contrary to the condition in the female, where hair folUcles were not observed at this stage in or close around tlie manunary pit.

The epithelial hood anlages do not appear in the male at this age.

Newboni. At birth, as in the twenty day stage, all external appearances of mammary glands are lacking in the male. .\t this time the glands of the female may be observed with the unaided eye in li\ing and fresh specimens as light areas in the positions of the future nipples.

Microscopic sections likewise do not reveal even the shghtest trace of a nipple or an epithehal hood in the male at this time. In the female, however, the nipple almost fills the manunary pit and the epithelial hood is well formed.

The primary iluct attaches to the deep surface of the epidermis, but does not reach the outer surface of the body. Leaving the epiilermis, the duct passes deeply through the corium into the tela subcUtanea where it turns at right angles, passes parallel to the surface of the skin anil jire.^jents secondary ducts. Immediately beneath the corium the primary duct gives off a


2r2 J. A. MYERS

short l)ranch (as in tlic prccoding stage) which courses parallel with th(> surface. No similar l)raiich was observed in females. The ducts at this stage, iuchiding representatives of secondary, tertiary, (juateniary, etc., present the same general apiiearance and the same method of branching as those found in females. The second ing\iinal glanil, however, is an exception as it is still very rudimentary.

.\ continuous lumen is present in each sj'stem of ducts observed. Tlie lumen of the i)rimary duct, however, does not extend as far as the attached end of the duct. The short branch of the primary duct immediately subjacent to the corium likewise

Fig. 4 Drawn from a cleared preparation (internal view) of a male albino rat one week after birth (Imiiy weiKliI S griimsi to show distrilnitinn and relations of diirts of riRht abdominal pland (.11; and right first inguinal gland (li). X 5. The second inguinal glands were absent in this specimen. Nipples and epithelial hoods are alment. Otherwise the glands of the two sexes are similar in their <levelopment. Compare this figure with figure 7 of an earlier paper (Myers, '16, p. 379). L., lymph node; p.d., primary duct; s.d., secondary duct; t.d., tertiary duet; tr.d.. terminal duct; e.b., end bud; co., collateral duct.

possesses a lumen. The limiina, like those of the female at this age, have not reached their definitive stage of development.

One iveek. One week after birth the milk ducts of the glands in males apparently do not present many more branches than at the time of birth. The glands occupy the same positions as those of females and the branching has reached ajiproximately the ,><ame stage ffig. 4). The only exception to this statement is found in tlie .second inguinal gland. This glatid has become so rudimentary that it sometimes fails to appear in cleared preparations. \\'hen it is present it presents a very small number of branches.

Tiro weeks. At the end of the second week in each of the first thoracic glands in males its duct is seen to extend ce|)halad


from its attachment to the epidermis. No branches were observed to take a caudal direction, as is the case in the female. The ducts of the second and third thoracic glands take a more, lateral direction and arc fully as well developed as those of females. In the abdominal and first inguinal glands a few lateral buds are seen on the main ducts. The ducts of corresponding glands of the females show a large number of such buds. The second inguinal ghiTid when \isible in cleared preparations presents a primarj' duct which courses caudad. Its branches are very few. At the free ends of the ducts of all glands at this stage are found growing end buds, as were previously described in the female.

Three and four weeks. At three weeks the glands show very Httle increase in development over the two weeks stage. The glaiuls of four weeks' male rats, however, show further development. Each of the first thoracic glands presents a larger number of branching ducts, the majority of which are directed cephalad from the point of surface attachment. A few branches, however, take a caudal direction as was found to be the case in females in younger stages. The ducts of the second and third thoracic glands have sent out several collateral branches which in some individuals nearly obHterate the space between the ducts of these glands. The abdominal glands now present a large number of lateral buds which, as in females, are not so numerous on the primary and secondary drifts as on the tertiary, ([uaternary and terminal ducts. In the male rat, the first inguinal gland of one side is occasionally lacking. Frequently the second inguinal glamls are absent. Neither the first nor the second inguinal glands are absent in any female observed. It is not uncommon, however, to find the second or third thoracic gland undeveloped in either sex.

Five weeks. It will be recalled that at the fifth week the glands of the female were observed to develop very rapidly (Myers, ('16). The ducts are much longer than in pre\'ious stages. Many new branches spring from the more distal ducts of the glands. The interval between the ducts of the second and third tiioracic glands has largely disappeared on both sides, tiiere being a

214 J. A. MYEKS

slight ovorlappiiig of tlic liucts. The ducts of iIk- abdominal and first inguinal glands likewise overlap, thus obhterating the interval between these two glands. Furthermore some of the ducts of the second inguinal gland have gr<iwn cephalad, have met and overlapped the caudally directed ducts of the first inguinal. The ducts of the second inguinal gland branch very jH-ofuscly.

At this stage the mammary glands of the male have failed to keep pace with the raj^id development of the glands of the female. While in some individuals the ducts of the second and third thoracic glands are nearly in apiio.siticm, there still exi.sts a definite inter\al in most cases. There are likewise verj- marked inter\-als between the abdominal and first inguinal and the first and second inguiiuil glanils. The secontl inguinal gland shows no I)rogress in development over the preceding stages. Not only are the milk ducts of the nmle shorter in length, but they are less numerous than those of the female.

.S')'j, seven and eight weeks. At six, seven and eight weeks the differences between the male and female glands are becoming still more obvious. The most noticeable difference is the absence or very rudimentary condition of the seconil inguinal gland of the male. A mere glance at any of the remaining glands will enable the observer to distinguish between male and female. The most prominent distinguishing character is the large number of collateral ducts that have grown from the chief milk ducts in the female. The collaterals ha\e branched antl each terminal l^ranch presents a large end but!. In the males a much smaller number of collaterals a])pear. Thus the arborization of the male gland is much less dense. In no case have the ducts of two adjacent glands in the male overlajiped so as to form a continuous mass of gland tissue.

Xine and ten wecLs. In an earlier stuily (Myers, '10), it was shown that the maimnary glands of female rats show a tremendous increase in growth and development about the ninth or tenth week. The overlapj)ing of the ducts of the thoracic glands is so extensive that a continuous mass of gland tis.sue extends from the cephalic ends of the tlucts of the first thoracic gland to


the caiulal ends of tho ducts of the third thoracic glands. No coiiii)lct(' intervals exist between the glands. Furthermore ducts from each of the first thoracic glands have grown so near the mid-line that only a small space now separates them. A considerable space exists between the last thoracic and abdominal glands. The ducts of the abdoniinal and first and second inguinal glands also overlap so as to form one continuous mass of gland tissue, extending from the most cejihalic of the ducts of the abdominal gland to a point somewhat caudad to the genital orifice. Medial branches of the second inguinal gland, extending near the mid-line, nearly surround the vagina.

A comparison of figure 5 with figure 13 of an earlier paper (Myers, 'IG, p. 385) shows that in male rats of nine and ten weeks there is no progressive development of the mammary glands corresponding to that just described in females. In figure 5, the second inguinal gland is seen to present very few branches, thus offering a marked contrast \nth the same gland of the female. In most of the individuals examined, cleared preparations failed to reveal the presence of the second inguinal gland. The first inguinal and abilominal glands are sejiarated by a definite space, there being no overlapijing. It should be mentioned, however, that in one male of nine weeks the ducts of the first inguinal and abdominal glands were found to present a rather dense arborization. The ducts of the two glands in this case overlap to some extent. Even in this indi\-idual, which apparently represents an unusual variation, the glands are still distinctly less extensively developed than in the females of corresponding age.

Microscopic sections reveal large lumina in the milk ducts of the male. As was found in females, the end buds now possess .somewhat larger lumina than the duets into which they open. The primary duct when traced toward the surface of the body is seen to pass from the tela subeutanea into the corium where it makes a very oblique angle with the surface. Before reaching the surface, however, the lumen is seen to end in the duct within the corium in all glands studied at this stage. In one case the jirimary duet ended in very close relation to the at

2 in


tachcd ciul of a hair follicle. Tlu" hmicn of tlic duct could not be tracml into tho hair follicle, but the relation of the duct to the hair follicle is strongly suggestive of that existing in lower forms of mammals.

Fig. 5 Drawn from a cleared preparation (internal view) of a male alliino rat nine weeks old (body weight 05 grams). X 5. Compare this figure with figure 13 of an earlier paper (Myers, '16), p. 385), showing the much more extensive and complicated branching of the ducts in the female of corresponding age. .4., left abdominal gland; H., left first inguinal gland; C., rudimentary left second inguinal gland. The drawing.-) do not represent the glands in their natural spatial position with reference to each other.

It is interesting to note that in several of the jiostnatal stages studied it was po.ssible to obtain a few intiividiuils of appro.ximately the same age whose gross body weight was somewhat above or below the normal for that age. In general it may be said that in both sexes the mammarj* glands convspond somewhat


with the gross body weight of the animal, i.e., in those indivi(hials noticoably iindor woight the glands are somewhat smaller than in those of normal and above normal weight. When the body weight is far below normal the glands of the female do not show such a dense arborization at the age when puberty normally appears. Perhaps this is due to the fact that the arrival of jjuberty is slightly delayed.

A few male rats between the ages of ten and sixteen weeks were examined. In all observations made the glands of the sixteen week rats were found somewhat better developed than in those of nine and ten weeks. Apparently considerable growth has taken place and new ducts have developed as the arborization of ducts appears more dense.


The nipple

In the fetus at eighteen days in the male albino rat the surface of the future nipple area appears somewhat lighter than the surrounding epidermis. It is very rare to find a developing manunary pit in the male at this time. In nineteen day male fetuses an occasional gland can be recognized from the surface, but with difficulty, while at twenty days there is no external indication of the mammary glands in the male. About two weeks after birth, when the nipples of the female become so conspicuous owing to the lack of hair around the nipples (Jackson, '12) and to the very rapid growth of the nipple itself (Myers, '16), there is no trace of a nijjple in the male. The hair develops quite uniformly over the entire abdominal and thoracic wall. From this stage tlirough the ten weeks stage no nipples were observed in any of the male indi\'iduals studied (cf. figs. 6 and 7). Microscopic examinations of all stages likewise show the absence of nipples.

The absence of nipples in the male rats is contra.>*ted with their presence as observed by many workers in a large numl>er of animal species, including man. \Miile the nipples of men are rudiment ar\', they are absent only in anomalous cases.



It is a well known fact tliat nidimontary nipples also appear in the males of many of our doniestie animals. Acrording to Schiekele, Kschrit-ht CW) found rudimentary nipples also in all males of the whale family.

In tlie monotremes, Owen ('32) noticed that mammary glands are present in both sexes. Later, Bresslau COS) and ('12 b) in monotremes observed mannnary glands in males which agree in shape and size with those of feniales. While nipples have been

P'ig. fi DrawinR «f tlip vcnlral view of a young adult foiiialc albino rat in lartation. Hair removed from ventral wall. This specimen shows the absence of the left third thoracic gland. This is one of the most common variations in the Dunilier of nipples in the female.

rig. 7 Drawing of the ventral view of a young adult male albino rat. Hair removed from ventral waft. No traces of mammary gland nipples are present.


doscrihod in inulc marsupials by Laurent ('39) and Katz ('82), yet Lechc ('97) and Bresslau ('12 b) believe that nipples do not ordinarily appear in male marsupials of the Australian species. According to Schickele ('99), Rapp ('52) states that generally no trace of a nipple appears in the males of Edentata.

Schickele ('99j in studjing the number, position, and arrangement into gi-oups of the nipples in the mouse, found it necessary to exclude adult males since in them he was never able macroscopically to recognize even rudimentary nipples with absolute certainty. In his macoscopic observations on adult rats he likewise found in males no nipples, not even rudiments. In one young male rat, however, he found 12 nipple anlages arranged the same as in female mice.

Our present knowledge therefore leads to the conclusion that in monotremes, the mammary glands of the two sexes reach about the same stage of development while in the higher forms of mammals the mammary glands of the male appear in a rudimentary condition. The existence of mammary glands in the males of some species of marsupials seems doubtful. In some forms (chiefly rodents) the nipples in the males have entirely disappeared.

Milk dtids

In the present investigation the primary milk duct of the male albino rat was observed in eighteen day fetuses to resemble somewhat a large developing hair follicle. .\t nineteen days the duct has largely lost its resemblance to a hair follicle. It has grown in length, and in some glands presents secondary and tertiary ducts. The twenty day stage shows the ducts in approximately the same stage of development as in females of the same age. (^ne difference, however, is in the appearance of a branch from the primary duct of males inunediately beneath the corium. From birth to five weeks, the glands of the two sexes are aliout parallel in their development. .Vt five weeks, the glands of the male l)egin to lag behind of the female, the chief difference being that a smaller number of collateral branches appear on the ducts of the male. .Vt nine weeks (about the time of puberty), the ducts

220 J. A. MYEU.S

of the foiiialc show :i vast nuinluT of l)ranchf's \vl>il<' those of tlu> male arc but shRhtly furtl\er ilevt-lopi'd than at five weeks. The marked contrast at nine weeks can be best ai)[)reoiated by (•omi)arinK figure ") with figure \'.\ of an earlier paper (Myers, '16).

The resemblance of the early iiriniary duet to a hair follicle nuiy possibly have a phylogenetie significance. IJresslau ('02) found in marsupials that in the regions of the future mammary glands large hair anlages develop. From one side of each anlage a sebaceous gland buds off, while from the other side a milk duct develops. As time goes on the free end of the hair anlage and the sebaceous gland atrophy, but the milk duct develops very rapidly and retains its opening into the attached end of the hair follicle. In squirrels (Sciurus vulgaris), Bresslau ('12 c) found that the mammary gland anlage in both sexes divides into a medial and a lateral part; from the lateral part the milkducts develop, while from the medial part a bristle-like tactile hair develops, which in adult animals far surpasses in length and thickness the other hairs of the body.

Develojiing hairs and sebaceous glands closely related to the mammarj- gland anlage have also been described in man by Kggeling ('05), Brouha ('05) and Lustig ('16).

My figure 7 somewhat resembles Bresslau's figure 5 (Miinchener Med. Woch., .Jahrg., 59, S. 2794). In a later stage, it will be recalled, a branch was seen to spring from the side of the primary duct anlage near its attached end. It is possible that this branch may correspond with the sebaceous gland found in marsupials by Bresslau ('02) and in man by Lustig (see Lustig's fig. 10). Therefore the manimarj^ gland anlage of the male rat at the time when it appears similar to a hair anlage, is perhaps repeating the conditions found in lower forms. The develojjing niannnarj- gland of the female probably through a similar stage, but owing to the fact that it is much more highly specialized than the gland of the male it has more stages to pass tlirougli, and the earlier stages are doubtless passed over very rapidly.

The fact that after the fifth week of postnatal life the glands of the female are a little better developed than those of the


male albino rat aprccs with the findings of Kiilliker f'70). He states that in huinau from one to ten years after hirth the milk ducts and end buds increase only a little, branching somewhat more richly in the female than in the male sex. Krause ('06) also states that there is no essential ditTerence in the glands of the two sexes at birth; the differences belong to later perirxis of life. Broman ('11) states that in the human the glands of the male do not develop after l)irth. In the rat. however, my figures 4 and 5 show that in the male there is considerable growth in the ducts after birth. This growth apparently merely keeps pace with the general body growth.

The tremendous differences which appear between the glands of the two sexes of rats about the ninth week correspond to the differences which have been described as occurring in the mammary gland of many species at about the time of puberty. The extent to which the mammary gland of the male develops is somewhat variable. At nine weeks as shown above most of the male rats show a small amount of branching of the ducts while others of the same age present a fairly dense arborization of milk ducts. The male albino rats examined after puberty by me show that slight development continues after that time.

While examining the manuiiary glands of many men, Cooper {'40) found that there is considerable indix-idual • variation. Ordinarily it is stated that the milk-ducts of males cease to develop after puberty and very soon pass through the stage of involution (McMurrich, '15, Jordan, '1(5. and many others).

Kiilliker ("T'J) in men of twenty years, found manmiary gland vesicles which were not found at about the time of puberty. Schenk found end vesicles in the glands of men, but makes no mention of their age. It was not until about the thirtieth year that Kolliker found regressive metamorphosis beginning in the manunary glands of men. Furthermore, Merkel ('99) and Kerr ('1(3) state that occasionally at the time of puberty some swelling and soreness are experienced in the manunary glands of men. Therefore it must be concluded that the mammary glands in the male may show some (although slight) growth after puberty.

'2'2'2 J. A. MYKIIS

In the (loscriplivo part of this paper it was shown that the nuinlHT of nmniniary glands is more variable in the male than in the female rat. In the female rat it has been pointed out by Schickele ('99), Henneberg ("00), Frank and Unger ('II) and Myers ('!()) that an oeeasional supernunu^rary gland develops. The second or third thoracic glands may be lacking, but the others are always present in the female. The first and second inguinal glands as well as the .second and third thoracic glands may be missing in the male. In all males examined the second inguinal gland when jiresent is in a very rudimentary state.

As preA-iously shown (Henneberg, '00, Myers, '16) in the female rat the second inguinal gland of each side is located laterocephalad to the urethral orifice. The tlucts of this gland for the most part take a caudal direction from their epidermal attachment. Ultimately the branches extend caudad to the vagina for a considerable distance. In the male the second inguinal glands have approximately the .^ame location. Caudal to the genital eminence, however, the integimient becomes the outer wall of the scrotum. The skin of this region, being specialized so as to form a jiart of the scrotum, apparently forms a poor medium for the development and ramification of milk ducts.

The fact that the ducts corresponding to several of the mammary glands are often absent, together with the fact that the nipples are very rarely developetl indicate that the mammary glands may be gradually disappearing in the male rat. On the other hand, the ducts present in the male so closely resemble of the female until the time of puberty that it is to believe that the male milk glands were at one time functional, and with the proper stimulus might again develop into functional glands. Steinach ('12) found that the manmiary glands of male guinea-pigs sometimes become functional after a pair of ovaries have been transplanted beneath the skin. Owing to the absence of nipples, Steinach regarded the rat as unfit for such experiments.


Occurrence of mammary glands in both sexes

Several theories have been advanced regarding the occurrence of the nianiniiiry gland in hoth sexes. Barkow ('20) pointed out that in forms lower than inannnals the brooding may be done by either the male or the female, or the two sexes may share equally in this work. Ruge ('95), Gegenbaur ('98) ami Wilder ('08) believe that the mammary glands first appeared in the female and were later transmitted to the male. On the contrary, Haacke ('93) thinks the mammary glands were transmitted from the male to the female.

A more logical view is supported bj- Paul ('84), W'estling ('89), and Rresslau ('08, '10, and '12) who after carefully stud\ing the mammary glands of lower mammals concluded that from its beginning the manunary gland was developed to the same extent in both sexes. The female gi-adually took the responsibility of nourishing the offspring and the mammary glands in the male became correspondingly nonfunctional.

The present work on the albino rat lends support to this theory. In the earlier fetal stages the manunary glands appear the same in both sexes. As development progresses the glands of the male lag b(>hind those of the female, and no nipple is formed. When the adult condition is reached, the glantls of the male possess no nipple and the milk ducts are in a somewhat rudimentary and variable state. Such a condition, together with the relations fouiul in lower forms, lead one to conclude that the manunary glands were in ancestral forms eiiually developed in both sexes. The individual male rat in which Schickele('99) found a complete set of nipples probably represents a reversion toward the earlier ancestral type in which the mammary glands of the two sexes were more nearly alike in structure.


The results of the present study of the development of the manunary glanil in the male rat may be summarized briefly as follows :

224 J. A. MYERS

1. Wliilo the SOX glands arc in the inilifforont omliryonal stage, there is no apparent ditTerence between the anlages of the niaininary glands in the two sexes.

2. In fetuses of eighteen tlays the mammary gland anlages of the male differ from those of the female in that they possess no Miaiiimary i)it. In twenty day f(>tuses, when the nipjilc anlages are present in females, no trace of a nipjile was observed in males. In fact, nipples apparently fail to develop in the male at all, ineUuling postnatal stages. Hair develops uniformly over the ventral surface of the body of the male leaxnng no macroscopic indication of a mammary gland area.

3. Like the nipple, the epitheUal hood is also absent in male albino rats.

4. The primary milk duct in eiglitcen day male fetuses resembles a hair anlage. At twenty days the branches appear ap)proximately as were found in females of the same age. Until about the fifth postnatal week the milk ducts of the two sexes are approximately parallel in their development, .\fter the fifth week, the ducts of the female present more branches than those of the male. The difference is not very great, however, until the ninth week (age of puberty). Then the ducts of the female branch very profusely while the ducts of the male show but little change.

5. In the male the lumina of the milk ducts develop in the -same manner and at about the same time as in the female.

6. The second inguinal gland of the male is very rudimentary and may be entirely lacking. This condition is evidently due to the development of the scrotum in the region which the second inguinal gland normallj- occupies.

7. The number of manmiary glands is more variable in the male than in the female. No supernumerary glands were observed in the male. The second and thinl thoracic and the first and second inguinal glands are sometimes absent in the male. Usually not more than one or two of these glands are ab.sent in a single indi\-idual. The other manunary glands were present in all rats examined.



Barkow, H. C. L. 1S20 Anatoiiiisch-physiologisclif UntersurhiinKcn, vorzOj?lich iiluT das Schlagadersystom dcr Vogcl. Meckel's Arch. f. Anat. u. Physiol.

Bresslau, E. 1902 a Beitnige ziir Entwickelungsgeschichte der Mammarorgane bei den Bcutelthicren. Zeitschrift f. Morphologic u. Anthropologic, Bd. 4.

1902 b Weiterc Untersuohungen Uber Ontogenic und Phylogenic des Mammarapparatcs dcr Siiugetiere. Anat. Anz., Bd. 21. 1908 Die Entwickelung dos Mammarapparatcs der Monotrcmen, Maraupialior und ciniger Placentalier. I. Entwicklung und I'rsprung des Mammarapparatcs von Echidna. Seraon's Zoolog. Forschungsreiscn, Bd. 4. Liefcrung 5.

1910 Der Manimarapparat (Entwicklung und Stammesgcschichte). Ergebn. d. Anat. u. Entw., Bd. 19.

1912 a Uel)er Hyperthelie. Miinchener Med. Wochcnschr., Jahrg. 59, S. 2793-2795. "

1912 b Die Entwickelung des Mammarapparatcs der Monotrcmen, Marsupiolier und einiger Placentalier. II. Der Mammarapparat des erwachsenen Echidna-wcibchens. Scmon's Zoolog. Forschungsreiscn, Hd. 4 ( Jcnaischc Denkschr., Bd. 7,1.

1912 e Die Entwickelung des Mammarapparatcs dcr Monotrcmen, Marsui)ialier und eiiiigcr Placentalier. III. Entwickelung des Mammarapparatcs dcr Marsupialier, Insectivoren, Xagethiere, Carnivoren und Wicdcrkauer.

Brouan, Ivak 1911 Normale und abnorme Entwicklung des Menschen.

Brouha, Dr. 1905 Recherches sur les diverses phases du developpement et de I'activite de la mamelle. Archives de Biologic, T. 21.

Cooper, A. 1S40 On the anatomy of the breast. London.

EoGELiNG, H. 1904 I'cber cin wichtigcs Stadium in der Entwicklung der foctale .Mamma beim Menschen. Anat. .\nz.. Bd. 24. 1905 I'cbcr die Driisen des Warzenhofs beim Menschen. Jen. Zeitschr. f. Xatunv., Bd. 39.

EscHRicHT, 1849 Die nordischen Walthicre.

Frank, R. T., and Unoer, A. 1911 An experimental study of the causes which produce the growth of the mammary gland. Archives of Internal Medicine, vol. 7.

Geoenbaur, C. 1898 Vcrgleichende Anatomic der Wirbclthiere, Bd. 1.

Haacke, \V. 1885 On the marsupial ovum, the mammary pouch and the male milk glands of Echidna hystrix. Proc. of the Royal Society of London, vol. 38.

1893 Ueber die Entstehung des Saugethiercs. Biolog. Centralbl., Bd. 8.

Henneberg, Bruxno 1900 Die erste Entwickelung der Mammarorgane bei der Ratte. Anat. Heft, Bd. 13.

Jackson, C. M. 1912 On the recognition of sex through external characters in the young rat. Biological Bulletin, vol. 23.

220 J. A. MVEHS

JoKUAN. H. K., AND FKIlcrsON, J. l!»lpnorKiine. IlertwiR's EntwickoluiiKslchre dcr Wirlicitierc, Hd. 2, Tcil 1-2.

Lane-Clay I'o.N, .Mis.s J. K., and STAUi,iN(i. 10. 11. llXMi .\n experimental enquiry into the factors which determine the (srowth and activity of the mammary glands. Proc. of the Uoyal Society, Series B, vol. 77.

Latrent, M. 1839 Kecherches anatoniiques et zoologiques sur les raaramifercs marsiipiaux. .\nnales d'anatomic et dc physiologie, T. 3.

LrsTiG, Hilda 1916 Zur EntwicklunKSRCschichte der menschlichen Bnistdriise. Arcliiv. f. mikr. .Vnat., Bd. S7.

McMrRRicii, J. P. 1915 The development of the human hrxly. Kifth edition, Philadelphia.

Merkel, Fr. 1899 Handl)uch der topographischen .Vnatomic. B<1. 2.

Myers, .1. A. 1910 Studies on the manunary gland. I. The growth and distribution of the niilk-<lucts and the development of the nipple in the albino rat from birth to ten weeks of age. Am. Jour. Anal., vol. 19. (Abstract also published in the Anat. Rec, 1916, vol. 10, p. 230.) 1917 a A comparison of the mammary glands in male and female albino rats. (Abstract^ Anat. Uec, vol. 11, p. 391. 1917 b Studies on the nuinunary gland. II. The fetal development of the mammary gland in the female albino rat. Am. Jour. Anat. fin press).

Owen, R. 18,'{2 a On the mammary glands of the Ornithorhynchus paradoxus. Philos. Transactions, vol. 122.

1S32 b On the mammary glands of Echidna hystrix. Proceed. Comm. .Science and Corr. Zoolog. Soc, London, p. 2.

•Pail, H. 1884 Ueber Hautanpassung der Siiugethiere, Jena, II. Pohle.

Rapp, 1852 Die Edentaten.

Rl'OE, G. Die Ilautmusculatiir der Monotremen und ihre Bezichungen zu dera Marsupial-imd M.'inunarapparate. Semon's Zool. Forschimgsreisen, Bd. 2 (Jenaische Denkschriften, Bd. 5).

ScHIcKELE, G. 1899 Beitriige zur Morphologic und Entwickelung der normalen und id>crz;ihligen .Milchdrliscri. Zeitschrift f. Morphologic und .\ntliropologie, Bd. 1, Heft 3.

SciiE.VK, S. L. Istologia normale dell'uomo (ital. transl. by Monti and Golgi).

Steinach, E. 1912 Willklirliche I'mwandlimg von Siiugetier-Miinnchen in Tiere mit ausgepriigt weiblichen. Geschlcchtscharakteren und weiblicher Psyche. .Vrchiv fdr die gesammte Physiologic (Pflilgcr's).

•VVestlino, Ch. 1889 .Anatomische I'ntersurhungen iiber Echidna. Bih. Akad. Hanill.. Stockholm. Bd. 15.

Wilder, H. H. 1909 History of the human body.

•The original pajK-r was not available for this work.



II. !•;. HADASCll AM) J. I. FANZ From tlw Diiiiitl Haugh Inslilulc of Anatuiny of the Jefferson Medical College


This apparatus was originally dcsignod for use in the preparation of a sludcnts' loan cnlioction of normal iiistoifipic specinicns, in order to enaltle llie technician to carry twenty to folly sections at one time, tliroucli iIk' various s1e|)s of cleparaflinization 1o clearing, williout mani|)iila1inK tiie indi\i(lual slides. 'I'hc ;;rea1 i)ractical utility and efficiency of the device, liowex'er, weic soon evident and the wide range of use to which it niiKht he api)lied, .seemed 1o warrant its description.

It is clearly manifest that in the preparation of SOOO or more slides that anything would minimize the routine labor to thi- extent of 20 to 30 per cent should have an important place as a labor-saving device. By using several ba.ske1s, many slides can be pn^pared in a short time, insuring uniformity in stain reaction, etc.

'rii(> basket can be used most satisfactorily in the following work:

1. In the |)reiiaration of students' loan .sets of normal histology, jiathologic histology, botany, zoology and wherev^r many slides are to be UKule.

2. In serial section work and in reconstruction of embryos, small organs, tumoi-s and botanical or zoological specimens.

3. In rovitine (hospital) pathologic histology, where a variety of tissU)' si)ecimens obtained at necropsy is to be studied histologically.

As the basket is adjustable and will hold any sized microscojie slide, its is exten<led over those ty]>es that are limited to one size. The slides today are not of uniform length, as formerly and adjustable featiu'c has been very ojjiioiluiu' in th(> pres(Mit work.

In the constiuclion of the apparatus alumimnn was sel(>cted ami used throughout i\s most suitable from a .standpoint of lightness, durability and workable (lualities. .Mumimnu is effected to a destructive degree by strong alkalies, acids and mates. These fortmiately do not enter into the ordinary tcchnic of histology or pathology. Strong iodine solutions coriode this metal slightly, but in the strength useil for the removal of bichloride of mercury crystals, if does no harm in the time re(|uired for this process. The selection of this metal was the result of the expericiu'e of one of the authors who devised a ten to twenty slide basket of thin sheet alumimnn which answers very well for serials, of a hundred or .so, of 2 by 3 inch slides.'

'Anat. Rcc. vol. .3, no. 2, February, 1909.



Conslruclinn. Figure 1 sliows llic iiiiassciniilcil parts ami illiistralos the wnil«rlyiii(j piiiu-iplc of siinplicily and case of const nicl ion. Tlic haw and top plates l.l and li) arc oj indies lon^, I ,'\i inelies wide and 1 ineli tliiek. These are K<)oved to liohl the slides. The ^^I■ooves arc j^ inch wide and J inch deep and a full inch loiiK, leavinM; a Icdjic at the l):u-k to prevent the slides from slipping IhrouRh. The pattern for these plates was prepariMl l)y making a pla.ster of paris of the side of an old pilLsbury slide l)ox, which liad uiuisually wide and ileop

Fig. 1. Unassemljiod parts of the Imskot.

grooves. Twenty of t hese urooves were utilized and sufficient additional lenRth allowed for the support rods.

The platct* arc drilled for anil supported l)v four aluininuni rods each 4 inches lonjf and J inch thick. These are threaded at, hoth end.-* for about 1 mch and arc pro\-ided with two aluminum nuts for each end. The nuts are made out of J inch sheet aluminum. The handle is made of the .s;ime size aluminum ro<l and |) throush only the top plate. It evtends :il>oMt -i inches above this plate, thereby makinn it convenient to nr^'-'^P "' handle ami support the back of the top plate with the index finK<'r while washing the si>c1ions with alcohol or water from Wiish bottles. The plates were cast a( a foundry, but could, no



doiiht, ho readily stamped from sheet aluminum witli a suitable 8t«el die.

The lieif^ht of the hjifiket over all is (ij iiielies, the weight is 1G5 grams. The a.sseiuhled ajjparatus is shown in fij^ure 2.

One inipoitant consideration in the dimensions of the apparatus Wius the jar to be u.sed. In order to have as little wjuste Hpace ;i.s po.ssible and yet accommodate the handle so that the jar could be clo.sed to prevent evaporation of the various solutions, a specimen jar was chosen. The iiLside dimensions are as follows: height, 7 J inches;

Fig. 2. Parts assembled auci filled with slides.

wiikh, .").', inches; dei)th, 21 inches. The top edge of the jar and the glass cover are ground so that a close fit is insured. To the top three pieces of cork are cemented with asphaltuni. The one at the midille of the exp<)s«Ml surface .serves as a liandie. wiiile the other two (thinner) are undeiiieath an<l near the ends. The latter .serve !us stops and prevent the cover from slipping off accidentally, .^^|>haltum answers for all covers exce|)t that for the xylol y.w. The xylol softens the asphalt um and causes it to lun. I'"igure A shows a set of slides in a jar of alcohol before staining. Forty of these thin slides couhl readily have btvn placeil, back to buck in the basket, but oidy twenty were



HM'd in till' illiistratidii. An .•iliiininuni wrcncli for tinlilcninp; the mils is :ils(i shown in llir illustration. This prevents hiiirinn pri>ihiced liy a wrench of harder metal.

Nine of these jars con.slitule an ordinary .set and are lahelled (both eover and jar) for the following reaK<'nts: xylol, alcohol, iodine, alcohol, water, iieinatoxylin, eosin. picrofuchsin and alcohol. If other stains as pr.racarmin or safraniii are used, extra jars are required.

Fig. .'}. The l>a.><kcl in an iipproprialc jar and alinniiniin wrcncli for tlu- nuts.

The sections :ire placed in the basket and (I) transf(>rred to the xylol jar for about live minutes to remove the paraffin. The bjisket is then removed, dr.'iiiicd anil washed with !).") |)er cent alcohol (by wash bottle) and placed (2) in the ioilin jar for five minutes to remove the bichlorid of mercury cry.stals. Tlie .s«>ctions are again Wiu*hed with alcohol and placed in C.i) the alcohol jar to remove the remainder of iodiii.


If llic fixinn jiKcnl coiilains no liiclilorid of mercury, ilic-c nvn >ic|)s arc iiiiiicccssary. 'I'lic lias^kct of slide-* is (lien placed in a luicket of wafer llo remove llie alcolml) and llien transferred for five minutes io (4) llic licmatoxylin jar. Tlie hematoxylin is llicn washed olT hy placinji till' l)asket in a bucket, or hatlery jar of water and then tratisforred to (.")) the jar of tosin (0.5 i)er cent for two mimites), or picrofuclisin (one and one-half minutes).

The slide- are then washed in water a.s before and allowed to stand theriMU for several minutes. This is especially impoHant after an a(iueous eosin has heen used, as the water seenis Io fix it in tin- tissties. If the^e are transferreil U> alcoiiol too soon, (lie eosin is very rajiidly removed by the deliydr.aiiiin alcohol.

The liaskei of slide- is next washed willi alcohol ('.•") per c<'nl) to remove as nuicii of the water as possible and then transferred to a flat dish of alcohol to ((>) dehydrate. The writer pn-fers to clear each .slide individually an<l not the basketful at once. The use of a flat tray for dehydr.ition en.ables the technician to pick out each slide without liflinf; the baskei. The sections are taken out one at a time, cleane<l, blotteil and coveicd with (7) creosote to clear. This is left on until all Iwenty have been so treated and then lliey are taken up in sequence, diained. blotied and (S) mounted in b.-ilsam.

The ;iicohol used for dehsdralion and reuKjval of iodin is saved, filtered throunh animal charcoal and u.s<'d in the embalming mixture for the preseivalion of cadavers, so that there is very little wa.sted. The creosote is also saved and used a second time, ius its use has not been imjiainMl.

.\s the writers are at present in the prep;iralion of a stud>' in which the sections are ai least 1 inch l)v '. inch, anoiher apparatus .and sei of slides had Io be devised Jo meei this s|)eci;d re(|uirenient. li}iure 4 shows the adapiabiliiv of ihe prece(lin}; apparatus in which the slide..are 10 inches loiiy; and 2 inches wide. For .a serial coiisi-iiTifi of about li.'iOO Io 1000 .(•(■lions, each one by J inch, it is out of the (|Ue-(ion to consider 2 l)y '.i incli slides. For this reason larfjer slides had to be considered and apparaiiLs suitable thereto con.structed. The ])rincipl(> of construction is the same, the Mippoil rods being cut to any desired lennlh and ihreaded in the same manner as before. The slides were m.ade fioni S by 10 inch and 10 by 12 inch and larner waste negatives from ihe PhoiogiMiiliic anil X-ray Depanment of ihe .lelTerson Medical llospilal thiough the couiiesy of Frof. Willis I. Man':(\s. The emulsion was easily and ihoiounhly removed: ihe plales were then cut into the (lesire(l sizes and Ihe edges filed. ■ The ])rep, (ration of 200 such slides recpiired a comparaliv(>ly short time, ("over nla.s.s(\s for these can readily l)e made in the same manner from tliin 8 by 10 inch negatives ii>- they suffice where general low power objectives are to be used. I'or higher power work ph()t(>gvapliic celluloid is lH>ing ex|H'rimeiiied wiih to test its .applicability.

Suitable reagent dishes for large biuskets can be obtained at Ihe various instrument dealers as the trays handled by them are


H. K. RAnAsrn wn j. i. fanz

stniinht -sided. (Uass plates cut to the proper size ami Mipplicti with corks as in tin- jirecedin^, servo very well for covoi"s.

FiKUre ."> shows three sli(U>s 2 by 10 inches. A, shows a plain side; li, shows tile paraffin sections freshly nioinited; C, shows a finished slide. These sections are thronnh both ventricles of the heart of an adult cat.

Fig. 4. Slide basket adapted to 10 inch slides.

Fig. 5. Three lit inch slide.*!, .t , philii; H, puniflin sections mounted; C, slide completed.


nr Till: i)iiiLK>i.)i(Ariiic hkrvkc ocrnDcn 13.


J. A. H(J\VKLL From the Anatomical Laboratory of the University of Missouri


That tho growth of tho bono is dcpoiulent, to 8omo extent, at least, upon the amount of stress and strain exerted upon the bone, has been well established. It is strongly indicated by the remarkable mechanically adapted arrangement of the trabeculae and com]iacta, which shows most strikingly in the vertebrae and bones of the leg in man, and which has been analyzc<l chieflj' l)y II. \nii Meyer, .1. Wolf, and (luite recently by R. Koch ('17). It is i)roven by studies which have been made on bones in which, as a result of injury, new directions of lines of force have been introduced with a resultant formation of entirely new sets of bone trabeculae — studies which have been made especially by W. Roux and .1. Wolff. This literature has bi>en fully renewed recently by R. Koch to whose article those particularly interested are refeired.

Undeniable as it is that mechanical factors exert some governing influence over the growth of bone, there still remains open the question as to how far-reaching this control is. The extreme mechanical view is held by R. Thoma ('07), that all bone formation, even including the first bone formed in the embryo, takes place as a response of bone forming tissue to the action of stress and strain. He holds that growth in thickness, in human bones, begins when the stress and strain exceed a certain minimum, which he estimates roughly to be equivalent to a weight of 0.6 gi-ams per square millimeter of cross-section, acting over twenty-four hours. If stress and strain becom" increasingly greater, new bone forms with increasing rapidity until a certain (undetermined) optimum is reached. Beyond tliis. new growth



234 .1. A. IIOWKI.I.

Inctuno slower, until, after a eertnin (iiiuleterniiiied) inaxinmin is passed, hone resorption takes place. Hone resorption also takes plaee, if the stress and strain fall helow the nuninuiin.

The prow til in length difTers, aeeording to Thonia. in that no diminution lakes place, no niatti r liowhigh orhow low the aninunt of stress and strain. 'I'lie miniinuni i)n ssure, however, needed to stimulate growth in length, he jilacts at the same point as for gi-owth in thickness. One gains the imin-ession that Thoma considers the regulation of hone formation hy the mechanical effect of stress and strain to be absolute.

There is, however, already nuich evidence against this extreme \-iew. Thus Oilier ('()7), in his classical trans])lantation experiments, found that periosteum, transplanted from the tibia to the comb of a rooster, forms a shell of bone — clearly without the action of stress or strain, .\gain, K. H Weber (cited in Herbst, '01) found, in a new-born calf which had no spinal cord below the cer\'ical region, ami no muscle in the ]iosterior half of the body, the skeletal iiarts well ileveloped \\'hilc tin y were only half as heavy as normal, and the joints were ankylo.'^ed, still they had difTerentiatetl and undergone a considerable amount of growth, while deprived of the stress and strain exerted by the normal intraembryonic movements It is, moreover, a matter of common observation, that considerable growth takes place in a paralyzed arm, particularly in length, though it lags well behind the healthy arm, in thickness

There seems, therefore, to be need for an experimental analysis, in order to determine whether growth of bone takes place in the absence of mechanical stress and strain, and, if so, how much, and to how great an extent it compares with the normal growth.

A beginning was made by I'ottorf ('Hi), working with i)uppies. He performed two experiments: in the first he diminished the stress and strain exerted on one of the fore-legs of a three weeks puppy, by holding the leg in a sling, so that no weight was borne by this leg. .\t the end of three weeks he foimd the bones of the two fore-legs to be about the same length, while the bone in the legs walked on were considerably thicker than in


the other. In a socoiid oxporiiiipnt, one log was paralyzed by euttinp; the nerves of the Inachial plexus, thereby eliminating, in addition to the stress and strain due to the weight of the body, that due to inusele pull. After twenty days it was found, again, that gi-owtli in length had tjeen nearly the same in the bones of the two legs, while theie was a very gicat difference in thickness — the bones in the unoperated leg l)eing, in some places from two to four times as thick as those in the immobilized leg. Ossification had taken places in on tlie operated side, although the size of the epiphyses was slightly than on ilic unoperated siile.

In the present study it was suggested by Dr ("lark that the work started by Pottorf be caiTied to older stages. In order to obtain the maximum information from the minimum of experimental material it was decided to employ the X-ray. There would thus be obtained a series of records on the same animal, before operation, and at regular intervals afterwanl. \\'hile X-ray pictures give a somewhat confused record, as compared with studies of the bones them.selves, and do not show the cartilages, still, for comparative purposes, they give sufficiently accurate information as regards length, diameter, shape, size of epiphys(>s. union of epiphysis with diaphysis, and a fair indication of thickness. This method, moreover, has the advantage that the recorils of the different ages are all taken from the same animal.


The method of ope rating was the same as that used by Pottorf. Two fox teni( r i)Ui)pi( s, aliout four weeks old, were used for the experiment. In both animals the operation jierformed consisted in cutting the main nerves of the brachial plexus on the right side, in order to j)roduce paralysis of the muscles. In both ca.<( s the operation was carrii d through with very little hemorrhage — the axillary artery and v(ins and their chief branches were undisturbed. In on(> of the jnippics a limited amount of movement was regained after a few weeks, showing that the ojuration had not been complete This animal was, therefore, discarded. Ill the other, control was maintained over the pectoralis and



lntissiii\us ilorsi nuisclos, so that a certain amount of muscle ]nill was exeit(il on th(> iipjM r part of the huinenis. This served to hold the liiiil) close to the iiody. and made the use of a sling unnecessary. All voluntary conlrol over the muscles moving the hoiu s of the elhow and carpal joints was lost, the contracture of nnisch s produced a giadually increasing fixation of the elbow joint in a sharply flexed position. No weight was borne on the leg of the operated side.

Ill aling by granulation took place within a few days after the operation, and the atTictid limb nMnaiiu d Iwalthy throughout the exi)(riment. Especial care was taken in fci dinji. :ind the wtight. recorded weekly, is shown in tal)le 1:








poll nda




•NoVCIIlIxT 11.





9 4


Novoiiil'cr 17.


3 9






Novi-mbiT '2o,








•December 1,






11 1


Decomlier 7,


6 35




II 41


December 1.").






11 G


•December 21.


7 3




11 Gl


Detemlier 2'.*.





10 .

11 Gl

.I:inu:irv ti.

11II7 .

S 5





-0 01

•.l!iiiii:iry l.'t,

1917 .





11. S


• Dates at \vhich X ray plates were taken.

The table shows that there was steaily increase in w( ight, at an average rate of seven-tenths of a pound per week for the fu'st twelve weeks. For the last seven weeks th(> increase tiropped to an average of one-trnth of a pound per week. The general hialth, however, was < xc* Ih nt throughout, and measurements of the i>ones from the X-ray platts show that giowth continued throughout th(> period of the (XperiiiKiit, including the last four weeks when the weight was practically stationary.

X-ray pictures were taken just befem* the etjie ration, tit approximate'ly three week intervals throughout, anel imnu'eliately after ele'ath. In orele^ tei e)btain e'le-ar pictui-e's light e'the-r ane-sthe.sia


was produced. The animal was killed with cliloroforin, nineteen weeks after the operation. Immediately afttr death the bones of the two fore-legs were remo\ed, cleaned, and various comparative studies made on the fresh bones. The bones were split longitudinally in order to study the thickness, architecture, epiphy.seal lines antl arrangement of Irabeculae. Later they were preserved in alcohol. Before splitting the humeri a cylindrical section 1 cm. high was taken from the center of the shaft of each bone to test for compressive strength.

In taking up the results obtained, the tlata furnished by the X-ray records will be considered first, and later the data obtained from the studj' of the bones after death.

In taking the X-ray plati s, the attempt was made to have the limb as n( ar the plate as possible, in order to have the minimum of distortion. A partially successful attempt was also made to have the bones in the same approximate position. \Miile measurements made from the plates are not so accurate as measurements from bones, still the probabihty of error is not great enough to interfere with the main results, particularly as regards the length of humenis, ulna and radius and the diameter of the humerus.

In figures 1 and 2 are given drawings, made from the X-ray plates, of the bones of the operateil and imoperated side at various ages. A comparison of the two in the successively older stages shows, m many ways, a considerable uniformity. The shapes of the bones in the unused leg do not differ greatly from those of the bones in the used leg. In length, again, the difference is not gi'eat. In diameter and thickness of compaeta. however, there is a marked difference.

In order to find out the exact lengths, etc., various measurements were made. The more striking of these are reproduced in table 2 and some of the results plotted in the form of curves, shown in figure ',i. It must be remembered that the lengths do not include the cartilages at the joint ends of the epiphysis, since these do not show in the plates.

A study of the tables and chart .shows that in length, growth of the bones in the operated side nearly kept pace with those in the



uiioppratcd side for six weeks after the openition. This agrees with the results of Pottorf, who found very Httlc difference in length on the two sides after twenty days. At nine weeks, however, the unused bones had lagged behind the used bones, and the difference increased slowly but steadily as long as the




FiRS. 1 and 2 DrawiiiK-', made from the X-ray plates, of humerii.s, radiii.i and ulna from the two 1ck«. Xovcmljer II shows the left leg just before operation. PicturcD of March 28 were taken after death. Two-fifths natural size.




FEB. 24 OP.


Fia. 2


J. A. llOWtl.L.


irNOTii or



LCNom or

t LN.\ IN CKN TiMrreHS

LENom or


DiAUKTcn or








S. o






s a


L o 11







Novi'inbcr 11

Dicembcr 1

nci-omber 21

J:iiiU!iry 13

i i'l)rii;iry 3

Ki-I>ru!iry 24

March 2S

Mcasurc'iiipiits of cleaned bones. .

6.2 7.3 8.2 8.7 9.1 9.3 9.6


6 2 7.35 8.3 9 05 9 5 9 8 10.15


0.05 0.10 .35 0.4 5 55


6.35 8.55 10 10 8 11.55 11.8 12 4


6 35 8,55

10 125

11 1

11 9

12 3 12 95


0.125 3 35 0.5 55


5.2 7.0 8.1 8.8 9.35 9.8 10.25

10 05

5 2 7.0 8.15 9.2 9.85 10.3 10 8

10 55

0.0 05 0.4 5 0.5 55


6 0.65 65 025 0.55 55 0.55


0.6 70 75 SO S75 92.5 10


05 10 0.175 0...25 375 0.45



cm. 13. 11.

11. 10.

9. 8. 7. 6.


rn cm.

10 0.9 0.8 0.7

































  • "•«.








eeks 3 6 9 12 15 18 21

Fig. 3 Chart showinK Rraphicaily the actual lengths of the two humeri and ulnae, and the diameters of the two humeri.


cxporiinent continued, until, at tho last record, tlio l)on('s in the used l(g were about ").") nun. longer than the corresponding bones in the umis(>d leg an excess of approximately 5 per cent. It should be emphasized, however, that, to the end, growth in length coiitinucil in the uiuised bones. Separate measurements show that this giowth in length took, jjlace both in th<> diaphj'ses and in the epiphyses.

ciia.\(;k in diamkikk

The measurements of the diameters at the c<'nter of the shaft of the humerus are given in table 2. Measurements of the diameters of the radii and ulnae were somewhat unsatisfactory, because these bones are not cylincb-ical, and sliglit diff(>rences in position introduced errors. This was not true of the humeri, however, which are more nearly cylindrical in the middle thirtl, and their diameters are represented in the lower curves in figure 3. The difference between the usetl and unus(Hl bones is very striking, and is noticeable even after tliree weeks. The measurements indicate that, in the unused humerus, there was first a slight increase (from 0.6 to 0.65 cm.) and later a diminution (to O.'iiy cm.) which remained unchanged for the last month and a half of the experiment, wliile the iliameter of the u.sed humerus increased steadily from 0.6 to 1 cm. The unusrd radius and ulna also showed but slight increase in iliameter, while in the corresponding bones in the healthy leg, the diameter nearly doubled.


M( asurements of th(> thickness of the compacta, as shown in the X-ray plates, were the least satisfactory of all the measurements made. An appreciable difference (0.025 cm.) between the two humeri, however, appeared at three weeks. At the (Mid of the experiment the difTerence had increa.sed to 0.1 cm., the thickness of the two bones, at corresponding jxnnts, mea.suring 0.15 cm. and 0.25 cm. respectively. The ulnae .showed similar differences, while the radii (owing partly to difference in posi

242 J. A. nowKi.i,

tion) showed a difTerencp of only 0.025 cm. a( tlic end of the oxpcriniont— as shown in tlic X-ray jilatcs.

Tlic difToronces in thirknc ss found l»y Pottorf at the end of twenty days were much greater than in this study at the end of three weeks, for he found the tliiekncss of tlie used humerus from two to four times as Ri't'at as that of the unused, at corre.spontUng points.


The X-ray plates show that the ossification of the epiphyses, and the union of epiphyses with diaphyscs progress uniformly in the l)ones of the two legs, the only apprceialjle ilitTerence being that the epiphyses are larger and slightly different in shape on the active side. In the distal end of the humerus the line between epiphysis and diaphysis is obliterated between February 3 and February 24, in both legs. The proximal epiphysis of the ulna united with the diaphysis between February 24 and March 28 in both legs. The remaining einphj'ses of the humerus, uhia and radius were still ununited in both legs at the end of the experiment.

The thickness of the epiphyseal cartilages was uniformly greater in the used than in the unused bones.


As already stated, as many studies as possible were made after the death of the animal on the fresh bones — they were cleaned, weighed, drawings and measurements were made, etc., they were spht longitudinally, and blocks were taken from the humeri for testing the crushing force. They were then preserved in alcohol, and further studies made of the architecture as shown in the split bones.

The finding at the site of operation and of the muscles were as follows. The jM'ctoralis, latissimus, serratus magnus and cephalo-humt ral muscKs were found to be healthy, and account for the fact that the shoulder was drawn close to the body. The othfr muscles of the leg were much atrophied. At the operation sit(> tlure was a mass of scar ti.ssue, into which ran


tho largo norvrs from tho proximal side, and from which on the distal side, remnants of nerves emerged. Portions of the nerves Inyond the scar and the atrophied muscles were placed in formalin. In contrast to the healthy structures from the other limb, they remained floating, showing the presence in them of fat.

In order to obtain an appro.ximate idea of th(> size of the arteries, an injection of red-lead was made. This showed an abundant arterial supply. The subclavian arteries of the two sides were laid open and their circumferences measured. The artery of the operated siile was found to have a circumference of 5 mm., as compared with 6 mm. for the one on the unopcrated side — making the diameters 1.6 and 1.9 respectively. Such a difference is to be expected, when the difference in activity and size of limb are considered. Clearly no objection can be raised on the score of lack of sufficient blood-supply to the bones of the operated side.

The gross appearance of the two sets of bones is shown in the photographs of the fresh bones, reproduced in figures 4 and 5. As may be easily seen, there is no striking difference between the two sets, except in diameter, and general size. Similar processes and grooves arc present iu both sets, both show approximately the same curves, the difference in length is not great, but the bones from the unused leg are much more dehcate than those from the used leg.

The measurements of the bones themselves, some of which are given in table 2, show, for lengths of bones and diameter of the humerus, slight differences from the measiuvments, made from the X-ray plates of March 28, which were taken after death, and therefore serve as a partial control to the X-ray measurements. In nearly all measurements, those from the X-ray are sliglitly higher (about 3 per cent) than those from the bones them-selves. In the comparison between the bones of the two sides, however, the difference between the two sets of measurements is very little, thi- gi-catrst divergence being in th(> lengths of the ulnae, the measurenu iits of which show a difference of 0.7 cm. a.s cf. witli (1.55 cm. in the X-ray measurements. The percentage amounts by which the unused bone has lagged behind the used


J. A. IldWKI.l

bont' ill l( ngth arc, for huimrus, ulii:i aiul riulius: 4.4, 5.5 and 4.7 per ctiit n sp( ctiNcIy.

.\s imiicatt'd l>y the X-ray plaits, tlic difference in diameter of the bones from tlu> (wo sides is vexy great. The measurement of the hunu ri agree very elosely with those made from tlie X-ray ]>hites, and are given in tabh* 2. The measurements of the diameters of the radii and uhiae, however, show that those fi-om the X-ray records are not reliable, owing to the flattened shape of these two bones, which causes nioilifications according to sHght dilTcrences in po.sition. These bones are found to Imve the following greatest diameters, at the miildle of tlie bone :

rina on operated side

Kudiu.s on operated side..

cv%. cm.

0.4.5 On unoperated side O.'JS

57 On unoperatcd side 0.95

T.\BLE 3



mm. mm. mm.



fUnopcriited leg

\Operated !eg

I.S + 2.5 =43 1 3 -1- 1.4 = 2.7

5 2

2 8


rUnoperated leg

\Operatcd leg.

2.5 + 1.25 = 3.75 2.25



f I'noperated leg

\Operated leg

1 25+ 1 75 = 3.0 4 + 1 25 = 1 65

2 0.6

Thus the radius and humerus on the unused side have diameters slightly gi-eater than half- the ulna a diameter less than half — the diameters of the coiTes])onding bones in the used leg.

In order to study the thickness and the strength and arrangement of the trabeculae, the bones were split longitudinally through corresiKinding ]>arfs. In the radii and ulnae the cuts were made througli the line of narrowest diameter. .Measurements of thickness, taken at corresponding points at the middle of the bone, are, for the two thicknesses, shown in the cut surface.

coxTKoi, or honm; cuowtii iv nons


KiK.s. I iiiul .'i rii(it(i);ru|>li« of llic Imhics of the two logs after deatli. ApproMiiuitcly natural size.

'J4() J. A. IIOWKI.I.

Tlnis tlic (hicknessos of the lionos on tlic operated side are found to l)e less than those on the nnoperated side l)y 'M), 40 and 41.t) per rent for tlie luinieriis. iihia and raihus respectively, whiU' the size of the nunhillary canal shows a still greater difference — being practically obliterated for a considerable jjortion of the unused ulna, and, in the unused radius, less than one-third as wide as in the used radius.

The difference in thickness conies out conspicuously In the scapulae. The unused scapula is much more delicate than the used scapula, and, as shown in the photograi)h. has on the operated side an area in each of the supra- and infraspinatus fossae, which is covered only by a membrane, bone being entirely absent.

The trabeculae, particularly in ilic humerus, show interesting differences (fig. G). In the bones of the used leg they are decidedly larger and stronger than on the operated side. They show, at both the proximal and distal ends a definite arrangement — somewhat similar to that of the femur of man -for transmitting the body weight. Delinite bars arch upward from the lateral and medial sides of the compacta toward the epii)hysi al line, while in the epiphisis there are bars and sheets in the line of greatest pressure. In the unused bone there is a .suggestion of a similar arrangement, but there are no sheets of bone, and the bars are very delicate, with relatively much larger spaces lu'tween. A similar difTerence is present in the ilistal extremities of the lunneri, where the epiphyseal lines have disappeared. In the middle third of lh(> shaft there arc more trabeculae in the used than in the unused liumerus.

The study of the split bones corroborates that of the X-ray l)lates as regards the union of epiphyses with diaphyses — the epiphyseal line has disajipeared in the distal ends of both humeri and in the |)roximal ends of the ulnae. The other e])iphyseal cartilages are still present in both bones, and are thicker on the nnoperated side. The joint cartilages are also thicker on this side.

The gross differences in the bones of the two .sides are shown in the comparative weigJits of the fresh bones, which are given in table 4:



Hudius. ..





10 G55 7 525 7.115



2 520 3.225

Km. 5


.1. v. TIOWKM,




Fi(t. 6 Drawinitii of the proximal iind distal ends of the two humeri. The actual IrriEth \x somewhat greater than is shown. P^nlarged approximately 1.6 time."



From this table it appoars that the huinorus, ulna, and radiu.i from tlic oju'ratcd sido are 5S.4, (iti.,") and 'AXi per rent less respectively than the corresponding bones of the unoperated side.

It was thouRht that interesting data might be obtained by testing the comparative amounts of force necessary to crush the bones from the two legs. Professor LaKue of the Engineering Department of the University of Missouri very kindly placed at my disjjosal one of the testing apparatuses of the department of engineering, and assisted in its operation. P'ollowing his suggestion, a block was taken from the shaft of each humerus, care being taken that the height was at least as great as the diameter, anil that the two cut surfaces were ground parallel. In order to determine beforehand the total cross-sectional area, a drawing of the cut surface was made at an enlargement of 1 to 25, and the total area occujiied by bone estimated by means of a planimeter. The bones had undergone partial drjing (several hours) before testing.

The results of the tests, and the estimations made from them are shown in table 5:




Cross sectional area of bone

38 01 sq. mm. 1075.0 lbs. 12.82 18,251.0

14.78 sq. mm. 340 lbs. 10.43

Crushing force

Kilograms per square millimeter

Pounds per square inch


The relative cross-sectional areas (fig. 7) of the two bones give a better idea of the relative amount of bone than the diameters given above, and they agree fairly closely with the differences shown by weigliing. Thus the cross-sectional area of the compact bone of the unused humerus is 60 per cent less than the other, while its weight as shown above is 5S.4 per cent less.

Possibly the most surjirising of all the findings of the present study was the amount of force :^4() pounds — necessary to crush the bone from the leg which, for four and one-half months,

Tnr AN.\TOMICAL RECORD, Vol.. 13, NO. 5



had lifcii subjected to the action of a iicKliKible amount of stress and strain.

Most interesting is the comparison l)etween the two bones, when the figures are reduced to inch-pounds, as shown in the tabl(\ Wliile tlie amount required to crush the unused bones was 19 per cent less than for tlie used l)oue, it is surprising to find that the resisting strength of the unused bone is as high as 14,845 pounds per square inch. This is not far below the

Fig. 7 Projections of the proximal surfaces of the cylindrical blocks taken from the two humeri for testing. They arc placed so that, in each case, the top of the picture corresponds with the anterior (cephalad) surface of the bone. Enlarged 6.25 times.

compressive strength of human bones which lies between 18,000 and 24,000 pounds per square inch (Rauber, quoted by Koch, '17, p. 282).


The results brought out in this stuily allow certain definite conclusions to be drawn, while on other points they give data which are suggestive, but which need further testing.

The answer to the fjuestion whether all bone growth is dependent upon the amount of str(>ss and strain is deiinite and conclusive, and is in the negative. The growth in length of the bones deprived of the action upon them of all but a negligible amount of stress and strain, a growth of such an extent that the humerus


becomes 5G per cent longer, while the radius and ulna nearly double in length during four and a half months, does not admit of any question on this point. On the other hand, the answer to the question whether bone growth is entirely independent of the action of stress and strain is equally definite and is also in the negative. This is shown conclusively by the very much smaller diameter, thickness of compacta, size of trabeculae, and greatly reduced weight of bones largely deprived of the action of mechanical stress and strain.

It is clear, then, that the growth of bone is regulated in part by the action of the mechanical factors of stress and strain, and in part by other factors. So far as shown by this expmment, which covers a period in a growing dog, between about four and twenty-three weeks after birth, the part of bone growth dependent upon the action of mechanical stress and strain includes most, if not all, growth in diameter, that is, most of the formation of bone by the apposition of new layers. This applies to the growth in thickness of the trabeculae, once they have been laid down, as well as to that of the compacta. It is probable that a part of the growth in length is also dependent upon the amount of stress and strain. This appears to be increasingly important as the animal glows older, for, while growth in length in the unused bones practically kept pace with that in the used bones, for .six weeks after operation, thirteen weeks later the increase in length of the bones in the used leg had been 20 to 25 per cent greater than that of the bones in the paralyzed leg.

The jiart of bone growth not regulated by the action of the mechanical factors of stress anil strain, so far as shown in this experiment, includes most of the growth in length, that is, the new bone which is formed at the ends of diaphyses, including extensions to the compacta and the trabeculae, as well as bone formed in epiphyses. The time at which the epiphyses unite with the diaphyses is also a factor wliich appears to be uninfluenced by stress and strain, for the same epiphyses have united and the same remained united in the two hgs, at the end of the experiment.

252 J. A. now KM.

This cxpcrimont does not give a clear picture of the exact changes in thickness wliich occur in hones deprivrd of thc> action of stress and strain, lu-cause of the uncertainty of tlie X-ray l)ictures. The amount of apposition and resorption which takes phice can not be estiniateil, and to determine* this other experiments must be carried out. That some resorption takes place is shown clearly by the windows found in the scapula, in the supra- and infra-sjiinous fossae. It is planned to carry the experiment over still longer periods, and also to interrupt the experiment at various shorter intervals, in order that certain studies not possible from the X-ray plates may be made.

I wish to acknowledge my indebtedness to Dr. E. R. Clark of the .\natomy I)ei)artment of the University of Missouri for his kind advice and assistance in the development of this problem. I also desire to thank Mr. Harry A. LaRue, Instructor in Civil iMigineering and Dr. H. C. Rentchler of the Physics dejiartnunt of this university for their aid in some of the technical work of this experiment, and Mr. G. T. Kline for assistance with the dra\\'ings.


Herbst, C. 1901 F'ormative Reize in dor Tierishchen Ontogcnesc. Leipzifc. Koch, J. C. 191" Tlic laws of bone architecture. Am. Jour. .\nat., vol. 21,

p. 177. Oli.ikr 1867 Traitc? cxperimentale sur la regeneration dcs os. Paris. PoTTOHF, J. L. 1916 An experimental study of bone growth in the dog. Anat.

Rec., vol. 10, p. 2.34. Thoma, R. 1907 ^^yno.stosis suturae sagitallis cranii ein Beitrag zur Histo mechanik de.s Skelcts und zur Lehre von dem inlerstitiellcn Knochcn wachstuni Vircliow's .\rchiv., vol. 188, p. 248.








'I'wo jiapcis in this series have already dealt with the interstitial ceils in the reproductive organs of the chicken. (Boring, and Pearl and Boring, '12). Further work has substantiated our former conclusions except that a few interstitial cells have been found in just hatched niale chicks. We give these further facts here, together with a brief review of some of the literature on interstitial cells. These cells have been much under discussion, not only as to their origin and function, but even as to their very existence. They have been studied much more in mammals than in birds, but they have been recorded in some species of all classes of vertebrates. We are not concerned here with the question of their origin, but with that of their nature and distribution.

The term interstitial cells is used by most writers to designate cells lying in the connective tissue of the gonatl and functioning as gland cells to form some sort of an internal secretion. Some writers use the term indiscriminate!}' for all the tissue that fills in between the germ cells. Linion and Bouin used the term interstitial gland for the combination of these gland cells in any one reproductive organ. The reason for so doing is that the cells are foiuid fille<l with granules of fat and protein in the manner of gland cells. Many investigators have attributed to this internal secretion the function of controlling the development of

I Papers from the Biological Laboratory of the Maine Agjicultural Experiment Station, No. lU.



secondary sex oluu'actcrs. Liinon, Bouin, Rcgaud, Cesa-IMjinchi and Hanos are sonio of the names associated with this theory. Their evidence is l)ased on observation, ex])erinient and certain pathological cases where reproductive organs with degenerate genu cells and increased interstitial cells were coupled with sterility and normal sex behavior. Kegaud records a decrease in the interstitial gland in isolated rabbits. Cesa-Bianchi records a decrease of these cells in hibernating animals. Ancel and Bouin state that roentgenization of mammalian testes destroys only the geiin cells and results in steriUty, but does not effect the interstitial cells or the secondary sex characters. ^^'cichelbaum describes a case of alcoholism in which the seminal tubes were degenerated, but the hiterstitial colls increased in number. Des Cilleuls claims that the interstitial cells first make their appearance in the chick at thirty days after hatching, an<l that this time coincides with tl\e appearance of the secondary sex characters. Regaud and Hanes both coimect the time when these cells are most abundant and secreting most actively with the time of greatest sex activity and the development of the secondary sex characters. So various observatiimal and experimental facts seem to substantiate this theory.

However, there is a great body of facts which do not fit in with this theory at all, and there are many other theories proposed as to the functions of these cells. Pfiiiger, van Bencden, Plato and Winiwarter, interpret the material mth which these cells are packed a.s food material destined for the nutrition of the germ cells. Wallart finds them best developed from birth to ])uberty and suggests that they have a trophic efTect on the circulation. Jardy credits them with a general trophic effect on the nutrition of all tissues. Winiwarter finds the interstitial cells in the human testis better developed in the foetus than in the matiu-e individual, and even found them entirely degenerating in an indivitlual forty-one years of age. He speaks of their secretion as the expression of an overactivity of cells destined for nutrition. Marshall ajul Jolly suggest that an internal secretion is formed by "the follicular epitheUal cells or the interstitial cells of the stroma," in the mammahan ovary and is the cause of


heat. Pardi clauns that the interstitial gland is necessan' for the control of the growth of the embryo in conjunction with the corpus lutcuni, during the first half of the jjcriod of gestation, and entirely during the later part of gestation. He finds that the interstitial gland in the female increases in size at puberty and decreases in old age, and claims that this is connected with its influence on the nutrition of the embrj'o, rather than on the development of the secondary sex characters. Pearl and Surface describe interstitial cells in the normal cow ovan.' and in a cystic ovary of a cow which took on mule secondary sex characters. In this case the interstitial secreting mechanism was the same in both animals. Therefore the change in secondary sex characters could not reasonably be ascribed to a secretion from these cells.

So much for the hj'potheses about the functions of the interstitial cells. The facts about their very existence are even more at variance than the statements and theories of the above authors indicate. Fraenkel studied the ovaries of 46 different mammals and found interstitial cells present in only 22 of these species. Schaefer continued this study, working on 50 ovaries, 11 of which belonged to species already studied by Fraenkel. She also found great variation in the presence of the interstitial gland. It is not consistently present in any one order of mammals, being absent for instance, in out of 11 Carnivora studied. Further, it ie not consistently present even in one species, as Schaefer found an interstitial gland in 3 species where Fraenkel found none. Aim^ reports it lacking in the sheep and pig. Buscjuet makes the general statement that the interstitial gland does not exist in the ovary of all species of manunals. Kingsbury studied the ovary of the cat, and reports great variation in the number of interstitial cells present at all ditTerent ages. He calls them "modified free lipoid-containing stroma cells." He fails to find "any peculiar relation of the interstitial cells to either blood vessels or lymph vessels" and concludes that the "interstitial cells possess no morphological indiviiluality and hence do not deser\'e recognition as a distinct kind of cell." Mazzetti has studied interstitial cells in the testes of many


kinds of vertebrates and finds such great variation in the quantity of thorn present that he eonchides that they ran have nothing to do witii secondary sex characters. Kirkl)ride records a wide difference in quantity in human testes of the same age. These facts woukl seem to justify Fraenkel's conclusion that "a ghind existing for the specific purjjose of forming a substance of very distinct value to the organism as a whole would be more constant in its presence and development."

So much for the general conflicting theories and facts about interstitial cells. Thc^^e apply more particularly to manunals, but interstitial cells have been described in both male and female birds. Waldeyer first mentioned them in 1870. In hens, Ganfini describes them as found sparsely in the ovarian stroma, but abundantly in the theca interna of the follicles. He says they are characterized by granules, some of which stain with osTuic and some with safranin or fuchsin. He figures two different kintls of interstitial cells, some packed with granules which stain easily, and some with protoplasm which appears clear and vacuolated. His drawings show only the nests of clear cells in the theca interna of the follicle. From the work of the present authors (cf. X and XI of these Studies) it would appear that these two kinds of cells have an entirely different function and fate. The clear cells are never loaded with acidophile gi-anules and are therefore not the glandular interstitial cells. Loisel describes interstitial cells in the connective tissue between the follicles of the ovary, containing fat granules which stain black with osmic. He thinks the oviiries are primarily glandular in function anil have acquired their function of forming eggs secondarily. Sonnenbrodt figures nests of interstitial cells in the follicular theca of the hen's ovary. These ai)|>ear to have a vacuolated cytoiilasm, there being no indication in the figures of stained graimles. These resemble the clear cells described by Ganfini. In female hybrid birds Poll describes mas.'^es of epithelial interstitial tis,>^ue. which is glandular, while there is no trace of egg formation. Smith finds groups of interstitial cells in hybrid female jiheasants, but no ova. No investi


gator, ill fact, .seems to deny the presence of interstitial cells in the female bird.

In male birds, however, there is a difTerciico of opinion as to the presence of interstitial cells. Ceni records an interstitial gland in male ducks. In birds from which he removed the cerel)ral hcmisplicres, the niterstitial tissue increased coincidently with the decrease of the seminal tubules, but the secondary sex characters roinained un(l('vcloi)ed. He does not state whether this hyiiertrophied interstitial tissue is connective ti.ssue or gland cells. It is interesting to compare these results with those of Myers on white rats with the vas deferens ligated. He finds that the .seminal tubules degenerate, while the interstitial ti.ssue remains uiiafTectod. He states that the interstitial tissue apparently increased in quantity, but this apparent increase was merely apparent, due to the degeneration and shrinkage of the tubules. Poll records one case of a male hybrid with a large interstitial gland. Loisel says that the cells of the testis .show no glandular activity, in fact, even disappear during spermatogenesis. Mazzetti pictures a small gi-ouj) of interstitial cells in a triangular space between the seminal tubules of the chicken, but says they are rare even though this bird has very marked secondary sex characters. Des Cilleuls records them as being first found in the chicken at about thirty days after hatching, and claims that this time coincides with the time of the first appearance of secondary characters. It may be of interest to recall here that Barred Plymouth Rocks can be correctly sexed at hatching on the basis of external appearance only. Reeves studied cliicken testes of three, five and a half, nine and eighteen months, and claims to find interstitial cells in all of them, but more at three and five and a half months than later. He tested them by staining the gi-anules witii Mallory's connective tis.sue stain, iron haematoxyHn and Sudan III.

In a previous study, one of the present authors (.\. M. B.) has claimed that no cells of a glandular nature could be found in the chicken testis. In view of these c(mflictiug statements it has seemed worth while to take up this study again on both male and female birds.



In the prp\'ious study of testes, material preserved in Gilson was stained with iron haoniatoxylin or Delafield's and orange (\, anil some material put up in Flemming was cleared in ehloroforni and mounted in chloroform balsam in order not to dissolve the fat blackened by the osmic. Other staining methods have been useil this time. Several of the differential stains for secretion granules were tried out and finally two adopted which have been used by \Mntehead, Benthin and other workers on the interstitial cells of mammals; Mann's methyl blue and eosin and Mallory's connective tissue stain. Both of these give a striking differentiation. With Mami's stain granule-laden cells stand out as reddish pin-ple spots in the midst of a pale blue or lavendar background, while blood corpuscles take the red eosin color. With Mallory's stain, the gland cells are purplish red, the corpuscles orange or bright red and the backgioimd is a bright blue. These stains react with gi-anules of a protein nature, and the reaction is iM'rfectly well tlefinetl. There is no possibility of missing secreting cells when these stains are used, if there are any in the material. Tests for granuh^s of a fatty nature were not made at this time, as this had been tested in the preA-ious study of interstitial cells in the chicken testis and had given negative results. "WTiitehcad finds the fat content in inann)ials much more varialile than the protein, there being no fattj- secretion at all in the pig.

The ovaries used for this study were from four actively laying birds, one of which was a bantam and the other three Barred Plymouth Rock, one old C'ampine which had stopped lajing some time jm'viously, and several just hatched chicks and guineas. For most of these any small piece of the ovary was sectioned, but with one Barred Plymouth Rock the discharged follicles of various sizes were sectioned separately.



'I'hc (lilTcrciitiul stain showed iiiiiiu'diatcly that there is an abundaiici- of ghiiidular cells in all the mature ovaries. No such cells, however, were differentiated in the just-hatched chicks or guineas. These cells may be scattered through the stroma but occur chiefly near the periphery'. They may occur a few near each other (fig. 1) or they may be -jjacked so thickly in one place that they color the whole region red to a glance with a low magnification (fig. 2). In the sections of discharged follicles of different ages they show especially on the surface of the stalk of the follicle. There are no more of them on the old discharged follicles than on the recently discharged ones. In all these ovaries, the clear cells described by Ganfini were quite conspicuous, but they seemed to be an entirely distinct type of cell from the interstitial cells and were seldom found in the same location. The one place where they come in contact with each other was in the mass of cells filling up some of the old discharged follicles.

A study of the cytology of the indi\'idual interstitial cells (fig. 3) shows them to be ess^entially the same as those described in the cow ovary by Pearl and Surface. There cannot be the slightest doubt of the homology of these cells in birds and mammals. They are roundish cells scattered indi^^dually througli the coimective tissue, that is, never arranged in nests together. The nucleus is small and stains intensely. The structure of the cytoplasm is completely obscured by the mass of gi-anules in the cells. These granules are fairly large in projiortiou to the size of the cell as shown in figure 3. The gi-anulcs take any acid stain, eosin, orange G, or acid fuchsin, and stain black with iron haematoxylin.

It is difficult to compare the ovaries as to the quantity of interstitial cells without sectioning the entire ovary of each bird, for they are more abundant in irregular regions of the surface. But in the old ("'ami)ine which had stopped lajnng, there were more interstitial cells to be found in the more central parts of the ovary than in the younger birds, while there were as many on






  • ^

Fig. 1 From periphery of ovary of laying hen. Glandular interstitial cells sparsely scattered. X 264.

Fig. 2 From periphery of ovary of same laying hen as figure 1. Glandular interstitial cells packed close together. X 264.


tho surface, so that it sppms probable that the total quantity is greater in the old bird (fig. 4). This is in aeeord with the findings of Kasai in human material. The results of Harms, however, appear to indicate that it can by no means be a general rule that these interstitial cells are more abundant at higher ages.

Fig. 3 Interstitial cells from ovary in figure 2. X 0.50.


The same staining tests for gland cells were applied to the chicken testis. It will be recalled that 21 male birds had previou.sly been studied for interstitial cells. These were of var>'ing ages, just -hatched, six, eight, ten, and twelve months. More sections were made of some of this same material, and they were stained according to Mann's and Mallory's methods. Also preparat ions were made of a whole new .series of birds from six to eigliteen months old, and of a few more newly hatched chicks. This new series was prepared as part of an extensive investigation which one of us (R. P.) has under way in which data regarding the testis in monthly i)hases of its cyclical changes for an entire year will be studied. The material compriscil sections



of the testes of 4'A birds. Three or four of these birds were killed ever}' month, th(< youngest group at six months, and the oldest at eighteen months. This entire .series was .stainetl for gland cells and examineil carefully, and none of the mature birils, eitluT the new ones, or those used before, showed any secreting interstitial cells. However, in the testes of four

Fig. 4 From periphery of ovary of old hen past laying. Interstitial celU numerouB at periphery and into central stroma. X 264.

newly-hatched cliicks some granule-laden cells were found, staining reddish purple as in the ovarj'. These were perfectly easy to recognize although they were found in only a few regions of the testis. They were usuallj' grouped, several near each other, but sometimes a single cell stood by itself, (fig 5). Studied under higher magnification, these cells appear a little larger and rounder than those in the ovary (cf. figs. 4 and 6).



They contain fewer granules and these granuler* var>' in size. The nuclei closely resemble those of the connective tissue in which they lie embedded. This investigation, then, shows interstitial cells present in the newly hatched chicks studied, but not in any of tJU mature l>irds though they were of various ages.

Fig. 5 From testis of just hatched chick. A few interstitial cells in connective tissue between seiiiiual tubules. X 264.

Fig. 6 Interstitial cells from testis in figure 5. X 950.


V. niscrssioN

'riuTc aro several jiossiblc explanations for tlic connictiii^ results of Mazzetti, des C'illeuls, Reeves, aiui tlie present authors. Interstitial cells are reported present and absent for male birds of praetiealiy every age. Des C'illeuls says they first appear at thirty days, while the only birds in which the present authors lind them are just-hatched chicks. Reeves finds tlieni at all ages, but most abundant at three and one-half and five months. Mazzetti reports them in mature birds, liut says that they are rare. So they possibly may be present in male birds of any ages, but are much more frequenltly absent, and are, therefore, not a necessary component of the testis structure. However, it must be emphasized that the only way to be sure of interstitial cells is to use a differential stain, and then one can not fail to identify them. Neither Mazzetti nor des Cilleuls did this. It is very to be misled by small pieces of tubules being surrouniled by connective tissue. We have olj.served several such places in our preparations, which resemble some of the illustrations of groups of interstitial cells. Reeves, however, used Mallory's and Mann's stains. It is interesting that he fouiul more such cells in young birds than in mature ones in view of the fact that we have found none in such a large series of mature birds.

It appears to be definitely established that true interstitial cells are always present in the ovary. The situation appears to be very different than in the case of the testis, not only with reference to l)irds but to manunals also.

A point which we would emphasize, on the basis of our studies of these cells in mammals as well as in l>ir(ls, is that there appears to be no doubt that the true interstitial cells are not merely honu)logous but indeed structurally (i. e., cytologically) identical in the male (when present) and the female. This \ lew was exjm'.ssed by Plato ('97) and we are in entire agreem( nt with it. This point has an iini)ortant bearing on the question of the influence of tliese cells in sex determination (,cf.

KKritoDrcTivK ()1u;an'.s ok tiik ( iiicken 2(».")

Harms) and in the control of spcondar>' characters. It will he fully discussed in a later paper in this series.

We hope to study a series of testes from the time of hatching up to six months, in order to complete the series, and clear up these points of dilTcrence finally. The discrepancy hetween our results for nuiture birds anil those of Reeves is absolute. We have no desire to question the accuracy of his observations, and on the other hand we are entirely sure of, and indeed are prepared to demonstrate, the absence of these characteristic interstitial cells in our material. For just hatched chicks we agree, interstitial (u'lls are present in the testis. The next step is to find out when between hatching anil six months, they disappear. This we hope to do later.

For the present the es.sential point which we wish to emphasize, and which we belie\e that our results demonstrate, is that the characteristic, true interstitial cells are neither a necessai'y nor a constant element in the make-up of the testis in the male of the domestic fowl They may be, and usually are, totally absent from the testes of males over six months of age and of full sexual normality, both in respect of primary and seccmdary characters. It suffices here to draw attention dearly to this fact, which emerges with great definiteness not alone from oiu- own work, but also from a review of the literatur(> of the .subject Later we exjjcct to make constructive use of tliis evidence in connection with the problem of the causation anil control of .secondary sex characters Without entering upon any discussion of the matter now it is e\ident that the facts regarding the occurrence and distribution of interstitial cells are such as to make it very difficult to suppose that they havo any causal inllucncc ui)on secondary sex characters.

TIIK AN\Tnulc\t Ht:loRI>. VOL 13 NO. S

2(i() Al.KK M. H<iUIN<; .\\I> UAYMON'I) I'KAUI.

\ I. I 1 1 i:i! \ ri itK (TIKI)

Ma^ i.i^-4 ami Stahi iNii I'JlKi DiT C'tipinisclu' Koordinntion dor Fiiiiklionpn dcg

KOrpers. Krg. d. I'liysiol., IW. .'>, p. 1)64. Bentiiin, \V. 1911 I'cbor Follikclatrosie in Siiugcticroviiricii. Arcli. f. Ciynii koioKu- ill, p. rm.

HoKiNc, A. M. KUL' 'J'hc intcrslitiul cells and ilii- supposed inlfrnal secretion

of the rliirkrn testis. (Sex Studies, III i. Mini. Hiilletiii, vol. L':{. i)p.

lU-l,).}. HoriN A.VD Anckl IIKB Heeherclics sur los cellules inlerst'tipllcs du testicule

de« mniiiniiffres. Arch, de zool. exp.. T. 1, p. 317, Series ■}.

1004 Hecherchrs sur In siKnificntinn physiologique de la (jlande in torstiticlle du tosticule dcs m.niniMifi^res. Jour, de physiol. et de

path, (ton., T. C, p. 1012. B^grKl, H. 1910 La .secretion interne de I'ovarie. Hiol. .Med., T. 9, p. S9. Cexi, C. 1909 L'inHu?nza de! corvello suUo sviluppo e sulla funzione depli

irE;ini s('s.-u.ili iiiMschili. Hivista sper. di Freii.itria, S. 35, p. 3.

1909 L'influenzii di aleune sostanze d'azionc prcvnlcntcmentc ccre brale siilla funzione dci teaticoli (Caff(5, Veronal e Absinthe). Riv.

sper. di Krenatria, 35, p. 3. Ce.^a UiANTiii, D. 190S Die alcunc particolarita di siru'turi dei fenomeni di

secrezione del ct-rpo lutoo. Internal. Monatschr. f. .\nat. u. I'hysiol.,

S. 25. PI). 1-43. DBS Clu.Bri.>s, J. 1912 Interstilinl testicular cells and secondary sex characters. C. K. Snc. H'ol.. vol. 73, pp. 371-372. Fraknkel, L. 1905 Verg. histol. I'ntersuchungcn iilier das Vorkommen

drifigen Formationen im inter.stitiellen Kicrstock.sgewebe. Arch. f.

(iyniik., Bd. 75., Cahi.o 1913 Su! comportaniento delle cellule intcrstiziali del testi colo negli stati iiiorbosi (lenerali dell orijnni.sino. Arch, per le Sci.

Med., S. 37. (iANFlNl. C. 1908 Sulla strulturo e sviluppo delle cellule intcrstiziali dell'

ovajo. .Arch, di .\nat. e. di Knib., S. 7. p. 373. Hanes, F. M. 1911 l{elations of interstitial cells of Leydip to production of

an internal secretion by mannnalian testis. Jour. Kxp. Med., p. 338. ilAKMS. W. 191! lOxperinientelle rntersuchun<;en iiber die innere Sekrelion

der K'iindrusen und deren Be/.iehung zuiii nesamtorpiinismua.

I,ei|i:.i(r. Kasai, K. 1'N)S Teller ilie Zwischcnzellen dcs llo.lens. \'irchon's .Vn-hiv.,

Hd. I'M, p. 1. Kl.N(i.--uiKV, B. F. 191 I 'I'lie interstitial cells of the nianuiiidi'tn ovary: Felis

donicsfica. Am. Jour. Annt., vol. 16, p. .59. KiKKHKiDK, -Makv B. 1911 Embryoftcnic disturbance of the testis. Arch.

Knlw. Org., Bd. 32, p. 717. LuisEL, GfSTWK 1903 F.laborations itciodique dans le testicule

dcs oiscaux. C'onip. Bend, de I'.Vfsociation des .\natomistes, 5th

Session. Liogr, T., p. 222.


L018KL, CJrsTAVE Lea phOnonii^nt's des t^'-cri'lion duns less glundes f(<^Ditales.

Ucvue Kcupr.ilc ot fiiits nou»'paux. Jour, do 1' Anatomic et de la

I'liysiolonio Wtli Anne. Xo. .">, p. 530. Mausiiali., I'". H. A. AND Joi.i.v, \V. A. 1905 Coiitributions to the pliysiolopy of

niaminuliun reproduction. Part I. Oestious cyc'c of a dog. I'art

II. Ovary as an orpan of internal s<'cretion. Phil. Tr.ins. Hoy. Soc,

Serie.s H. lOS, pp. m-U\. Mazzetti, LonKTTO 1911 I caratteri ses^uali secondari e le cellule intcrstiziali

del tosticolo. Aniif. \uz., S. 38, p. 301. MvERS, B. D. 1915 HistoloRicil chanpes in testes foUownnj; v.ascctomy. Annt.

Rec, vol. 10, p. 22s. Pakdi, I'. 1911 ()l).servazi'jni sullc cellule intcrstiziali ovarchc della coniglia e

suugli dementi della teca interna dell' ovaio niulicbre al di fuori e

durante la .^tato di pravidanza. .Vrch. di Biologic normalc et Path.

yr. 68. Pakdi V. 1914 Siir la fonction endocrine dc I'ovairc durant la gestation.

Arch. Ital. de Biologic, T. 01, p. 1. Pkahi., U. and Bokixo, .\. M. 1912 Fat deposition in the testis of the domestic

fowl. (.Sex Studies IV). Science, X. .S., vol. 36, pp. 8.33-835. Pkahl, H. a.nd Si-iiFACE. V. M. 1915 Sex Studies. VII. On the assumption

of male secon.lary oh.iracters by a cow with cystic degeneration of the

ovaries. Ann. Kept. Me. .\cr. Expt. Stat., 1915, p. C'l. Pl.ATo, .). 1S9(< Die inlerstitiellen Zellen des Hodens und ihre physiologische

IJcdeutung. Arch. f. Mikr. Anat.. Bd. IS, p. 2S0.

1,897 Zur Kcr.ntnis der Anatomic und Phj-siologie der Geschlcchts organe. Arch. mik. Anat., Bd. 50. Poi.i., H. 1911 Mischling.studien VI. Eieistock u. Ei bei fniehtbaren ii. unf.

Mischlingen. Arch. f. Mikr. Anat., 7.8. Pt. II, p. 63.

1900 Der Geschlechtsapparat der Mischlinge von Cairina moschata

cf und Anas boschas <?. Sitzber. der pes. nal. Freunde, Xr. I,

Jahr. 19U(i, p. l. Poll, II. and Tikfexskk, W. 1907 Mischlintr.stadien: die Hi.<tolopic der

Keimdnit^on bei Mischlinpcn. Sitzber. der pes. Xat. Freunde, Xr.

0, Jahr. 1!X)7. p. 157. PoroFF, X. 1911 [Interstitial tissue and corpus lutcum of the ovary.) Arch.

Biol., vol. 26, p. 483. Ref.\t:s, T. p. 1915 On the presence of interstitial cells in the chicken's testi?.

Anat. Rec, vol. 9, p. 383. Regaid, (". DruHEtiL. (i. 1909 X'ouvelles rccherchcs sur les modifica tioiLS dc lii glande interstiticlle (|e I'ovaire, consecutives a I'isolcmcnt

et a la cohabitation avec le male. Comp. Rend. .Soc. de Biol., T. 67,

p. 34.8. ScHAEFFEH, A.N.VA 1911 Wrplcichend histologische rntersuchunpen Ober die

interslilielle Arch. f. (iynakolopic. Bd. 94, p. 1. SoNNENUHODf 1908 Die Wachstunisperiode der Oocyte des Huhnes. .\rch.

f. Mikr. Anat., Bd. 72, p. 415. Van dek Stuicht 1912 Corpus luteuni and interstitial gland of the ovary.

.Vrch. Biol., vol. 27, pp. .■).S.')-722.


WEICllEl.DAr.M AM) Kyiii-e 1012 IChanRps in li'sli-s in rlironic iilcoholism.l

SU Aknil. wiss. Wicn., vol. I'Jl. pp. ol-fi". WlllTEllEAU, U. H. 1!H).") SlU'lics of inltTNtiliiil cells of I^-ydiR. II. Their

postenibryonic dcvclopnirnl in !i piR. Am. Jour. An:it., vol. 4, p. 193.

1912 On the chemical nature of certain (tranule.s in the interstitial

cells of the testis. Am. Jour. Anat.. vol. H, p. Ki. WiMWAKTEii, H. VON 1912 Observations cytoloRiquc sur les cellules inter Btitielle du testiciile humain. Anat. Anz.. Bd. 41. p. 309.



LESLIE B. ARKV From the Analoiiiical Laboratory of the Northu-egtern University Medical School'


The iHull inuclcaic giant (•cUs associated with developing bone have usually been considered a.s the causative agents of bone res()r))ti()ii. and for this reason have been termed osteochists. There is, however, no du'ect evidence that these i)olykar}'ocytes so function. The writer ('17) has contendetl that they may be interpreted equally well as degenerating, fused osteoblasts.

Kvans and Schulcniaun (T4) were able to .-^how that various cells of the body are vitally 'stained' by certain benzidine dyes. Although anatomically distmct these cells hold in common this staining response, which involves a phagocytosis of the dye gramilcs. difforiTig in no essential resj)ect from the ordinary ingestion of <l(l)ris by such cells in the exercise of their customary scavenger actiA-ities. It has been suggested that cells responding in such a fashion to these dyes constitute a i)hysiologically homogenous gi-ouj) and can be designated macroi)hages.

Recently Shipley and Macklin ('10) have attempted the vital staining of osteoclasts. It was thought that since CJoldmann ('12) foimd the ])athological giant cell of experimental tuberculosis to react avidly to benzidine dyes, the anatomically similar osteoclast, with its alleged bone resorptive function, would i)resumably ingest the dye also. Contrary to exjiectations the osteoclasts of young animals, vitally stained witli tryiian-i)lue, did not take uji the dye at all. The conclusion is drawn that since the osteoclasts are non-trypanophilic they are likewise nonphagocytic, any office they ma>' perform in bone destruction being through the pressure of their bodies or by the proiluction of a dissolving secretion. Foreign inclusions in the cytoplasm

' Contribution No. 60, May 30, 1917.


were iiiit ()l)S('rv('(l. 'I'lic onliTiaiy rcliculo-i'iulothcliiil cells of the nuuTow, cm the contrary, did stain with the dye and arc identified as the elements which ])han()cyt().se the residiimii of hone dissolution.

It is inij)ortant to koe]) sharj^ly differentiated questions as to the alleged relation of osteoclasts to hone resorption from those concerning their phagocytic activity. The ah.sence of coarse dehris in the cytoplasm docs not of itself imph' that the osteoclasts do not rcsori) hone; presumably bone destruction is essentially of a dissolving and dig) sting nattirc, the end products not improbably being rather tincly divided particles.^ On the other hand, the presence of cytoplasmic inclusions does not indicate that the osteoclasts are actually responsible for the occurrence of the material ingested.

In a former comnumication (17) 1 have held that the osteoclasts may be interpreted as sjTicitial masses derived from dejileted osteoi)lasts whose staining reaction has become oxy])hilic. In that study it became evident that the osteoclasts in size both by incorporating new osteoblasts with which they come in contact and by ingesting the bone cells laid bare by resorptive processes. ^Vhether or not the first method of growth is one of j)hagocytosis. the seccmd unquestionably is such. The bone cell, usually enveloped by an acid-resistant capsule, constitutes a foreign body of relatively large size, and all stages of its inclusion by the osteoclast may be found in favorable material.

Figure 1 shows an elongated process of an osteoclast merely in contact with one side of a bone cell. The engulfment has proceeded to .such an extent in the stages depicted l)y the two photographs (figs. 2 and '.i) that the encapsulated bone cells are half within and half without their respective osteoclasts; these reprodvictions illustrate but poorly the striking apjiearance of such j)rei)arations. Kncapsulated, stellate cells are not infreciucntly found in the cytoplasm of osteoclasts (fig. 4); these are interpreted as bone cells whose capsules are still resistant to cytoplasmic digestion. From the relative infrequency with

' It is. nevrrtholc'.s, poasihlc to find eroded .spicules .ipparentlv entirely free and completely enveloped by large oHteocla.sts.



wiiicli persist cut (•ai)siilcs arc seen it is probable that the enclosed cells are eventually liberated.

It is apparent from the foregoing account that the osteoclasts exhii)it all stages which we are accustomed to accept in microscopical preparations as evidential of phagocytic activity. For this reason the failure of cells to absorb such vital dyes as trypan






9 ?


V « 




P'ig. 1 .\ first staf,c in the <'iijjulfment of a bone cell hy an ostooclaKt. \ proccs.-! of the jtraniilar tistcoclasl is in contact with the bono cell jc. X "So.

FIr. 2 The encapsulated bone cell i is half ingested by an osteoclast which lies on a spicule of bone. Only a portion of the entire asteoclast appears in this section. Photopraph. X (150.

Fip. 3 .\ stage in osteoclastic phagocytosis similar to that of figure 2. The bone cell is half within and half without the osteoclast. Photograph. X 650.

Fig. -1 Two eiicapsulaied, ^^lelIate cells lie within the cytoplasm of the osteoclast. These inclusions are interpreted as phagocytosed bone colls. X 7S5.


blue wouUI seem not to warrant the donying to them of phagocytic" activity; hero in the ()stcocla.><t is a type of cell which dem()nstral)ly engulfs large hodies yet fails to ingest the finer dye particles. Evans and Sehulemann ('14) have likewise pointed out that none of the blood cells, vvvu those that are nf)tably phagocytic, take up the vital dyes; Downey ('17), on the contrary, finds that such dye particles are phagocytosed by leucocytes under especially favorable conditions.'


InTspective of questions as to the causative agents of bone resorption it is demonstrable that the osteoclasts jihagocytose i)one cells lai<l bare by th(> resorptive jirocesses.

Hence the reported failure of osteoclasts and leucocytes to stain after ordinary treatment with vital dyes, such as trypanblue, indicates that this negative reaction can not be relied upon to furnish final evidence concerning the absence of phagocytic activity.


Arey, L. B. I!tl7 On the origin and fate of the osfeocla.sts. Anat. Rec., vol.

U, no. G, |)|>. :51f>-322. Downey, H. 1917 Relictions of blood and tissue cells to acid colloidal dyes

under cxpcrinicnt;il conditions. Anat Rec. vol. 12, no. 4, pp. 429-454. EvA.NS, H. M. AND ScHULEMANN, \V. 1914 The action of vital stains belonging

to the benzidine group. Science, n. s., vol. 39, pp. 44.3-451. GoLDMANN, E. E. 1912 Die iiusscre und innere Sekretion des gesunden und

krunken Organi.snius im Lichle der vitalen 'Fiirbung.' Teil II.

Beitriigc z. klin. Chirug. Hd. 78, pp. 1-108. Shipley, P. C!. Mackiin, C. C 1916 Some features of osteogenesis in the

light of vital staining. .Vmer. Jour. Physiol., vol. 12, no. 1, pp. 117 12X

' It is worth considering whether the factors emphEksized by Downey as responsible for the ordinary failure of leucocytes to react to the dye are not at least in part operative in the case of the osteoclast.

AtrriiniiM AimTKAfT or tim« i-atkii ikmiku




S. H. WILLI.V-MS AND U. W. U.VLCII Miami University, Oxford, Ohio


In 1895, u douhlc ])ig was born doad on the farm of Craig lirothcrs, Oxford. Ohio, brpodors of Poland China stock. The specimen was collected by Prof. A. L. Trcadwell and left in the lal)()ratory of Miami University. Except for opening the body cavities no work was tlone on the pig until in the fall of 191(). The appearance of the article on a monster pig by Carey in The Anatomical Record on February, 1917, renders a detailed description of our sjiecimen unnecessary. We present, thej-efore. a brief outline only, describing the pecuharities of this double monster.

Figures 1 and 2 are j^hotogi-aphs of the animal from the dorsal anil the ventral sides. .Vs in Carey's specimen there has been a separation of the growing points of the embryo into two parts, probablj' two primitive streaks formed side by side, and then a fusion of the adjacent parts of these two gi-owing points to form tile monstrosity.

Accnrdinf^ to Adaini and MeCrac. p. tiS, in double monsters there is Hrst of all a comjilete cleavage and then a secondary fusion of contiguous jiarts. This results in this case in the production of a mono-symmetrical janiceps with a perfect ventral face and a rudimentary dorsal face not recognizable as such. The fusion continues through the thoracic region but the bodies are sei)arate above the abdomen and as nearly as we can tell from the specimen (which was opened in 1895) there are two unil)ilical cords. This, however, is not certain. The reception of two blood supplies from the mother is exccptioiuil.


■J<4 S U. WILLIAMS AM) It. W . ItAlt'H

Tho urinogpnital orRaiis arc paired and norinal in oach V)ody. Tlic specimen is (>f the inalo sex. In tho pis dcsnilxd l)y Carey, tile iieads are fused side by side witli one liead turned to the left and one to tiie ri^lit. If tho head of our specimen has been formed in the same way, there has boon a mueh {p(>ater degree of njrulatioii. the essential sui)iiression of the face turned to the




Ao.R., rieht aorta

Ao.L., left aorta

A.y'.C, anterior vena eava

Card., rardiae end of sloinaeh

Curl., eartihiKinous ootid between the two liodies in the region of llie anterior end of thorax ibo ealled dorsal side)

Ce., cceum

Cleft., division of anterior end of tongue

L., left body

Par., partition in roof of mouth

from 'dorsiil' to 'ventral' Pul.L., left pulmonary artery from

left aorta f'ul.If.. right pulmonary artery from

right aorta I'.WC, posterior vena cava /';//■. pylorie end of stomach f{., right l)o<ly Hud.F., rudimentary face Sp.C, spinal cord, left Vert., atlas and axis of right vertebral




right (dorsal) and tho dovcloiJiiicnt of tho oxtoriial form of the hoad witli reference to tho face turned toward tlie left (ventral).

AI.lMlA'IAin S\STi;.\I

Figure '^ is a diagi-am of the structures of the alimentary system. .Vt the tip, tho tongue is double, a cleft extoniling for a short distance back in tho mid line. Further back behind tho



tongue, there is a connective tissue partition which extends from the roof to tho floor of tho pharynx. The two spaces formed by this i)artition unite again, when tho latter stops, into a small cavity which is continued back as the single esophageal lumen.

Tho slender tube of tho esophagus extends back to an opening on tho posterior and dorsal side of tho largo stomach. This organ has a longitudinal constriction in its middle indicating its double origin. The intestine loaves the anterior ventral side of the stomach and tm-ns posteriorly as a single slender coiled tube 42 inclit s in length. This connects with a transverse tube.





at each end of which is a rpcuiii, the l)f'Kiiiiiinf^ of thf two hirge inti'stincs. The largi' iiitcstino for the left body is lOj inehes long and the one for the right smaller body is 9 inches in length. There is one spleen found in the body of the left pig. Possibly the fact that the liver mass in the right body is two to three times the volume of that in the left may have to dr) with the absence of symmetry in the spleen. This explanation will not hold, however, for the pancreas, which, though jjoorly i)r(served is single. It lies by the duodenum and in the right body cavity. The liver is all comiected dorsally but free ventrally and, as stated above, most of it is on the right side.


The heart is a fused structure with two large thin walled auricles and two large ventricles with eciually thick walls. The auricle of the right side receives directly the single anterior vena cava from the head and anterior extremities. The auricle on the left side is connected \\'ith a dorsal sac, possibly a sinus venosus, which also has a passageway into the right auricle. All the blood from the ]K)st(rior jiart of both bodies comes into this sinus througli a vessel we have called the posterior vena cava.

There is a passage way from each of the auricles into the V( ntricle of its own side. The ventricles are distinct with a thick partition between and each ventricle .sends a large aorta to the body on the opposite side, these aortas crossing anterior to the heart. The aorta from the left sitle supplies blood to the head while that from the right side sends no branches farther forward than the fore legs. Possibly this may explain the difference in the size of the two bodies. If approximately half the l)lood was pumped by each ventricle, the left body would have Ijcen better fed since none of its blood is sent to the luad.

The aorta to either .side .sends a branch artery to each fore leg of the botly it sup])li( s. These legs are at different levels (figs. 1, 2) and between the two branches of each aorta there is given ofT a small branch to the lung, a pulmonary artery (fig. 4, Pul.R, Piil.L). These jnilmonarj' arteries pass to the lung comi)lex which is fused dorsally and free ventrally. On the right side there are three lobes and on the left four lobes of the lung.






There is hut one inulKii not two as in the specimen described l)y Carey.

ski:m;i'()N ik;. :,)

Tlie skull is shajx-d iiuicli likt- that of a uonaal pi^- Uii the dorsal side is a fontauelle covered with a mass of wrinkled skin representing the rudiment of a face and hound to the hrain hjconnective tissue strands which are part of tlie dura mater (Rud F. Figs. 1, 2, (i, 7). There are two separate vertebral columns with three points of attachment to the skull, the central one being common to each colunm (fig. (>). 'I'lu're is a central bony mass between the two atlases and probably derived from them, which makes part of the middle articulation to the skull. The second and following vertebrae in each Ixxly are nearly like those in normal pigs.







There are thoracic ril)s on the dorsal side some of which may connect to the cartilapinous mass [Cart. fig. ti) which connects till- two hodies dorsally. It is ))ossihle that this is |iart of a sternum. Al any rate, the Neutral lilis and stci-num arc more nearly' uoi-uial.

M;KVt>l S SVSIK.M (.I'KiS. c, 7i

The singh' cfrel)rum is slightly larger than that of a i)ig of the same size, ha\-iiig a 'disturl)ed' area in the region just below the dorsal rudimentary face. In dissecting off the skull ca]) shown in figure 7 no identifiable brain tissue was cut. simiily connective tissue such as composes the dura.

The cerebellum is normally placed and jjrescnts three lobes, one median and two lateral. The nudidlas are fused to each oth< r and from each a normal sjiinal coril extends ])osteriorly.


The ction of this nionstir jng is ])rcsentcd because it differs in many respects from that deseribrd by Carey, in some points being more fused, in others being more distinct.

1. The head is nmrr nearly normal with but two ears, two eyes and two nostrils.

2. The alimentary canal has a divided pharynx, a grooved stomach, a single pancreas and a single spleen.

'.i. The trachea is single, connecting to a ventrally lobed lung mass.

4. The heart is fu.sed, with a single venous system. The arterial supply of the head is from one side of the heart only.

5. The nervous system is fused to a nnich greater degree, there being complete se])aratii)ii in the spinal cords and fusion from the medullas anteriorly.

We wish to thank Dr. \. 1.. Ti-(>adw('ll for the ]>rivil("ge of dissecting the pig. and Dr. J. .\.. Culler for making tiie skiagraph.


.\i)A\ii. .1. (!. AND McCrak, J. K12 .\ text book of patholog)', pp. 759. Cahev, K. 1017 The .inulnmy of ;i doublo pig, Synccphiilus thoracopagus with

especial consideration of tlie genetic significHncc of the circulatory

apparatus. .\nat. Uec. vol. 12, pp. 177-191.



KOGEKS K. OLIVKK Anntomical Department, ViiiveTsxty nf Colorado


There is no place in medical literature where more confusion exists than in the discussion of the incmhranc s and other folds of |)eriton('uni found in the al)di)Miinal cavity atljaccnt to the ascending colon and caecum. Embryologists offer only brief suggestions as to their origin, and anatomists, after describing various caecal bands, and mentioning the ascending meso-colon. are silent. The existence of other so-calletl membranes had long been recognized by Virchow, Jonnesco, Treves, and others. Professor Biiuiey in 1005, under the name of pericolitis dextra, first described the structure which is now called Jackson's Membrane. Three years later Jabez N. Jackson recognized it as a causative factor of pathological conditions, and wrote his initial article, Membranous Pericolitis, in which he ascril)ed this condition to a memlirane which involved the ascending colon.

Other investigators, notably Pilcher and Brand, have compiled statistics and thce)rized in regard to its development and etiology. In 1912, Professor Flint of Yale, offereel the first satisfacteiry theory of its development and it is now generally accepted as a fact that it is a true embryological remnant, which, due to patliolofrical ciianges in its nnrnial structure, may kinking of the ascending colon and angulation of the hepatic flexure. The late recognition of this meml)rane as a elistinct anatomical entity is due to its superficial siniiliarity to the adhesions which are always present in acute antl clironic inflannnations of the peritoneum, so that earlier writers attributeil to it a i^urely inflanunatory origin.

Till; ANMOMICAL IIBCOHD, Vol.. 13, No.



Jackson's Moinlnaiio fonsists of a thin voil-liko strurturo which oripinatcs from the post oro-hit oral abdomiiia wall, and in some instances from the lateral parietal wall. It is attached to the ventral surface of the ascending colon and caput coli, and larcly involves the caecum or appendix. It usually resents the app(aranee of a transparent, highly vascular peritoneal fold. In older subjects it is thicker, more fibrous and avascular, and closely resembles an adhesion. Often its point of attachment is irregular and gaps are left between tendinous slips, thus forming digital fossae. In the majority of cases examined in the dissecting room, it seemed to originate a few centimeters above the iliac crest, and to continue for a variable distance along the body wall, finally merging on the inferior surface of the liver with the hepato-renal ligament. Radiating fibers ext( nd both downwards and upwards from this wide origin and usually blend with the anterior longituilinal band of the colon. .\ free lower margin can be demonstrated in most cases, althougli sometimes it forms a sac in which the ascending colon is placed. It has !)(■( II described as extending over to the transverse colon and binding it down so that the latter parallels the ascending colon, formii\g the so-called, "shot gun bowel." Its blood vessels are uiuisually large, dilated and unbranched and parallel each other, accompanying glistening tendinous slips. The veins at tlu ir points of exit from the wall of the colon are often tortuous and ( ngorged wiili l)|(i()il. This characteristic vascular appearance is more marked in young subjects and imi)li('s from its arrangement a distinct blood suj)ply. Dissection on the cadaver has substantiated this, and the origin of the ve.s.sels has been traced to the second lumbar, and more rarely the renal artery.


Only one microscopic examination has been made so far of a colon to which a membrane of this character was attached. Dr. I. .1. Hall, the pathologist of the Kansas City fleneral Hospital, under the .supervision of J. N. Jackson, made such a report in


which hp stated that no pvidonoos of inflainmaton' professes were noted in the eolon and suKKested that tlie presence of many endotheUal-defts in the membrane denoted probaljly a chronic lymphstasis. A similar section was studied in this laboratory. Very sliglit indammatory processes were noted in the cf)Ion wall ami in the membrane itself. The case, from which thespecimen was taken at autopsy, was one in which the patient had died from other causes, but with a history of chronic constipation. This condition was found to have been caused by a sharp constriction at the junction of the descending colon with the sigmoid flexure, and it was evident that this would have been sufficient to the conditions already referred to.


From the above description it is seen why various right sided pericolic bands have been mistaken for the Jackson Veil. Broad membranous adhesions are the most fre(iuent source of error and can lie distinguished by their irregular distribution, nonvaseularity, and by the unreadiness with which they can be stripped from the colon. Caecal folds and ai)i)endiceal ligaments have each their own characteristic origin. \n ascending mesocolon which occur.s in 26 per cent of all cases is distinguished by its point of attachment. Two additional important points .should be noted in the identification of this membrane. Tirst, invariably a huge mobile, ililated caecum is associated with it. Second, it can always be demonstrated by ilissection as a fold of tissue distinct from the parietal peritoneum.

In relation to the caecum one more fold of peritoneum was noticed, a definite structure which was found in twenty out of twenty-four cadavers in the dissecting room. Its occurrence, in such a high percentage of eases, its regular origin and similarity of appearance makes it highly probable that this is another true anatomical structure. Review of works of anatomy yield no definite information, although other supra-caecal bands are described. Robinson, in \\H)2, stated that the caecum was fixed by the right phreno-colic ligament, to which this band bears a slight resemblance. It originates on the lateral wall of


the ahtloinrn at a point opposite the iliac erest about 5 to 10 em. aliovc the anterior superior spine of the ilium and extends transversely over li> the ventral surface of the caput coli. The width of its upper margin varies iroin 1 ..'> to I? cm., and when it is made tense by traction on the colon it forms the lower boundary of a depression several centimeters deep. It is e\-ident that when the bowel is dilated at or near the point of ins(>rtion, this ileo-colic band will form a definite dam across the right para-colic gutter, and this apjiears to he not an al)normal condition, for the ca(>cimi is freely mobile and dilatetl in ()7 per cent ot all cases. This may be a factor significant in relation to the drainage of this area.


The interdependence of the various develnpiiiciital jiroccs-ses plays a striking part in the formation of membranes of this nature. It seems probable that viscera, such as the liver, kidney, and ascending colon, aid in shaping the attachments and Knal i)osition of the .lackson \'eil. .\n element of the utmost importance is the body growth in later foetal life, which takes place princijndly in the region of the lumbar vertebrae.

Flint has described the Jack.son Veil as merely a more complete fusion of the visceral peritoneum with the dorsal parietal peritoneum. secondary attachments are drawn downwards by the descent and rotation of the caecum into a thin veil-like membrane. Moic rarely he attributes it to a lateral prolongation of the great omentum. This developmental theory has not been fully .substantiated in this aboratory by the dissection of a .series of foetuses and one infant at term. It will bear still further elaboration.

I>et us first consider the .so-call(>d upward migi'ation of the kidney and the descent of the caecum in relation to this membrane. Re\iewing briefly the change in position of tlu se structures during intra-uterine life, we find that the kidney in the first half of foetal life lies opposite the first thn-e lumbar vertebrae, while in the second half of the same period the cranial pole to the level of the eleventh rib and the caudal pole descends to the upper border of the fifth lumbar vertebra. This is merely


a p-owth in length. After hirtli, in children of loss than i»ne year of ago. in .")() per cent of cases the caudal pole is in the iliac fossa 'and not until the child is over two year sof a^e does it rise above tlic iliai' crest. This is purely a passive change which is <lue to a stronj^er f^iowtli of the posterior abdominal wall in the lumbar-region.

After rotation the large intestiiie lies transversely along the greater curvature of the stomach with the caecum in the right side in front of the duodenum and closely api)lied to the caudal surjace of th(> right lobe of the liver and resting on the ventral surface of the ri^lit kidney. The descent of the caecum takes l^Iace after the third month, and in the fourth month the ascending colon is formed ; at the same time»tixation of the ascending colon begins and proceeds lateral-wards. The line of attachment is usually an oblique one which extends from the caudal end of the right kidney upwards from right to left. The caecum and ascending colon are originally completely invested with peritoneum. The a.scending colon develops in the space made by the rapid growth of the body wall. The height of this area, being increased slightly by the decrease in size of the Uver.

The descent of the caecum is relatively a passive one. At the fourth month the transverse colon occupies an oblique position; the caecum (having imdergone complete rotation) is to the right, in close api)ro.\imatioii to the bodj- wall. At this stage it lies on the ventral surface of the kidney, midway between its two poles. The lower pole of the kidney is still in the iliac fossa. The posterior surface of the caecum has already become firmly attached to the ventral surface of the kidney, but before this peritoneal blending occurs the fixation of the caecum is accom])lished by delicate seeondiiry adhesions between the caput eoli and the lateral body wall. M the fifth month the caecum lies on the caudal pole of the kidney. Later it descends only to a little lower level, so that the caput coli in the adult lies on the inferior pole of the kidney. The final position and .shape of the caecum is determiiK d by the rate of growth of its walls from a fixed point. The po.stero-lateral wall outstrips the medial i>ne in growth.


The ascending colon is already completely formed but has not yet assumed its final position l)(>cause of the size of the liver. Its upper three-fourths is unfixed ami freely mobile. Extending from its ventral surface to the lateral body wall are secondary veil-like adhesions, closely sinmlating in their distribution the .laekson Membrane of the adult. It is i)robabie that these atlhesions sub.serve the important function of sliaping the hnal position of the ascending colon, as later, due to the rapid growth of the body wall they will tend to pull the fjut lateral-wards and upwards, so it assumes the hnal vertical position seen in the adult.

The Ja( kson Veil is therefore derived from delicate secondary adhesions between the \'isceral and parietal peritoneum. .\ series of four foetuses and one infant at term were carefully studies! and the above conclusions were drawn from this data. The developmental theory evolved presents some totally new viewpoints in regard to embryology of the caecum ami ascending colon in their relation to the adjoining viscera, and before complete acceptance they should be substantiated by other workers. 1 is shown that the early fixation of the caecum and the more rai):d growth of the posterior wall of the lumbar regions are the main actors involved in the formation of this meml)rane. The rotation and descent of th(> caecum are of only secondary importance This theory explains why the Jackson Veil possesses its i)eculiar vascularity and its common origin and attachment. This membrane does not attain its full development until the third year of the child's life or at even a later period when sufficient room is made between the twelfth ril) and iliac crest for the r, ception of the k'dney. .Vt this period ilevelopment of the vertebra centra the kidney to be pushed outward and with it necessarily the colon, so that a relaxation of any tension of this structure on the colon is accomplished.

Several facts suggest that an asceniling infection may often be transmitted from the lumen of the bowel to the kidney and otlu r viscer.a. In the first jilace, microscopic examination hints at a lymi)hatic stasis; in the sec<md place, there is a lack of free drainage of the right paracolic gutter; in the third place, there is an intimate relation between the blood supply of the


meinbrano and that of the kidney; ami in tho fourth place, it is known that a lar^c percent a^ie of pus infections of the kidney are due to the colon bacillus, and that the riglit kidney is more liable to infection than the left.

The occurrence of the Jackson Veil in sixteen out of twentyfour bodies suggests that it should be looked for in a higher percen age of cases in which there are vague abdominal symj)toms that are not reliLved by ordinary operative procedures.

Sl'MM \I{^

1. The Jackson Veil occurs normally in ai)out til) to 75 percent of all cases.

2. It is a definite anatomical structure with a regular distribution and characteristic vascularity.

'.i. An adrcjuate exi)lanation of its origin and development can be shown by the study of the foetus.

4. Ilco-coiic bands of the the character described in tliis article are an important factor in peritoneal absorption and draining.


Jackson, J. X. lOI.'J Ann. of .SurR. Transactions Western Surgical and Gync coloRj- .\ssoclalion, 1!K),S. Lane. 1912 Brit. Med. Joiirn. PlLCiiEK 1912 .\nn. of Siirn. .Mayo, \V. ,I. A. M. .\., vol. 63.

Kl.SENDKATII AM) .ScllNOOR 1911 .\lin. of .SufK.

Fli.nt 1912 J. 11. Hosp. Bull. lOlU S'urn. (!ynpc. ami Obst. Kki.hei. and .Mali. EiiibryoloKV, vol. 2. Hc.VTINGTON. .Analomv of IVriloneuiii :ind .Vbdonieii.

AI'TfloHn AlinTIM'T nr Tlllt r\i-».|{ innl RU

in Tiir iiiiii.inijitM'iiK nKitvui. iht iikh 13

j)i;iAii.i';i) sriDv of a .M(.).\sti:i{ with ( uamu RAC'HISCHISIS AND OTHER ANOMALIES

CAKHON <;iI.I,ASI'li: AM) lloWAIil) III I.I, lli:rsT()\

Depnrtmiul of .{nnluiiii/, I'tiiivrxitij of Coloratlo

SIX Kir;i-nEs

The subject of this report was a foetal monster of some eiRht months which was brought into the clinic of the Colorado University School of M((li<'iii(' last December. It pre.-^ented ([uite the usual appearance of the anencephalic monsters but it was thouglit that a detailed study of its general anatomy might prove of interest. The results of this study are embodied in the following description.

The general appearance of this monster is well shown in the l)hot(>giaphs of figures 1 and 2. Here the most striking feature is the upturned face, apparently set directly on top of the head. There is practically no neck, the head and face projecting upward from the shoulders. The upturned face is a result of the cer\ical flexure, the cer^^cal vertebrae being flexed sharply forwaril and thus throwing the face upward. This cervical flexure can be seen in figure '^. Another interesting feature is the greatly enlargeil arms and hands which also show well in hgures 1 and 2. Figure 3 presents the condition of craniorachischisis of which more will be said later.

Dissection of the upper and lower extremities revealed little of interest other than the abnormally large arms and hands. These made up in volume i)ercentage ai)out 18 i)er cent of the total body volume which was greatly in excess of the normal for a foetus of this age. The circulation of the extremities was apparently norn\al as was also the nerve supply of these parts. The brachial and hunl)o-sacral plexuses were normally developed.

The condition of craniorachischisis is shown in figure 3. The bony floor of the cranial cavity and the spinal canal has been

Ki(». 1 I'acf view of llic iiion.slcr.

Fiji. - Vontnil iis|ioct of tlic inoiistrr. Note upturned face and large arms and hand.s.

Fig. ;{ Dorsal aspecl .•.Ikiw.iik li..B;il lioni's of ^lillll ;iiid floor of .spinal canal. Cranioracliischisi.s.

cleamd in oriUr to briiiK '>ul iiinrc cU arly the conilition oi bone development. Thf re wa.** nn brain and lu) cord, the !)onfs being t'ovrrrd only with the nic nibrane which forinrd the floor of the einbryoloKic nrural ktoovc. The stiuanious parts of the bones of tlie skull and the parts of the v( rl( brae forininK the sides and dorsum of the spinal canal were lacking ditirely. Of the basal bones of the skull there are plainly s( ( ii the s])henoi(I with the


anterior clinoid processes, the petrous parts of the temporal bones with the internal auditory meatus, and the basal part of the occipital bone with tlie ju^!;iilar foramina. Immediately below the basal part of the occipital bone is a deep recess which is formed by the cervical flexure mentioned above. The spinal colunui .shows the bodies of the vertebrae with the transverse processes and the intervertebral foramina. It is interesting to note that while neither brain nor cord was formed there was a complete and apparently normal development of the peripheral nervous sy.stem.

Dissection of the thorax .showed the heart distinctly more to the riglit of the median line than to the left. The right side of the heart was much larger than the left due probably to an engorgement at death but otherwise there were no abnormalities. The lungs and thymus gland were normal.

In the abdomen a numlx r of anomalies were found. A general view of the abdominal features is shown in figure 4. Here the most striking thing is the single large kidney occupying the greater part of the space of the left lumbar region. There were two separate ureters and two suprarenal glands althougli only the left one of each is shown in the drawing. The peritoneum has been partially removed to better displaj' the kidney but the part remaining on the right side covers the right ureter and the right suprarenal gland. The grooves in the ventral surface of the kidney contain the left renal vessels. The uterus and bladder show a distinct displacement to the left which will be explained later. The stomach shows a bilobed fundus and a cyiindrieally shai)ed body. The spleen hail a large blood clot in its ventral aspect and two or three small accessory lobes not shown in the ilrawing. The liver and pancreas showed nothing unusual. The volume of the various organs and parts of the body were con\pared with Jackson's table on the jirenatal growth of the liuinaii body and none excepting the upper extremities found gn ally out of proportion to the normal.

Anomalies of the circulation were found in the aixlominal aorta and the vessels of the pehis. These are shown in figure 5 which is a diagram of the circulation as found. The inferior

mesenteric artery comes otT tlie aorta at a imich hij^lier level tlinii normal while the coeliac axis. sii])erior mesenteric, renal and ovarian arteries present their normal relationship. In the pelvis the aorta was continued in the median line to form a singl(> umhilical arteiy which caused a displacement of the uterus and bladder to the left. This median aorta or umbilical arterj' gave otT the uterine, right external iliac and the normal branches of the right internal iliac arteiy while the left common iliac

Ki({. 4 .\l)d<>iiiinal visotTii .'iftrr removal of livor ami intcslini's. .1. rJKlit iiiarcin <>f kidney covprrd with |X'riloiieum; li, normal ri^lit kidney .space; C, me.ienlc'ry drawn to t lie riKliI i IK liver space; E, diaphragm; A', vena cava; fl, stomaeli; //. spleen; /. lilood clot; ./. cut edge of le.s.ser omentum; A', left suprarenal Kland; /-. rut e<l({e of great omentum; .1/. kitlney ; .V, cut edge of ix-ritoncum; O, left ureter; /•", jjelvic colon; Q, ovary; H, uterus; iS', bladder.

FIk. 5 Diai;rnm of abdominal aorta and pelvic vessels. A, coeliac axis; li, sujierior mesenteric; (', inferior mesenteric; I), renal arteries; E, ovarian arteries; F. left common iliac; (!, internal iliac; //. uterine arteries; /. external iliac; J, branches correspondin(j to the remaining vessels of the right internal iliac; K, the single umbilical artery; //. right external iliac.

foniicd and tcnniiiatc d in the Irft intrriial and fxtornal iliacs with ni) iiniliilical artery. A i)ln)t(inii( rojjiapli of a cross section of tlic uinliilical conl with its v( in and one artfry is shown in linurc t».

Fig. 6 I'hotomirronrjiph of I lie iiiiilMtiral cord. Nolo vein :iii(l one :irl<>ry.

("onrcrninjr the causative factors uiuierlying the production of monsters and the various anomalies many theories have been advanced. Such are faulty implantation of the ovum, mechanical influences, diseased condition of the uterus with consequent al)sorption of jioisons, and the action of \arious salts as those of sodium, calcium and magnesium. .Vlcohol and anaesthetics have been viewed with .suspicion as to their influence. Malnutrition of the various ti.ssues and organs due to circulatory disturl)ances with resulting failure of development and atrophy of the i)arts afTectcd explain many anomalies. For a more complete discussion of these various theories the reader is referred to the current literature and to the many excellent textl)ooks of embryology.


A TEAt'JlLXCi -MUJ)i;i, Ol' A lo M.M. I'K, J^MBKVU

IVAN K. UALLIN Department of Aitatomy and liiology, Marquette University Medical School


111 an intniductory course in iiiainnialian riiil)iy()l()g>' a great deal of jirecious time is Keneraliy lost in teacliiii}!; studr nts the values of striietiircs in serial sections. While the standard einhryijloKical iiiodds in use are indispensable in the teaching of the subject they are of little assistance in helping the student to understanil the relationship of a single section to those that follow and precede and in giving him a mental iiicture of the entire structure.

The model which is described below (fig. 1) is so constructed that a student Working on the serial sections can go to the model and see any jiarticular se<-tion in its relationslii]) to the ones that precede and follow it. .\ ]>art of this model was used in the laboratory during the past aca<lemic year with gratifying results. Both simple and difficult conditions in the microscojjical .sections were readily understood when the ]virticular segment of the model had been examined.

I do not believe that a large model of a pig embryo has ever been constnieteel on this plan before. The value of such a model in the teaching laboratory it seems to me warrants a description of the construction.

The model was made of jiaraffined blotting paper at an enlargement of .')() diaiiK ters. Thi- method of using the j)ai)er differed a little from the method I described in a previous publication.' The drawings of the sections were made on tracing cloth which are placed in a permanent file for future reference and by using a carbon pai)er the original drawing was transferred by one operation to the i)lotting i)ap» r. The blotting |)ai)er is afterwards soaked in melted jiaraffin.- ^Io.•^t of the general (iiilline of the section can be cut with scis.sors and afterwar<l com|ileted with a knife cutfing over a glass plate.

In cutting the sections only that ])art of the mesenchynia lying next to the ectoderm is rejiresented. This is retained to give a sufficient thickness and firmness to the blocks of segments when they are jicrmanently stacked. Structures in the interior are held in place by means of bridges which are retained ]iermanently.

' .\m. Jour. .Vn.'it., .July, l'.»17.

'This mrthud, I believe, has been described before. However, I have been unable to find a reference lo it.




The iiHiiIcl w.'is luiilt ill scuiiH'iits. The thickness of the sfRiiiciits vnrit's and depends ui)iin tln' structures within. Certain scKinents, for <'xaniple no. 7 and no. liJ, are eoini)aratively thin. It was neeessjirv to stack them in this way so that certain iinjMirtant structiires would not l)e covered. The segment (Xo. 24) is mounted on a wood stand (stand doi s not show in photonrajjh). A block of wood was carved to receive the contour of this segment which was made fast to it with nails. The entire stand and segment were afterwards electrojilated with copper. The other .scKinents merely rest on toj) of this last segment when th<' entire model is set uj).

The structures in each s( pment were jjainted .so as to make them stand out more prominently. The tyjie of ])i<;mcnts used were the ordinary tulie )>i(;m( iits inixiii with rulibini; v;iriiish to the eonsistt ncj^

of thick s>Tup. This method wu.s suggested by the de]iartnient artist Mr. Leo Mas.sopust. This mixture has been found to be very satisfactory. When the mixture has thoroughly dried on the model it foriii-< a very hard coating which apjiears to give promise of good wear. .VII i)ainted surfaces are finally covered with a coat of French varnish.

Figures 2 and '.i are i)hot<>gra])hs of the cephalic an<l cau<lal face of .sigment 14. This segment contains a jiortion of the heart with very sm.all parts of the liver showing on the caudal face.

There may be certain advantages in building a model of tiiis t\|)e b}' the use of the wax plate method. Structures, like ganglia, could be rounded off nion' easily than when the blotting paper is used. The wax segments could be elect ro])lated with eo|)])(r and in this way be macle more serviceabli; for room haii<lling. The cutting of thi'


•J". 17

iviraffiuc'd hlottitiK i):ipir is also more diflRcdlt tluitithe cuttinRof wax plates.

I wish to take this oiipoiiiinity tn ackiiiiwiciJKr llic iiitiili}{riit assi.stanrr of our hiWoratory technician Mr. Paul Knrdes without whose help the completion ot this model would have iieen delayed many months.



or Tiu: iiiiiLlooHAPMir hiekvici: octodek 13.


J;D\VARD CARROLL DAY Depnrtmeiit of Jiiology, Bryn Maivr College



A li)iin-fclt need for a suitaMo and suhstantial laboratory voliidc led to tlic coiistniction of tlic car shown in the accompanying pliotograplis (figs. 1 and 2 1.

Tiic frame is of lialf-incli galvanized iron l)ipc, and measures 34 indies in length, 18 in width anil Hi in height between the top and bottom trays. The front and hind wheels are 12 and 8 inches respectively and are the rubber-tired regulation type made for go-carts. The forks as well as the wheels were obtained from a manufacturer in Piiiladclphia. The ]ilumber of the College did the work of cutting out and fitting the frames together and the mechanic attached the wheels.

It will be ob.s-erved that both i>air of wheels are carried on longitudinal bars inside of and jiarallel to the bai's forming the bottom • if the frame. The inside bar is, in each, 3 inches from the outside one, measuring from center to center. The vertical jiipes for the small rear wheels are set 4' inches from the rear end and the verticals for the large fnmt wheels are .set 2' inches from the front end, and are <'onnected to the long bars with T-fittings. The vertical i)ipes for the rear wheels had to be bored out a trifle in order to take the shafts of till' forks. .\ collar on the upiier end of the shaft iirevents the fork from falling out when the car is lifted. The axle of the front wheels is fastcncil witli screws into the vertical jiipes which have been plugged and nolciied at the ends for the purpose. The front wheels are braced by cros.sed stays running from the axle up to tlie long bars, while the iiind ones are braced by a horizontal |ii]ie just above the forks. The handle of the car is given the desired angle by a 45 degree elbow on either side.

The trays are of ,' inch hard wood and are held in place by cleats. The top one is 2' inches deep and has a hand-hole at each end. White enamel gives an attractive linish to the vehicle.

In ordei' that the e(|uilibrium of the car may not be u]>.set when the two hind wheels turn siileways as in rounding a corner, they nuist necessarily be situated near to the end aiul close to the sides of the frame. .\lt hough ', inch pipe was used, a % size would do equally as well for the given dimensions of the car.







I'or !i lalxiiatotv liMviii;; I'DiiiprcssiMl ;iir Jivailalilc an injection tank of tile type sliown in liiiurc A will (iicatly fai'ililatc the injection of the circulatory systems of a laijic nuinliei' of cats or i'ai>l)its for dissection pini)()ses.

'I'lie inlet for the compressed air is on the left; the outlet for tlie injection mass on the rinhl. The pulley in tiie center is connected to a shaft carrying a (lonl)le-lilad<'(l stirrinji-paddle at its lower end on the floor of the lank. 'Die slifrhl knoh just to the left of the [)ulley on the top surface is an inlet lilteil with a screw-plup where the injection mass may he p()ure<l in. The dial is that of a lO-pound |)ressure jjauge. To tile ritihl of this is an aii-valve for retiulatiiiy; the jjressure. An ann-linlc .'ii llie liolioni of liie can, filled with a hrass cover wliidi is

held III place with machine screws, facilitates the cleaning uf the tank after usintr. A small circular window in this cover the level of the liquid within. The is inserted from the insitie and is held in place with sealinfi-wax. The liottom of tlie lank, like the top, is convex in order heller to wiliisland the pressiuc and to concentrate Ihe liciuid toward liie center as its (|Uantily diminishes.

Willi a relalixcly ihiii injection-mass and a pressure of lliree to four pounds the ai)paialus works very satisfactorily when injection is made llii-oumh the heart of the animal. We made a i)uncture in the left ventricle, inserted Ihe cannula, holding it in place with the tinjiers, gently turned on the litjuid, injected until the line arterial capillaries stood out iii>on the surface of the viscera- wliich re(|uired ahout a minulc, then heated the dorsal aorta. The arteries were thus tilled with Ihe yellow starch while Ihe venous system hecanie surcharned with lilood; thus both .systems could he readily traced oni by the sludrnt.


AN in:in:i)i'r\HY case of coxcKxriwi. ahskxck


MAHCrS W. LYON, JR. George Washington University, Washington, D. C.

ONE Fir.rnK

The followiriK case of ahsciicc of a kidney is of interest in view of tho statement in the family liistory of tho suhjoet that the maternal (irandinother had heen operated on for some ahdominal eonditioii when it was found that one kidney was absent.

The anomaly oceiirred in a white female, ape H . weight \'M) pounds, who eame to necroi)sy fiiilowiiif; an operation for the relief of aseites due to cirrhosis of the liver. Associated with the ah.-'ent kidney there was an ahsence of the t)varv and uterine tube of the corresponding side.

The ritilit kidney was entiicly absent and no renmants of it could be found in the abdominal or pelvic cavities. The left kidney was in the usual position, and hyperti'()])hied. It weifjlied 270 grams and mea,«vu'ed 14(1 by 70 by 4") nun., with a cortical tiiickness of from 10 to 1.") mm. It was supjilied with a single ureter.

The left renal artery was of normal size. The right renal artery was very small .'ind distiibuted to the right su|)rarenal body. I'rom its a very fine Iwig came otT which probably went to the suprarenal also.

The left ureter was single and normal in every way. The right ureter was present forming a distinct tube in its lower portion and reduced to a fibrous cord in its upper portion which disappeared into a of loose fibrous and areolar tissue. Although the right ureter was connected with the bladder, no opening from it into the bladder could be found. The interior of the right ureter contained a small amount of fine granular debris.

Hoth suprarenal bodies were present. The left one was in its usual position above the kidney. Unfortunately it was removed and its connections destroyed before it was learned that the right kidney was lacking, .\ccordingly it does not appear in the illustration.

The uterus was very small and in its usual position. It had but one uterine tube, the left, which as shown in the illustration was bent over toward the right embracing a single ovary in its fold and between it and the uterus. \o other ovary could be found. There was no indi<'ation of another uterine tube. .Mtiiougli the single ovary a<-tu:dly appears on the right side, it is to be interpreted as the left ovary both on account of its relations with the left uterine tube as well as to the previously observed fact' that in i-ongenital absence of a kidney there is

• Radascli, H, K., Aiiicr. Journ. Med. Sci., n. .s.. 136, III, 190S.


fn'<iiifiitly ail iissociatnl lack of (icvrlopiiii'iit of tlic sexual ulaiids and apparatus nii tlii> saiiir side. The sexual history of this sui)ject is as follows; .Menstrual [x-riods lie^saii at \'.\, irieRiilar at first, hut horanip regular, no pain, no clots, ipiantity normal. Marrie<l at 1(1, had a |)icin;iliiic hirtli after heiiif; married fifteen years. Had another iiiiscarriii^e at 'Mi years and never iiiensi mated since. .\ small myoma, ••ilioiii I.") mill, in diameter, was in the uterine wall.

lig. 1 Case of conucnilal aliscnct- of ri({iit kidney, ovary, and iiliTinc tube. Pri'iiaration shows hyprrtrophii-d loft kidney, and it.s connections with the ad(lominal aorta and bladder; the riuhl .iiiprarenal Imdy and its connections with the abdoniinal aorta; the righl ureter lost at)ove in a small mass of fibrous and areolar li.-wue, and joined below with the liladder; the incised uterus with the left uterine tube eurvinR over to the right and between it and the f\indus of the uterus the ovary. 'I"he dark irregular area at the ripht is the liiiibriated end of the tube. For comparison a II) cm. rule is placed at the bottom of the preparation.






G. CARL HUBER Dcparlmenl of Anatomy, University of Michigan


A comparative study of the renal tubules of the different classes of vertebrates was projected nearh- twenty years ago, soon after publishing results of a study on the development and shape of the uriniferous tubules of the higher mammals.' In this study, as projected, it was purposed to reconstruct the excretory tubules of the pronephros and mesonephros of certain of the lower vertebrates, including amphibia, and the metanephric tubules of certain reptiles, birds and mammalia. In the reconstruction, by the Born wax plate method, of the pronephric tubules of a larval toad, and mesonephric tubules of an adultfrog and the metanephric tubules of certain reptilia no great difficulty was experienced, and such reconstructions were made, somewhat over fifteen years ago in conjunction with Profe.«sor Ward J. MacNeal, sometime Instructor in this department. On attempting to reconstruct the metanephric tubules of birds, it was learned after unsuccessful trials that this was beyond the limits of the method, so also with endeavors to reconstruct the metanephric tubules of adult mammalia. The projected study, therefore, was aljandoned for a time. The form of the adult, manmialian renal tubule was later ascertained by specially devised methods of teasing.' This special method of teasing has relatively recently l)een successfully used in an investigation of the form of the metanephric tubule of birds. I am in a position now, therefore, to present figures giving the morphology

' Hubcr, G. Carl. On the development and shape of uriniferous tubules of certain of the higher mummnls. .Vm. Jour. .\naf., Supplement to vol. 4, 1905.

' Huber, G. Carl. .\ method for Isolating the renal tubules of mammalia. Anat. Ree., vol. 5, 1911.




of type forms of renal tul)ules from the difTcrent cla,sses of vertelirates. Their presentation does not seem to me superflous at the present time, even though Zarnik' has in the meantime published an excellent study of the renal tuliules of certain reptilia. The duct system of the kidneys of certain of the forms studied does not lend itself readily to investigation by reconstruction methods, owing to magniliration necessary to make the requisite drawings, and the size of the resultant moilel. Therefore, an attcmiH was made, relatively successful in certain of the forms studied, to ascertain the form of the duct S3'stem by means of the celluloid, corrosion injection method devised by me and presented in connection with a study of the artcriolae rectae of the mammalian kidney.' The figures here included, based on ]ircparations collected through many years of work, therefore, are in part of reconstructions made, in part of complete renal tulniles isolated by teasing and in part of celluloid, corrosion l)reparations. It is hoped that the figures will be sufficiently clear to obviate the necessity of extended description. This account is intended more as a pictorial summarj^ of results obtained as concerns the morphology of the renal tubules and the kidney duct systems of certain type vertebrates, than a critical (iiscussion of the minute anatomj' of the renal tubules considered,

ince the literature is much richer as concerns the latter than

that which concerns the morphology of the tubules. Likewise, only the pertinent literature dealing with the morphology of renal tubules is here considered, otherwise this account would be extended far beyond the limits of these pages.


The reconstruction of the proncphric tubule here figured was ol.tainod from a series of frontal sections of a larval toad, having a length of 10 mm. The variety of Bufo I am unfortunately unable to give. The ova were collected in a neighboring pond

' Znrnik, B. VorRleirhondc Studion ul)er don Ban der Niere von Echinda und dor I{«'i)tilirnniorc. .IcnaiscliP Zcitsclirift, vol. 46, 1910.

  • lliil'or. (i. ("art. The artcriolae rcrtac of the mammalian kidney. Am.

.lour. .\nat.. vol. fi. 1>K)7.


and could not l)o spfoificallj' idcntifiod. Tho roconstruction was made at a magnification of two hundred diameters. The figure is added more with a view of making the series of figures complete than with a view of adding materially to current knowledge of the aniphii)ian pronephric tubule. The figure here given presents similarity to that given by Herbert H. Field,' figure 41, plate 4, a reconstruction of the right pronephros of a larva of Rana having a length of S mm., from anterior end to tip of tail. Field's account of the amphibian pronephros is the best available. In the l)rief statement here given, his nomenclature is used. The pronejihric tubule shown in figure l.pre

n&i «»— -.»ii»^ :n& ct

Fip. 1 Reconstruction of the pronephric tubule of a larva of Bufo, the larva having a length of 10 mm. from anterior end to tip of tail. X 80. tis., ncphrostoinc; rl., common trunk.

sents only two ncphrostomes, n.s. The two nephrostomal tubules join in a Y-shaped junction, hidden beneath the model as placed in the figure, in the neighborhood of the region indicated by a cross. The common trunk presents three main loops extending primarily in cephahc caudal direction, each presenting three to four secondary loops. The figure presents a sketch over an enlarged photographic print, the photograph having been taken from a direction which admits tracing the common trunk most completely. The arrow indicates the direction taken by the tubule in the one location where confusion may arise. The actual length of the common trunk reconstructed as measured on the model, is approximately -.."> nun.

' KieUI, Herbert II. The development of the pronephros and segmental duct in amphibia. Bulletin of the museum of Comparative Zoolog.v at Harvard Col-, lege, vol. 21, 1891.

308 Q. CAUL lllIlEll


Tho niosonophric tuhulc licrc linunHl was roconstructod from the kidiiov of an adult, fciiialf fro;";. Haiia catcshiaiia. The duct system is fipurod after cdrrosion pn^parations made l>y injecting through the ductus tleferens or ureter, a solution of celluloid in aceton, colored with alkaniii. Otdy now and then was it po.ssii)Ie to inject anyway completely the duct s^'stem. Usually, owing to the pressure necessary, the ureter ruptures before the desired injection is obtained. .Vfter injection, the parts remained in place for fifteen to twenty minutes and were then transferred to a 7") per cent solution of hydrochloric acid, in which they remained for about twenty-four hours. The macerated pieces were then transferred to a large dish of water and the softened tissue removed by playing water against the preparation with a dropper provided with a rubber bulb. After the corrosions were thoroughly cleansed, they were placed in distilled water for a few hours, then ilehydrated, and transferred to xylol and if desired mounted in xylol-i)alsam. The mesonephros — kidney — of the frog is a compound tubular gland, possessing numerous renal or me.sonephric tubules, terminating in a series of transverse collecting ducts which unite with the ductus deferens or ureter. The ductus deferens joins the kidney near the posterior end of its dorso-lateral margin, passing forward to the cephalic end, where it becomes partly embedded in kidney substance. Into this main duct there emjity at fairly regular intervals and approximately at a right angle numerous transversely coursing collecting ducts, which, with or without branching pass near the dorsal surface of the kidney to near its lateral margin. These transverse collecting ducts have a relatively straight course and are in the main parallel to each other. For the greater part of their course they are embedded in the kidney sub.stanee near its dorsal surface. In figure 2, there is presented a corrosion preparation of the ductus tleferens with the connected, transversely coursing, collecting ducts from the kidney- of Rana cate.sbiana. The injection in this was fairly C(jmplete and after niacera• tion it was possible to mount in one piece the entire corrosion.


The transverse colloc-ting ducts are represented for about one half of their entire length. About the full extent of their branching may lie observed, since scarcely any branching occurs except relatively near their termination in the ductus deferens. In figure 3, is shown a corrosion of one of the transverse collecting ducts with the terminal ends of the connected renal tul)ules, from a region in which the injection extended for a distance into the renal tubules. The transverse collecting tubules shown in figure 3, corresponds to one of the main side branches of the ductus deferens as seen" in figure 2. Tlic dependent tut)ules, hooked at the lower end (fig. 3) correspon<i to that portion of the recon

I'in. - Corrosion preparation of ductus deferens and transverse collecting tubules of kidney of Hana catesbiana.

structed renal tubule, shown in figure 4, leading to the segment of the reconstructed transverse collecting duct. The manner of termination of the renal tubules in the transverse collecting ducts is clearly shown in figure 3. They end along the ventral border of these collecting ducts in two somewhat irregularly arranged rows. B}' an imaginary combination of the portions of the duct system of the frog's kidney as detailed in figures 2 and 3, the entire duct system of the frogs kidney to and including portions of the renal tubules, may I)e projected.

The renal tubule of the frog -mesonephric tubule — was figured by Xussbaum' many years ago. In figure 28, plate 23, Arch. f. .Mik. .\nat., vol. 27, 1SS(), he presents a complete uriniferous tul)ule from the kidney of Hana esculenta, teased after maceration in hydrochloric acid. This figure has been repeatedly

• N'ussbaum, M. Ueber den Bau und die Thiitigkei^t der Xierenorgane. .\rch. f. Mik. Anat., vol. 27. I8SC.



copied and presents many points of similarity to the figure of a wax reconstruction here shown. In figure 4, is shown a reconstruction of a complete renal tubule of the frog, Rana catesbiana, beginning with renal corpuscle and ending in the transverse


Fig. 3 Corrosion preparation of transverse collecting duel with the terminal portions of the outcring renal tiilmles from the kidney of liana catesliiana.

Fig. 4 Keconstruction of a renal (me8unephric)tubule8of a female frog. Rana catesbiana. X 80. re, renal corpuscle; r.., neck; prox.con.. proximal convoluted segment; dis.con.. distal convolut«l segment ; con./., connecting tubule; <, transverse collecting duct.

collecting duct. The figure is placed so that the dorsal surface corresponds to the upper border of the figure. A cross section of the kidney of Rana catesbiana appears not imlike figure 64,


of Haiipp's' oditioii of I'xkor's and Wicdcrseheriu's "Anatomic (k's Frosclii's," litO."). The renal tubules traverse the entire dorso-ventral thickness of the kidney. The great majority of the renal eorjjuscles form an irregular layer sifualeil near the ventral surface of the kidney. Kach renal tubule begins at a renal corpuscle. The renal corpuscles are relatively large, of flattened oval form, with long axes extending in a dorso-ventral direction. The one reconstructed presents the following measurements: length, 0.12.3 mm,; width, U.l mm.; thickness, O.OSH nmi. The renal corpuscle is joined to the tubule by a short and narrow neck, n., directed dorsalward, constituting part 1, of the tubule proper. The neck is lined by relatively short, nonpigmented cells supporting long cilia. The neck is follcjwed by a tubular portion having a relatively large diameter and extending toward the dorsal surface of the kidney there to form several bold loops which may extend to about the middle of the kidney substance. This portion is comparable to what is designated a.s the proximal convoluted portion of the mammalian renal tubule. Its course is clearly (hscernible in the tigure: I have designated it prox.con., proximal convoluted portion. The actual length of this portion of the tubule reconstructed, as measured on the model, is approximately 8.S mm., and it has a diameter, on the average, of I'jm- It extends from the narrow neck, just above the renal corpuscle toward the dorsal surface where several bold loops may be observed, then returned toward the ventral surface ending near the level of the renal corpuscle. In the figure this j)ortion of the tubule is partly hidden by the small, coiled tubular portion lying to the left and slightly above the renal corpu.scle. The cellular lining of this portion of the tubule is not unlike that lining the proximal convoluted portion of the manunalian kidney, presenting an inner striated border and now and then pigment granules in the basal portions of the cells. The portions of the renal tubules of the frog here designated the proximal corrulated jjortion, taken collectively, constitute the greater portion of the dorsal half of

' Gaupp, F.. Ecker's und Wiedcrsheim's Annloinie di-s Kmsohes, Dritte .'VbtliciluTiK. 1!X)-I. IIiirnorKaiK', pjiRr. SiS.

•\\'2 (i. CAKl, III llKK

tilt* kidney suhstaiiop. Tliis prDxiinal convoluted portion is followed by a short and narrow segment eouiparalile, in position only, to the medullar}' loop (looj) of Henle) of the niainnialian renal tubule. In the figure given, this portion of the tubule reconstruct eil is practically hidden by the tulmlar coil complex above referred to. This .segment is .said to l)e lined by ciliated epithelium similar to that lining the neck; it constitutes Claupp's Dritter ('analabschnitt.' In the tubule reconstructed this portion presents a length of a little less than 0.2 mm.; however, in the model its delimitation is somewhat uncertain. In the .succeeding tubular segment, Gaupp's 'Vierter Canalabschnitt," the tubule again increases in size slightly, reaching a diameter of al)()ut 3(V, and presents numerous coils and loops certain of which reach the ventral surface of the kindey. The general course and relations to the renal corpuscle of these loops is well shown in the figure; thej* forming the tubular coil complex of the lowcT third of the figure. This portion presents an actual length of approximately 2.5 mm. It is lined by a short epithehum, presenting basal rodding. In cross sections of the kithiey this l)ortion of the tulmle is found in the ventral portion, extending from a little above the level of the renal corpuscles to the ventral surface of the kidney. This portion of the frog's renal tubule may lie compared to the thicker portion of the a.scentiing hmp of the medullary loop and the distal convoluted portion of the mammalian renal tubule and is here designated the distal convoluted ixirlion, flis.ron. of figure 4. This segment is succeeded Ijy the final segment leading to the transverse collecting duct. If may be designated the junctional or connecting tubule, con.L, dearly shown in the figure of the model as it joins that segment of the transverse collecting duct, Ir.col., reconstructed. In the tubule reconstructed it presents a length of 0.7 nmi. In figure 3, the dependent, liook-shaped tubular portions represent this fifth tubular segment, tlie junctional or connecting tul)ule. This portion of the tul>ule is lined by an epithelium similar to that of the collecting ducts. The entire renal tubule of the frog as recon.structed and presented in figure 4, has a total length of slightly over 7 mm.


mktanf.i'Huk; tcbulks

A .study of thc! iiictuiicphric tuljiile rutiuircs coiisidcnition of the renal tubules of roi)tilia, birds and mammalia. For this study the renal tubules of a number of species of reptilia were modelled after th(> Born wax plate method, in the hope that some data eoiicerniiig the ance.stral origin of the metanephrie tubule might be obtained. This hope was not realized. There is found a eertain resemblance in morphology between the amphibian mesonephric tubule and thc reptilian metanephric tubule, but beyond such resemblance in form the kindejs of the types studied present no definite data as to their developmental relationship. Models of renal tubules from kidneys of the turtle, (lie alligator, the snake and the lizard are here depicted. Zarnik has considered very fully the structure of the reptilian kidney, embracing in his study, among other forms, the kidney of the lizard, the snake the alligator and the turtle. His comprehensive publication appeared after the models here figured were completed, This article is valuable for the literature reviewed as well as for the observations recorded. These as concerns the form of the renal tubules investigated are based almost wholl}' on teased preparations. My own observations corroborate many of his results. Their publication at this time seems ncver-t he-less warranted in that they are based on models in which, as I have had abundant opportunity to learn, the relations of the different parts of the respective tubules are more accurately niaiiitaiiied than in teased preparations, and further, thc forms studied l)y me arc not identical with those investigated by Zarnik. In presenting my own results of the study of the forms of the renal tubules of reptilia it will be necessary from time to time to draw attention to Zarnik's observations, obviating \\w necessity of ;i complete review of his work.

For tlic study nt' tiie morphology of the renal tubule of the turtle several such tubules were reconstructed from the kidney of ("hrysemys marginata. The kidney of this turtle presents on its surface shallow grooves, in the main parallel, and having a direction which is in general at right angles with the long axis



of the kidiioy. Thoso proov cs aiiastomoso horo ami t here. They IiouikI iiiilistinct foliLs or psihuIoIdIiuIcs, drained l)y main branches of the duct system. The ureter enters on the ventral surface. Two of the several renal tul)ules reconstructed are reproduced in figure 5, A and B. In the folds or pseudolobules of the chelonian kiilney, as also pointed out by Zarnik, the branches of the renal artery as also the branches of the venae renahs revehentes coarse through (he centers of the pseudolobules, between two la\'ers of kidney sui)stance, which respectively meet iu the center

intseg. \ y



Fig. 5 .\ aixl B, two renal tuljulcs reconstructed from the kidney of Chrysomys inarginata. X 80. re. renal corpuscle; n., neck; prox.con., proximal convoluted scKinenf ; int.seg., intermediate segment ; rfi's.ron., 'distal convoluted segment; con.t., connecting tubule;, collecting duct.

of the pseudolobules and extend to its surfaces. The main branches of the duct system, arising from the primary branches of the ureter, radiate from the ventral to the dorsal surface of the kidney, coursing in tlie kidne3' substance, on each side near the surface of the pseudolobules. Each so called pseudolobule, therefore, is composed of two layers of kidney substance having between them branches of the renal artery and branches of the venae renalis revehentes, each layer ijeing composed of masses or renal tubules, the main axis of whose coil complex is approximately at right angles to the surface of the pseudolobules, and to the main branches of the collecting ducts.


The rrnal tubules of Chrysomys inarRinata bcRin in a rolativcly small renal eorpusele averaging about oom in diameter. They are situated on each side near the centers of the pseudolobules forming on eaeh side r)f the center an irregular layer. The renal corpuscle is followed by a short and narrow neck, lined by an epithelium bearing long cilia. The neck is followed by what may be regarded as the proximal convoluted portion. Zarnik's 'Ilauptstiick' which courses toward the periphery of the pseudolobuie, overlying the associated collecting duct and forms a large and conspicuous portion of the coil complex of each renal tubule. In tubule A, of figure 5, this segment, prox.con. is arranged in the form of a letter N, extending from the neck, n., to the region marked by a cross, found at the middle of the upper portion of the figure. The three parts of the N-form present each smaller secondary loo])s. In tubule B. of figure ri. the arrangement of the proximal convoluted portion has an arrangement which is essentially the same as in tubule A, though the final rising arm of this X-shaped segment is not as long. In tubule A, this tubular segment presents a length of 1.5 mm.; in tul)ule B, 1.2") mm. and a diameter of from 150 to Oop. It is lined by an epithelium not unlike that lining the proximal convoluted portion of the mammalian renal tubule, presenting also a striated inner border. Collectively these tubular segments form the outer border of the pseudolobules. This large and long segment is followed by a short and narrow segment, lined by ciliated epithelium and (•orresi)onds in position only to the thin arm of the medullarj- loop of the mammalian renal tubule. In tul)ule A, of the figure this short intermediate segment m<.se^. descends from the region of the cross to the region where the tubule again becomes larger, about the middle of the figure. It has a length of a little le.'is than 0..'^ nun., and a cHameter which averages 20^. In tubule B, this segment is represented by the dotted loop found a little to the left of the middle of the figure. This intermediate segment is followed by what I shall designate as the distal convoluted portion, forming in both tubules figured a coil complex in the region of the renal corpuscles, thus found near the center of the irsoudololiules. In both tubules



this segment prosoiits a length of very nearly 1 mm., and a (lianiolcr which varies from Mn to 4o/i. This sc'{i;ment is lined hy a low eolutnnar epithelium showing a striated protoplasms. It is followed by a junetional or connecting tulnile, con.t., which in both figures lies to the right, extending respectively from t e loop which lies at the lower right hand corner of each tubulo (igured to the .segment of the collecting duct which crosses each tubular complex at ai)pniximately a right angle, col.d. This tul»ular .segment has a length of 0.5 mm. This gives a length to tubule A, from renal corpuscle to collecting duct of 3.3 mm. The tubules here figured arc similar to those figured by Zarnick, both in form and arrangement, except that the duct segment

Kig. G Duct systoiu of pscudolobulc of kidni'v of Chrysemys mari;inata, corrosion preparation.

P"ig. 7 Di.stal collecting duct with connecting or junctional tubular segments of renal tubules emptying therein. Chrysemj's marginata, corrosion preparation.

figured by him crosses the coil coin|)l('x a little nearer its center than in my figures, see Zarnik's figures 35 to 37 of plate 7. Other renal tubules modelled from the kidney of Chrysemys marginata and not here figured are in their morphology not unlike the tuliules here figured.

I n figure G is represented a drawing of a fairly complete celluloi(| corrosion injection of the duct system of a p.seudoIobule. The main duct, bottom of the figure, arises from the ureter on the \entral surface of the kidney. The primary and secondary branches radiate toward the dorsal surface of the respective p.seudolobules. coursing in the kidney substance, on each .side


noaixT the surface of tho psou(l()l(»l)ulo than its ponter. Into oach of the branches empty numerous renal tubules. In fiRure?, is shown one of the radiating collectinR ducts, many of which are figured in figure ti, taken from a corrosion injection in which the injection mass in certain parts extended for a distance int(j the connected renal tubules. In this figure, each dependent, iiooiv-shaped tubule corresponds to a connecting or junctional tulnde a.s figured in the reconstructions shown in A and B of figure 5. The segment of the collecting duct shown in figure 5, rol.d. corresponds to a short segment of the distal collecting <luct uniting the several connecting tubules as seen in figure 7, or of the radiating and branching ducts depicted in figure (i. Figures (i and 7, give a better perspective of the duct system of the chelonian kidney than do figures 41 and 42 of plate 7 of Zarnik's contribution.

AUioator. For tlie material used in the study of the renal tubule and collecting chicts of kidney of the alligator, I am indebted to Doctor Greenman, Director of the Wistar Institute, who kindly furnished me several specimens of AUigator mississippiensis, measuring approximately two feet in length. All previous descriptions of the kidney of the alligator note its simple structure. It is found accurately described by Zarnik in his comprehensive contribution in which he discusses and in many points corroborates the earlier observations of Solger" and Szakall." In the kidney of the alligator the ureter divides near the caudal end of the kidney into a dorsal and ventral branch which course respecti\ely along the dorsal and ventral surface of the kidney. Into these two JM-anches of the ureter empty numerous lateral, primarj- collecting ducts, which after division extend on each side of the ureter to near the lateral and median liorders. In figure S. is represented the cephalic end oi the \entral ureter with the primary- collecting ducts termination

• Solger, H. Zur Kenntnis tier Krokodilniere und der Xierenfnrbstoffc niodcrcr Wirhpltiere. Zcitschr. f. wiss. Zool., vol. 41, ISSo.

• SzaknII, T. t'bcr don H;iu des rrogctiitalsystcms des Krokodils. Innug. — Diss, (iiescn, 1S99. .Seen l)y me only in review (I'age. 311. HcnvtiR's Handlmch der Vcrgleich. u. Kxperiment. Entwicklungsldire, Dritter band. Krster Theil, 1006.



therein, as obtained from a preparation made by the eelluloid injection corrosion nictliod. In the lower left of the fiKure is iiulitated the manner of termination of the initial collection duct comiiiR from the renal tubules. In figure 9 is given a drawing of a very completely injected corrosion preparation of one of the side branches of the ureter with the connected initial collecting ducts. These terminate in the respective collecting ducts in two fairly regular rows as is evident from the figure. A comparison of figures S and of this article with figures 3.S and 34, plate ti, of Zarnik's c()ntril)ution will show a close correspondence of results obtained by methods of teasing and corrosion injection

Fig. 8 Cephalic end of ventral l)ranch of the ureter, with lateral branches, collecting ducts, alligator nii.ssissi|)piensis, corrosion preparation.

Fig. 9 Lateral collecting duct, receiving the distal ends of the renal tubules terminating therein, connecting tubules, alligator mis.sissippiensis, corrosion preparation.

as concerns the duct system of the alligator kidney. The substance of the alligator's kidney is arranged in a dorsal and a ventral half, corresponding to the dorsal and ventral branches of the ureter, ."separated l)y a small amount of connective tissue and branches of the renal artery and accompanying veins. The dorsal and ventral portion of the renal substance are minor pictures, .xo far as tubular arrangements and ducts are concerned. This is clearly shown in text-figure 34 and figure .oi, plate 10 of Zarnik's article and this is confirmed in my own serial sections. This observer successfully teased the renal tubules of the alii



gator !is may hv seen from his figure 'Mki. plate (i. In figures A and J{, of figure 10, are reproduced two renal tubules of alligator mississipi)icnsi.s reconstructed after the Bon* method. They are taken from the thicker central and slightly thinner lateral portions of a series of cross-sections. The renal corpu.scles of the




Kig. 10 A and B, reconstructions of two renal tubules of alligator mississippiensis. X 80. re, renal corpuscle; n., neck; prox.con., |)roxijnal convoluted segment; int.nrg., intermediate segment; dis.con., distal convoluted segment; roii.l., connecting tubule; col.d., collecting duct.

several tubules form a fairly regular layer in each half of the kidney .substance, a little distance away from the connective tissue septum. The renal corpuscle forming the beginning of the tubule .\, figure 10, pre.sents a flattened oval form with a cross diameter of A^^l. The renal corpu.scle is followed by a narrow


nock, (lirootcd toward the periphery, which in the aUigator is not lined hy ciliated epithelium. The neck is followed hy a lonj;, slender looj), which extends from the renal corpuscle to the pheriphcry of the kidney, looping back again to about the level of the renal corpuscle. This loop, which is very simple in arrangement, forms the proximal convoluted portion, prox.con., and is lined l)v an epithelium having a striated inner border. Its length in this tubule is 2.3 mm. and it presents an average diameter of a little o\'er 40^. This segment is followed by a short intermediate segment, lined by ciliated epitheUum, extending from about the level of the renal corpuscle to the loop coil shown at the lower end of the figure. In the model, this portion is not clearly demarked; it presents a length of about 0.:^ mm. Tliis portion is followed by the distal convoluted portion, forming the coil complex with the .short ascending loop, forming the lower end of the figure, and leading to the long connecting tut)ule and initial collecting duct ascending, quite free, at the left the of figure. The distal convoluted portion presents a length of 0.8 mm. and the connecting tubule and initial collecting duct, together, a length of 1.1 mm., making the entire length of this tul)ule 4.5 mm. Tubule B, of figure 10, is very similar in form and arrangement of parts to tubule A, differing from it only in being slightly shorter.

Snake. For the investigation of the snake's kidney, the common stripped snake, Eutaenia sirtalis Baird, was chosen. Zarnik reports on the renal tubules of a number of ophidian forms, basing his study on teased preparations. The left kidney was studied; the duct system l)y the corrosion injection method, the renal tubules by means of wax reconstruction. In the stripped snake the left kidney is an elongated, slightly lobulated organ, the ureter coursing along the ventral and ventro-menial border to near its cephalic end. In figure 1 1, is .shown a corrosion of the ureter with the attached prinruiry collecting ducts. These, as may l)e seen from th(» figure, cMitcr the ureter at a very acute angle and after a short distance turn at nearly a right angle, to course over the dorso-lateral surface of the kidney; receiving the initial collecting ducts. In figure 12, is presented one of the



main side l)raii(h('s of tho foIIoctiiiK duct systprn, takon from a very fully injected i)rej)arati()n, in which the injection m.-iss extended into the initial collecting ducts. Their manner of teniiinatiuK in the superficially placed collectinf^ ducts is clearly shown in the figure. In figure A, and li, of figure 13, are presented two views of a niodle of a renal tubule of the common stripped snake. In B, the ruction as viewed from the side as shown in A, is presented as when viewed from the convex side. The renal tulniies of the snake, at least in the form studied,

c^k ■..'..-"'j^^^^.

Fig. II Ureter witli coilcctinR darts tcnninating then in from the left kidney of the coninion stripped siinko. Kutucnia sirf :ilis; rorrosion prei)ariif ion.

Fig. 12 Colleeling duet of the kidney of Eutaenia sirtalis, the eommon stripped snake, with the distal ends of renal tubules emptying therein, corrosion preparation.

present an arched form as ^(^n\ in figure 13, A, and a quite compact arrangement, .somewhat difficult to portray in a figure. In this renal tubule, the renal corpuscle is relatively small and of nearly spherical form, having a diameter of \'?nx. This is attached to a long .slender neck, lined by ciliated epithelium, which is followed hy the proximal convoluted portion, arranged in the form of a main loop extending from the renal corpu.scle to the region of the sm'face collecting ducts anil returning again to a little l)elow the level of the renal corpuscle. This main loop, which presents a number of secondary loops, has a length if l.() mm. and is lined l>y an epithelium having a striated inner border. The proximal cniiMiluted segment is I'ollowed l>y a





short iiilcriiiodiato port ion lined by ciliatod ('i)itlicliiiin and this ii) turn l)y the distal convolutoil portion. This later portion forms the coil complex well seen at the bottom in H.of fisure 13, forming in reality a long loop, directed away from the collecting

duct, toward wliicli it returns and with which it connects through a long slender connecting tubule and initial collecting duct. The



\ di&coa


Fig. 1.3 .\ and B. Two views of the s:iino renal tubule reoonstrueted from the kidney of the snake, Eutaenia sirfalis. X 80. Figure A, gives a sidcvicw, figure B, the same model as viewed with the convex side toward the observer, re., renal corpuscle; n., neck; prox.con., proximal convoluted segment; int.seg., intermediate segment; dis.con., distal convoluted segment; con.t., connecting tulmle; col.d., colleotinR duct.

• distal convoluted portion has a length of 1.") mm. and the connecting tubule a length of 0.75 mm., giving the entire tuljule a length of 4 mm. The figures here given showing a reconstruction of the ophidian renal tul)ule bears a close reseml)lance to figures 24 and 25, of plate 5, of Zarnik's article.

Lizard. The limited material at my dispo.sal did not enable me to attempt a study of the duct .system of the lizard's kidney


liy moans of corrosion injection methods. Several renal tubules were reconstructed from a series of cross sections of a kidney of a small lizard uliicli unfortunately was not specifically identified. The single figure here given, figure 14, is added more for confirmation of the results of a nuicli more complete stufly. by means of teasing, of the kidney of the lizanl by Zarnik. .\s may be ob.served from a study of figure 14, the renal corpuscle is of nearly spherical form having a diameter of .55^. The neck is short and is lined by ciliated epithelium. The proximal convoluted portion, Zarnik's 'Hauijtstvick,' forms a long X-shaped. double loop, with numerous secondary loops, ha\nng a length of 1 mm. This tubular .segment is lined by relatively clear epithelium




Fig. 14 Reconstruction of renal tulmle of the liz.ird. X 80. re, renal corpuscle; proi.con., proxiniitl convoluted segment; inl.seg., intermediate segment; dis.con., distal convoluted segment; con.l., connecting tubule.

having an inner striated border. Its relation to the renal corpuscle is not unlike that found in the mammalian renal tubule. The intermediate segment, 'Ubergangsstiick' and 'Schleifenstuck' of Zarnik's is Uned by ciliated epithelium and forms a short .segment, in close relation with the renal corpuscle. This is followed by the distal convoluted portion, consisting of a U-shaped loop with lumierous secondary loops, having a length of 0.6 mm. This segment joins the collecting duct by means of a short and relatively thick connecting tubule. The entire length of the tubule figured is a little less than 2 mm. The difTerent parts of the tubule are compactly arranged, forming a coil complex more closely grouped than in the renal tubules of the other reptilian forms studied i>y me. Figure 14, of this article should be compared with figure 12a, plate 3, of Zarnik's contribution, both figures are very similar in all essentials.



For tho study of the mctjiiicphric tul)ul(' of l)ir(ls I mado uso of the kiilncys of aiiult roosters, (iallus doinestiru. As stated in the introduction to this article an attempt was made quite a number of years ago to reconstruct by means of the wax plate method, the renal tubuh^s of birds, using for this work, series of cross sections' of the kichiey of the dove. After a miml)er of attempts and the consumption of much time this method was abandoned since success did not .seem attainable. It is only relatively recently that time and opportunity presented itself to again take up the investigation of the avian renal tubule, this time by methods of maceration and teasing. This latter attempt has proven more successful. The material was prepared for teasing liy injecting untler about 120 pounds of pressure, the fresh kidney, through the renal artery with a 7.'> per cent solution of hydrochloric acid. After this injection, the kidneys was removed and placed in a solution of hydrochloric acid of like strength for a period of several hours, then thoroughly washed in water, stained in an alum hematoxylin solution and teased under the steroscopic binocular.

While the isolation of coni|)lete renal tubules from the bird's kidney presents less difficulty than the isolation of complete renal tubules from the mammalian kidney, still it reciuires hours and hours of teasing and the passing over of many failures before one obtains a series of satisfactory preparations. The literature contains few references and fewer figures relating to the renal tubules of birds. One very brief reference is found in Chapter 21, tjtricker's Manual of Histology (English translation. Wood and Co., New York, 72) written by Carl Ludwig. In figure 171 D of this text is presented a diagrammatic outline drawing represent iiig a renal tubule from the kiilnev of a dove. A footnote states that this figure was designed by Iliifner. The text contains the following reference to this figure:

It will l)c iiotifcfl at oner iiow strong the rcscriihianco is to the >iriniferoiis tuluilc of Mammals. The several siilxlivisiDiis which we make in the latter are al.xo to be fowiiil here, ami tliey occur in the .-^aiiie order. A.s far a.s our present knowleiige goes, tiie uriniferous tubules of all birds are made on the same plan a.s those of the dove.


The kidney of the l)inl is an olonRatod, tlistinrtly, lohuiated organ, with the ureter on the ventral and me.scjventral side. I'riiininent branches pass from the ureter to the several lobules. The ventral or mesoventral portion of each lobule presents a short renal pyramid. The central core of a lobule presents an apl)earaiic(' uliich is not unlike that i)resented by the uianinialian kidney, namely, a shallow medulla and a cortex. The cortex, particularly, extends on all sides beyond the limits of this central core, presenting an appearance not unlike that pre.sented by the reptilian kidney, especially the chelonian kidney. It is perhaps fortunate for this work that it was necessary to resort to methods of teasing l)v reason of failure with the method of reconstruction in that the teasing method admits of taking tissue portions from all parts of a lobule, while with reconstruction methods one is limited, by reason of consumption of time, to a reconstruction of one or a small number of tubules. In this study it was early evident that the renal tubules forming the periphery of the lobules presented a form which is very much simpler than is presented by those forming the central core of the loi)ules. The former jiresenting a form and arrangement of parts which is not unlike that of the reptilian renal tubules, while the latter present a form and arrangement of parts which re.semlile the mammalian renal tulniles, with transition from one type of tul)iik' to the other in an intermediate zone. Vnnn the many renal tul)ules of l)irds, successfully and completely isolated, and permanently mounted, I have selected a small series showing this tran.sition from the reptilian type of tubule to the mammalian type. The nece.ssary drawings were made from permanently mounted, completely isolated renal tubules, at a magnification of 200, by aid of camera lucida. In order to bring the figures to a size making reproiluctions as text figures possible, and yet maintain the same magnification for all the figures of renal tubules of birds, the figures here given represent just half the magnification of the figures detailing the amphil)ian and reptilian renal tulniles. The measurements gi\-en for the renal tubules of birds are taken from the drawings, and are perhaps not as accurate as those taken from wax reconstructions.


Tlu'v nivo, howovor, very nearly the aetual lengths of the res|)t'«'tive renal tubules. In eonsidering the liianieter of the ilitTerent parts of the renal tul)ules of birds, as here figured, it should be understood that in order to mount permanently in plycfrine the eomplcteiy teased renal tul)ules it is neeessary to allow them to dry partially on the slide after the removal of the fluid in which they are teased. This fixes them to the slide suflieiently so that they are not broken and distorted when the eover glass and mounting fluid are added. However, in this process the respective tubule loses in diameter, though scarcely in length. The camera fueida drawing here reproduced, therefore, do not re|)resent the full diameter of the tubules figured. It is further to l)e understood that in order to l)e certain that the tubule under teasing is in reality a single tubule, it is necessaryto tease and separate all coils or loops and to be able to tease completely from renal corpuscle to collecting ducts the entire tubule. This makes it difficult and quite impossible to bring the several parts into normal positions, though there is a marked tendency for a given tubule to do this if allowed to remain undisturbed for a few moments in the (luitl in which it is being teased. However, the manipulations necessar}'^ to mounting permanently such a prejiaration usually leads to slight distortion antl movement of parts so that the respective tul)ule comes to rest on the sUde slightly spread out, admitting usually of readier interpretation and clearer representation in drawing, but tloes not i)resent the normal relations of parts. In this latter respect tubules isolated by teasing sufTer in comparison with such as are reproduced by wax reconstruction.

It is my purpose to discuss the several types of renal tubules of the bird observed, by beginning with a consideration of the simplest type and proceeding to the larger and more complex types. In figure I'l, A, H and (', are reproduced three renal tubules taken frcjni the more peripheral i)ortions of one of the lobules of the kidney of an adult rooster. lOach is complete from renal corpuscle to collecting duct. A comparison of these figures with the figures here given of the reptilian renal tubule, especialh' .\ and B, of figure 5, will show tliat there exists a



marked morpholot^ic similarity in form ami arrangompnt of parts of tlio reptilian and certain avian renal tubules. In each of the three tubules reproduced in figure lo, the renal corpuscle is relatively small and of nearly spherical form. Kaeh is joined to the respeelive tubule by a narrow iieek. This is followed by a tubular segment which constitutes the proximal convoluted portion, arranged in each of the three tubules figured in the form of a letter N, each part of the \ presenting numerous secondary


Fi(5. 1.5 .-V, H ami C. Three renal tubules of the repliliati type, teased front the peripheral protion of a renal lobule, from the kidney of an adult rooster, Gallus doinistica. X 40. re, renal corpuscle; ii., neck; prox.con., proximal convoluted segment; inl.seg., intermediate segment; dis.con., distal convoluted segment ; con.t., connecting or junctional tubule;, collecting duct.

loops, all partly extended in the preparations as mounted. The arrangements of the tul)ular segment is perhaps most clearly shown in tubule A, in which its general course is clearly followed. A study of tubule B, shows the same arrangement and in tubule C, the greater part of the coil complex, tilted accidently to the left, constitutes this tubular segment. In this part of tulnile C, the normal relation of the respective loops is better maintained than ill tlie oilier two tubules. If the coil complex in C, tilted diagonally to Ihe left, be brought in line with the rest of the

32S G. CAilL muEii

finiirc, fairly normal relations are obtained for all of this tubule. The actual length of the proximal ronv'oluted tubular segment in the three tubules shown in figure "», is — A, 3.8 mm.; H ^3.5 nun.; C— 3.6 mm. I realize fully that there exist several sources of error in the method used in oljtaining the actual length of the several tubular .segments considererd -measuring the tul)ule as drawn enlarged in the figure and dividing by magnification used — however, the length given may be regarded as approximately correct. Macerated and teased preparations, even where mounted in glycerine, present very little evidence a.s concerns the cellular structure of the difTerent segments, a sUght difTerence in the dejiths of staining may lie observed, but the nuclei are not difTerent iatetl. In this respect the method of teasing is less favorable than wax plate reconstruction, since in the latter a very complete series of sections is a necessitj\ However, the proximal convoluted portion of the birds renal tubule is readily recognized and presents a structure which is similar to that shown for this tubular segment of the mammahaii riMial tubule. In the three renal tuliules presented in figure 15, the proximal convoluteil segments of each tubule terminate at the upper limits of the figure, at the region indicated by a small This tul>ular segment is followed l)y a relatively thin tubular portion, comparable to the thin arm of the medullary loop of the manunalian renal tubule, though its epithelium is not of a distinctly squamous type. This segment, which may be known as the intermediate .segment, in the majority of the tubules completely or only partially teased presents a number of small loops, diflicult to unravel. This intermediate .segment is followed by the di.stal convoluted portion, consisting of a numl)er of compactly arranged loops and coils, also forming a jiortion diflicult to In the three tubules figured this portion is presented as unroled and forms the most dependent part of each of the three figures. By means of the connecting or junctional tutnilar .segments, the distal convoluted portions join the respective collecting ducts; the di.stal end of the collecting ducts in the peripheral jwrtions of the lobule, coursing nearly parallel to the dorsal surface of the lobule and receiving connecting tubules


at short intervals. 'I'his may in part he seen in A, f)f figure 15, in which a short segment of a terminal coUectiiiK duct, receiving four connecting or junctional tul)ulps is shown. The entire length of tubule A, figure 15, as mea.sured from the drawing is ().5 mm. In figure Ki is reproduced a drawing of a tubule completely isolated from a small portion of tissue taken from the periphery of one of the lobules. The parts indicated for the tubules described under figure 15 are present in this tubule, though it was found as a much more compact mass before teasing, and is mounted spread out in nearly a single plane. The entire length of this tubule is approximately 8 mm., of which 4 mm. falls

1 ig 16 tiiluilc from bird, Gallus domestica, teased from the peripheral portion of a renal lobule. X 40. The proximal convoluted segment, prox.con., and the distal convoluted sRemcnt, ilis.con., are hero clearly represented.

to the proximal convoluted portion. In the figure of this tubule, the two coils complexes constituting the proximal and distal convoluted segments, are clearly seen in the two distinct groups of loops forming the right and left part of the figure. In figure 17, is presented a drawing of a completely isolated^tul)ule with a portion of the connected duct sy.-Jtem, taken from the more peripheral portion of the central part of a lol)ule. This tubule approaches in form and arrangement of parts a manuualian renal tubule, having a short medullary loop which presents, however, a number of small secondary loops. The renal corpuscle is of sperical form and is relatively larger than in the tubules shown in figures 15 and Hi. The initial portion of the



proximal convoluted scKUK'iit oxtond.s toward tho ventral portion (if tho lohule toward the rop;ion of the ineilulla, then returns toward the peripliery of tlie cortex to form an extended loop with many secondary, smalltT loops. The length of this proximal convoluted segment, whicii forms the coil complex at the left of the figure and extends from the renal corpuscle to the region

Fig- 17 Uonal tubule of the l)ird, Gnllus domestica, completely isolated by teasing, showing transition from reptilian to mammalian type or renal tubule. X 40. re, renal corpuscle; prox.con., proximal convoluted segment; med.L, medullar}' loop; dis.con., distal convoluted segment; con.t., connecting tubule; col.d., collecting dutst.

where the tubule becomes suddenly thinner, is approximately 7.0 mm.; thus approaching the length of the proximal convoluted segment of the manwnalian renal tubule. The medullary loop of this tubule is relatively short, having a Icngth.of 1.8 mm. and ending in the region where the tubule again increases in size; about the middle of the figtire. This segment is followed by a



seooiul coil complex, forming the distal convoluted portion, practically unrolled as presented in this figure, in reality forming a cdiiipact coil complex, and leading to a relatively long connecting or junctional tubule ending in the collecting duct. The entire length of this tul)ule is approximately Vi mm. In figure

Fig. 18 Renal tubule of the bird. Gullus domestica, isolated by teasing. X 40. This tubule p(> a short nicduU.iry loop, presenting short loops in the distal arm. The (lifTcront sogincnts are so placed as to be easUy recognized. Figure legends as for tigurc 17.

IS is reproduced a tubule which was very successfully mounted with reference to exposing the several segments of the tubule, which is very similar to that .shown in figure 17, though it was broken in teasing just distal to the distal convoluted portion, so that its relation to the collecting duct cannot be shown. The


rehitivoly largo renal corpuscle of spherical form, the long proximal coiivdlutcd scumciit, arraiifiod in (he f(irm of a ionn loop with numerous secondary loops, the short medullary loop and the distal convoluted portion, are clearly seen in this figure. This tubule, to the extent contained in the figure, has a length of I'.i.ri mm., to which perhaps 1 mm. need he added to complete the length to the collecting duct. Of this length, S mm. falls to the proximal convoluted portion. Tubules very similar in form and relation of parts to those rejjroduced in figures 17 and IS maj- be found in the peripheral cortical portion of the mammalian kidney, especially in the human kidney, in which tubules are founil which do not pass the limits of the cortex.

In figure 10. is reproduced a camera lucida drawing of a renal tubule from the bird, taken from the central portion of one of the kidney lobules. This preparation was very successfully teased and is permanently mounted just as shown in the figure, in which the .several segments arc so placed that their course may be followed with ease. This renal tuljule of the bird is representive of the type of a\-ian renal tubules, which in general form, sequence and relation of parts, relation to cortex and medulla is in essentials very similar to a mammalian renal tubule. The figure .seems to me so clear that lengthy discussion is unncessary. The relatively large renal corpuscle, of spherical form, joins the proximal convoluted portion bv means of a short neck. The proximal convoluted portion is arranged in the form of a long loop, with numerous secondary loops, extending toward the periphery of the cortex and retin'ning to the vicinity of the renal corpuscle and for a distance into the medulla. The length of this tubular segment, from renal corpuscle to the region of the small cross, is approximately S.5 mm. The medullary loop (loop of Ilenle) extends for a distance into the medulla. In the teased preparation it was not jiossible to determine dehnitely the character of its epithelium, though the proximal arm appears stained slightly lighter than the distal arm as though lined by a thinner epithelium. The distal arm returns towartl the cortex and comes in close relation with the renal corpuscle of the tubule, this relation is not shown in the mounted preparation but wa.s




Fig. 19 Renal tubule of bird, Gallus domestica, isolated by teasing from the central jwrtinn of a renal lobule. X 40. This tubule is of the mammalian type of avian renal tubules, re, renal corpuscle; proi.con., proximal convoluted .segment; mcd.L, niedullnry loop; dis.con., distal convoluted segment: con. I., connecting or junctional tubule; col.d., collecting duct.


ascertained before the lul)ul(' was completely teased. The length of the medullary loop with the ascending liml> to heginiiiiifl of distal convoluted jiortion (coil complex, up!)er rip;ht of li^ure) is 3.(5 nun. The distal convoluted portion, upper right of figure, presents two more prominent loops each with several secondary loops. Before teasing, this portion was compactly arranged and has a length of '2.2 mm. It is followed l\v a long connecting or junctional tubular segment leading to the collecting duct, and adding 1.2 nmi. to the length of the tube. As may be seen from the figure a portion of the collecting duct system was isolated in connection with the tuhule figured. The branches of the collecting duct figured represent other renal tubules connecting with the same. The entire length of this tubide is appro.ximately 15. o mm., as measured from the drawing. In figure 20, is reproduced a camera lucida drawing of the longest renal tubule of the bird isolated. It comes from the central portion of a renal lobule. In trying to mount permanently this preparation, I endeavored to adjust slightly the position of the renal corpuscle, with the result that the renal tubule was broken in two [ilaces in the immediate vacinity of the renal corpuscle. The broken pieces, represented by dotted lines, were brought to place as shown in the figure. This tubule is not so favorably placed in the permanent mount as the one shown in the previous figure. However, it is thought that the different parts can be traced. The tubule shown in figure 20. presents the following measurements — proximal convoluted portion, 7..") nun.; medullary loop with distal ascending limb to beginning of distal convoluted portion — 0.5 mm.; (.listal convoluted segment — 2.2 mm.; connecting or junctional segment — 2 mm.; giving the entire tubule from renal corpuscle to collecting duct a length of IS. 2 nun., which is somewhat over half the length of the longest mammalian renal tubule isolated by me. In figure 21, is shown a camera lucida drawing taken from the most

Fig. 20 RonnI tiiliulc of the bird, Galltis domeatica. Mammnlian typo of avian rcnnl tiihiilrM; iHolatrd \>y trtising. Tul)ulo was broken in making pernianrnt inDUnl; l)rokcn parts outlined in dotted lines. X 40. Figure legends as In figure 19.






successful tciiscd jircparution of the duct s\'stc'iii of the bird's kidney made by nie. The preparation came from the central portion of a renal lobule and traces one of the main collecting ducts through its several branches to and including the end segments of renal tubules ending in this portion of the duct system. Many of the main branches of this duct arc not traced to their termination in this figure, but extend to the perii)heral and lateral portions of the renal lobule receiving the renal tubules which have been described as showing a reptilian type of renal tubule. It will !)(' observed that the renal tubules of the bird end in the several divisions of the duct system, those ending deei) in the medulla having long connecting or junctional tubules as shown in figures 19 and 20.

A study of the figure of the avian renal tubule as here jiresented may serve to show that in the birds kidney, there is found a transition, so far as concerns the form of the renal tubule, from the reptilian type of metanephric tubule to the mammalian tyjie of the metanephric tubule. A study of the vascular sujiply. with the use of the modern corrosion methods is very desiral)le. Zarnik gives a good account of the blood supply of the reptilian kidney, 1 believe as far as concerns the terminal branches of the renal artery and their relation to the renal tubule of the mammalian kitlncy, my own account gives correctly the essentials. It is desirable, therefore, that the details of the blood supply of the avian kidney be eciually carefully studied. It is hoped that opportunity will present itself to do this in this laboratory or that some other laboratory will undertake this investigation.

Fig. 21 Primary collcctiiiK diirt with l)r.anchcs. teased from tlie center i)f .a reniil lolmie of the hird, (iailu.s d<>me.stira. The tortuous tul)uh>s. upper part of figure, represent distal eonvoluted and junetional segments of renal tul>ules.

Fir. 22 Uenal tuliule of ral>l)it. i.solated l)y teasing. X 20. rr. renal corpuscle; prox.con., proximal ronvolulcd segmei\t ; nieil.l.. medullary loop, from X to X of this loop, the thin .Kcgmeiit lined liy .squamous epithelium, liis.con., distal convoluted segment. 'I'he luhule ends just distal to distal eonvoluted portion.

Tll»: A^-^T<)UlrAL nccORD, VOL.13, NO.

33S Ci. CAKI. Ill IIKK


It is not my purpose to discuss horp fully the riMial tubules of the in;iiMiii;ili:m kitliicy. This hits been done relatively recently l>y Karl Peter'" in his eomprehensive monograph, by Inouyr'" in an addendum to the Peter monofiraph, by Zarnik in his report on the renal tubules of lOehidna and by myself, my report being bused on completely isolated renal tubules of the rai>l>its kidney. However, in order to make this series of renal tubules of vertebrates com]il(>te in this one article, there is here gi\('ii ill figure '22, a new drawing of tubule H. of plate I, accompanying the contribution — "A method for isolating the renal tubules of mammalia" (Hubcr)- reproduced here at a magnification of 20 diameters, one-half of the magnification used in reproducing the renal tubules of birds, and one-fourth of the magnification used in the re{)roduction of the renal tul)ules of the other vertebrates studied. From the text of the article mentioned it may be learned that the maminaliaii renal tubule here shown, and wliidi may be taken a^^ a type tiilmle. has a total length of 2'A mm. It presents a mammalian renal tul)ule of metlium length. One, the renal corpuscle of which is situated at about the mitl level of the cortex. In this tubule, the proximal convoluted portion with the medullary segment has a length of 0.4 mm.: the thin jiortion of th(> medullary loop. ;i length of ti.T nun.; the thicker portion of the ascending liiiii) of the medullary loop to the level of the renal corpuscle a length of ti.O mm.; the distal convoluted segment a length of 0.9 mm. The tubule was isolated completely and mounted in the position as drawn" to just beyond the distal convoluted portion. The connection or junctional segment is thus mis.sing. This would add approxi " Peter. Karl. I'ntcrsurhung iilior Hnii iind Kntwiokclung dor Xiere. f'Karl Peter, Die N'ierenknniili'heii de.s .MeiiKc-hen iind einiRor SiiuRetiere.', Michio Inouye. die Nierenlviiiiiilrheii den iiindrs uiid de.s 'rniiiiinlers. (iiistiiv Fischer, Jcnii. inO!>.

" In the fijfure a.s here (jiveii. tlie looj) in the lower end of the medullary loop, as drawn in tiilmle H, plate 1. .Vnal. Kcc. vol. .5. is eliminated. This was done by first drawing the lower end of the medullary loop as found in the preparation, then placing a string over the IraeinK then extending the string in the form of u simple loop, as given in figure 22.


niatcly 1 nun. to tlw total length of this tuhulf. Ilic fi)itlicliutn rharartcrizinji tho dilTerent scKnients of the niannnalian renal tuhulc, I have discussed in the former publieation as also the position of the difTerent segments in the kidney substance and their rchitioii to each other.

.\ earcful investigation, with tlie u.-^e of appropriate maceration aiKJ tea.sing methods of the renal tubules of a number of nianunals is still desirable. IVter has ealled attention to the ditTerenee in structure of the two arms of the medullary loop and of the loop itself in difTerent mammals and in the same kidney for tubules of different lengths. This I have confirmed, but not in prejiarations showinji renal tubules, completely isolated, so that the comiJarative and total length of the difTerent segments of a given renal tubule could be given ; which is much to be desired. Other parts of the renal tubules of mammals vary in detail of structure. In the guinea jHg, for instance, the distal convoluted portion has a iliameter which is approximately the same as that of the proximal convoluted portion, though it is nuich shorter, differing in this respect from that of the rabbit. The relation existing between structure and function makes it desirai)le to be fully informed as to the details of structure of the difTerent segments of the renal tubules.

It is hoped that this series of figures, brought together in limited space, presenting type forms of the renal tubules of the different classes of vertebrates, may not be without value to both students of morphology and functions.



OF 'iiii; liAiiKir

G. CAKL HUUKK AM) AHNOM) II. EGGKKTH Department of Anatomy, University of Michigan


There exist in tlic literature numerous contributions dealing with the form ami distribution of the fjustatory papillae in vertebrates and with the structure and innervation of the gustatory buds. This literature is foreign to the problem here under discussion and will not receive consideration at the present time. Relatively few ()l)servers have dealt with the development of the gustatory i)a|)illac, and fewer still have considered the developm<>tit of the papilla foliata of the rabbit or of other forms in which this type of gustatory papilla is prominent, v. Wyss' who was the first to call attention to the details of .structure of the papilla foliata of the rabbit and the great number of gustatory buds there found, makes only incidental mention of its development, stating that it is (piite well formed in the new born rabbit and contains taste buds which have, however, a more rouniled form and are smaller than in the grown animal. Lustig- states that with the aid of a lens he was able to recognize the papilla foliata in rabbit embryos having a length of (i mm. to 7..") mm. and that they are more dc'veloped in 10 mm. ral)i>it embryos, however, that it was not possible for him so define the taste buds. A fairly comi)rehensive account of the development of the pai)illa foliata of the rabbit is given by Hermann.^ This ob.server found that

' V. Wyss. H.. Dio Ix-cliorformiKcn Oruaiio der Ziingo. Arcli. f. Mik. .Viiat.. vol. 6, 1870.

' Liisti);. \.. HoilriiKi- ziir Konntni.s.sdcr Kril« icki'luriK ilcr Ciosi'hmack.sknii.s|)rn. SilziinKsl)or. .Vkad. <1. Wi.s.sciisrh. \'ii'iia. Math. Nal. C'la.ssc, vol. S<). 1S.S4.

' II:inn;iiin. 1"., Mcitr.iK ziir lCiilwii'kluii(;--*Ki'soliicht<' (U's Gesrhiimrksorgiins boim Kiiiiiiiclirn. .Vrch. f, .Mik. .Vnat.. vol. 2\. 1885.


'M'2 <i. ( AKI. Ill Hi:U AM) AU.N'OLl) II. KCCiKHTII

the jKipilla foliata was first rocoKnizahlc in ral)l)it oiiil)ryos having a IfiiKlli of ■">-t mill- la^!:t^ alioiit twcnty-llircr days), as an oval, slinlilly clcvatcil area, witli lon^;; axis liorizoiital and witli faint grooves havinji a vertical <lin'cti(iii. In longitudinal sections, papillae formed l)y simi)le infolding; of the e])it helium were oiiserved. In embryos having a length of 70 mm. (aiiout twentyfour days) the foliate area is no longer (|uite horizontal in position, its posterior end being nearer the median line. The ei)ith(>lial pajiillae are longer and broader, the primary infoldinns presenting at about the middle of their dejjth, on earh side, a slifjht projection, most fully developed in the midille of the area. In embryos liavinn a leiifith of !!.") mm., thus only a few days before i)irth the sli^jht i)rojeetinfj;s of the jH-imary folds havedeveloped into secondary folds and the anlagen of the serous lingual glands are evident. In the new born rabbit the secondary folds of the papillae are longer, the whole area, however, presenting essentially the same ai)pearance as that fouixl in a !••"> mm. rabbit. In its main feature, we cnnlirni tlie oiiserxatioiis of I lcriii;imi. As a result of a comjiarative study Hriicher' concludes that the prototype of the gustatory papillae is the jKipilla fungiforinis, out of which developed both the p. vallata and the p. foliata, since transition forms are observed, (inielin^ on the other hand, also as a result of a comparative study, concludes that the p. vallata is not develojied from the p. fungiforinis, nor the p. foliata from the p. vallata. but tii;it iiotli p. vallata and p. foliata are independent in their origin, each having its special .>;eat of development, and that transition forms are not ob.served. Hoffmann,' Tuckermann,' and more recently, (iraberg." have pub ' Hr\irlicr. ('l)cr dip Vprlhcilnnn iiikI .Vnordiiiinp dor Goachmarks-papillpn auf dcr ZiiMKi' drr i^tilKOtliicrp, .sppc. ilrr lliiflliicrc. Drutscli. Zpit.xpli. f. Tliiormcd. II, vprg. Path., vol. 10. ISW.

  • (iinp|in, Zur .MorpholoKJc dpr Papilla vallata und foliata. Anii. f. .\lik.

,\nat., vol. 4(1, I.S<I2.

• llofTniann, t'lipr dip VorUrpitunK dpr {ipsrhniacksorgano liriiii Mriischen. Arch. f. Path, vol., 02, 1.875.

' Turkcrinann, F., On the dcvclopnipiil of thr taste-organs of man, .lour. Anat. Phy.s., vcA.Zi. 1889.

' (;r&l)crK. .].. Uoitriiup znr CJcneso dcr Opschmarksorgane des Mcnschen, Morphol. Arl.rit., vol. H, 1808.


lishc'd concerning the development of the valhitii papillae «)f man, hut since their ohservat ictus h(>ar only incidentally on those here recorded their jtubiicatioiis will not receive review here. M. Ilcidcnliain,"'"' has in two relatively recent communications considered very fully certain aspects of the papilla foliafa of the rahhit. ("crtain points discussed is these articles may here be touched on. He finds the foliala area of the grown rabbit to be of jiear or egg-shape, willi tlic narrower end directed caudaliy, tlie rounder end forwards. The folds, of which there are about l(i, run from the ventral to the dorsal Itorder, diverging somewhat dorsally, with the longer ones having a slight curvature. Bifurcation of the folds was observed. This division of folds he recognizes as indication of a developmental process. This thought he exi) as follows: "Wir halten es chiher fiir inoglich, ilass die Leisten Histosysteme sind deren Anlage, im Laufe der Kntwicklung (lurch spaltung vermehrt werden, so dass gelegenthch bei unvollkonuneuer Sonderung (iabeluiigen entstehen."

Though familiar with Hermann's observations on the develoi>ment of the papilla foliata in the rabbit, before our investigation was undertaken, it occurred to us that a careful study of the morphogenesis t)f the foliata area was still desirable, especially if free use was matle of the Born wax plate reconstruction method, and a eomjjlete series of timed stages, ranging from the time of its first appearance as a definite area to the time of birth could be commandeil. It seemed to us desirable to determine whether a tlefinite pattern of epithelial folds was here prescMit. whether the area had its anlage as whole or si)read by jteripheral growth with increase of the numbers of folds, as well as to determine more definitely the morphogenesis of the folds and the relations of the serous lingual glands to them. To do this it was found necessary to reconstruct by means ot llie Born plate method the

• llcidi'nimiii, M.. riilcrmu-lniiiKcti (iIxT die Ti'ilk(ir|)crnutur di-r Cir.schniackskiio.M|H-n in licr I'lipillii foliiitu dcs Kaiiiiiclictiti. .\n:it. .Viiz., vol. 45, 1914.

'" llcidcniiain. M. Clicr die .Sinncsfi'ldcr uiid die Ooschm.irksknosiK'n dcr Piipillu foliata drs KaiiincliciiH. HcitraKc zur 'IVilki)r|MTtlicorii', III. .Vrch. f. Mik. Aiiat., vol. S.">. 1914.


entire foliate area in a iminljcr of stages of development. The models here li^ured were prepared by Mr. l']KK('rth and eonstitutcd a part of the work recniired toward a .Master of Arts degree with major in Anatomy.

.Ml of the observations here recorded are based on timed material. Tor the IS. lit and "JO day stages, frontal and sagittal series of heads of embryos rai)l)its were u.sed, fixed in (,'arnoy's fluid and stained on the slide in iron hematoxylin and Congo red. For the stages covering 21 day to 80 day embryos, the lower jaw with tongue and larynx (for younger stages), the tongue and larynx (older stages) were removed and fixed in Yoshii's fluid. The matf rial derived from the 21 day and 22 day embrj'os was sectional so as to obtain frontal sections of the tongue; approximately longitudinal .sections of the foliata areas. For all other stages the foliate area with a block of underlying muscle was excised, oriented in the paraffine block so that longitudinal sections, at right angle to the main folds, might be cut. Complete .series of sections. 4ai or Ofi in thickness of all stages, cut by the water-on-the-knife method with .sliding microtome, were at our tlisposal. These were stained, some of the series in iron hematoxylin, others in Minot's hematoxylin and countcrstained in Congo red. For complete series of relatively thin sections, the water-on-the-knife method with .sliding microtome is especially recommended in preference to series cut on a rotary microtome, esjx'cially if the series are to be used, for reconstruction purposes. The models, though very time consuming in their prejiaration, presented no special difficulty in execution. They were made at a magnification of :}()() diameters with thickness of the |)lates adjusted to meet the needs. The figures of the models, greatly reduced in reproduction, present the configuration of the epithcliuni of the foliate jsrea as viewed from the under or mucosal side. .\ series of drawings of the epithelial border, taken from approximately the center of the jiapilla foliata of the several stages studied, from longitudinally cut series, is presentetl is figure 1. These figures may with jirofit be compared with the figure of the model for the ecjuivalent stage, since for the stages modelled, the drawing for the respective



,*v v;- '^ ■


Kin. 1 DrawiiiKs i>f llic cpitholiitl hordcr with the uiidcrlyinK mucosa of the foliiito arpsi of a ral)l)it cinljryos of dilTcrciit slant's "f di'volopiiu'iit. .1 ami H, X 100; r to J. X 50. .1. from rabbit cml.ryo of 18 days; /<. JO <lays; C, 21 days; D, 23 <lays;'£, 24 days; F. 20 days; G, 27 days; //, 28 days; /. 29 days; J. 30 days.

Uli C. < AIU. IIIHKK AVI) AKNol.l) II. K(i(iKKTH

stanc \\;is lak(>ii from one of the sections of the series used in tlic prt'paralion of llu- inotlcl covcnnn that stiige.

In presenting our own observations we will hefiin witli tlic consideration of the sta^e in wiiieh the foliata area is first recogni/ed and proceed in serial order of stages to the time of hirth. It has apiu'ared to ns that the areas in which the foliate papillae develop, may he recognized Iiefore the appearance of the primary epithelial folds characterizing these areas. In a series of frontal sections, including the posterior jiortion of the tongue of a ral)l)it embryo of IS days, a slightly elevated, oval area may he fairly clearly delimited on the lateral and dorso-Iateral surface of the posterior part of the tongue, in front of the de\cloping larynx and in front of the arcus palatoglossus of each side. Ill .1. of figure 1. tln'ic is i)resented a drawing of the epithelium of ai)proxiniately the middle section ])assing through the right pre-foliate area. A study of this series permits the conclusion that the epithelium of the anterior part of this area is cut in fairly accurate cross-sections, wliile tliat of the posterior part is cut slightly oblitiuely. This area, as the ligure may .serve to show, is co\ered by s(|uamous epithelium, which in tlie jirefoliate area is slightly thicker than in the region anterior anil j)osterior to this area. In a .series of frmUal sections of the tongue of a rabbit embryo of 20 days, the foliate area is clearly recognized, as may be observed from a study of section li, of figure 1. The foliate areas present at this stage, on the under surface of the epithelium, a series of ei)ithelial folds which clearly characterize these areas. The drawing was made from approximately the middle of the series of sections passing through the right foliate area of this stage. The section drawn presents a fairly accurate cros.s-section of the eiiithelium, but passes slightly oblifpiely through the area, not ([uite jiarallel to its long axis. That the 10 or 11 fairly well formed ei)ithelial paj^illae, as seen in the sections, are in reality sections of epithelial folds is readily determined by graphic reconstruction and is corroborated i)y a study of a series of sagittal sections of the head of a rabbit embryo of 10 days in which the foliate areas are cut tangeiitially with reference to the epithelial surfaces, .so that the epithelial folds arc


cut lonKitiidirially for a short distance. Tlic primary folds of the foiiala areas, therefore, make their a|)|)earaiiee al)out the Utth day of the rul)l)its' development and arc dearly indicated on the under surface of tlie ei)ithelium l)y the "iOth day. although thr-re is no definite indication of such ei)ithelial folds on surface insj)ection, excii \\h<ii viewed with the aid of a stereo.scopic binocular. The impression is gaineil that these folds spring into existence' relatively suddenly as if hy a foldinti of the under lay< r, the fit rminal la.Ncr. of the epilheliinii. .\< the liniire may serxc to -how. the folding does not inxolxc the entire thickness of the epithelium and there is oi)served no unusual number of mitotic ligures in this foliate area. The nicchani-;m we are unable to determine positively. In a study of the series of stages .shown is figure 1, it should be recognized that .1 and Ji of this fifrure were drawn at a magnification of (i(H) diameters, while all the othei' sections were drawn at a magnification of '.MM diameters, the whole Hgure being reduced to the same extent in reiiroduction. For the next stage, that of a 21 day rabbit embryo, we are abl(> to jiresent a figure of a model showing the under surface of the ei)i'helium one of the foliate areas as w«>ll as a drawing of oiw of the series of .sections fnitii which this model was made. In figiUT '2 is reproduced a photograph of the under or mucosal surface of a nKulel representing a cast of the ei)ithelium of the right foliate area and immediate surroundings, made at a magnification of 'M)0 diameters. The right foliate area was cut from the tongue of a rabbit embryo of 21 days, oriented in the parafline block so as to enable obtaining cross-sections of the major folds, shallow grooves on the surface, clearly evident under the stereoscopic binocular indicating the direction of fokls. l-'rom the finished reconstruction it is learned that the line of the .sections is not parallel to the long axis of this foliate area. This is to be given cognizance in interpreting C, of figure 1. a drawing of the epithelium of one of the sections of this area, which does not present a section through its longest axis, though taken from al)out the middle of the series and as shown by the model, passes almost exactly at right angles to the long axis of the major folds. In figure 2. the foliate area jn'opor




is readily (icliniited. The dorsal border of the area is directed upwanls and the anterior border to the left. To the left and left upper portion of this figure the presence of epithelial pits, which received mucosal papillae may be observed, ventrally and posterior to the foliate area, the mucosal surface of the epithelium is fairly smooth. The foliate area as reconstructed presents an elongated egg-shape and in general contour, except for width, and in configuration is not unlike the foliate area of an adult rabbit as tigured by Heidenhain'" in his text figures d, e and f. In this area, as the figure of the model may serve to show, there ;ire |)resent at least 15 primary epitheUal folds. It may further be ol)serve<l that there is at this stage division or bifurcation of certain of the primary epithelilal folds, both in a dorsal and in a ventral direction as also figured and discussed l)y Heidenhain, for the foliate area of the adult rabbit. The fact that about as many jirimary epithelial folds are present in the foliate area at this early stage in development a.s in the adult rabbit would seem to indicate that these primary epithelial folds may be regarded as Tlistosystems" in Ileidenhain's terminology; further, the fact that about as much splitting or bifurcation of the primary epithelial folds is to be obser\-ed at this stage as in the adult rabbit may serve to indicate that the number of the primary ejiithelial folds is not materially increasetl by longitudinal si)litting of folds of earlier stages of development. As may be ob.served from the figure, the primary epithelial folds converge toward about the middle of the ventral border, and th(> major folds present a slight curvature directed toward the front of the tongue, ver}- much as describetl by Heidenhain

Fin. 2 I'liotonrapli of Ihi" mii<'<i!<:il surfarp of a wax plate reconst motion of the epitlieliiiin of the rinlil foliate area of a raMiit embryo of 21 days. .\Iagiiilicalion approxiiiialely M\ iliameters.

KIr. .'! Pilot (>({rapli of llie rnueo.sal surfaee of a wax plate reeonsi ruction of the ppitlieliuin of tlie riglit foliate area of a rabbit embryo of 23 days. Magnification approximately -V) diamteres.

FiR. 4 I'hotoRraph of the intieosjkl surfaee of a wax plate reconstruction of the epitheliiuu of the left foliate area of a rabbit embryo of 24 days. MaK>'ifieation approximately M diameters. Note the aiila(;en of the serous liuf^ual glands in the center of the foliate area.

350 G. CAHL ULUEK AM) AKNOI.l) 11. K(!(;EK'rH

for tho adult rahhit. Sect ion C, of fisiurc 1. i)r('s(Mi(s a drawiriK of tlic opitlu'lium of ai)|)roxiiimtcly the middle of the serii'.s from which the model of li^ure 2 was made. The .section is so placed that the epithelium hordering the anterior part of the foliate area i.>^ directed toward the left, a.s with all the sections of the series shown in Hgure 1. Perhaps I'.i definite primary epithelial folds as seen in cross-section, may l)e ol)served in this tlKurc. It .should he observed that the shallow grooves, descrihed as seen on surface view of this foliate area, fall over the center of the epithelial folds, and are the grooves into which empty the serous lingual glands in later stages of development. The primary mucosal folds interdigitate with the primary epithelial folds. The series of sections of the foliate areas of a ral)i)it eml)ryo of 22 days are in all respects so similar to of the 21 day stage just described, that a special consideration of the 22 day stage may here be disi^ensed with. The 23 day stage was reconstructed from an excellent series of sections which pass almost at right angle to the major priinaiy I'liithclin! folds and give the epithelium of this area in \('ry good cross-section. The under surface of the model of the epithelium is jiresenteil in figure 3, the accompanying section drawing in /-*, of figure 1. Figure 3, thus .shows the mucosal surface of the right foliate area of a rabbit embryo of 23 days, with the surroimding epithelium. There are found It) well formed primary epithelial folds, three of which show division in a dorsal direction. These folds are well formed and clearly defined and are so clearly portrayed in the figure that a detailed consideration of them is deemed unnecessary. At the lower border of the figure there may be observed the anlage of a mucous gland, budding from the epithelium ventral to the foliate area. Three such gland anlagen are observed in the model only one of which is seen in the figvnc. The mucous glands of the posterior and posterior ventral part of the rabbit's tongue thus have their anlagen at an earlier stage of development than do the serous lingual glands of the foliate area. When .section D. of figure 1, (23 days) is compared with section C (21 days) of this figure, it will be ob.served that the jirimary epithelial folds as seen in sections, are deeper in the older stage,

THK r\l'll.I,A KOMATA OK THK lt.\BBIT 351

thouKli of ossonfially the saino form. Tho priiniiry murosal folds arc likewise better develciped in the older stage. It may he ob.servod from a study of the figure that nearly every primary iimeosal fold carries near its crest a capillary loop, which cai)illary looi)s, for relatively l(»ng distances, course parallel to the epithelium. We have included a reconstruction and drawing of a section of a foliate area of a ral)l)it embryo of 24 days, in that in this .stage there is oljserved the first indication of the anlagen of tlie serous lingual glands also characteristic changes in the primary epithi-lial folds an> seen. In figure 4, is reproduced the mucosal surface of a ruction of the epithelium of the left foliate area of a rabbit embryo of 24 days. (In comparing figures 4, .") and ti, with the figures 2 and '.i, it will be noted that figures 2 aii<l 'A. are of the right foliate area, while the other three figures, 4, ."> and (i. are of the l(>ft area. The pitted under surface of the e])ithelium, mucosal pipillae, indicates the region anterior to the foliate area. In the .sections of all the .stages figured in figure 1, the anterior border is to the left of the figure, so as to make a comparison of sections more easy.) This area l)resent perhai)s Iti primary folds, several divided folds may be seen, one of which divides ventrally. The g(>n('rai arrangement of the primary epithelial folds is clearly iiulicated in the figure. Kspecial attention is drawn to the .serated edge of the middle portions of the longer, centrally placed folds. The folds, in this region present on their crests short bud-like protru.sions recognized as anlagen of the serous lingual glanils. .Vs is evident from the figure, these glands appear first in the more central region of the foliate area. Hermann has previously stated that the central region of a foliate area is most fully developed. Po the right and ventral in this figure there may be ob.served the anlagen of several nnicous glands, distinctly longer than the .serous lingual glands, presenting bulbous ends. ha\ing ;in epithelium which enal)les their recognition as mucous giantls. In se\-eral of the more centrally placed primary epithelial I'oKls. as seen in the reconstruction, there may be observed about miilway down the resjiective foUI, a narrow shelf passing along each side of lh(> r(>s|)(>ctive fold. These represent the first indication



Vig. 5 PhotoRraph of the mucosal siirfaco of a «ax |ilalc ruction of the ri)itli('liuiii and ihicts of tlie serous lin);ual gland of the foliate urea of a rabbit embryo of '2i'i days. MaK'iiliealioii appro.ximately oO diameters.

KiR. (') I'hotopraph of the tiiueosal surface of a wax plale reconstruction of the cpitlielium and duels of the .serous linRual ulands. with the primary and secondary epithelial fold.s of the foliate area t)f a rabbit endiryo of 30 days. Magnihcation approximately 50 diameters.



of the socdiidary cpiUiplial foMs and arc much riiDrp clearly discerned in (ho tlrauin^ of one of the sections of the seriiw from which this model was made, K, of figure 1. in this section, the epithehum horderinK the anterior part of the foliate area is directed toward the left while in the model, this rcRion, recognized by the epithelial pits, is directed towards the right. This series of sections is cut almost parallel to the long axis of the foliate area and i>resent good cross-sections of the epithelium of the more central part of the field. It will l)e ol).<erved on comparison with the former sections, that the primary ej)ithelial folds are deeper than in earlier stages and that they are thicker and broader in their basal portions, this thickening extending to about half the depth of the respective epithelial fold. The lower border of this thickening on each side of a primary epithelial fold marks the aniagen of the secondary epithelial folds, much more clearly seen in older stages. It may be observed that certain of the primary epithelial folds extend for a greater distance into the mucosa than to others. These longer 'folds' indicate regions in the preparation in which the plane of section included serous lingual gland aniagen. The two series of sections of the 25 day stage were both cut (juite obliciuely to the long axis of the respective foliate areas, and thus alscj the primary epithelial folds, giving these folds a broader contour than is warranted from u study and part graphic reconstruction of these series. A figure of this stage is not included. This stage shows only a very slight advance in development over that shown for the 24 day ral>bit enibryo. The stage of the 21) day rabbit is presented both in model and in section. The model reproiluced in figure 5, is of the. left foliate area of a 2() day rabbit embryo. There are present in all 15 well developed primary epithelial folds and two less fully developed ones. The secondary epithelial folds are more fully developed, ext<Miding deeper into the mucous, than in the 24 day stage, though the.<e secondary epithi'lial folds are somewhat difficult to discern in the figure of the moilel in that they are hidden by reason of the compactness of the folds; on study of the model itself, they are clearly recogniz(>d and may be seen in the section of this



stage to be discussed presently. The serous liiimial glands show marked inorcasc in Iciipth, many havinp; picrciMJ the iniisculature of the tongue for (|uite a distance, ha\ing hullious, l)ut as yet unliranded ends. There is also observed a marked increase in the number of the glands, no less than 140 gland ducts are to be counted in the reconstruction of this area. The ilucts of glands arise from the crests of the primary ejjithelial folds, A number of large mucous glands are evident to the left and lower left in the figure, just outside of the folds limiting the foliate area. A typical section of the epithelium, taken approximately from the middle of the series from which the model was made is shown in drawing /■', of figure 1. The thickened bases of the primary epithelial folds, with distinct indications of the secondarj' epithelial folds, are easilj' recognized in this figure. In this figure many of the primary epithelial folds extend as gland <lucts into th(> mucosa, and certain ones beyond the limits of the drawing, which includes the mucosal layer to the u])per level of the tongue musculature. It is requested to note the positions of the capillary loops in the crests of the primary mucosal folds as seen in figure 1, F, and compare the same with the positions of the capillar}' looi)s in the crests of the simpler mucosal folds as seen in figure 1, D. The relative position of these capillary lf)ops remains constant niul thus enable an intcrprc^tation of the morijhogencsis of the epithelial and mucosal folds. In the ral)l)it embryo of 26 days the foliate areas are readily recognized with the unaided ej'c. Since, as may be observed from a study of the section as drawn in F, of figure 1, the surface grooves have ac<iuired depth, givuig the upi)er border of the section a distinctly scalloped appearance; it will also be observed that these surface grooves are situated over the middle of the j)riniarv epithelial folds from the crests of which bud the serous lingual glands. Drawings G, H and /, of figure 1, are from series of sections of the foliate areas of rabbit embryos of 27, 28 and 29 days, respectively. They are inserted mainly to give details of the development of the secondary ei)ithelial and the secondary nmcosal folds, developing — the epithelial folds from the of the primary epithelial folds with which they run parallel. In each figure


the positions of the capillary loops is to bo noted. It will ho scon that tho sooondary opithclial folds an* dearly iiKlicatcd in G, of linurc 1, they arc longer ami more distinct in //, of this figure and in section /, in limine 1. ap|)ear in cross-sections as distinct, slender pajMllae, divcrfiiiif;; slit;htly from each side of the bases of the several primary epithelial folds; these slender papillae representing cross-sections of the secondary epithelial folds. The serous lingual glands have accjuired length during these stages extending now deeply into the tongue musculature, showing l)ull)()iis or branched ends and evidence of the presence of a lumen in many of them. In figure (5, is shown a reconstruction of the left foliate area of a rabbit embryo of 30 days, removed from the uterus just prior to birth. It represents the oldest stage modelled and gives the muco.sal view of the epithelium of a foliate area which in many respects is like that of a foliate area of an adult rabbit. This area presents Hi primary folds and toward the right of the figure several small folds which have the same general directions as do the foliate folds, but without presenting their morphology in that they are not accompanied by secondary folds and have not given origin to serous lingual glands. This area has spread out to such extent, in course of development, that the secondary epithelial folds are clearly seen, as running parallel to each side of the respective primary epithelial folds. The positions of the ^}rimary and seconilary mucosal folds can easily be depicted in that they fill the clefts bounded by the epithelial folds. Approximately 200 ducts of serous lingual glands are counted in this reconstruction, probably about the numi)er present in a foliate area of the- adult rai)l)it. The backward turn of the ducts of the serous lingual glands of the most posterior portion of the foliate area, to the left in figure 0, is to be noted. \i this stage the ducts lead to the surface of the epithelium and do not empty at the bottom of the deep clefts as found in the grown and fully developed area. It seems reasonable to conclude although we have not at our disposal the neees.sary stages to substantiate this conclusion, that the deep clefts separating the surface folds of the fully developed foliate area, are in part formed through a desquamation of the

356 o. cARi. iit'nKu a.vd arvoi.d h. eggertii

epitholimn separating tho dvict luiniiia as they pass throui^h the epithelium to the surface. In section .1, of fipiurc 1 is given a drawing of tho epithelial horder and underlying mucosa of approximately the middle section of the series from which the model in figure (5 was made. This section is in direction nearly parallel to the long axis of the foliate area sectioned and cuts ♦he major folds at right angle, presenting also a good cross-section of the epithelium. 'Ihis figure is in a large measure self explanatory, primary and secondary epithelial and mueo.sal folds are clearly discernible as are also the ducts of the serous lingual glands. It should he noted that in the fully developed foliate area, the capillary loops found in the crests of the primary mucosal folds, separating the primary epithelial folds, /) and E, of figure 1 fall to the center of the surface folds as found in the fully developed area foliata, the clefts separating these surface folds falling in line with the rows of duets of the serous lingual glands as seen in figmv (i, each such surface ff)ld emhracing the tissue found between two rows of ducts of the lingual serous glands.

The fjuestion of the histogenesis of the gustatory buds has not formed a special portion of this investigation. The methods of fixation and staining used, have enal)led us to confirm in a large mea.sure the early observations of Hermann' as concerns the development of the gustatory buds of the rabbit's foliate area. Relatively few gustatory buds are present in the foliate areas of the rabbit at the time of birth. These few are nearly all found scatteretl here and there near the mouths of the ducts of the serous lingual glands. A study of the histogenesis of the gustatory buds of the foliate areas of the rabbit is projected in a future investigation, with the aid of special technical methods, of which the pyridine silver impregnation method shall receive special eon.sideration. It seemed to us advisable to first accjuire familiaritj' with the morphogenesis of the foliate area, before a study of the histogenesis of the gustatory buds was undertaken. The severance of official connection with the Department of .\natomy of one of us has prompted us to place on record the results of this investigation as thus far attained, namely, the


ni()i(ih<)gonosj.s of tlic ftjliate areas of the ral»l)it from time of its first apjiearance to the time of birth. Our observations may be summarized briefly as follows:

1. The foliate areas of the rabbit's tongue may be recognized as sliphtly elevated, oval pre-foliate areas in rabbit embryos of the isth day.

2. The primary epithelial folds characterizing these areas, begin to appear on the lOth day, and are well formed an<l present in ap])roximately full number on the 21st day. These primary epithelial folds are to be recognized as 'histosystems,' .since they are found in fairly constant number througliout the period of morphogenesis of these areas; jiraetically the entire foliate area on each side being in anlage at the same time.

.'i. The serous lingual glands, associated with the foliate areas, make their ai)pearanee, in the central region of the foliate area, during the 24th day. The secondary epithelial are in anlage during the 24th day, in the central region of foliate area, and show appreciable development over the entire area by the 2(jth day.

4. The epithelial and nmcosal folds, both primar}- and secondary, as viewed from the under side of a foliate area of the 30th day, show an arrangement and state of development which is essentially the same as in a fully developed foliate area, although the deep clefts or grooves which separate the surface folds of the adult area are as yet merely indicated. At this period of development the gustatory buds are as yet few in number and are found in dose relation to the mouths of the ducts of the .serous lingual glands, which reach the surface of the epithelium.

AiTimirH' AiMTHArr or thim PAi'r.n IMUKD






Jiihiis lliipkinx Mnticat Srhoiil

f'oncorninf!; tlio tiiiio oloinent involved in tlio mitotic division of the soniutic cells there is at present scarcely any data.

Retzius' found that the total time involved for mitosis in the ectodermal cells of triton punctatus was from 2 to 3 hours, s(jmetinies more, sometimes less. He has nothinji to say reftardinp the duration of the various phases nor whether he considered the process as ending with the division of the cell.

C'lark'.s= observations on the division of mesenchyme cells in the tail of the living tadpole were carried on in connection with his studies on the living growing lymi)hatics. Unfortunately there is a fairly complete record for oidy one cell. The prophase and metaphase together lasted ahout 1 hour and 1") minutes; the duration of metaphase was not given. During anaphase which lasted about 4 minutes the movement of the chromosomes could be seen. Telophase followed immediately and constriction except for a slender strand was complete in 3 to 4 minutes; 2 liours elapsed however l)efore .separation was complete. Not until about 3 hours after the division of the cell (except for the slender strand) diil it ajipear to be in a normal resting conilition. It is interesting to note that during a period of 23 days when the tail was under observation only about onehalf of the cells underwent division and none of them divided moic than once.

' l{i-tziii8, C!. ISSl. .St udicn liber die Zelleiitlieilurij;. Hiol. fiitersuchiiiigcn. 'Clark, K. U. 1012. Fuither oliservntion.s on liviiiK Krowiii); lymphatics: their rclaliun to I lie mesenchyme cells. -Vm. .Four. .\nat.. vol. l:!.


I aiiilxTt and Hancs' found in plasma cultures that the connective tissue cells of the cat divide in lo to 30 minutes at 37°C., while those of the rat n>(|uire from 2') to 4o minutes. They found in the cat that (lie time involved Ijetween the separation of the daughter chromosomes and the complete division of tlie cytoplasm was from 4.5 to (l.o minutes. 'I'his corres|)onds \'ery closely with the duration of in the chick eml)ryo cultures given below. It is difficult to compare with our results the duration of prophase, metaphase and since they are not Riven separately by these observers, and since it is not stated in precise terms whether the observations on projihase were begun with the resolving of the nuclear membrane or after the stage was already iiiulcr way. It is very difficult to find cells that are just entering upon the stage and our cx|)erience has been that one recognizes the prophase after it is well under way and \ery rarely does one have under observation a restjng cell and watch its initial changes into the prophase stage. There is no indication in the paper by Lambert and Hanes that they were particular about this point, so that in all probability their observations were begun on cells already in the prophase stage. If so, it will be necessary to add from 15 to 30 minutes to their estimate of the duration of prophase,, and anaphase.

We infer from their paper that they consider mitosis to end with the division of the cytoplasm into the two daughter cells. .So that if we include under mitosis the long period of reconstruction, after the cell has divided, during which various changes take place in the cell until it arrives at the resting stage, it would probably be neces-sary to add 1 or 2 hours to the observed duration of mitosis as given by Lambert and Hanes. This would ^ive from 2 to 3 hours for the total duration of cell division, which «orres|)on(ls fairly well with our results.

Lambert* in another paper takes up the rate of cell division in rat connective tissue cells. He finds in plasma cultures from

' Lambert, R. A. and Hanes, F. M. 1913. Beobarhtungen an Gewebskulturcn in vitro. V'irch. Arch., Bd., 211.

  • Lamlicrt, R. A. 1913. Comparative studies upon cancer and normal cells.

II. The character of growth in vitro with special reference to cell divisions. Jour. Kxp. Med., vol. 17.


10 to 40 days old that the time required for mitosis at 38° and 39'(". varied from 21 to 29 minutes, while at M" to mT. the time varied from 'Mi to 50 minutes. In the cells dividing at '.iH° to 39° the intervals from the first appearance of constriction until eomi)Iete division of the cytoplasm were as fr)llows; 3, 5, 5, 3, and 3 minutes. It appears from his tahle that a pi-riod <if 2 to 3 mimitcs intervenes between the separation of the (iauti;hter chromosomes and the beginning of cytopla.smic constriction. The duration of prophase, metaphase and taken together varied as follows; 22, 15, IS, 23 and Itj minutes. Here again we are unable to determine just what Lambert considers as the beginning of mitosis. He does not state in precise terms (hat his observations began at the liegiiming of, so that we have the same diHieulty here in making comparisons with our oijservations on chick cultures as we had with the observations of Lambert and Hanes.

\\'hile the observations recorded below were proceeding, there aj)peared an article by Levi'^ on almost the identical line of work. Levi's ob.servations were on embryonic chick tissues cultivated for the most part in normal plasma and we are thus enabled to compare the behavior of cells in plasma and in Locke's solution while undergoing such a highly complicated process as mitosis. Since so many adverse criticisms have been made against the use of Locke's solution as a culture medium, it is of especial interest to note that our observations are on the whole very similar to those of Levi both in regard to the general behavior of the cells during mitosis and in regard to the time element involved for the various phases.

In order to record accurately the time involved during the various of mitotic division, it was necessary to have clearly in mind a sharp line of distinction between each phase following as nearly as possible the customary usages of the terms prophase, metaphase, anaphase and telophase.

The observations recorded below are on mesenchyme cells from embjyo chicks 4 to 11 days old and cultivated in Locke's

» Levi, G. 1916. II ritmo e le modallita dclla mitosi nolle rollul.- v!v.-nti coltivate 'in vitro.' Arch. Ital. di .Vnat. e di Embr., vol. 15.

3«)'J \\.\liKi:\ II. I.KWIS AN'I) MAUr.AIiET R. LEWIS

sdlutioii with or uillnmt ilic ailtlitioii of bouillon. I'lir ciiitiire? were kept at al)oiit :V.l ('. uitli slifiht variation to :iS or 40°C. Tho cultures varird in an<' fnnii '_' I to 72 hours.

The i)roi)' is usually dclincd as the preliminary sta^e in mitosis including all tlie phenomena prior to the division of the chromosomes. Since it is not possible to see exactly when the chroinosomes divide, the term as used in this article involves the processes from th(> time of the first changes in the cell until the chromosomes arc arranged in the equatorial plate. The latter condition is easily recopiized.

The interval between the first observation of the cell in prophase and the end of prophase varied from r> to 50 minutes, and is as follows for the various cells observed: 5 + , 8 + , 8 + , 10 + , 10 + , 12 + , 15 + , 17 + , 30 + , 38 + , and 50+ minutes. This indicates that prophase usually occupies from 30 to (iO minutes.

Le\i gives the time involved for prophase and metaphase toeether as from 12 to 30 minutes but he does not specify the e-xtent to which the prophase had advanced when the observations were l)egun. Levi had only three observations on the duration of metaphase, namely 8, 11 and 13 minutes; and six on and metaphase together, namely 28, 12, 20, 22, 25 and .30 minutes. The duration of pro])hase in Levi's plasma cultures was probably between 5 and 20 minutes, to which must be added a considerable ])eriod of time on the assumption that his observations were begun after the was under way. \\'ith these corrections, the iluration of the projihasc in plasma is similar to that in Locke's solution.

The metaphase is usually defined as embracing the period of the splitting of the chromosomes. We have considered the metaphase as extending from the time th(^ chromosomes are fii-st arranged in the ecuiatorial jilate until they I)(>gin to move toward the poles. During the first jiart of this period, the ecjuatorial plate is thick and it is impossible to distinguish the separate chromosomes, while during the latter part of the period the plate splits into two parallel plates with only a very fine line between them. These two plates do not usually begin to move apart for a minute or more after the sjilitting. Metaphase


varies from 1 to lo minutes, occasionally it may last lonRor. in one case 2") and in anollicr 45 minutes. The intervals involved for this stage are as follows: 1+, 1.25 + , 2 + , 2.5 + . 2,5, 3 + ,

< + , 3, 3, 4, (j + , 7, '.), 11.25, 12, 12.5, 15, 25 and 45 minutes.

The metaphase is shorter on the average than it was in I^vi'.s three ohscr\ations for plasma cultures.

1 he anaphase involves the moving apart of the divided chromosomes to the poles of the spindle and their later union into sjiiremes to l)uild up the daughter nuclei. As used in this article the anaphase includes only the time during the migration of the chromosomes to the poles and varies from 1 to 5 minutes. The periods are as follows: 1. 2, 2, 2, 2. 2, 2, 2, 2, 2, 2, 2, 2, 2.5, 2.5, 2.5, 2.5, 2.5. 2.5. 2.5, 3, 3, 3, 3, 3, 3, 3, and 5 minutes. The time averages from 2 to 3 minutes. Levi's observations on plasma cultures show a variation from 1.5 to 19 minutes, most of them lasting from 3 to 7 minutes.

Telophase is usually considered the final stage in mitosis: the cytoi)lasm divides forming the two daughter cells and the nuclear membrane reforms. As here used, it extends from the end of anaphase to the separation of the cell into two daughter cells (except, as often happens, for a thin .strand of cytoplasm which may connect the two daughter cells for many minutes after the divi.sion is practically complete). .\n appreciable interval elapses between the end of anaphase and the first appearance of the indentation, during which the cell elongates and flattens out on the sides. This period may vary from 1.5 to 7.5 miiuites but usually lasts only 1 or 2 minutes. The observed intervals are as follows: 0.5, t).75, 1, 1, 1, 1, 1, 1. 1. 1, 1. 1+, 1.5, 1.5, 2; 2, 2, 3, 3, 4, 5 and 7.5 minutes. This period we son\etimes designate as (a).

The time involved from the first api^earance of indentation, usually on one side first, until complete division lexcejU for a thin narrow stajk) varies from 2 to 10 minutes; with an average 2 to 4 minutes. The individual observations are as follows: 2. 2, 2, 2, 2. 2.5, 2.5. 2.5, 3, 3, 3, 3. 3, 3.5. 3.5. 3.5. 4, 4. 4. 4. 4. 4, 4, 5,5, (i, and 10 minutes. This jieriod may be designated as telophase (b).


'1 he duration of the entire telophase stage, that is. from the end of anaphase to tlie division of the cell, varies from 2 to 13 minutes. The observed durations are as follows: 2, 2.5, 2.5, 3, 3. 3, 3. 3, 3, 3, 3, 3.5, 4, 4, 4, 4, 4.75, 5, 5, 5, 5, 5. 5, 5, 5, 5, 5, 6, (i.5, 7, 7, 7, !>, 11.."). ami 13 minutes; average 3 to (i minutes. Levi's (il).servations show a duration of from 1 to 10 minutes and are as follows: 1, 1.5, 1.5, 2, 2, 2.5, 2.5, 2.5, 2.5, 3. 3, 3, 3, 3, 3, 3, 3, 3, 4, 4.5, 5, 5, 5, 6, 6, 6, 6, H, 6, 6, 6, 9, 10 and 10 minutes. The time is about the same, therefore, in Levi's plasma cultures as in the Locke's solution.

.\lth(jugh mitosis is usually supposed to end with the division of the cell into two daughter cells there remains a considerable interval of time diu'ing which the daughter imclei pass through various stages before they arrive at the normal resting condition. For convenience, we have called this the reconstruction period. It is difficult to determine just when the cell arrives at what might be called a normal resting condition. The chromosomes are first gathered together in a small mass, around which a clear substance gradually collects and as the chromosomes fade away from sight, the nuclear wall and the nucleoli become visible. Ihis nucleus is much smaller than that of the resting stage and in fixed specimens stains deeper with iron hematoxylin than does the larger resting nucleus. The intervals involved from the end of division until the first appearance of the nuclear membrane are as follows: 8, S. 8, 10, 10. 10, 12. 12, 13, 17, 18. 20, 20, 21 and 23 minutes. '1 he nuclear membrane usually appears about the same time in the two daughter cells.

The period from the first appearance of the nuclear membrane until the resting stage is difficult to determine exactly. Our observations indicate that it is as long if not longer than the; i.e.. 10 + , 10 + , 15 + , 30, 35 + , 40, 40, 55, t)0 + , 80, 85. 110 and 110 minutes. During this period there is a gradual increase in the size of the nucleus and of the cytoplasm.

Ihe records of a few of the mesenchyme cells from cultures of a 7 da J' embryo are shown in table 1.

Other observations on 4, 5, 6, 9, 10 and 1 1 tlay embryos do not indicate any marked differences in the rate of mitosis nor is



TABLE I Rale of mitons of mesenchyme cells of a 7 day chick embryo





Tcliiphase (a)

'IVIopliase (b)

Kccunst ruction (a)

Keconstruction (b)



4 3

1.5 3.5










92+ 122+


12.5 2.5 1 2


3 +

2 1 3


3.5 3.5 10




8+ 3 3

1.5 3 5 13




2.5+ 3 1 2.5


8+ 2 5 2 5


6+ 2 5 1 4 8




12 3 2 4

there any noticeable increase or decrease in the rate in the cultures that are 24. 48 and 72 hours old.

The time recfuired then for the complete process of mitotic cell division would lie within the following limits: Prophase, 30 to ()0 minutes; metaphase. 2 to 10 miiuites; anaphase 2 to :■? minutes; 3 to 12 minutes and the reconstruction period from 30 to 120 minutes: total 70 to 180 minutes. From the present data 2 to 3 hours would be a fair estimate of the time involved.

An attemi)t was made to conipare mitosis in tissue cultures with that found under more normal conditions and for this purpose a four day chick with the amnion intact was put int<i a small dish of Locke's solution and the rate of division observed

T.\BI.E 2 Rale of mitosis of the smooth muscle cells in the amnion of a i day chick em'-ryo



.Vnaphase ■

Telophase (a). . Telophase (b), . .

Reconstruction (a)





















2 5








2 5





2.5 5






2 5 9





wahrex u. i.kwis avd Margaret r. lewis

in tho smooth muscle rolls of the amiiioii. The results are Riven in tal)le 2. The process is more diflicult to follow under these con<litions than in cultures. The proi)hase was olj.scrved over most of its tluration in the seventh case and was '.V.i niiiuites. The inetaphase lasted from 2 to 15 minutes, the anaphase 1 to ■S minutes, the telophase A to ") minutes and the first part of the reconstruction period from 4 to U) minutes.

If we compare the data at hand on the various phases of mitosis as collected together in table 3, it seems very probable that the dtwation of mitosis is not so very different in the various types of







Mcsench>Tne tadpole tail, Clark

< 75+ >




Connective tissue cat, Lambert and Hanps

< 25-45+ >

< 13-23+ >

4.5-6.5 3-5

Connective tissue rat, Lambert

McsenchjTne chick, plasma, Levi



3-7a 1.5-19

2J-6a 1-10

MesenchjTne chick, Locke


2-12a 1-45

2-3a 1-45

3-6a 2-13


Smooth muscle chick


. 2-15


3-5 5


Duration in minutes of the various phases of mitosis. + indicates that more time should probably be added, a. Average.

cells, if we make the necessarj' allowances for the fact that the observations were usually l)egun after proi)hase was under way. Not until we have more accurate data however can this be decided with any decree of satisfaction. Concerning the duration of prophase the data is too meager for compari.son. Metaphase is on the average somewhat longer in Levi's plasma cultures than in our Locke cultures or in the smooth nmscle of the amnion. The variations are greater in our observations than in Levi's. More extended ol)servations in plasma however might show as great variation as in Locke's solution. .\ naphase also is slightly


Iniincr DH the average in the plasma than in Locke's solution and the variations are nioro marked and greater in the latter solution. rill' cxtrcnu's in anaphase as in metaphase must i)e in the reaction of the individual cells since these great variations occur in cells side hy side in the same culture. The duration of telophase which can be more accurately recorded than the other ])hases shows a striking similarity in all types of cells and less variation. It is rather remarkable that the telophase of the mesonchyme cells in the tadpoles tail, connective tissue cells of the cat and rat in plasma, mesenchyme of the chick in plasma

in(l Locke's, and smooth mascle should all have about the same

IcMfith of time. Variations in the duration of any one is one of the most characteristic features of cell division and is aiiparcntl}' dependent on individual cell differences.


UT T)i»:\i'iii< neitvicc, ocrfiur.R 30


I'AIL B. EATOX Friim till- Anatomical Laboratory, Johns Hopkins Medical School


The classical description of the coeliac axis states that at the level of the upper Ixudir of the pancreas it divides into three branches, the left gastric, the hepatic, and the splenic art<'ries. A number of studies of this vessel have been published within the last few years, notably those of Rossi and ('ova( 1), I^eriche and \'illeniiii (2), Descomps (3), Picriuand (4), de Rio Branco (5), Robinson ((>), and Lipshutz (7). Of these Leriche and ^'ilI^nlin alone report that the}' found the classical "tripus" in a iiKijiirity i unspecified) of cases. At the suggestion of Prof. W. H. Lewis this study was undertaken to determine as definitely as possible the normal type and to get some data on the frequency of tleparture from this norm.

The sketches which form the basis of this report were made by myself from dissections made for the most ]iart by freshman medical students (to whom 1 am indei)ted for many courtesies) and the amount of material studied and its distrii)ution are shown ill tlic fdlldwiiijz; tal>l('.

S'umhrr Dinsrcling Rvum^ of aubJKis

Johns Hopkins University Medical School. 47

I'niversity of Maryhmtl Medical School H4

CicorRe WnshinKton University Medical School. IJ

University of Pennsylvania Medical School.... II

JefTerson Medical College -10

Woman's .Medical ColIcKe of I'hiladclpliia. . .'{

Tcniplc University Medical School IJ

I'liiiadelpliia Policlinic I

Medico-Cliiiiir^tical CollcK*' of Philadelphia... S

CieorRctown University Medical S<'lioi)l. .">

Total... JOti

■ I wish here to make acknowledgment and express my thanks to those whodo

courtesy has made this work iwssible and very pleasant. In particular to Pro 309




The SOX iukI skin tint of 2(t(> of these subjects was recorded. Tlicv were (listrihutod as follows.







26 36







'I'hc nunilicr of cases is not sufficiently large to warrant any positive statement hut there does not appear to be any great sexual variation in the variability of the structures under consideration. The same may be said with respect to race. A numbei- of ()l)servers liave reported that negro subjects seem to show a greater tenilency to variation than white ones but our use of the term negro is too loose to be of any value from an anthropologic standpoint. Dr. Hrdlieka (Division of Anthropology, Smith.sonian Institution) informs me tliat

There is ground for l)clicvinK that the iicjiro uf pure i)lood siiould show a greater staljility, anatomically, than the white man, who is generally a pretty tiiorough mixture and who has been subjected to more varied environmental, occ-upational, and other influences: hut our knowledge of the details of heredity is not sufficient to enable us to predict the etTect of hylnidizMtion njion this stability.

If our dissecting rooms contain the i)roportion of hybrids that has been alleged of our large cities (fifty to .seventy per cent) a certain tlegree of anthropological skill will be nccessarj' to make the separations necessary to settle this question.

fessors V. I'. .Miill, iiml \\ . II. Lewi.s, of .lolins Hopkins fnivcrsity. iind their a.<<i:iiits; I'rofc.-i.sors .1. lloliiics Smith ami .loscph W. Ilolhiml of tlic ruivcrsity of .Miirylaiid; rrofcssor Carl K. Davis of (rcoTKC WasliiiiKtoii rnivorsily ; Professors (Jeorge A. I'iersol and Ceorge FetterofT and Drs. (r. .M. Dorrann- and P. (t. .Skillem of the I'nivorsity of Pennsylvania; Profe.ssors .1. P. ShaelTorand I), (t. Metheny, and Drs. lliifTiiian an<l Lipshuiz of .lefTerson .Medical Collene; Professors Henry Morris and .Mary Hicking Thornton of the Woman's Medical College of Philadelphia: Profe.-isor .Vddincll Hewson of Temple I'niversity and Philadelphia Policlinic: Professors .lohn ( >. lleisler and II. II. Cushiiig of MedicoChiriirgiral College: and Professors IVank L. Baker and W. I'. Heiider of (ieorgetown I'niversilv.




The history of the discover}' and naming of the Coeliac Axis is gone into very tlionnifilily l»y dc Kio Hraneo (')). thceinhryolojjy by Tandlcr (Sj and I lie comparative anatomy l)y Rossi and ("ova (\). These monographs are accessible ami inakr further reference unnecessar}'.


Table 1 gives the simplest elassifieation possiljle. The first l)art is taken bodily from liranco and the remainder atlded to show that extending the scope of the incjuiry to the point of













cotxrir AXIS





Rossi and Cova

Lcricho ami Villoniin.


55 ,5() 50


49 44 45

12 5

6 4




de Rio Branco





27 (10.5-11%)






50 20C




7 21 19







Totals ...










doubling the numlx^r of subjects examined makes no great change in the percentages. The inferior phrenic arteries have been disregarded and will be treated separately.

The sketch on page 372 is an attempt at a logical or geometrical classification. As is to be exjiiH'ted this breaks down if examined too critically. It may l)e extended on either side to show the sei)arate origin of the three branch(\s of the axis from the aorta or the absence of the si)lenic artery, which is much more stable than either of the others. Most of the spaces might be filled from the literature, but extension would make the dia



gram unwieldy ami it is siil)iiiitti' Hepatic artery; H.A., Xccessory hepatic artery; 5., Splenic artery; /'., Pancreatic artery.

aorta antl linally as in Type 1\' entirely olT tlie axis. .Movement to the left from the same point shows the same thing with respect to the hepatic artery. From the same point moving down one rolumn shows the addition of a pancreatic hranch. In the third horizontal column the tirst is repeated with the addition of a liei)atic arter\' from the sup(>rior mesenteric artery. The



figures in tho lower rip;ht hand conipr of each section of the diagram indicate tlie iiuniljcr of times that particuhir arraiiKCment was found in this series of cases. Those figures in parentheses in the lower left hand corners indicate the number of cases in which the left gastric artery furnished an access<jry hepatic artery. If any argument were needed as to the necessity for the investigation of a larger number of case.s it could easily be found in the fact that in this series that particular arrangement occurred 14 times or a fraction less than 7 per cent while Lijv shutz (7) found it in 3") j)cr cent of .S3 cas(>s. So far I have been unal>lc Id lind any (ith( r way to get it into the diagram.


Lcrichc and \'illeinin

Rcssi and Cova



do Hio IJranco




a m I

102 55 50 50 50 83



lyPE I






10-4 9%



26 5 16 13 21 47-22.8%


60 37 28 30 41 140-68%


3 »•> 4 5 1 12 9-4 4%


It will be noticed that there is a distinct grouping in one section of the diagram. This is as it should be, for, as urged l)y Ruge (Dj and Lipshutz, arterial variations group themselvi>s into distinct types in numbers inversely ])roportionate to the amount of their departure from the nmnial.

It is of course inii)ossil)le to compare these results closely with of jirevious investigators. Table 2 is however an attempt at such a comparison and I believe that such errors as may have crept in will not greatly affect the percentages. The columns for Types II and III are incomplete for the reason that Leriche and N'illemin merely state that the three i)ran(hes came off at the same level in the majority of cases. Their cases are therefore not counted in calculating the percentages for columns.

',Vt4 I'M I, ». EATOX


1. Thf iioniml type of cooliac axis is that which gives off the loft gastric artery as a collateral branch l)efore the hifurcation into the hepatic anil i^plenic arteries. (02.1 per cent of 541

2. The classical 'tripus' occurs somewhat less than half as frequently. (24 per C(>nt of .")41 cases.)


(1) Rossi ed Cova Arcliivio italinano di Aii:it. v ill Fmbriol. Florence, t. III fas. 2, pp. -IS-VSL't;.

(2) Leriche et Villkmi.v 1907 Bibliogr. Anatom. Paris, t. 16, fa.s. 2, pp. 111-125.

(3) Descomps, p. 1910 Ix- Tronc Coeliaque, Paris.

(4) PicQCAND, G. BiblioRr. Anatom. T. 19, fas. 4, pp. 159-201.

(5) DA SiLVA Rio Buanco 1912 Le Tronc Coeliaque, Paris.

(6) Roui.sso.v, B. I'MiS .Vincrioan Practitioner and News.

(7) LlP.sHCTZ, B. 1917 .\nnals of Surgery, February.

(8) Tandleh. J. I'.HM Anatom. Heftc. Wiesbaden, vol. 25, pp. 473-500.

(9) Huge, G. 18S3 .Morph. Jahrbuch, vol. 9.



F. I'OYALES De la Clinica O/ldlmica del Prof. Marqucz; Madrid, ExpnUa


La posici6n cstrdbica convergentc que presentan los ojos de los rorion nacidos os un hccho cuya intorprctaci6n im rstii dofinitivaiiicnU' fijada y para la cual sc han invocado no pocas tcorias cuyas tesis se puodon reducir a dos grupos: falta dc cvoluci6n c'crcljral 6 desordcnos (mi la inisina por taras horcilit arias 6 lesiones inflainatorias (Pcrinaud) ; hipcrnu'tropia constantf del ojo del nino acoinpafiada de reduccion en la ainplitud del canipo pupilar y de la convergencia.

Ni una iii otra cxplicaii satisfaotoriainente el iiuiieado hecho ya que la priiiina, iiiuy coiupleja, no da euenta de porque las propias causas degenerativas 6 inflaniatorias que ocasionan el estrabisnio eonvergente no produeen el divergente; y la segunda no i)untualiza, eonio, persistiendo la hipernietropia, no persisLe t anil mil el estrabisnio (jue en algunos easos ciueda reducido d la priniera seniana.

La troria (\\U' a mi ver expliea el exjiresado heeho, de niotlo mils ajustado a los proeesos biol6gieos y eon arreglo a la eual realic6 mi modesta labor de invest igaei6n, me fu6 sugerida jior la eonsideraci6n ilel proceso evolutivo histogenetieo del .Vnfioxus. Cr(M) indieada una breve reeapitulaeii'm del desarroUo de dicho animal, sobre todo en lo ([ue afeeta a la formaei6n de los nu'iseulos oeulares, ])ara que el lector pueda seguir el hilo que ha guiado mi traliajo ya ([ue solo ])or eonjeturas filogt'iu'tieas. deilueidas de estutlios jiraetieados en vertebrados infer lores, es como nos es


370 r. I'OYAi.Ks

dado iiulucir alguno; hochos primitivos dc la ombriolonui huiiiana, tan ctiinijlcja on sus priincrns clapas.

Dinanios prmianuMitr (jue cl sistcma 6culo-ni()t()r, cdnstituido por la scric dc miisciilos rectos y oblicuos todos dc fil)ra ostriada, tiene su origcn en cl nicsodcrmd. Cuando cl enibrion do Anfioxus ha llopado al ostado do gastrula so vo naoor dol ontode rino a los dos lados do la linoa inodia una sorio ilo div( rtioulos kpuosos, produoidos por evaginacioncs de esta hoja; estos divertfculos repnsontan la mayor i)arto do la motamorizaoi6ii do! nu-odonno. Los nu'tanioros iiuo nos intorcsan son los corrospondicntes a a caboza y rcgi6n branciuial.

No hay acuerdo respccto A la finalidad y niimcro de los segmontos primordial s f|Uo constituyon la oaboza do los omlirionos \'an Wijho aoojita inievo; Habl los n duoo a cinco n(^;aii(lo la existencia de los ouatro primcros designados por \ an \\'ijho; Dohrii admito dioz sogmontos y on esto modo do prnsar ooincidon K lliaji y otros onibriulogos. Nosotros oonsidoramos la sogmontaci6n admitida por ^'an Wijhe y la signifioaoion dada d oada segmento conio la que mcjor oxplica el complicado dcsarrol o de la cal)(>za do los vortobrados.

El inimcr segmento do \'an Wijho, esta situado pov delante de la invaginaoi6n bucal y el segimdo detrds; los otros continuan en serio conolativa, por lo cual ol primcro corrrspondo li la cavidiid promandibular. El primer segmento esta constituiilo por un gran miotoma ([uo ocupa toda su regi6n dorsal, oareeiondo do parte ventral; es el mioton\a origon de los nuise.ulos del ojo, rectos intrrno y supi rior y oblicuo monor, inorvados i)or ol teroer par craneal ; la parte ventral no oxiste. La mision dol miotoma dorsal al que es debida "a mayor parte dol sistcma Aeulo-motor absorbe la funci6n dol primer segmento y d prsar de su indopendeneia por su parte inferior so eonumiea por un tubo ei)itolia' con los demas segmentos, tubo (juo (juoda comprendido ontro las bolsas branquiales consecutivas y va d abrir.-e a la i)art(' and rior dv la cavidad parietal.

El scgundo segmento constituido tanto por una parte dorsal como por otra ventral es el linico dondo paroce marcarse un etiuilibrio entre los dos elementos que lo constituyon: d la parte dorsal


corrrspondo un iiiiotoiiia del cual prooodo el im'isrido ohlicuo iDuyor y d la jjartc vt'iitial so dcbt'ii el arco inainlibular y los im'isculos quo do 61 sc dorivaii.

lOn ol tercer seginoiiti) ouya constituci6n varia por completo do la del priincni y la del scf^iiiidi), oxisto un inarca<lo dcsarrollo do la |)artc voutral, ciuoilaudo la rrgi6ii dorsal rcdurida a un pcquouo niiotoma casi atrofiado; es el segmento nids poqueno de todos, puos su parte ventral so fu.siona con la parte ventral del cuarto soKinrnto y ayuda a eonstituir la cavidad del areo hioideo; a su niiotoma dorsal corrosiJondo la f()rmaci6n del niuseulo reeto externo. La breve descripci6n ((uc antecede relativa al desarro'lo de est OS tres srgmentos y a los eleinen'tos fjue de ellos nacen, nos liaee p(>nsar que cstos han de coiTor la suerte de sus sul)stancias progenitoras y de ello so deduce que existe una diferencia de origen entre el musculo recto interno y el musculo recto externo, en atenci6n a fine el gran miotoma dorsal del jirimer segme to da origen al interno y el easi atroliatio miotoma dorsal del tercero origina el externo. Por otra parte, estas diferencias de origen dan lugar :i diferencias ("struct urales en la misma epoca emhriol6gica. In chos anat6micos (lue homos ile utilizar para hacer ver, que el recto externo es inferior al recto interno en edail, mimero y Iongit ud de sus fibras, deterniinando, por la mayor potencia del recto interno, el estrabismo cong(?nito convergente que tienen los ninos al naccr.

Las diferencias embriogenicas indicadas ratificalas Wiedersheim al hacrr estudios comparativos entre la disposicion do los n(M\-ios cranrales y ra(juuleos fundando su distribuci6n de los metameros de la cabeza, en las iileAs de \'an Wijhe. Asi los nervios corrcspondientes 6. cada segmento los di\ide en rafces ventrales y dorsales; el nervio motor ocular comun y la rama oftalmica ]irofunda del trigemino correspondi rian al jirimi'r metilmero; ol patotico y el trigc'^mino, menos la rama oftalmica profunda al mctamero srgundo. Al estudiar la distrilnici6n de los nervios en el tercrr segmento, duda de la existencia de la ])arte dorsal, no negando. sin embargo, (iy(> el diminuto miotoma atroHado colocado en la parte sui)erior pueda ser origen del recto


externo y considora al ii(r\-io motor ocular oxteriio distribuy^ndose ]wr cstc spgiiieiito.

La o|)ini6n dc Wi('di'r.";lK"ini \-\vnc a eomprobar la iuiposibilidad dp ncRar la misi6n ilcl tercer scginonto y establece una diferencia do origcn entre ambos nuisculos.

ruiulandonios on ol dosarroUo quo da a la motaniorizaci6n niosodorniioa \'an Wijho, vonios quo la parto dorsal dol rosto de los scgmentos, cuya atrofia se inicia on el tercero, siguo en todos: en ol ouarto y quinto no hay parto dorsal; on ol soxto, ()UO(la algo sin finalidad dotorniinada; on el s^ptimo, octavo y novono sigue el d(>sarrollo de la parto ventral y las porciones ventrales de estos segment OS producen las ca\'idades y miisculos de los arcos viscoralrs llogando a unirso al jirimor sogmento del tronco.

Frorioj) ha \oni(l() a nioditicar la toon'a do \'an Wijho con sus intcrcsantes estudios de la cabeza del embri6n del torpedo. De dlos deduce la oxistoncia do varios soginontos, (juo comiJjoiiden el lado craneal y la jirimera hondidura jjrancjuial, segmontos a los (|ue no asigna misi6n dotjcrminada dandoles una duraci6n muy efimera; mas tarde aquellos serian sustituidos por el mesoblasto iiisrginonta<lo do la caboza, situado por tanto, dolanto do la bolsa brantiuial priniora. Al principio, osto mesoblasto os insogmentado, mds poco d poco aparece la formacion de las bolsas branquiales y se detorminan las ca\'idados do los arcos. Es especialmente & esta segmontaoi6n a la quo corrospondon los metdmeros de Van Wijhe. Con\iene no olvidar que en el mesoblasto cefdlico se ha do considerar tambien una parto dorsal hom6loga & la parte dorsal de los .sogmentos do \'aii ^^ijho, iloiulo tionen su origon los miisculos del ojo. La teorfa do Froriop aunciuo mds moderna, es d mi juicio, mds incompleta y solo una modificaci6n de la anterior que con raz6n porsiste conio cldsica.

Establecidas las diforoncias do origon onil)ri()g(nico ontro los miisculos recto intorno y recto oxtorno, doscriljire las estructurales que he observado dosde el cuarto mes de la vida fetal hasta el naoimiento. Como ya ho dicho, ol hooho a cuya ooinprobaci6n dcdi(iu<5 este estudio nie fuo sugorido al loer las diforoncias de origen embriol6gico anteriormente expuestas y advertir la frecuencia, en casi la tot alidad do los niiios recien nacidos, de la


J)<)sici6ii (■str:il)ic!i coiivornciitc dc los glohos ociilans. Dcsdc el piiiiicr iiioinciito jx-iiso (juc tluranto la viila iutra-utcriiia, el (losaiTollo del recto iutcrno y tlol recto exteriu) habri'a de sj-r diverse, de tal forma, que el mayor desarrollo del interno deteriiiiiKuite de la i)osiei6n forzada, mantenida durante todo el jxiiodn tie la ge.staci6n, siria al iiacc r los nifnts la causa del estrahismo.

lie ])racticado autojisias en fetos desde el cuarto mes de la vida fetal, i)Ues la ad((uisici6n de fetos de menos tiempoes nuiy dificil; hien fjue es de jjresumir que si las diferencias existen desde el (uartii iiics en adelante deben existir en los meses anteriores. Ai)untare ((uc las (lifcrencias de sexo no iin))lican inayores diferencias en el desarrollo de diclios ini'isculos y ([ue los fetos tie aspccto sifilitico parecen ser el material mas apropiado para estat)lecerlas, hien (jue sobre este punto no se pueda hacer una afirniaci6n catrg6rica.

I'-xponffo ;i continuation las iliferencias estructurales nmsculares apreciables en las preparaciones cuyas fipuras van atljuntas.

En el corte (lo\ recto externo(fig. Dsolo aparecen niarcadaniente tenitlos los niideos tiue st)n numerosos, elijisoitles y orientailos paralelamente a la fibra muscular; sin embargo parecen cstar en un magma uniforme, el protojilasma tie catla ct'-lula no csta delimitatlo, lo ((ue manifiesta un estado rudinientario. embrionario, estatlo en tjue los prt)toplasnias se tinen mal y cs tliticil hacer resaltar la fibra nuiscular.

En el ctirte del recto interno (fig. 2) ajiarecen los ni'ideos ocupando una posiciAn semcjante a la de lt)s nudeos del otro corte; se ven los prf)toi)lasmas musculares bien limitadt)s, eilindricos, gianulosos, carecen de estrfas y parecen como entrelazatlos, es tlecir, (lue Ins elenientos tlel nuisculo recto interno son tie mas etlatl, las substancias coloreantes retlt)ntlean los limitcs i)roto])lasmatict)s y las tliferencias en el cuarto mes de la vitla fetal colocan al recto interno como nuisculo mas tlesarroUatlo, sobre totlo en It) tjue se reiiere a la etlatl tie las iibras.

En las prpparacit)nes con-espontlientes d las figuras 3 y 4, tlemu<strase la tliversa cantidatl de zonas de elenientos mas embrit)narit)s que existen en el recto externo on coniparaci6n con el


recto interne lo cjue hace suponer fjue siendo aiiiljos miisculos aiitanonistas en esta epoca de la vida fetal, la nieiior contraccion (lei recto iutenio lialmi de arrastrar ii la fonvcrgeiicia al globo ocular, ya (luc el recto externo por su inferioridad no puede conlrarn>star su accioii. I'^n el recto iutcrno se veii reducidns las zoiias einbrionarias d unas niasas blancuzcas, aisladas; eti el recto externo se ve ilentro de una zona central enil)rionaria algunas iiiasas jjanlas de elementos musculares nijis desarroUados que tieiideu poco a i)oco a invadir la zona central scgi'in avanza el desarrollo del niusculo. A iiuis aiuiiento, las fihras de ainljos rectos aparecen perfect aniente tenidas, sefialandose muy bien los liinites protoplasniaticos, pero en las fibras del recto interno se uiarca una ligera corteza estriada, eleniento que no aparece en la libra del nu'isculo recto externo de esta epoca.

lOxaniinando las figuars 5 y (i. se pueden establecer diferencias en el nunu to dv iibras, nienor en el recto externo que en el int( rno. En e) corte del recto externo las tibras estan, ademas. separadas por bastante tejido conectivo laxo, enibrionario, tejido que existe en mucha nienor cantidad en el recto interno. En la misnia fibra nuiscular adviertese ademas, otra diferencia y cs cjue las del recto externo son cortas y siguen con aspecto cilindrico, en tanto que las del recto interno, son mas alargadas, caracter que se va Mianifestando segiin avanza la evoluciAn : sus nudeos adoptan esta forma tamljien, como si estuvieran oprimidos por el desarrollo mayor del material estriado; las fibras del recto externo aparecen nuis negras porque se tinen bien y resaltan sobre un

Fig. I C'orto lonKiliiiliiial ilcl recto pxtorno on rl cuarto mcs de la viila fetal (270 (li.'imetro.'i I.

Vif.. 2 Corto lonjjiliKlinal ilel recto interno en el ciiarlo mes tie la viila fetal (270 diiirnetros).

Kig. 3 Corte lonKitiidiiial del musculo recto externo en el quinto nies de la vida fetal (4.") didnietros >.

Fig. 4 Corte longitudinal del miu.-«cu1o recto interno en el quinto mes de la vida fetal (45 didnietros).

Fig. .5 Corte longitudinal del intisrulo recto externo end sexto mes de la vida fetal tlSO diAinetros).

Fig. () Corte longitudinal del musculo recto interno en el sexto mes do la vida fet.'d (ISO di:lmetros).


fondo ainhariiio const it uido por todas las cdlulas muscularcs y tf'jido coiijuiilivo laxo in;is nidiincntario.

Las difcrencias vn fctos do sicti- a nuovp nicsos no son ya tan fticilos de establecer como se jjucilo obsorvar en las figuras 7 y 8; sin embargo cl niiniero do fibras siguc siondo aun monor en el rtH'to oxtorno quo on intoriio e igual ocurre con la longitud y cantitlad dc tcjitlo conjunlivo laxo. A mas aumcnto del ([uc ostd hccha la micro-fotografia se observa en cl recto interno la multil)licaci6n del m'ldeo, cl engrosamicnto de la corteza estriada y muchos midcos colocados dcbajo del sarcolcma, lo ([uc hacc pensar que aquellos, oprimidos, huyeran del protoi^lasma por las discontinuidades que estc presenta, en Aortud de la invasi6n crcciente (Id material estriado; en las fibras del recto cxterno cl nucleo siguc sicndo central y ocupa la dircccion del eje de la libra.

Las figuras 9 y 10, corresponden A cortes transversales que tiene como dotalle interesantc la divcrsa manera dc asociarse las fibras muscularcs constituyendo fasclculos .sccundarios perfcctamente indi\idualizados por tabiqucs oonjuntivos, fasciculos que parecen estar constituidos i)or menor numero de fibras en el recto cxterno ([uc en cl interno; en cstc musculo se obscrva la forniaci6n de haces tcrciarios result antes dc la integi-aci6n de los fasciculos secundarios a otras unidades mayores, halldndose aciucllos limitados por scptos concctivos, ba.stantc c.spcsos e infiltrados en cl tejido adiposo. Estos haces tcrciarios .se cncuentran tambien en cl recto cxterno pero en nuicho nicnor niimcro.

Las diferencias nucro.sc6picas descritas, ocun-idas durante el transciu'so de la Anda intra-uterina en el dcsarrollo d(> los dos rectos, cxterno 6 interno, son causa de ([Uc al naccr cl feto y ser su retina impresionada por la luz, se ponga en mo\'imiento el sistema 6culo-motor, cl recto interno mas potcnte y desarrollado ((UC el cxterno habra dc determinar la ])o.sici6n estiabica del globo ocular con un estrabismo convergente no completamente transversal, sino ligeramente oblicuo, en direcci6n de arriba & aliajo y de fuera a dentro con duraclAn distinta de tresa diez dfas bicn marcada. Poco a poco va dcsai)arecicndo d estral)ismo y solo cuando se hace fijar la mirada al nifio aparece de nuevo la posici6n estrabica momentanea.



Tin. ~ Corlp lonKiliidinal del musculo rcclo pxicriio en el .s«-ptiriii) iiics dr l:k vidii fetal (90 diduietros).

FiK. S C'ortc loii(;itudinal del musculo redo intcnio en id s6ptiino ines dc In vida fptal (OO didinclros i.

ImK- " Corto trausvcrsal del musculo redo cxicrno en ol noveno mi-s dc la vida fetal ("10 diduietrosl.

Kin. 10 Corte Irausvei-sal del nuisculu recto internu en el noveno mcs de la vida felal (!tO diiiinetros).

?S4 K. roYALES

La acci6n combiiiaila dc todos los nivisculos del sisteina 6cul()niotor. cuaiido las lilmis inuscularcs dc todos olios cstan igualnu'iilc dosarrolladas y la accioii dc la luz sohrc la retina, son las causas que hacen dcsajiarecer cste estrahismo cnu' dobe llevar el ealiHeafivo de fisiol6{.Mco, teniondo prcsente (luc ninKim musculo del ojo sea recto u ohlicuo, tiene acci6n independieute earacterfstica principal (}ue sienijire csta a>iulada ]ior la de otro nu'isculo; todos unidos constituyen el delicado fisiologisnio ([uc niantienela \-isi6n binocular A rsto hcinos d<' afuulir i\\w los ^;l(il)os oculares al rccibir las prinieras inipresioncs luininosas, ticnden li que por las eondicioncs caracterfsticas de que esta dotada la macula, la imagen se forme en ella y por esto es ueccsario el iiaralelismo de los ejes 6pticos y visuales para que se realicen las condicioues de la \isi6n dara y distinta.

He practieado 21 autopsias

En fetos de t(5rniino "

En fetos de siete nieses 5

En fetos de seis nieses 4

En fetos de cinco nieses 3

Kn fetos de rviatro nieses 2

Punto intercsante, antes de terminar, es la delicada extirpaci6n de los rectos para que no se confundan con los demas miisculos y a mas de esto procurar obtenerlos enteros. Para esto, abriendo una ventana en la conjunti\a y con gancho de estrabotomia, se busca la inserci6n escler6tical de ambos rectos, practieado lo cual se coloca en cada uno de ellos una pinza pequena de Pcan disecandolos despues liasta donde se inieila, seccionando el oblicuo menor, que abraza al recto externo, y a continuacion, en vez de ir a eiegas d buscar la inserci6n posterior, se procede a cortar la eaja craneal en una secci6n circular, se s(>para la masa encefalica y (jueda al descubierto la pared superior de la6rbita que se incinde y una vez ilentro tie atiuella, se disecan los musculos rectos de la inserci6n posterior.





JOHN' A. KITTEL.SO.V Inslilute of Anatomy, University of Minnesota, Minneapolis


Introduction 385

Miitcrial and methods 386

Data and discussion on postnatal growth of the kidney . 392

1. Weights of the kidneys used . . 392

2. Growth of cortex and medulla 393

3. Growth of the renal corpuscles 396

Summa ry 405

Literature cited 407


-Main- questions concerning the postnatal growth of the ki(hie}' are still unsettled. In the rat, the postnatal growth curve of kidney weight in relation to body weight was worked out hy Hatai ('13) and Jackson ri3). Waschetko (T4) estimated the number of renal (Malpighian) corpuscles per cubic millimeter of cortex in the albino rat at various ages from birth to five weeks. The relative growth of cortex and medulla and the total number and volume of the renal corpuscles in the kitlney at various ages are the special prol^loms for which data have been collected in the present investigation.

The work was done in the Anatomical Lalioratory of the University of Minnesota. The problem was suggested by Prof. C. M. Jackson and carried out under his supervision; I wish to express my thanks for his valuable aid and criticism. I also wish to thank Dr. E. T. Bell, of the Department of Pathology, for suggestions and aid in some related work on the human kidney, a few observations on which are incidentally included in the present work.





The material used included the kidneys of seven alljino rats (Mus noivegicus all)iiuis'i from the colony in tlio Anatomical Laboratory. The general data are given in table 2.

The itlentification numl^ers of the rats are used as follows. The letters 'V (Vaughn) and 'K' (Kittelson) indicate the series, the number preceding the decimal point is the litter number, wliile the number following the decimal point designates the individual rat. Three of the rats were from one Utter (VI), and two from another (V2). Three female and four male rats are included. Their ages were newborn and one, two, three, seven, twelve and thirty-five weeks.

The rats were killed by chloroform and weighed, their noseanus and tail lengths also being measured. Both kidneys were weighed together. The weight of each kidney used was assumed to be one-half of the combined kidnej' weight, as the indi\idual kidney weights had not been recoriled.

One kidney (no record made of whether right or left) in each case was fixed in Zenker's fluid for eighteen to twenty-four hours. It was then embedded in paraffin and cut in serial cross sections ten micra in thickness. The sections were then mounted in complete .'^eries and stained with haematoxylin and eosin. In the adult kidnej* (Kl.l), only a few sections were mounted serially, for the most part every fifth section being taken.

The method used in obtaining the total volumes of medulla and cortex was the paper method similar to that used by Hammar ('14) in his volumetric work on the thymus, and by Jack.son ('17) in his work on the hypophysis. The sections were enlarged by means of a Leitz-iMiinger projection apparatus and the outhnes of medulla, cortex and pelvis were drawn on "American Linen Record" paper fsheets 18 by 2.3 inches, 3() pounds per ream). The paper outlines of cortex and medulla were then cut out and weighed, the results being shown in table 1.

It was necessary, however, to find the area corresponding to a given weight of the paper. To find this, samples were taken from 100 sheets of the paper u.sed. Ivich sample consisted of




a •a






■M fl

tc -3 1.5


e c ii 000

r; s 2


."- ~ .Tt





p c^ S 2 17< :*: '"


u o =




s t: !.-:



n « 






c _



c c c

00 =


.2 .2 .2

■- •- .0




U V i^

o u







- - *


U U L.

^ ™ _—

- S

k. t. 'B'.


3 3 C






c t !:•

>. ^ >>



k. u u



^ c^

CJ &



> > >

> >• >




a a a

a fe. <

iS ° ^

3 •3

g f §

CM « ?i

5^1 a 'a a





M t>^ US


a => -.

-^ C^l -^

r i"-'

a w



ri g) Q





-r « X i2 C: 1


•■•7 iC SO




Ci •* f>

CM 3 « 



o t. 2 B


Q 00 M 3 «^ U5

5 CC w

^ ,*< O H

a uS



00 —

M C^l M








'-; ?; .^


S S =


=: a



S = 3

-!• X






s s 3"

? s 3






a o


i 1




CM « •








cc :c





%• — «^

CM fC "

r^ 00 -^

H •

■ ■ M



Ti — —




.^ ^^ w


four pioros (one from each corner of the sheet), each five centimeters square, amount iiip to 100 sq. cm. The 100 samples weighecl 122.85 grams or one sample (100 s(\. cm.) weighed 1.2285 grams, on the average. From this it follows that one gram of paper corresponds to 81.4 sq. cm. of area. The total weight of medulla or cortex in paper in grams is therefore multijilicd by 81.4 to obtain the corresponding area. This magnified area in square centimeters was then divided by the square of the magnification to reduce it to actual area. The actual area nuiltiplied by the thickness of the .section (or sections, since every other section, or every fourth section, or every tenth section, was used) gives the actual volume of medulla or cortex in each case (tables 1 and 2).

In table 2, the volumes of the kidneys ajjjjcar much smaller than might be expected from the corresponding weights. Tliis difference is due to three factors: (1) the volume of the capsule and of the renal sinus (pelvis), both of which were excluded; (2) shrinkage of material in the process of preparation; (,3) the specific gravity of the kidney, making the volume (in cubic centimeters) somewhat less than the corresponding weight (in grams).

In counting the number of Malpighian corpuscles of the kidney by means of the mounted serial sections, a two-fold problem is presented. The corpuscles of varjing sizes must all be counted once and none nuist he counted more than once. In order to count them all at least once, it is only necessary to cut the sections so thin that they are thinner than the diameter of the smallest corpuscle. In the newborn rat, the smallest corpuscle was found to be about 40 micra in diameter, hence sections cut at 10 micra would be certain to section every Malpighian corpuscle.

In order to be certain that no Malpighian corpuscles were counted more than once, Miller and Carlton ('95) used the following method: The average diameter of the corpuscles was determined, and sections of a piece of cortex, cut at about the average diameter obtained. Section 1 was then projected by a camera lucida and the corpuscles marked by dots. Section 2



ta , ji 2

M —









3 ^ L u w z a a a

= d

£ 2

V* —


U5 1* 'O












g 8











? g


u u






s 5








C-. -r




— w


>< n


'^ w




S 3 a J s














•-s c






o w




u .

£ S X n








s i

g ■§ « 15







o i*


n u5



H ^

h, B










o t5



r *!


Cl 1^


3 Q









« —




















(N O




o o



CI r^










X u



X ^









^ '■'5









^ o














"' !5


— .Z




-^ ^^



— *^



^, "^






~ «?

■a a



3 =




X —


= 1







— •-"







> Is:


•• ••


was then drawn upon the drawing of section 1 and the corpusclas ooincitlinf: in Ixitli sections noted and eliminated.

In my work this method was modified as follows. The work was done bj- the use of a Leitz-Edinger projection apparatus. A projected image of any given section, say section 1, was drawn and the Malpi^hian corpuscles indicated l>v circles (^hlack). When the drawing was finished, the next .section to be drawn, section 2, was projected on the drawing of section 1. Drawing 1 was then made to fit projection 2 and, this accomplished, drawing 1 was fastened securely. A sheet of carbon copying paper (similar to that used in typewriting) was then placed over drawing 1 and a fresh sheet of drawing paper placed over the carbon paper. Section 2 was then drawn on the fresh sheet of drawing paper, at the same time being automatically transferred to drawing 1 as a carbon copy in different color (blue). Now everything that coincides in the two sections also coincides in the two drawings, and the color of anj' isolated corpuscle tells to which section it belongs. By using this method, the Malpighian corpuscles were seen in the drawings in three ways:

(1) as double circles, one circle blue the other circle black:

(2) as single blue circles: and (3) as single black circles. All the double circles w'ere drawings of corpuscles which occurred in both sections and were not counted, since they had already been counted once. A single blue circle represents the first section of a corpuscle and a single black circle represents the last .section of a corpuscle. Hence, the total number of all the single blue circles in all the drawings woulil erjual the total number of renal corpuscles in the kidney. The total number of all the single black circles in all the drawings would also equal the total number of corpuscles in the kidney. The method, therefore, gives a check on itself in obtaining the total number of corpuscles in the kidney. In all, the two counts, for single blue and single black circles, were very close together.

Imperfect sections were avoided when«>ver possible. When a wrinkled section had to be u.sed, the resulting displacement of the drawings wa.s measured, the displacement of the renal corpuscles being established by the displacement of vessels and


section outlines. The result was also ehock('<l up under the microsenpe.

The pn)l)lpnj of determining the size and total volume of the renal (Malpighian) corpuscles was more diftieult. The range in size (diameter) was directlj' determined by using a Leitz filar micrometer. In each case, corpuscles which appeared to be the largest or smallest were followed through serial sections and measured to obtain their maximum diameter. The diameters of the largest and smallest corpuscles thus found an- given in table 5 (columns, a, b, d, e). representing in each case measurements from ten selected corpuscles. The cortex was arbitrarily divided into an outer antl an inner zone of equal width, and the measurements for each zone kept separate, as shown in table 5.

The average diameters of the renal corpuscles in the two zones were estimated by the modification of the paper method used by .Jackson ('17) in measuring the volumes of cells and nuclei. In every case, fifty corpuscles (rej)resenting all those occurring in a given area) from each zone were drawn in outline at a magnification of 150 to 200 diameters. The paper was the same as that used in measuring cortex anil medulla. The outlines of the corinisclcs were cut out, weighed, and the weight (in grams) reduced to area (square centimeters) by nudtiplying by 81.4 (as previou.sly explained). Dividing this area by the square of the magnification gives the actual area, and dividing again by oO gives the average area of a renal corpuscle, jis appearing in the sections. If the corpuscles be considered as spheres, this average area should represent two-thirds the area of the corresponding great circle (since according to the rules of solid geometrj' the volume of a sphere equals two-thirds of the volume of the circumscribed cylinder).

From the average great circle, the average tliameter and \()lume of the renal corpuscles for each zone are easily obtained, as shown in table 5, columns c and i. To calculate the average size of corpuscles for the entire kidney, however, two additional factors nuist be observed. In the first place, the corpuscles are in general more numerous per unit of volume in the outer than in the imicr zone. By counting the number of corpuscles ap

392 jonv A. KirrKLSo.v

pcariiiK in equal areas in the sections in the two zones, the average ratios were (letcrniincd, as jjiven in cohiinn g of table o. The average diameter of the renal corpuscles for the whole kidney, (-orrected for this factor, appears in column h. Finally, it is evitlent from the shape of the kidney that although the width of the two zones is etiual, the outer will he much larger in volume. Since it was not practicable to estimate this ditTcrence accurately, the final average diameter of the corpuscles for the whole kidney (tal)le 5, column i) was estimated arbitrarily as lialf way between the figures oiitained in table 5, colunm h, and the average diameter for the outer zone (column c). While notabsolutely accurate, this may be considered a fair approximation.

From the average diameter of the renal corpuscle, the average volume is easily obtained. This, multiplied by the total number of corpuscles, gives the total volume of the renal corpuscles in one kidney (table 5, column j).


1. Weights of the kidneys used

The gross body weight and the weight of both kidneys of the rats used (table 2) correspoiul fairly well with those of normal rats of similar weight and body length as worked out by Hatai ('13) and .Jackson ('13). The differences in weight between both kidneys in the rats used and in normal rats of the same sex and body length, as shown by the tables of Donaldson (,'lo), are as follows: The kidneys of the newborn rat are 0.003 gram above normal weight ; those of the one week rat 0.002 gram above; those of the two weeks rat 0.001 gram above; those of the three weeks rat ().03() gram below; those of the seven weeks rat 0.012 gram below; those of the twelve weeks rat 0.082 gram above; and those of the thirty-five weeks rat 0.227 gram below. These ditTcrences are comparatively small, with the exception of the adult at thirty-five weeks. . Even the latter is probably within the limits of normal variation. Thus the kidneys used may be considered as normal in weight, since the relations between kidney weight (and volume) to body weight at the various periods agree fairly with the established norms.

2. Growlh of cortex and medulla

The data showiriK the absolute growth in volume for cortex and inodulla of the kidiiey at the various apes arc pivcn in table 2. 'ihc sinnilicaiK-e of these figures may be more readily comprehended in terms of relative growth, as given in table 3.

The ratio between the volumes of medulla and cortex (table 3) is somewhat variable. In the newborn, the ratio is 1: 3.05; that is, the cortex is 3.05 times greater in volume. At oiie week, the ratio has changed to 1:2.91, apparently indicating a relatively smaller increase in the cortex. This continues at two weeks,


Relative growth im volumes of cortex and medulla in the kidney of the albino rat at various ages. Data derived from table i






INDEX or onoaa boot WEioirr (ORAUS) divided








V2 1





1 91







1 86


VI. 3





2 52

1 25







1 35

VI. 7







VI. 8







Kl I







the ratio then being 1 : 2.02. The cortex apparently reaches its minimum relative size (compared with medulla) at three weeks, the ratio then being 1: 1.50. At seven weeks, however, a relati\(^ly greater increase in cortical volume has changed the ratio of medulla and cortex to 1: 2.10, which remains nearly the same ( 1 : 2.0(1) at twelve weeks. The ratio in the adult rat is 1 : 2.S7, indicating that the relatively greater growth of the cortex continues during the final periotl.

The volumetric relations of cortex and medulla are similarly cxidcnl from llic ])ercentages which each forms of tlieir combined volume at the various ages (table 3). Thus the cortex decreases in relative volume from 75 per cent of the total at l)irth to tM)


per cent at three weeks, iiicrea-siiig afi;;i'm to 74 per cent in the adult. Tlio nu'dulla sliows a corn'spondiiif!; increase from 25 per cent in the ncwijoin to 40 per rent at three weeks, decreasing again to 20 per cent in the adult.

The relative growth of eorte.x and medulla may also he studied from a different point of view by comparing the volume of each with the correspondinfi total ixidy weight. An arbitrary index of their relative growth is thus obtained, as shown in the two columns of tal)le 2. In general, the growth index of the cortex appears more nearly constant, indicating that the in the volume of the cortex conforms more nearly to the growth of the body a.s a whoi(>. There is a slight relative increase in the cortex between birth and two weeks, followo(M)y a slight decrease thereafter. The medulla, on the other hand, shows a more varied relative growth rate. It doubles its growth index (compared with the entire l)ody weight) between birth and two weeks, continuing its relative increase to a maximum at three weeks. Thereafter it shows continuous and rather marked decrease, indicating that its growth lags behind in comparison with that of the whole body.

In tlie ailult I at thirty-five weeks) the ratios in the volumes of cortex and medulla, both in comparison with each other and with the whole body, have thus apparently returned nearh' to those found in the newborn. On account of the possil^lity of individual variations, no great emphasis can be laid upon the individual data, but the general trend of the relative growth rates in the cortex and medulla appears clearly.

Jackson ('V.i) has shown that kidn(>ys of the albino rat (postnatal) at first increase in weight more rapidl> than the body, increasing from ().{)() jier cent of the liody at birth to a maximum of 1.44 per cent at twenty days, decreasing thereafter to 1.03 per cent at ten weeks, and to 0.93 per cent at five months. The data from the i^resent investigation indicate that this early rapid increase in weight (up to thr(>e weeks) is due chiefly to growth of the medulla, the relative increase of cortex being less marked during the corresponding period. The later decrease in the relative weight of the kidney likewise appears to be due


chicfiy to decrease in the growth rate of the medulla, since the growth rate in the eortex is more uniform.

Miller and ("arltoii ('!)")) found the total cortical volume in the fresh kidney of the adult cat to be 12.5 cc. and the volumetric ratio between medulla and cortex to be 3:7 (or 1:2.33). In hardened kidneys the ratio was more nearly 1:2.

Schweigger-Seidel ('Go) found the weight of the cortex in an adult pig's kidney to be 102 grams (volume 1)9,000 cu. mm.) and the weight of the medulla 12..") grams. This gives a ratio of medulla to cortex in the adult i)ig of about 1 : S by weight. The pig has also been characterized by Peter ('09) as having relatively an extremely large cortex.

Toldt ('74), Kiiiz (■()<)), Hauch i'03) and Felix ("12) have studied the relationship of medulla and cortex in human material at various ages, making linear measurements of the thickness of medulla and cortex from cross and longitudinal sections of the kidney. The ratio of medulla to cortex (in thickness) appears to be four or five to one at birth and about two to one from the it^o of twelve years on. That is, the cortex appears to increase much more than the medulla. Hauch ("03) noted great increase in th(> cortical mass of the human kidney between birth and seven years. Felix ('12) states that the cortex grows regularly between birth and seven years, with but little growth in the meilulla; while after the age of seven years both ct)rtex and medulla grow eciually and double their diameters by the age of pulxrty. Such linear measurements, however, are not conclusive in showing the actual ratio between the volumes of medulla and cortex, and are apt to be mi.-^leading as an index to total volumetric or weight relations.

As a matter of fact, in work now in progress in an adult human kidney, 1 find the ratio between medulla and cortex to be about 1: 1.9 (by volume). The data concerning this kidney are given in the following table.


Obtrrvalioim on a uormnl kidney of an ailult irhite male, age about thirty-five yearg, who trns injured by a blow upon the head and died six hours later

Holy length 1('>" cm.

Body woiglit (osliiniitpd) 75 klliii. (170 lbs.)

WcikIiI of right kidney iit autopsy (Left kidney not weighed) 1 15 grams

Weight of cortex (after hardening) !K).2() grains

Weight of medulla (after hardening) 43.00 grama

\olume of cortex (after hardening) 93 cc.

Volume of medulla (after handening) 50 cc.

Estimated total number of renal corpuscles in entire kidney. . .1,040,000

After weighing the kidney at autopsy, the fibrous capsule was removed and the kidney was hardened l)y placing it for forty-eight hours in a 10 per cent formalin solution. The kidney was then cut in gross transverse sections two or three millimeters thick, llach section was then divided carefully with a fine scalpel along the border line sejiarating cortex and methilla. The renal pelvis, fat, vessels, etc., of the sinus renalis were then removed. After weighing the cortex and medulla (not inchiding capsule, pehis, etc.), tlie volume of each was determined by water disi>lacement in a graduated glass cyHnder. « 

Ip general, therefore, it appears that in the kidaey of the rat the volumetric ratio between medulla and cortex decreases from about 1:3 in the newborn to 1: 1.5 at three weeks, increasing again to nearly 1:3 in the adult. In the human adult kidney the ratio is about 1:1.9, the earlier relations being uncertain. In the adult cat the volumetric ratio of medulla to cortex is about 1:2.3 (IMiller and Carlton). In the pig the cortex appears relatively enormous, the ratio of medulla to cortex being about 1: 8 (by weight), according to Schweigger-Seidel.

3. Growth of the renal {Malpighian) corpuscles

a. Total tntmher of corpuscles at various ages. As shown in table 2, the total number of (fully developed) renal (Malpighian) corpuscles apparently increases from 10,465 in the newborn to 28,,S()3 at twelve weeks. In addition to these, however, ."lOdS developing renal corpuscles were observed in the newborn kidney, 691t) in the kidney at one week, about thirty at two weeks, l)ut none in process of development in the' older kidneys. This


soonis to show that ahhough the process of new formation had not yet entirely ceased at two weeks, it had beeonie very slow. Since a develo|)in}i; corpuscle is usually so large and distinct that it can hardly be overlooked when searched for, it is evident that the relatively sniall apj^arent increase in the number seen after the tliird week is probably due to individual variation. If this be true, the permanent number of renal corpuscles in the albino rat may be placed at about 27,S()(), which is the average of llie nunilicrs given in table 2 for the rats at three, seven and twelve weeks. In round numbers, this would lie about 2S,(M)(), the number assumed to be present in the adult rat (Kl.l. thirtyfive weeks) in which no actual count of the corpuscles was made.

Some data cited for compari.son are shown in table 4.

Miller and Carlton ('95) estimated the number of renal corpuscles in the cat's kidney to be lt),0()0, which appears far too low. Peter ('09) estimated the number of renal tubules for the cat to be 250,000 (2()(),(»(I0 to ;^()(),()(H») in one kidney. Schweigger-Seidel ('65) estimated the total number of renal corpuscles in the kidney of the pig to be ^OO.OOO. Huschke ('28) calculated the total number of corpuscles in the human kidney to be 2,1()(),()0(). Schweigger-Seidel ('(55) considered this estimate impossible, since the cubical content of 2,000,000 corpuscles, at an average diameter of 0.200 mm. would ecpial one-eighth the total kidney volume. Sappey ('8!') estimated the total numlier of corpuscles in one human kidney at 560,000.

In the work on an adult human kidney previoush' referred to, I have measured the volume of the entire cortex (by water displacenjent) and have counted the renal (Malpighian) corpuscles in mounted serial sections of five carefully measured pieces taken from different parts of the cortex. The same method was used as explained for the rat's kidney. From the average data thus obtained, I estimate that the number of corpuscles in the entire cortex of this one human kidney is about 1,040,000; or, in round numbers, one million. This number is intermediate between the estimates of Huschke anil Sai)pey. There is doubtless normally some individual variation in the number of corpuscles, even in kidneys of the same size.

The time when the renal corpuscles cease to appear in the kidney of the rat, according to the data above cited, is evidently some time during the third week after birth. A few (about thirty) newly-developing corpuscles appeared in the kidney at two weeks, but none at three weeks of age or later. This would place the date of cessation in the new formation of renal corpuscles slightly later than that given by Riedel '(74), who stated that in the animals born blind (dog and cat), the new formation of corpuscles continues during the first two weeks of postnatal life. Golgi (date?) states that in the dog, cat, rabbit, guineapig and man he has observed a continuous formation of renal tubules and glomeruli up to several days after birth. As to the human kidney opinions may be divided into two groups. According to Toldt ('74) and Felix ('12), the new formation continues during the first week or ten days after birth. According to llckardt ('88), Herring ('00) and Stoerk ('04). on the other hand, the new formation cea.scs during the late fetal period, about the eighth or ninth month.

In the following table of data quoted by Policard (09) there is a series of apparent errors. The names of Huschke and Schweigger-Seidel are evidently interchanged. A zero Has been added to the number of Malpighian corpuscles given by Miller and Carlton ('95). changing the number from Ki.OOO to 1C)0.000. Peter ('09) found the number of uriniferous tubules in the cat to be 200,000 to ;^00,000 instead of 1024, and I cannot find his estimate of 300,000 uriniferous tubules in the dog as quoted by Policard.

Table by Policard ('09)







llDIllIlilV . .

( 'hion

2,000,000 .5tM),000 300,000 500,000 160.000 1.024




I'orc .




Miller et Carlton Peter


I). Xiiinbcr of renal cnrjmsclcs per cubic rniUirncUr of corlex. The nuiiil)or of corpuscles jx-r ciihic inillinictor of cortex in the rat at the various ages is shown in table 2. There is a steady reiluction in the number per cubic millimeter from 1057 at birth to 7") at thirtj'-five weeks, considerinK only fully formed corpuscles. If the developinp; corpuscles are counted with those fully formed, the number jier cul)ic millimeter at birth is loSli.

These results are in general agreement with those of Waschetko ('14), who counted the number of corpuscles in a microscopic field of known area, antl estimated therefrom the number per cubic millimeter in the albino rat. He found a decrease in the number of renal corpuscles per cubic millimeter of cortex from 1125 at birth to S14.5 at one week, 70().5 at two weeks, 1017 at three weeks, 571.5 at four weeks, and 5(57 at five weeks. Aside from the exceptional figure at three weeks, the general decrease is evident. Eckardt ('8S) first noted this reduction in the human kidney, where he found, in the same microscopic field area, an average of 122.1 corpuscles in the newborn and onlj^ 9.35 in the adult. Schweigger-Seidel ('65) estimated about 5 renal corpuscles per cubic millimeter in the renal cortex of the adult pig and about in man. Thus in the pig and man the renal corpuscles are apparently far less abundant relativelj^ than in the rat.

/vi»"-vteady in the number of renal corpuscles per cubic millini,..tpj. ^f cortex from birth to adult life is due partly to the gi-owth of tij-o corpuscles themselves (as will be shown later), but chiefly to the gro ^^.^j^ ,^,„j ,>xpansion of the convoluted tubules, resulting in a wider sei ,.^ration of the corpuscles.

c. Diavieler and wlunu^ ^j ^f^^ ^^^^^ corpuscles. The diameters of the renal corpuscles of t.v^^ ^^^^^^^ ^^^^^ ^^^^j^j^ 5 polumns a, b, c) show l)Ut little change ^l"i jng the first two weeks, while new corpuscles are forming. Tl. ^^ maximum diameter shows (in this series, at least) an api)an ^^^ decrease from 78 to 70 micra. The minimum diameter remam. ^t.^,i„„.^ry at about 40 micra, representing the diameter of th.^ j^^^^.,^. developing corpuscles during this period. The average diameter of the corpuscles in the outer zone thus shows an app^j.^^^^ decrease from 59 to 49

anti .")1 micra, prulmbly <lut' to a relatively greater nuinlxr of small corpuscles present at these periods. From three weeks to the adult (35 weeks) there is a constant incroas(> in inaxiniuin diameter (8U to 140 micra), in mininnnn diameter (.")1 to I'il niicra) and in average diameter (56 to 127 micra). If we compare the number of corpuscles in a field containing eciual areas of the outer and inner zones of the cortex (table o, column g) there ajipears a continuous decrease, however, in the relative number of corpuscles in the outer zone area from 57.5 per cent at birth to 50 per cent in the adult. That is, the corpuscles are more closely crowded together in the outer zone, which contains the younger corpuscles, a uniform distribution not being attained until the adult stage is approached.

In the inner zone of the cortex, the maximum, minimum, and a\erage diameters of the corpuscles in the kidneys of the newborn and on(> week rats are accordingly considerablj- larger than m the outer zon(> (table 5, colunms d, e, f). The maximum diameter in the inner zone shows an apparent decrease from 104 to 89 micra between birth and two weeks, and, in the same period, the minimum diameter decreases from 5(i to 51 micra. The average diameter accordingly decreases from 73 micra at birth to 5S micra at two weeks. From two weeks on they all; the maximum diameter increasing from 89 micra at three weeks to 140 niicra in the adult, the mininium from 51 micra at three weeks to 123 micra in the adult and the average diameter from 58 micra at three weeks to 127 micra in the adult. As com])ared with the outer zone, the relative number of corpuscles in the inner zone (in a field containing ecpial areas of each zone) shows an from 42.5 per cent in the newborn to 50 per cent in the adult.

'1 he average diameter of a renal corpuscle in the entire cortex (table 5, colunm h and i) is obtained from the preceding data, with corrections as explained previously under "Material and methods." This average diameter likewise shows a decrease during the first two weeks from (14 micra to 54 micra (column h) or from 02 micra to 53 micra as finally corrected (colunm i).


From the second week onward there appears a continuous ineroasc up to anavcraRO of 127 micra in the adult (at :}5 weeks).

Althoiif^li the average size of the renal (•or|)uscle decreas<*s sli}ihtly during the first two weeks, the total volume of the forpusclcs shows a steady increase during this time (table o, column j ). The smaller average size of the corpuscles is evident!}' more than counterhalanced by the increase in number of corpuscles between birth and two weeks of age. The increase in total vohnne of the corpuscles continues, increasing from 1.29475 cu. nun. at birth to 2'.I.S950U cu. nun. in kidue}' of the adult rat (at thirty-five weeks).

Some data for comparison are cited in table 4.

Miller and Carlton I'Oo) found the average diameter of the renal corpuscles in an adult cat to be 102.7 micra, which is somewhat less than that found by me for the rat (127 micra). Owing to their extremely low estimated numljer of corpuscles, their resulting total vohnne of the corpuscles in the cat is only 9.04 cu. nun. Peter ('07), hpwever, found a somewhat larger average diameter (124 micra) for the corpuscles in the cat's kidney, which is nearly the same as my average for the rat. He estimated the average number of coriniscles in the cat at 250.000, however, corresponding to a much greater total volume of the corpuscles (141.37 mm.).

Eckardt ('88) found the a\erage diameter of the human renal corpuscles to increase from an average of 84.77 micra in the newborn to 195.80 micra in the adult female, and to 213.49 micra in the adult male. Ktilz ('99) in the cortex of the human kidney at birth found the average diameter of the corpuscles in the inner zone to be 138 micra, while those of the outer zone averaged only 99 micra. Those of the inner zone showed no increase in size up to two and one-fourth years of age, but those of the outer zone during this time increa.sed to the same average iliamcter, 13S micra. He found the average diameter in the adult to be 23S micra.

Schweigger-Seidel ('05) estimated the average diameter of the corpuscles in the human adult kidney at 200 micra. I'sing the two estimates of the total number of corpuscles in the human


kidney, Ihischke's ('28) of 2,10(),()()() and Sappey's ('89) of 500,000, \vc obtain total volumes respect ivoly of 87i)().4S and 2:^4"). 72 cii. nun. If the nuuil)er of corpuscles as estimated 1)V me (1,04().(KI(>) is used, a total volume of 4350.35 cu. mm. for the corpuscles is obtained.

d. Ralio of the total roliimc of the renal corpuxelca to that of the renal cortex, and to the total kidney voluine. As seen in table, 5, the ratios of the total volimie of the renal corpuscles to the total volume of the cortex and of the whole kidney (cortex and medulla) are somewhat variabl<\ Thus the ratio between corpuscles and cortex varies from 1 : 7.0 in the newborn to 1: 17 at twelve weeks. In other words, the percentage of cortical \-olume formed by the corpuscles \aries fmm a maximum of aliout 13 per cent in the newborn to a minimum of about per cent at twelve ' weeks. At the other ages, the results are intermediate. In the adult rat (thirty-five weeks) the corpuscles form about 8 per cent of the cortical volume.

The corresponiliiifi ratio between total volume of corpuscles and total kidney volume (cortex and medulla) varies from about 1 : 10 (10 per cent) in the newborn to 1 : 25 (4 per cent) at twelve weeks. In the adult rat, the ratio is about 1:17, the corpuscles forming about per cent of the total kidney volume.

It is therefore evident that in total volume the renal corpuscles are relatively greatest in the newborn, the older stages showing an irregular decrease in their relative volume. To what extent this iiTcgularity is due to individual variation it is of course impossible to say.

.•^ome data for comparison are cited in table 4.

In an adult cat, Miller and Carlton ('95) found the ratio between total volume of the renal corpuscles and the total cortical volume to be 1 : 990 and the ratio between total volume of the corpuscles and total kitlney volume to be 1: 1427. extreme ratios are due to their low estimate for the number of corpu.sdes. Peter ('07), however, as before mentioned, estimated the total volume of the renal corj)uscles in the cat to be 141.37 cu. mm. If we use this estimate, together with the total cortical volume and total kidney volume as given by Miller and


Carlton ('95), \vc find the ratio between total volume of corpasdes and total cortical volimie to l)0 1 : (»4, while the ratio between total volume of corpuscles and total kidney volume becomes 1:1)1. These ratios are more nearly comparable to the corresponding ratios for the rat, though still verj' high (see tables 4 and 5).

In order to make a similar estimate for the human kidney, it is necessary to have data for the total cortical volume and total kidney volume in the normal adult kidney. Accohling to my observations on one kidney, the human renal cortex mca.sures about 93,000 cu. mm., the total volume of the corresponding kidney (cortex and medulla) being 143,000 cu. mm. (table 4). Taking Schweigger-Seidel's estimate of 0.200 mm. for the average diameter of the renal corpuscle, the ratios between total volume of renal corpuscles and total cortical volume will vary according to the estimated total number of corpuscles. The ratios accordingly are 1: 10 for Iluschke's ('2S) estimate; 1:39 for Sappey's ('89) and 1:21 for my own estimate. The corresponding ratios between total volume of corpuscles and total kidney volume become 1:10 (Huschke); 1:01 (Sappey) and 1:33 (^Kittelson). It may be further noted that Buschke's data give a ratio of 1 : 10 between total volume of renal corpuscles and total kidney volume instead of 1 : 8 as stated by SchweiggerSeidel ('05).

Even excluding the aberrant data of Miller and Carlton, it would appear that the relative total volume of the renal corpuscles, as coui[)are(l with total cortical or kidney volume, is in the rat much nearer to the human than it is in the cat.


A careful volumetric study of the cortex and medulla, together with complete numerical counting of the renal iMalpighian) corpuscles, was made upon serial sections of the kidney of the albino rat at birth, one, two, three, seven anil twelve weeks of age. The volumes of cortex and medulla were also meju^ured in an adult rat (thirty-five weeks). The volumes of cortex and medulla were also measured, and the number of corpuscles


cstiniat('<l. in an adult liuman kidney. Tho principal results may ln' sunimari/i il liricliy as fdihuvs.

1. Tlic postnatal urowth of the cortex in the rat's kidney is fairly uniform, showing in comparison with the entire body a relative increase between l)irtli anil two weeks of age, decreasing slightly thereafter.

2. The growth of the medulla in the rat's kidney is more varied. Its rate of growth between birth and three weeks of age much more rapidly than that of the cortex or of the entire bod}', anil decreases more rapidly thereafter. The characteristic curve of relative growth for the entire kidney, increasing to a maximum at three weeks and decreasing therefore is therefore apparently duo largely to the varying rate of growth in the mcilulla.

'^. The volumetric ratio of medulla to cortex in the rat's kidney changes from 1 : 3.05 at l)irth to 1 : 2.91 at the age of one week; 1:2.02 at two weeks, 1: 1.50 at three weeks, 1:2.10 at seven weeks and 1 : 2.00 at twelve weeks, and 1 : 2.87 in the adult. Linear measurements are inadequate to indicate the volumetric relations.

4. Tlie total number of fully formed renal corpuscles found in the kidney of the newborn rat is 10,465, or 15,533 including those incompletely formed. The number to 10,()S2 at one week, or 2(1,508 including those which are not fully formed. At two weeks there are 24,0()1 (plus about 30 incompletely formed); 25,030 were counted at three weeks; 28,683 at seven weeks and 2S,S(i3 at twelve weeks. .

5. The total numlier of corpusch^s in the rat's kidney is apparently reached during the third postnatal week, as but few coulil be found in process of formation at the age of two weeks and none at three weeks or thereafter. The apparent slight increase in the total munber of cf)rpuscles at seven and twelve weeks is therefore probably due to individual variation.

(). Tlie niinilxT of fully formed corpuscles per cul)ic millimeter of cortex in the kidney of the rat decreases from 1057 (or 15St), including the incomplete) at birth to 131 at twelve weeks, and to 75 in an adult at thirtv-five weeks.


7. The avorapc diaineter of the renal eorpuscles increases from ahout (i'i iiiicra in the nowhorn to about 127 niiera in the adult lilt. Tiicre is an apparent dcerease in the average sizr of the corpuscles at one and two weeks, due perhaps to a larger proportion of the small young corpuscles at these periods.

S. The total (relative) volume of the renal corpuscles forms alxnit 13 per cent of the cortex, or 10 per cent of the kidney, ill the iicwhnni. Later the relative volume of the corpuscles decreases irregularly, forming at twelve weeks only al)out (i |)er cent of the cortex or 4 per cent of the total kidney (both cortex and medulla).

9. In an adult human kidney, the cortex measured about 93,000 cu. nun. and the medulla 43,000 cu. mm. The number of corpuscles was estimated at 1,040,000, with a total volume of 435t).3.") cu. nun., forming 1 21 of the total cortical volume of 1 33 of the total kidney volume (both cortex and medulla).


Donaldson, Henry H. 191.') The Rat. .Memoirs of the Wistar Institute of

Anatomy, Xo. 6, I'liiladolphia. 1'x:kardt, C. T. 1888 I'ebcr die compensatorische' Hypertrophic und ila.s

physiolopische Wachstum der Xtbro Arch. f. path. .Vnat.. cviv, S.

•217-21.5. ' ' '

I'ki.ix. W. I(J12 The (ievclopinent of the urogenital oreans. Kcibcl and

.\Iair.s Human EinbryoloRV, vol. 2. OoLGi, ('. (undated) In Schcnck's Elementi di istologiu normale dell 'uomo.

Italian translation bj- Monti and (loljfi. Hammau. l.A. 1911 .Mot hode. die Monpe dcr Hindcunddcs Marks der Thymus,

sowie die -Vnzahl und die (Irosse der Has.sarachen Kiirper zahlen ma.«!sig fc,>!tzu.slellen. Zcit.schr. f. angew. .\nat. u. Koiist.. Bd. 1. II.4-.1.

Merlin. IIatai, .S. 19I:{ On the weiRht of the abdominal and thoracic viscera, the

sex Kland.s. ductless glands and the eyeballs of the albino rat (.Mu.s

norvegieus albinus) according to body weight. Am. Jour. Anat..

vol. 15, no. 1. IIaicii. K. 1903 I'eber die .Xnatomie und Kntwickelung der Nicrc. .Vnat.

Ilefte. Hkuuim;, p. J. HKX) Development of the kidney ai^ its relation to pathological

changes which occur in thcin. Jour, of Path, and Bact., vol. 6, p.

I.i9. Ile.HcnKK. 1S2S I'eber die Textur der Nieren. Isis, vol. 21, S. .ioO-oW. (Cited

by Schwcigger-Seidel.)


Jackhon, C. M. 1913 I'oKlnatiil Krowdi and variability of the body and of the

various organs in thi- alljino rat. Am. Jour. Anat., vol. 15, no. 1.

1017 Effects of inanition and refccdinK upon the Rrowtli and structure

of the hypophysis in the albino rat. Am. Jour. Anat., vol. 21, no. "2. KVlz, Li'DwiG. 1899 rntcrsuChunRcn iiber das postfotalc Wachstum der

mcnschlichen Nicre. InauR.-Diss. Kiel. Miller, W. S. and C.milton, K. V. 1S9.5 The relation of the cortex of the cat's

kidney to the volume of the kidney and an estimation of the number

of glomeruli. Transactions of the Wisconsin .Vcademy of Science, 10,

pp. 525-">3;i. Peter, Kakl 1907 I'obor die Nierenkaniilchen des Menschen und einiger

Stiugetiere. \'crli.'indl. d. anat. Ges.. 21" Vers. Wurzburg. pp. 114-124.

1909 Untcrsuchungen iiber Bau und Entwickelung der Niere. Policard, J. 1908 Le tube urinaire des mammifdrcs. RiEDEL, B. 1874 Kntwickelung der .Siiugethierniere. Untcrsuchungen aus

dem anatomischen Institut zu Hostock. Sappev, C. 1889 Traitc:- d'anatomie descriptive, T. 4, p. 499. Schweiooer-Seidel, F. IStio Die Xieren des Menschen und der Siiugetiere

ini ihreni feinerern Bau. Halle. Stoerk, O. 190-1 Beitrag zur Kenntnis des Aufbaus der mcnschlichen Niere.

Anat. Hefte, no. 72. ToLDT, C. 1874 Untersuchungen iiber das Wachstum der Nieren des Menschen

und der Siiugetiere. Sitz. Ber. d. k. Akad. d. Wiss. Wien. 69 Bd.,

3 Abt. Aprilheft. Waschetko, N. p. 1914 Zur Frage von dem physiologischen Wachstum der

Niere. Centralblatt allg. Pathol., Bd. 25, no. 14.





R. S. OUTSELL From the Department of Histology and Embryology, Cornell Universily


Among the slides in use in the Department of Histology and Embryology of Cornell University, Ithaca, N. Y., my attention was caliod some time ago to sections of sciatic nerve from a newl)()ni ciiild, in which were cells apparently of the hematopoietic series in the interfunicular connective tissue. A very little study inuler high magnification shows bej'ond a doubt that here in the connective tissue surrounding and very doseh' associated with the smaller vessels of the nerve, certain phases, at least, of blood cell formation were in progress when the tissue was fixed.

A number of questions immediately come to mind. What blood cells are being formed? From what are they derived? Are the hemoblasts arising from local mesodermal cells, or are they 'sown' by the blood vessels? Do the cells e\er get into the circulation? To what other blood-forming organs are these groups comparable? What special condition caused these to go on?, etc. Of course all of these (|uestions may n<it be answered fully and surely. The preparations themselves stand in the way of this, and also the fact that vcrj- little is known concerning the individual from whom the tissue was taken, liut even with the little which is positive, the probabilities and jK)ssibilities in the >;hould be of sufficient interest to justify a .short description.

As above mentioned, the child died at l)irth. The data card shows no account for the tleath. but a remark, added after some


410 K. s. (irTs»;Li.

of tlio tissue had hopii cxiuiiiiicd sunncst-^ tho oxistcncp of leukoniia. Ilxaminations of (issue at liaiid shows Utth' or no evidence of this effect. Sections of the Uver, I am told, show acute parencliyniatis hepatitis. In preparations of trachea and esophaRus, bladder, and (liyn)us, no pathological condition was noted. There is no data to tell what part of the nerve wa.s sectioned, but as a good-sized artery appears attached to it in many of the sections, it would seem that this is the inferior gluteal arterj-, and that the section of the nerve which was prejiared was taken close to its pa.s.sage through the sciatic notch. Other portions of the nerve have been recently sectioned, but as little or no hematopoietic tissue appears, one concludes that the region in wliicli tlie activity was going on is restricted to only a small portion of the nerve. However, the fact that similar groups of cells are found surrounding intrinsic vessels of the externa of the near-by artery, and also, to .some extent in the adventitia of the bladder leads one to the belief that the hematopoiesis in the connective tis.sue is genend and not nien>!y of local origin.

In the nerve the connective tissue is abundant and not very dense, so that the connective tissue spaces between the nerve bundles are large. Practically no fat appears, and the blood vessels are more abunilant, or at least more in evidence that in similar preparations from other indi\iduals. In close connection with the smaller of these vessels (arterioles, capillaries, and vemdes), are found the groups of blood-forming cells. One notices two characteristic appearances to these grouiis which is striking, although tran.sitional forms are abundant. One is small, compact, and confined to a small (round) area, closely surrounding the small vessels, and here in .some the uniformity and arrangement of the cells would at first lead to the belief tliat only ]yini)hocytes are being produced. The other extreme shows the cells diffused out into (he connective tissue .so that a large interfunicular space is lUled with cells, densely massed around the blood vessels and scattered through the connective tissue at the periphery. Here the cells are always of a variety of types whose character and arrangement will be dis


cussed later. AlthouKh the soptions are not serial, it would seem tliat these groups extend longitudinally along tlie course of the vessels.

After a good deal of study I have come to the condu.sion that only <ine process is going on here to any noticeable extent, and that the dilTerently appearing groups inerelj' represent different stages in this process, namely erythrogenesis. .As all the larger groups and some of the smaller ones contain erythrohlasts, and there seems to be no production of granulocytes, the presence of erythroblasts and their degree of differentiation gives some kind of a clue as to the stage of the separate groups of cells, and what seems to be the relative age' of the groups as well. Whether or not the stem cells migrated from the blood vessels, 1 beUcve that very few erythrocytes or erythroblasts find their way into vascular channels, and as they have not been observed degenerating in the tissues, it is fair evidence for believing tliat the process has not l)een going on long. .\s 3'oung stages are ahvaj'S found, there is no reason for beUeving that the activity has cea.sed in any of the groups observed.

The tissue was prepared under the direction of S. H. dage in 1901 or 1!)()2. It was fixed in .MiiUer's fluid, and afterward put for a short time in Hermami's fluid. The .several pieces were all imheilded in paradin. The most favorable ])()rtion, 1)V mi.sfortune, was sectionetl for class use and ver^- well stained with Dela field's haematoxylin, and eosin, However, some of the ti.s.sue I have cut, and .stained with eosin-methylene blue blood stains. Nuclear detail is naturally not well preserved.

There are a number of different types of cells present. The smaller groups in which there an^ no late erythroblasts or erythrocytes present, contain two tyi)es of cells which are found to intergrade. One resemijles a small lymphocyte, with tlark staining nucleus. The other type shows a large, pale basophilic nucleus with a narrow rim of cytoj^lasm. latter are at first small with basophilic cytoplasm, but in the older groujis where definite ery(hn)l)lasts arejiresent. they are larger and their cytoplasm is acidophilic to a varying degree. In the rather l() connective tissue aniuiid the ])eriphery may often be found


fil)robl;i.sts whose proci'sses arc appiireiitly Ijciii^ iwii in and whoso cell bodies are roundinR up. This might nie^that they were contributing to the formation of the group, or at some agent emanating from the l)lood vessels caused thei to so transform. One can easily imagine the small cells to o'elop from the modified fibroblasts, but the cells are not pressed well enough, or the stains specific enough to definitely deterine this. Often the appearance would lead one to the opinion tR the mother cells were derived from the blood vessels, so closcb is the group formed around their walls.

.\t any rate the begimiing seems to he the small cells, for they are numerous in the smaller (younger; groups, where early hemoglobin-containing cells are absent, and hardly to be found in the older stages. Forms transitional to the large cell type are very evident in the younger groui^s. Some groups are found consisting almost entirely of these larger cells, with perhaps one or two of the smaller type and a few in which the cytoplasmic margin is becoming thicker and is seen to contain hemoglobin, and whose nuclei are smaller and more darkly staining. Other places show every degree of change between the large cells and typical erythroblasts.

The larger groups are irregular and contain various transitional forms from the large cells to adult erythrocytes, although the latter are not very conspicuous or abundant. Around the periphery fnow much farther from the vessels) may still lie fovmd the rounding-otT fibroblasts. The hovmdaries of these jiroliferating centers are not well defined, the cells occupying spaces between connective tissue fibers. The different cell types are found nearly everywhere. It is interesting to note, however, that the large cells (erythrocytoblasts?) are most numerous (|)roliferating most) in close relation to the blood vessels, while the definitive erythroblasts and erythrocytes appear to be crowded out into the periphery of the cell group.

The production of erythroblasts and likewise erythrocytes is somewhat irregular, for the former vary greatly in size, most of them being larger than normal, while the erythrocytes found in the connective tissue spaces are as a rule larger than those in


(lie iicinliboriiiK blood vessels. It would seem that thi.s wholi' process of red i)lood cell formation is still at an earl}' stage or that it is somewhat aljortive, for if it was progressing to its normal conclusion, noii-nueleated red cells woukl be more abundant. Xowlicre, however, docs there seem to be a supcrabwiidance of (i\ throblasts; nor does it seem that the mature erythrocytes

Fig. 1 YouiiK Rroup showing compact arrftngeiiicnt around small vessels. -Vpparcntly later stage than shown in figure 4, but no cells ol type 5 (see fig. 2) are present. This is detail of group 1) figure .3. Fb, fibroblast; Ji.V., small ves.scl; ('n/j, cai)illary; Ec, erythrocyte in blood vessel. For cell ty|x?8 1, 2, 3. and 4, .see figure 2. 3ii, form transitional from type 3 to 4.

Fig. 2 ;. fell type found in early cell groups and sometimes in margins of older stages. Nucleus dark, cytoplasm does not slain easily, apparently quite amoeboid. These cells very much resemble the transforming fibrol>lasts (figs. 1, 4, and 5). £. Cell similar to /, but found more udundantly in early cell groups. .Vlthough no line can be drawn between them, this would represent a cell of more regular character, with less cytoplasm and presenting a less amoeboid apjiearancc (figs. 1, 4. and 5). S. Type found abundantly in early groups (figs. 1 and 41. Forms transitional to i are easily found, (fig. 1, Sa). 4- Cell with large pale nucleus and scant amount of slightly basophilic cytoplasm. .Vbundant in figures 1 and I, and transitional to definite erythrocytoblasts (.51 so extensively found in

ill later or 'mature' localizations of erythrogcnctic activity described here, ifigs.

.") and ti). 5. Krythrocytoblast. Nucleus similar to that of numlier 4- I'ells somewhat larger. Cytoplasm variably acidophilic. 6. Transitional to definitive erythoblast. Nucleus more darkly staining than the former, cytoplasm markedly acidophilic. 7. Large erythroblast. S. ."'mall erythroblast. 9. Krythroblast with irregularly divided nucleus, which suggests the cxtrustion of the smaller portion. Other somewhat similar cells were observed. 10. Krythrocvtes found in the connective tissue. //. Krvthrocvte from blood vessel.

were fiiidiii)! their way to the blood fhanncls, for in wi placv have I found the vessel walls to he ineoiiiplete, and the older cells an; apt to be found somewhat removed from the vessels.

Although nothing is eonelusive, there is a probability that the lilir()i)Iasts foiitribute to the formation of these erythropoietic fell groups. The}' without doubt are transformed morphologically, and the query arises, into what are they developed. In at least one place, cells of the larger type are developing between

Fin. 3 Tho whole nerve is shown, although the (inferior gluteal) artery, wliicli occupies a position on the left, has been omitted. The abundance of loose connective tissue around the nerve bundles and their sheaths, should be noted. The comparatively large arteries entering the nerve at this point may account for the abundant supply of smaller vessels occupyinR interfunicular po.sitions. The latter are two small to be easily demonstrated in this drawing, althoudli their position in most cases has been indicated. .\t A, note the large );roup of hematopoietic cells, and their proximity to an artery of considerable size. This proup is shown in some detail in figure C. U, (', and I) show less extensive accumulations of hematopoietic cells, similar in their size and arrangement to figures 1 and .3.

Fig. 1 Here is represented a very early group of cells. The close relation to tho small vessels (,BV and Cap) will be noted as well as the space in the connective tissue in which they arc found. Normal fibroblasts arc shown in the tissue surrounding the group {Fb), and also various stages of transforming fibroblasts (Fb'). Most of the cells present are typified by large pale nuclei and a narrow rim of slightly basophilic cytoplasm. (,ty|>e 4). Cells with smaller darker staining nuclei are also shown (/ and 2). One erythrocytoblast with acidophilic cytoplasm is seen (5). Xo definitive erythroblasts arc present.

Fig. 5 A stage later than the one in figure 4 is shown. Erythrocytoblasts predominate. These grade evenly into definitive erythroblasts tEbi. The nuclei at first show a deeper staining power. As this increases both cell and nucleus become more regularly spherical, .\ccompanying the above changes the nucleus becomes smaller, while the cell remains the same size, and hemoglobin becomes concentrated in the cytoplasm. Most of the erythroblasts arc about lialf the volume of the larger ones. .\ division is implied, but goixl examples of division stages have not been found. Xo erythrocytes are present here, and the size and character of the group shows the stage to In- comparatively early.

Fig. r> Region showing an advanced stage. Detail of group shown at .4 in figure 3. Xote the abundance of erythrocytoblasts in the region surrountling the small blood vessel. Here definitive erythroblasts arc comparatively few, while farther out in the connective tissue they are the most abundant cell form. Krytlirocytes (AV) are seen scattered through the connective tissue. They are considerably larger than tlio,se in the blood ves.^els. Xerve bundles occupy positions at the left. top. and right. .\ very small strand of nerve fibers is shown in the lower left liaiid corner (A'), and beside an artery is seen (Art).

4 111 K. s. CiLTSELL

the definitely arranged eoiuieetive tissue fibers of the perineurium, ami are oval, as if formed from libroblasts without in any way ehaiifiing their position. Normal fibroblasts in this region are absent, except at the margins where they seem to intergrade.

Only four or five polymorphotuielear eells Cprobablj- neutrophiles) antl two eosinoi)hiles were ol)served in all the seetions, and their presence seems of little significance.

No clue has been found regarding the reason for this abnormal localization of erjthropoietic activity.

Of the well established seats of hematopoiesis, the relations are most comparable with the localizations in the fetal liver, which is recognized by Maximow to be within the connective tissue and is largclj' erj'throgcnctie. The bone marrow differs greatly in that the acti\'ity is granulogenetic as well as erythrogenetic, and this, together with whatever other processes may be going on, makes the picture very complex, instead of simple. If in the case here described, the stem cells are derived from mesodermal cells in loco, the comparison with early formation in the mammalian yolk-sac might be made, but here the mother cells would be fibroblasts, not mesenchyme, or the less distantly diiTerentiated reticular cells.

The most direct comparison I have been able to find in mammals, is the condition in the interlobular connective tissue of the fetal thymus described by Badertscher.' Even here the process of erythrogenesis is accompanied by granulogenesis. It may be of some interest to note that, contrary to the general rule, no megakaryocj'tes were found.

"lymphocytes' present I believe to l)e lymphoitl hemo\so that these cells do not correspond in any direct way to the lymp*;hocytes of the blood, unless we concede the latter erythro, or "inolygenetic potentialities.

I have fou\^nd no reference to blood formation within a nerve. That it shoulc, J in this case have localized here seems to be merely because, with the general stinuilation to erythrogenesis, the

' J. A. Radcrtsclu r. 19I.'). Development of the thymus in the pig. II. Histogenesis. .Vni.Jour. .\ nat., vol. 17, no. 4. pp. 437-J.87.

^'Uie -1 blasts^, sr


coiulitions here were favorable. The blood supply is rich, the fdiiiioctive tisFue spaces are ample, and there is no interferenrr' wilh any rapidly growing organ. What iniglit lie found if various parts of this body could have been examined might very possibly show similar localizations in similarly favorable locations.




P. E. LI.NEBACK Atlanta Medical College


Probably the most beneficial part of a student's laboratory drawings is the close sttuiy he applies to minute stniclures in order to correctly reproduce them on paper. These two factoi-s of insix-ction and reproduction nive rise to the problem of how detailed his work should be. When,has he .«een the e.^^sentials in a given piece of work and when has he made a good drawing of them, arc the gist of the problem. Usually, as in a general course in embryologj", he is confronted by a great amount of detailed material, numerous textbook figures many of them elai)orately overdrawn, anil very limited time in which to cover the course. Tliis means that the instructor must give him constant and special attention or he is bewildered. It is imi>o.ssible for an in.structor to give each student this amount of personal oversight consequently many students may go ((uite through the course before they have the sligiitest ray of iiglit upon what they are doing. Even with the best efforts of the instructor many of the .sections are selected at random and the result is a collection of drawings disconnected and imrelated. Often they adhere too closely to textbook figures rather than to original studies, and represent a vague and incoherent grasp of the subject on tiie part of the student. In trying to overcome some of the.-^ difficulties the writ(>r has gradually worktnl out a scheme which he finds to be ([uite satisfactory as a definite, uncomplicatt'd, an<l rounded otit plan. The details of the scheme are not given here but oidy the principal points wiiich it is hoped may be of u.>ie to others.

.\fter the student has studied the introductory subjects, primitive cell, germ layers, and foetal membranes, and has gone over more in detail his chick material, whole mounts and ,<erial sections, he begins a detailed study of pig and human embryos by systems. His first step is to make a '.areful study and drawing, from his own specimen, of the lateral sm-face of the pig. In this he gives special attention to contour lines of the body, and the general position, shaiM^, and relationship of the larger structures such as branchial clefts, eye, heart, liver, mesonephros, and limb buds. Then he makes a similar drawing of a mid plane .sagittal section in which he again empluisizes the general aspect of the structures. These two drawings are finished in detail


420 p. E. LINEBACK

only to the extent that llicy show tlie general position, shape, and relationship of the important structures, consequently thov are more easily un<lerstood and arc of greater value in explaining his cross sections later. By a glance at these drawings he can easily see why he may hav(> a portion of the fore brain and hind brain in the same section, and why the hi-art's apex may appear in a section with the liver. For making these and all other drawings he use<l a simple projection box, a modification of Bcgg's apparatus (.Vnatomical Record, '15) and by this means his work is not only more accurately done but is easier for him.

He now takes up the study of the cntodermal or digestive trsict. This sj'stem is used as a center or buikling line about which all other structures are studied and drawn. Earlier attempts at the formation of this plan made use of other .systems as the building line, but were rejected in favor of the eiitodermal tract. The nervous .system came more nearly giving the desired results, in that it extends through the entire series anil is easy to iilentify and study, but on strict test it proves to lie more difficult for the student to grasp its details, especially in the brain region, and its uniformity througiiout th^rest of its length does not permit of .selecting ilefinite important levels for study, such as are afforded by the intestinal tract and its diverticulae. His procedure is to trace the tulie down the scries and note certain definite places where important changes or developments are discovered. Beginning with the toiigu(! he locates the distinctive part of the organ, (he tubercuhun impar, which he finishes in detail with any other .sirucdnes connected with the iligestive mechanism. ]w!ate, pharyngeal pouches, etc. The section in which he foimd the tuberculum he draws in outline, except the structures just named, and mnnl)ers it iis it comes in the series. Next he finils the section showing the pharynx jiroper where it app<'ars as a slit anil draws it in outline, nundieiing it .serially, but finishes the phar>-nx and related structures in detail. Tlien he stiulies the tube on down to where the tracheal diverticidum apfxvirs and rejieats the procedure. In this way he contiinies the slu<ly of the tube, sto|)ping at different places and making the outline of tile sections, numbeiing them serially, but working out in detail all cntodermal structures. The jilaces he selects after the above named are, the right apical lung bud. the bifurcation of the trachea (fig. 1), the oiilii|uely placi'd slom:ich (fig. 2), the entrance of the dorsal pancreatic iluct into the duodemnn (fig. '.i) the intestinal loop in the body stalk, point of exit of the allantois fmm the cloaca, point of entrance of me.sonephric ducts in(o cloaca, and tail gut.

.\fter the completion of the digestive tract he .studies the circulatory system. Tor this he <loes not make a new .set of drawings, but goes over the outlines in which he studied the cntodermal tract and draws the blood vascular mech.aiiism as it appears in each. In the tongue section he draws just sui'h asjiect of carotid arteries and jugular veins that appear here. Then in the next section where he .studied the pharynx he again draws the ves.sels which appear, and .so on successively to


the last section, coinpletinK in (ictail the blood vessels in each (sec duplicate A). Ho next pocs over the urogenital system after the same manner, selectinR the sections where he first studied the digestive tulx;, and later the circulatory system, and fills in the details of this system in each drawinn (see duplicates b). He follows this scheme in studying all the rest of the s\stems and structures, nervous system, ductless glands, face, limbs, until his drawings contain the whole of the body. In all the systems he must study carefully his own series for no textbook figures will show him structures properly adjusted to his entodermal drawings with which he l)egan.. If there are important structures or places in any system which are found in sections coming between any two given sections, he draws those features as inserts, placing them at the side of an adjacent figure (see dujilicate hi.) In his final study of the nervous system he will have several new .sections of the head end to draw which may now be more easily selected bj' working forward from his starting point, the tongue. These .sections are the seventh-eighth nerve mechanism, the origin of the third, and the origin of the fourl h nerves.

The advantages of tliis scheme may now be briefly stated. The student has a definite plan of work; he knows where to Ixigin and how to proceed. He studies and draws only one .system at a time and at the time he is being given lectures and rat ions upon that system. He will have the greater familiarity with all the .systems since he is continually reviewing each bj- taking up a new one, going over the same series of drawings. The number of drawings is reduced to the minimum, fourteen to eighteen, but these are .selected for a definite purpose and are structurally closelj' related. The collection of drawings shows a systematic, related, and completed succession of steps by which the stuilent has gained a more concise and clearer grasp of his subject. One objection has been made to the plan that after the first system has been studied at given levels all other systems are arbitrarily studied at these same levels, thereby the student loses the selective search over the series in these sj-stems, and he may «ven leave out some important feature in each. This last point may be disposed of by referring to the insert method (see duplicate bi), and as for the first point, while he makes his studies and drawings at arbitrary levels, selected originally for the entodennal tract, still he finds most all important features in the other .systems at levels. .\nd as for not studying each system conlimiously through the series but jumping immediately to the .selected levels, this still compels him to draw each system at mon; than one level, thus giving him greater familiaritj- with each under dilTerent aspects, and dtK>s not exclude such a careful scanning either, rinally it must be said that no claim of superiority is made for this scheme over the reconstruction metlxxl. in the hands of special students, but in general cla-ss work, with limited time, the students of ordinary ability will have better results with the former method.

422 p. K. LINEBACK

Figs. 1. 2, and 3 arc selected from the wliolc number of drawings to illustrate how the student first makes the outline of sections at given levels, numhering them serially, but finishes all entodermal structures.

Duplicates a and 6 are here used to show that the student, by the aid of his serial numbers, has gone back over the same sections and has studied and drawn the vascular system in a and later the urogenital system in 6. It must be understood that he does not make these duplicates in his drawings but they arc used in this paper to illustrate the repeated study over the same outlines.

Insert A' shows how he would draw an insert, in this instance the ventral pancreas attached to the hepatic diverticulum near the junction of that duct and the duodenum.




II T Tin: nmi-ionRAPnir nr.n\icx. octobek 27

ANOTllKK CASK OK ( ;^ .\ ANDlJi »M()1{I'1IISM'


Kansas Stale Agricultural College, Manhattan, Kansas



Morgan ('13) has defined gj^nandromorphisni as a condition ill wliich one part of the body is like the male and the other Hite the female. This difference niaj' l)e one of the sex glands or nierel}' of the secondary sexual characteristics or both.

In many groups of animals the differences in .size, color and shape l)et\veen the male and the female are very striking. This is particularly true among the insects where sexual dimorphism is so marked that not seldom have the male and the female been placetl in different species and even in tlifferent genera. These differences are all the morenoticeable when this condit ion appears in distinct parts of the same animal as, for example, a moth with one side very dark aiul without distinct color jjattern and the other light with conspicuous color pattern, or a i^heji-sant having the long brilliantly colcTcd feathers of metallic luster of the male l)ird on the one side and the short somijcr feathers of the female on the other and one foot with a large spur, the other without a spur.


Mori- cases of gynandromorphism have been reported among insects than among any other class of animals. Kxcept for the famous 'I'higster Hees' perhaps the greatest lunnber are found among th(> Lepidoptera. (iyiiantlroinorphs are foiuiil so fre(piently in insects and so rarely in other forms that it is not sur ' ('iintrili)itinn from the Znoloftiral I,iiboriitory. K' State Agricultunil ("illoKc No. 20.


Till \SATOMIi Al. HEIOHD, \Ot.. 11. NO. 7

DKrruHi.ii, fir


prisiuK thiit Mornuii shoiilil say tluit they occur almost exclusively in insects and rarely or never in birds or niaininals. Although this statement approaches the truth, yet anionp; other Arthoiidda. Bertkau >!;ives examples of various Crustacea and two of Arachnida. Tlie i)resence of the male and female sex elements in the same animal is not uncommon in the fish and not infrequent in the Amphibia. In some Amphiina (King '10), I his condition is accompanied l)v differences in external sexual characteristics. Amoiif; the birds, Bond ('14) gives an account of a gynandromorpiious jiheasant and Poll ('09) of a bullfinch. The latter author also refers to a case of a flicker and a challinch.

In mammals gyuandromorphism has been reporteil to have occurred in goats, pigs and man. Pick ('14) describes testicular tissue in difTerent regioas of the same gonad in the jiig. It has been suggested that the free-martin among cattle afTords a close parallel to the gynandromorphs in insects, in which the secondarj' sexual characteristics of the two sexes are segregated to the opposite sides of the body. It might be added that this suggestion is based upon the thecjry that the twin calves of which the free-martin is one, arise horn a single zygote which separated into two parts during the early segmentations. I.illie ('16) throws considerable doubt upon this theory of the cause of freemartins. His eviilence indicates that the free-martin is a true female and that the twins are not from the same zygote but from separate ova. The writer would raise the ((uestion whethr-r or not the cow described by Pearl and Surface ('15) whicli took on the secondary characteristics of a bull should not also be considered as a case of gyuandromorphism in mammals.

Many of i)artial segregation of secondary sexual cliaractei-s to opposite parts of the l)ody have been known in man. (Jarrigues in the Medical Hecord, New ^drk, states that he knew an individual who looked like a male on the left side and a female on the right side. Seldom, however, is the .segregation perfectly halved, (ludernatsch ('11) .says that herma])hroditism in the sense of separate o\aries and te.stes has not been demonstrated in man nor even in other mammals beyond a doul)t. He described a case of an individual forty years old who had the ex


tcriiul Kciiifaliii nf I lie female tyjte. The clitori.- was eiilargeil uitli the opeiiiiin of the uretlira <>ti tlie ventral side. No uteru> was present and the vagina endo<l lilindly. The Mullerian duct> had been lost. An abortive prostate-like Khi"<' ^^i^s attached Ui the urethra. .V liiinor was removed from the right inguinal (tanal wliieh had the form of a testis with an epididymis attaehed. Histological examination showed that it was an ovo-testis. In the other inguinal canal was another smaller gland, hut it was nut removed. 'I'here is little doubt, however, but that it was a sex gland. Nd other sex glands were present.

The discovery in our laboratory of a cat with a testis on the left side and an ox'o-testis on the right side adds another instance of gynandromorphism among mammals. The secondary .sexual characteristics as well as the structure of the external genitalia are unknown. However, each sex gland was in the normal |)osition for a cat of that sex.

DKSCUlI'IKl.N (»!■ IIIK ( I VNANDKt i.\I( )I{I*II( )fS CAT

The cat referred to was about two-thirds grown. It had a testis on the left side and what jiroved to be an ovo-testis on the right side. The animal hail b(>en killed, .skinned and partially dissected before its peculiarity had been discovered. The scrotal sac and the external genitalia had been removed. The drawing .shows a frontal-lateral view of the urinogenital sy.stem. The urinary bladder was pulled .somewhat to the right side in onler to show the entrance of the vas deferens and uterus to the urethra.

The testis w hich is about the size and shape of a navy bean is entirely on the outsiile of the body cavity ventrad and to the left of the ventral border of tlie pubis. It has all the api)earances of a normal testis. The spermatic cord extends from the testis through the obliciue nuiscle. where it divides into the vas deferens and the spermatic vein anil art(>ry. 'i'he vas deferens extends anteriorly until it curves over the ureter where it contiimes caudad dorsal to the neck of the urinary bladder. It pierces the prostate gland and enters the urethra about half between the base of the urinarv bladder and the extniur. The



pr<>>tati' >il;iml of tho loft side is hvrjior thiiii the one on. the right sid«', although there seems to be one on the right side.

The nvo-testis is located on the right side of the ixxly a little posterior to the kitlnej-. Anterior to the ovo-testis and partially surmundingiit is the dstiuni tuhae abdominale. The ovarian

Fir. 1 The urinoRcnituI systom of a Rynandroiiiorplious cat. /. kidney; t, ureter; 5, large intestine; 4, siMTinatic vein and artery; 5, vus deferens; 6, prostate Kland; 7, urethra; 8. urinoKenital opening; 9, testis; 10, ovarian vein; //, ovarian artery; IB, ostium tubat: abdominale; IS, ovo-testis; 14, uterine tube: I, round liRament : 16, cornua of iitoru.'); 17, urinary bladder: 18. neck of bladder.

artery and the ovarian vein extend to the left from the gland. Kxtemiing from the ostivuii tubae alidoniiiiale is the uterine tube which continues caudad almost paralk'l to the vas deferens. It enters the urethra through the abortive right prostate gland. The sex gland and the uterine tube are held in place by the broad ligament and the round ligament.


Althoiinh the fixution was poor, the material l)einK preserve<l merely for the frr()s< aii;itoiny, a histolop;ieal exaniiiiation reveal.that a portion ot the ulaiid is arranged in strands similar to the arrangc^ment of the semeniferous tubules of the testis, but no lumen is present. There are also parts of this gland which are made up of cells similar to the normal stroma cells of a cat ovary. .Vmong these stroma cells are two nests of cells each of which is composed of a large central cell surrounded by a number of smaller cells. The whole group has decidedly the appearance of a young (Iraalian follicle.

The reproductive system of the left side of the animal has the appearance of a normal male in every respect, while the right side has the appearance of a normal female with the exception of the shape of the ovo-testis and the almost uniform size of the uterine tube and the cornua uterus. This latter difference is not significant since the cat is a young one. The angularity of the ovo-testis may be due to the pressure of the other internal organs at the time it was preserved.


Gynandromorphs and gynandromorphism are of interest beyond the mere fact of their occurrence on account of the bearing they have uixm other mooted questions .such as, sex determination, heretlit^-, the incUviduality of chromosomes, the r61c of chromosomes in heredity, sex dimorphism and others. The fact, that of the gynandromorphs of known ancestry a large number are among hybrids, is significant.

From the evidence obtainetl from the Kugster bees, Boveri ( '8S) suggested that gynandoniorphism may be due to the uniting of the male proiuicleus with only one pole of the egg nucleus which hatl begun to (li\ide parthenogcnetically,and that the fused male and female nucleus gives rise to the female part of the animal. I^ater Morgan ('05) suggested that the Kugster gynandromorphs might be the result of polyspermy. One of the spermatozoa might unite with the egg nucleus and another one or more give rise to another segmentation nucleus. The former would give to the female side of the body. .\s a cort)llary

4:ill M AIO 1. II \KM \V

t<) liis hy|M)th('sis MorKHii sliilcs that if malc-tlclcrniiiiing and fpiiialc-clctcrmininn >i)«'riii;iln/,oa exist, that holh kinds havo ontort'tj I 111' cfrg ami llir iVmaio-dctoriiiiMing spcrinatozoon has united with the egg nucleus. He f^ivos as his reason for suggesting this hypothesis that as is generally admitted hut not |)rov(Kl heyond a douht that, in hees, fertilized eggs produce females and unfertilized eggs produce males, that polyspermy is not assumed hut is known to occur at times in hees and that such animals as the hee which produce pai'thenogeneticall,\ . exix-rimr'nt should easily show which i)arl of the gynandromorph is paternal and which maternal.

Unfortunately von SielH)ld ('ti4) in his description of the Kugster hees did not state which side of the gynandromorphous hees was like the Italian or (lueen hee nor which side was like the (Jermaii bee or father. Later, Boveri ('15) obtained from .Munich a jar of alcoholic specimens which turned' out to he which von Siebold had obtained. On account of their having been in alcohol for .so long a time the color was mostly gone, but careful examination showed that the mjile parts of the gjiiandromorphs are like the Italian bees while the female jiarts are like the (lerman bees. .\s Morgan ('Hi) sa^'s, "this conclu.sion gives a decisive answer in favor of his (Boveri's) hypothesis and sets my own (Morgan's) asitle for this case at least."'

Previous to the re-examination of the Kugster bees Morgan cited the evidence obtained from a cross between two silk-worm moths as proof of his hypothesis. The female belonged to a J'airopean breed having strijied caterpillars, the male belonged to the conmion .lapaii'-se breed having unstriped caterpillars. Two of the hybrids had the left side of the body striped and the right side unstrii)eil. According to Morgan, the striped side is due to the combined nuclei, the other side liue to the sperm nucleus alone, therefore paternal, .\cconling to Boveri the paternal side is the result of the combii\ed nucleus. If this be true, since it is known that the striped condition is dominant both sides should be striped. Furthermore if is known that both males and females among these moths may be produced parthenogenetically.


liridnes ('IG) suKgcsts that th«Ti' arc not male and fcnialispcriicitozoa. Imt tli<> sex is (lot«Tinin('<l hy tlic [)ri"<cnce or aln sencf of certain sijofifio chroniosoincs. When two or more X-chromosomes aro present in a zygote, no matter from when<;e thoy came the resultant individual will be a female. If there are not at least two X-ehroniosonies in a zygote the resultant individual will he a male. He further accounts ft»r the pre-senee of more than two se.\-ehromo.somes in a single zygote by nondisjunction. He suggests that if the .same sort of primary nondisjunction which has l)een assumed to give rise to XX'X' cells in an XX female should take place at a cleavage stage gynandromorphs would result. If this non-disjunction should take place in a ver\' early cleavage, particularly the first one, the individual would be a lateral gynandromorjil) but if it should take place in later cleavages, the individual would be a mosaic. This theory is based upon the work on nros()])hiIa.

Scojjoli .suggested that this halved gynandromorjjhous Phalaena pini was formed by the fusion of two pupae in ojie cocoon.

Donctvster has suggested that a gynandromorph is produced bj' the fertilization of a binucleate o\inn by two spermatozoa, one a male-determining and the other a female-determining spermatozoon.

M(>nzel supposed in case of the Kugster bees, that the queen was malformed and that the progress of the ova were so slowthat they began to segment parthenogenetically and later were fertilized. (12) has obtained a great number of gynandromorphs in crosses between Lymantria dispar and L. japonica, which were always of three kinds, female and male gynandroniorphs and 'weibchenmiinnchen.' These were always absent in certain crosses and present in definite proportions in others. He accounts for these on the hypothesis of the relative potencies of tlie sex factors in different rac(>s. With certain assumptions they occur in a perf(>ct Mendelian ratio.

^^'hen the male factor is dominant it is more j)otent than the female factor, but when two such female facti)rs are present they


are more potent than the sinRlo nmlo factor. T'ao factors for the .s»'Con<l;iry sexual characters Nehave in Hke manner.

A.s to fjynandroinorphisni in l>inls, Poll ('(W) sufigosted that the occurrence of the testis on the right side of the body was associated with the rijiht ovary which normally atrophies and the testis of the left side is jussociated with the socoiularv male characteristics. Cockayne ("lo) says that gynandromorphous birds are lateral genetic hermaphrodites and are all alike in having a testis on the male side and an ovary on the female side.


Both the theory oi Boveri and Morgan assume that the nucleus is the bearer of hereditary characteristics which is less generallj' accepted than that unfertilized eggs produce male bees. AMiile the weight of evidence is that the nucleus is, at least, a large factor in determining the hereditary characteristics it is far from being conclusively proved that it is the only factor. In the cases cited as evidence for each of the hypotheses little is known of the behavior of normal hybrids in heredity which would be nece.ssarj' to account for the behavior of such abnormal hybrids as gynandromorphs. Both hypotheses assume a very early localization of the tissue, as early as the first cleavage stage, and that all the cells arising from one blastomere must give rise to one side of the body, those from the other must give rise to the other side of the body. Moreover, each hypothesis accounts for bilateral gjnandromorphs only and does not account for the mosaic or mixed gj'nandromor])hs. In addition the of the bird, the pig, the cat and man having an ovo-testis can scarcely he explained in accordance with these theories.

Like the two theories previously discussed, the theory of nondisjunction assumes the all-important r61e of the nucleus as the bearer of the luTcditary cliaracteristics and further more that • the chromosomes are (|ualitatively dilTerent. The author has failed to explain the presence of secondary sexual characteristics of the sex on the side of the animal which has the gonad for that sex. In the example which the author cites (Bridges, '1(>, pp.


13()) the male .side of the aninial, according to his statement, should have the sex-Hnked characteristics of tlic ftinale, then-foro, the secondary sexual characteristics of tlic female, if the secondary sexual characteristics are sex-linked. If the secondary characteristics are not sex-linked the arrangement of the secondary sexual characteristics has not been accounted for. The writer fails to see how this theory can explain the gynandromorph having an ovo-testis and bilateral with reference to the secondary sexual characteristics.

Doncaster's theory assumes that the important factor is the spermatozoon. However, he has shown that binucleate ova do actually exist and further he has shown the conjugation of a spermatozoon with each nucleus and has seen the resultant mitoses. Yet, he has had no of gynandroniorphism in the race in which he found binucleate ova.

Goldsehmidt's theory is based upon the a-ssuniption of relative potencies of male and female factors and as an assumption may explain cases of lateral gynandromorphs, but could scarcely explain mo.saics and those cases with ovo-testis.

The case of the pheasant as described by Bond seems to be contradictory to Poll's suggestion as well as Cockayne's statement for the secondary male characters were on the left side and only one sex gland was present.

It seems that no theory thus far advanced serves to explain all cases. The solution oi the cause of gjiiandromorphism rests upon the solution of at least two fundamental biological phenomena, the det(>rmination of sex and the determination of the secondary sexual characteristics.


1. (lynandroinorphs are comjiaratively Cf^mmon among insects and rare among other groups of the animal kingdom.

2. Of the gynandroniorphs of known ancestry many of them are among hybrids.

H. The cat with an ovo-testis on one side and a testis on llie other adds another gynandromorph to those recorded among mammals.


4. No niii» of the thporios ntTon'd as an cxplaiialiiiii i>f the (•aus(> of ^'viiaiKlroniorphisin scoins to explain all cases of recorded gy iiand in 1 1 lorphs.

5. 'I'lie solution of the (^ause of fiynandromorphisin rests upon the solution of at least two fundamental biological phenomena. the determination of sex and the determination of the secondary sexual characteristics.


BoVKiu. Til. 18S8 (,'biT part idle HcfniclituiiK. Silz. Her. (i. (ics. f. Morph. w. Phys.. Miincli.. Bd. 4.

1915 t'ber die Kiitslchunn dcr Kugsterschcn Zwitlcrbicncn. Arch. f. Entw. Orgiinismen. Bd. -11.

Bn.Mi, ('. J. 1914 On a case of unilateral development of secondary male characters in a pheasant, with remark.s on the influence of hormones in the production of secondary sex characters. .Jour. tlen.. vol. 3.

Brandt, Alex. 18S9 .\n:itomisches und .MlRemeines iiber die gogenannte Hahnen-fedriRkeit und iiber anderweitifje fJeschlecht Anomalien bei Vogeln. Zeitschr. f. wiss. Zool., Bd. 4S.

BRiDOEii, Calvin B. 191G Non-disjunction as i)roof of the chromosome theory of heredity, den., vol. 1.

Chapix, ('.\THARINE I. 1915 A casc of hermaphroditism in Spelerpes bislineatus. Biol. Bull., vol. 29.

Co:;kayne, E. a. 1915 'Oynandromorphism' and kindred problems. With descriptions and fiRures of some hitherto undescribed examples. Jour. Oen.. vol. 5.

CoRiiY, II. 1905 Removal of a I umor from a hermaphrodite. Brit. -Med. .lour., vol. 2.

DoNCASTER, L. 1914 On the relations between chromosomes, sex limited transmission and sex determination in Abraxas prossuluriata. Jour. Oen., vol. 4.

Fi-ss, A. 1912 Cber die (ieschlechtszellen des Menschcn und der .SiiuRefiere. .Arch. f. mikr. -Vnat., Bd. 71.

t;(>LDs:ii.MiDT, U. 1912 l'>blichkcits8tudien an >Schmetterlinge>i. I. Zeit.f. ind. Abst. und Vererb., Bd. 7.

(;l-dek.vat.sch, J. F. 1911 Hermaphroditism verus in man. .\m. .lour. .Anat., vol. 11., Helen Dean 1910 .Some anomalies in the genital organs of Bufo lentigcnosus and their probable signihcance. Am. Jour. Anat., vol. 10.

Knapi'e, Emil 188(i Das Biddcrschc Organ. Morph. Jahrb., Bd. 11.

LiLLiE, Frank R. 191<) The theory of the free-martin. Science N. S., vol. 43.

.Morgan, T. H. 1905 An alternative interpretation of gynandromorphous insects. Science, N*. S., vol. 21.

1907 The cause of gynandroinorphism in insects. Am. Nat., vol. 41. 19(Kt Hybridology and gynandromorphism. Am. Nat., vol. 43.

AN'OTllKU <'.VSE OK (;VV AVhki iM()HIIII-\I 4^rt

MnlKiAV, 'I'.H. I!M.{ llrrcdity und aex. ( (jluiiil)i:i I niviiMiv I'rciu., Ni » ^Dri..

1910 The iMinstiT Kynandroiiiorpli l)cen. Am. Nat., vol. 50. .N'AriiTSHKIM, H. I!M.'{ CyloloKischf Stlldicn iilirr dir (icxrhlrrlitHbciitiiniiiuiiK

boi dpr lloniKbic-no fApis iiirllitica L.I Arch. f. Zcllf., Hd. 11. 1'karl, Raymond a.nd SriiK.\CE, Fka.nk 1915 The :i88uniptinn of male BCCon<l ury ^ex characters by a cow. Science, N. S., vol. 41. I'iCK, LuDwio 1914 tlbcr den wahren Ilcrmaphroditismus des .Nfcnschen und

der SiiuRetierc. Arch. f. mikr. Anat., Bd. S-1. Poll, Heinricii 1909 Zur Sehre von den Sekundiircn Sexual ("haractern.

Stzgsher. (ics. Naturf. Freunde zu Berlin. SiEUOLD, U. C. Th. 1864 I'eber Zwitterbienen. Zeit. f. wi.K.scnsch. Zool., Bd.

14. ISWJ KrtiHiz <|pr abgestorbenen Zwitterinutler des KuKStcrchen

Zwiltfrstiickcs in ("iHLslanz. Hioncnzucht. .Iiihre.. ]HIH'). !St. Gkoroe, La VALLETrK 1.S95 Zwittprbildung beim kleiiicn WiusHoriiinlch

(Triton taeniatus). Arcli. f. inikr. Anat.. Bd. 4.'). Tdyama, K. 1906 Studies on the hybridolugy of iniiects. I. Un some silkworm

crosses, with special reference to .Njendel's LaAv of heredity. Bull.

Col. .\Kric. Tokyo Imperial University, vol. 7. WiiKKLER, \V. M. ISKW .Some new eynandromorph antd. Mull. .\iii. Mu.teuni

Nat. Hist., Xew York.

1910 A gynandromorphous Multillid. Psyche, vol. 17.


Till: iiiiiLinoMiPUir «t:Hvici:, ukccubbK S

'IHK FLNKH SIKl ( rrj{i;()F Tin: ( ll.lAin (,AN(iLI().\



From the Lahoraloriu de Analomia e Histotogia da Facutdade de Medirina e Cirurgia dc SAn Paulo: Director, Prnf. A. Borcro

Numprous Raps exist in our knowledge of the histology of the various structures of the reptilia, and among these we may include our lack of knowledge of the structures of the ciliary ganglion or ganglia of the Ophidians. At the suggestion, therefore, of Professor Bovero we undertook a series of histological researches upon the ciliary ganglia of these vertebrates; not only on account of the need of more exact knowledge of the structure of the ciliarj' ganglia in these forms, but also because we found it verj- easj- to acquire the necessary material for this study, since we are located in the neighborhood of the Instituto Hutantan, which possesses the largest collections of living Ophidians in the world.

The researches directed to the knowledge oi the morphology of the cellular elements of the ciliary ganglion and its connections are of a relatively recent date. After the few researches made by (i. Retzius, Michel and d'Krchia, with the (lolgi silver-chromate method, which ditl not throw nmch light on their cellular morphology, the number of researches increased with the introduction of the photograjihic method of Ramon y Cajal, a method which has permitted us to extend our knowledge of this subject, especially after the investigations of (1. Sala, v. Lenho.ssck, Car|)enter, Kuntz and Pitzorno.

The sympathetic nature of the ciliary ganglion has been shown by physiological methods and has also been demonstratetl histologically in a large series of observations and researches, some of a very recent date (Chiarugi. ("ariienter, Kuntz, Gantiiii. Hiciuier. Reccari, Rruni).


4XS .1. M. VVK'>S\ (iAI.VAO

'l'lu> piiri'Iy liisli ill laical stutlics on the ciliary fiaiiKlio" of the Ucptilc- are much Icwcr in iiuimImt than those dealing with the developineiit of this and other jjiiinKlia of the cephalie sympathetic on the same vertebrates. Indeed they are limited to one contriliution of V. I.enhossek.' in uliich lie studied the ciliary fianglion of the i.acerta agilis, !.. niuralis. !,. \iri<lis anionj!; the l.acertidae; ot' the Tropidonotus, ( 'dluIxT, and Zainenis among the Ophidians; ol the Testudo graeca and ]*]mys lutaria among the ("helonians; and another one of Pitzorno^ who studied the said gangHon a,mong the following ("helonians: Thalassoehelys earreta, Testudo nemoralis, Testudo graeca.

If we limit ourselves more specially to the study of the Ophidians, we will then see that the only ilata that we possess are the observations of \-. I.enhossek and this one limited to the European specimens of small size.

For this reason, when we undertook our researches upon the ciliarv ganglion of the Ophidians, we endeavored to direct our attention to the greatest possible number of individuals of the most variable sjiecies. \arving in size and age, giving especial attention to the specimens of large size.

The following are the species investigateil by us in the present research :

Lachesis atrox (.lararacu.ssii), L. lanceolatus (.lararaca), 1.. alternatus (Urutu), Crotalus terrificus (Cascavel), Driniobius bifos.'iatus (Jararacussu d'agua), Hoa constrictor ((iiboia), Klajis tCoraes of various kinds).

It is very necessary to use a large nimil)er of indi\iduals on account of the extreme difficult.y of obtaining good staining by the method used, the reduced silver method of Ramon v ("a,jal in all its modifications. \. Lenhossek has previously deplored I)recariousness of tliis method and such inconvenience due to the refractiveness of the cells of the ciliarv ganglicm to the methods of silver nitrate reduction.

' V. Ijcnhoswpk 1911 Oils oiliarKunglion dor lieptilicii. .\iial. .VnzeiRcr, Bd. W, no. 2 u. 3. .Vrchiv. f. mikrosk. .\nat. Bd. 80, s. .SO-IOO.

'I'itzorno 191.3 C'onlriliiitn ullii cono.sconza dolla slriittur:i del Kanglio ciliiirc dci C'hcloni. .Vrrli. It. di Anat. <■ Kiiitiriol., v. I'J. f. .30.

fll.lAKV (;A.V<iL,l<)N «)K OI'IIIDIAVS 43JI

I'lic closun'ot the i)rl)ital ciivity on the side of < ho cranial cavity iiiakps it rather ditlicult to isohitc tho contont^ ni the orbital cavity and cvoiitually to isolate the kji'ih"" <"r uaiijilionic romplcx of the ociiioniotor nerve.

In the iarf;e.s|)eciniens. specially in the Hoa consirictorand (><>talus terrificus, it was sometimes rather easy to dissect a larK<' and evident Kanglion in close proximity to the third nerve corresponding \ery prolmhly to the proximal RanRlion of v. i.etdioss(>k.

\Ve cannot state definitely as to the existence in the several species examined by us, of the two K-'^'ifllin described by v. T<enhossek, however, although we had at our disposal complete series of sections and these were carefully studied we were able to observe l>ut one ganglionic formation.

In order to lie reasonably assured of success it i> ad\ i.-able to use only fresh material and such as is in perfect condition; for the heads in which we found .some inflammatory |)roce.d in the ciliary ganglion of ('helonians by Pitzoinn, said lobes not being characteristic and exclusive of the sympathetic ganglia of the C'helonians, but also in the cells of the spinal ganglia of these same animals. Neither did we ol>ser\ (• in the cell^ of our specimens, the fenest rationed cells nor the ppIIs with prolonpiations forminp loojis. Tho colls shajx' is noiieniliy >|ili('ric;il or pyriforiii. Imt iii:tiiy times specially in the regions (il the gauKlion where bhe cells are compactly arranged, these .It tain an elonnateil form.

Our prei)arations do not give ilistiiict e\i(lence of a neiirotiijrillar reticulum in the interior of the cell body, even after use of a numlier of the suggested variations of the Ramon }■ C'ajal method; we have nevertheless ol)serve(l an apparently homogeneous structure ilemonstrated !>>■ v. Lenliossek.

Each cell, no matter what its form, is enclosed in a very thin tissue capsule, which cdntinues near the pole as a conic process.

Contrary to what has i)een stated by v. Lenhossek, the amphycites nuclei are to be observed here and there.

The single axone process detaches itself from the cell body as a small cone, which with its place of origin is uniformily turned toward the oculomotor nerve. The neuraxis filament follows a retrograde root in a more or less proximal direction and then leaves the fiber of the third nerve forming with them various angles and then taking a distal direction goes to form the ciliary nerves.

Not rarely and specially in certain species, we could very frenuently observe the origin of the neuraxis in the distal extremity of the cell, that is, in the opposed extremity to the habitual point of access or the oculomotor fibers to the ganglion cells.

The neuraxes of the ciliary ganglion cells may be easily distinguished from the afTerent fibers of the ganglion, not only where leave the nerve bundle, but also in the neighborhood of the cone of origin, by their smaller diameter and by reason of the fact that they take a paler tint than that taken by the afferent fibers.

The initial course of the ciliary fiber is essentially as described by %-. Lenhossek, that is, coursing within the afferent fiber through the capsule cone. In the great majority of the cases, the ciliary fiber is parallel to the course of the oculomotor fiber, however, the ciliary fiber are observed in some instances to form wide spirals around the oculomotor fil)er: the ciliary fiber also becomes very thin when it is in relation with the oculomotor fiber, i>ut it enlarges al)ruptly as it recedes from it.


As was observod hy v. Lcnhossok in siicpes>lHl (jrcparalioiis, we ohscrvrd hftwccii tho ovpn ronfour of tho cell ami the inner surface of tho capsule, a fissure which cannot be inter()reted as an artifact, but as lieiiiR occupied by a thin sj'ncytial sfratumineludiuK the aniphycite nuclei of the pole and those for the rernaindfr of the peripherj' of the RanRlion cell, thniich the number of these latter may be scarce.

The afferent fillers of the cihiry ganglion are finer than those of the oculomotor nerve destined to the muscles; these fine fibers penetrate into the cone of the pericellular cap.sule, that is in the great majority of cases, at the same point from where the ciliary fihers have exit; the conic process is formed rather by the apposition of the capsule to the afferent and efferent fibers.

The afTerent fii)ers, as a rule, divide within this conic prolongation, among the amphycyte nuclei in three or four very thin slightly contorted diverging branches, with few ramifications and, therefore, very rarely constituting a basket formation, which is very simple, when it exists, as has already been demonstrated by V. Lenhossek. Some of these ramifications end very quickly in the amjihycite cone, but others, which an- much more numerous, on as far as the .surface of the i)olar extremity of the cell, where they terminate as free endings.

The ramificati<ms of the endocapsular fibers, by rea.son of their relatively straight course and their .scarcity, are far from showing the very rich and elegant pericellular net, characteristic nf the ciliary ganglion cells of the mammals.

The ending of the endocap.sular fil)ers takes j)lace immediately on the cell surface as thickening or small dilations; we had occasion to see repeatedly that terminal expansions presented themselves as voluminous rounded or oval balls and generally placed between the cell and the capsule, or even in relation to the amphycites cone.

\\'hereas the fine endings of the oculomotor fiber are inunediately adjacent to the cellular body, the larger ramifications of the same afTerent fibers run in the space between tlie cell and the capsule, that is, in the thin syncytial protoplasmic stratum to wliich we h;i\ e already referred.



The majority of the details described by us are in essential very similar to those (lescnl)ed by v. L(Mihossek.

This present coininuiiication is to be regarded of the character of a preliminary note. I'urtlier details it is expected will be ascertained on completion of more extended investigations with more abundant material. It is then hoped to pul)lish the more extensive data accompanied by the necessary figures.


Tii»: iiiiit.inriM\Piii( mkiivk » , itKCrMiirR H




From Ihe l.nliiiratoriu de Anatiimia e Histologin da Fariildade de Medirina e CiruTgia de Siiii f'nnlii: Director. I'ruf. .1. Hopero

While stationed at the Laboratory of Histoh^gy of the Sao Paulo -Medical School, we had the opportunity to apply to the hearts of Ojihidia, the methods of elective staiiiiiip; and subse(|uent clearing of the cartilage, as |)roposed by l.undwall.

It was shown by the researches of Favaro' which hail been in I)art published when we began our studies on the heart of Snakes, that there exists in the 'septum intermedium' of the Tmpidonotus natrix, a small Opliidium which is not poisonous and very common in Europe, an accumulation of a \esieular connective tissue, known as mucous tissue, by (!reil.

This was known with reference to tlie hearts of other reptiles. This tissue is composed of large, vesicular, polyhedral or irregularly rounded cells, among which is found a small amount of chondroinucoid ground substance. This ground substance surrounds the cells, forming more or less complete se|it;i, which make up as a whole a s|)ecies of alveolar stroma.

The api)lication of Lundwall's methods for the staining of cartilage (fixation with formalin or with alcohol, remaining in alcohol for more than 4S hours, staining with acid solution of methyl green, or of toluidine blue, tliscoloration with alcohol 95 per cent, dehydratioji, clearing with benzol), was suggested to us by Profe.s.sor Bovero.

We had in mind when we applied this method to the hearts of some Ophidia, to verifj- the existence of cartilaginous ti.s.sue,

' Kiivaro. G. RiciTche cmbriologiclic o .Vnatomiclif intorno al cuorc dei vcrtebrati. I'arlr prima 191.!. Parte sceonda 1911. Padova. Fratelli Dnickor. Edilori.


444 A. I)K I.KMOS TdHltES

wliicli li;iii already l)ooii done for (ithcr reptiles, and to note especially il there was any modification of this cartilage so far as the size nr the age of our Ojjhidia was concerned.

Thus, thanks to the kindness of Dr. \ital Brazil we had an oi)p(>rtunity to make these researches on the hearts of a great nunilier of snakes, belonging to the following species: Lachesis Neuwiedii, Crotalus terrifieus, Lachesis alternatus, Lachesis lanceolatus, and, two more species of the Oxirhopus variety.

\\'e obtained invariably the same result in all the hearts prepared bj' us according to Lundwall's method (we used always methyl green, Cogit), that is, at the basis of the ventricles corresponding to the 'Sulcus atrioveiitricularis' there is always a zone not very well marked at its periphery, which absorbs the colouring substance very readily, but the colour which is obtained is always much paler and less intense than the solution u.sed.

This zone which ai>pears as a green spot does not discolour, even after a prolonged lapse of time, but is only clearly evident after the clearing by benzol.

We did not observe any dilTerence worth mentioning among the species of Uphidia which we used, we noted, however, that the larger the heart was in conipari.son with the size of animal, the more intense was the colouring, and the more distinct was the outlines of the coloured zone.

We made microscopical sections from .several hearts, in order to demonstrate the exact position of the area which absorbs the methyl green, while the rest of the heart was discoloured. These sections were made in .series perpendicularly to the a.xis of the heart.

A study of serial .sections enabled us to demonstrate that the zone thus colf)ured in green by Lundwall's methods corresponds exactly to the .se|)tum internie<lium, therefore, was the vesicular connective tissue described by (Jreil and by'Favaro in the Tro|)idf)notus natrix. This tissue jiresented the characteristics described at the beginning of this article.

Such a form of connective vesicular tissue corresponds to one of the stages of ontogenic develojiment of the cartilaginous tissue. We observed that the substance of the septa interposed


ht'twecn the vesicular cells has u groat resistance and is very refractive. This suhslaiice ai)i)ears always hasophile, although not intensely so, this lias(ij)hilic characteristic explains t(^ us the fact that we obtaineil a green colcturing with I-undwall's method, elective for cartilage, so that this sub.stance gave us the same reaction of tlie groiMid sul)stance of the cartilage, liut much less intense than if ol)tained by tlie same tiirtlioil a|i|)lied, for instance, to hyaline cartilage.

Wv have thought these facts worthy of reccjrd, since this method gives a rapid macroscopic demonstration in the Ophidia of the exi.stence of a particular form of cartilaginous tissue (l'rr>tocartilage, parenchymatous cartilage of Schiifer), the presence of which cartilage has only i)een demonstrable by microscopic examination ((!reil and Favaro).

These experiments of ours, prove that this particular form of ti.ssue with chondromucoid substance repeats in the phylogenesis the evolution of the cartilaginous tissue of the individual, for corresponding to the same point where we found this vesicular tissue, there exists in animals of a higher class of development a tissue which is typically cartilaginous.

We come therefore to the conclusion that lAmdwall's method, which is eminently elective, demonstrates in the heart of Ophidia, as cited above, the existence of this primitive form of cartilaginous tis.sue in the thickness of the intermediary septum which is evident by way in which it absorbs the colouring substance.

It would be certainly very interesting to be able to continue with Lundwall's method or with such modifications as experiences has shown to l)e op|)()rtune (Hovero). in the various classes, orders, and species of vertei)rates, perhaps even at various age of rejiresentative of the various types, the existence and distril)Ution of the cartilaginous tissue of the heart. In this way the phylogenetic modifications of the various parts which represent or constitute the cartilaginous skeleton of the heart would l)(' demonstrated.

A I TIIOH f« ilMTItALT til Tlll» r«l*KII IWIUEU

lit Till Hiiir.i'ti.ittiniir •«» Hvii i . oiirrunieii *(




From l.ahoraloriis nj Aualiimy and Histulogy da h' lur uliiiuU dr Mrdiriiui e

Cirurgia de Shti I'aulo (Brazil): Director, Dr. .1. Horcru

111 thf .\ii!itomipal Si'ction of the Primciro Coiigrcsso Mi-dico Fiiulista (Di'ccinbcr '). lOMJ), I h:i<l the (>i)j)()rt unity to present ;i series of oxeellont prei);irati<)iis from various orj^ans of the Ixitly i>ossessing eartilaginous skeleton, which \\ereol)taine<i hy met hods based uj)oii the ehromatic elcftivity of th<' cartilape and the ch-arinn of the prejjaralions.

In the same month of Decemher. C!. Xohaek (Anatomical Record, vol 11, no. 5, IQlti), called attention to the excellent results which may be obtained in the endiryonal cartihiRe, by the method propf>sed by Van Wijhe, which according to the author is very little knowii.

As We employed in the execution of our j)reparations of the ailult cartilaginous skeleton, onlv the methods jiroposed bv I.undwall (Anat. Anz., vol. 2.'), 1!)U4; vol. 27, 19().j; vol.U), 1912) for the study of embryonal cartilage, we think that, sjx'cially after the appearance of N'ol)ack's note, it will be timely t(^ give some considerations to this .subject.

In 1914, Profes.sor liovero' had the occa.sion to present to the S<icieda<le de Medicina e Cirurgia de Sao Paulo, a great ninnlHT of prcparation.s of the cartilage of the jilica seiiiihuiaris oculi (third eyelid) of man and other mammals, obtained with the methylgrcH'n .staining method projjosed by Lundwall ('()")) for the embryonal cartilage, having at that tiiiii' called attention to the fact that the ])rei)arations were not only and deiiion.strable but |)ernianent.

Such ])rei)arations ar(k unstable oidy when pre])ared of embryonal cartilage, it being noticed that their degree of stability is in the dire<!t ratio of the degree of maturity of the cartilaginous tis.sue.

Kollowing .such steps in the laboratory, tlirected by Hovero, we luidertook to jirepare a series of cartilaginous skeletons of the larynx, trachea and bront^hi, of man and other vertebrate animals and of the cartilages of the nasal alae and sej)lum: of the fibrous cartilage of the te)ii]>oro-iiiandibularand sterno-<'lavicular joints; of the semilunar cartilages of the knee and of the calcaneous-:istragalus-scaphoid joint;:; and of the cartilage of the auricle, all from man. but from different races and ages.

Some of the i)reparations were obtained by the methods projKisi'd by Luudwall for the .staining of cartilage alone and others were obtained by sxich modifications of the method as experience showed us to lie of ad%'antage.



The first mtthod i)r()i)()S((l liy Lundwall, very well known nowa-days, whii'h n.ay Jic ftniiid in iiliiiost every text-book of histoloKieal techiii(iui (Hhrlich, Mosse and Krause, '10; Carazzi and Levi, '11, et eiter:i>. as also in some sjx'eial memories (Briini, '08), is hased specially iipiMi the staining with a 1 to KKKJ solution of methylgreiii in 70 per fi ni alcohol (latily the methylgreen has been substituted by , toluidird>lue) to which has Ijeen added a 7 to lOOtJ solution of aeetic acid, with sTibscquent washing in 95 per cent, alcohol until the prejjaration does not lose anymore stain; dehydration by absolute alcohol; clearing by benzol changed many times; and finally conservation in a nuxture of benzol, carbon sul])hide and essence of peppermint. The chaiiRes made by Professor B(tvero in this method, does not alter in any way the fundanuiital features of the method as devised by Lundwall, but they are to a certain jxjint an incorporation in Lundwall's methods of analogous methods proposed by Bakay and by \'an Wijhe.

We observed many tines the fact, that the methylgreen obtained from ililTerent sources or even that obtained from the same source but at different times, does not react the same in the presence of the ground substance of the cartilage, specially in regard to the ekctivity of the staining arid also as regards its stability. The causes of such variations in the actiDn of the methylgreen, may be found in the fact that it is not a single stable compound, but a mixture of staining substances, and in fact that the diff.Tent kinds of methylgreen employed by us, and with which we had n 'g'ltive or unsatisfactory results, had suffered som(! chemical decomposition.

Having alwaj's in mind the idea of the basophilic character of the grountl substance of the cartilage, we used, besides the methylgreen, other substances such as toluidin blue, methylenblue, methylviolct and safranine. In so-ne instances, we substituted the acetic acid by hydrochloric acid. The duration of -the staining was largely modified as was also the t nnpi-ratiu-e at which it took place, this, according to the sizi' of the preparati ms and according to the vari.ties of cartilage that we trieil to show. We us>d the benzol in very much less quantity than that advised by Lundwall, not only as a matter of economy but also becaiise it was quite difficult to obtain this reagent here of the neces.sar>' purity. Thus, three, four or even more preparations were pas.sed successively through the same amount of benzjl. For this reason, it was necessary to leave the preparations inmersed for a longer tine in the bonz)l, to compensate for the less frequent change of the reagent.

We may say, as a general rule, that no matter which staining substance is ased, its electivity for staining and its stability are in the direct ratio, as we hav- already said, not only to the maturity of the amorphous ground substance of cartilage, Init also, what may be readily imderstood, to the quantity of the ground substance.

Therefore, we mav establish a priori, that not only the adult, cartilag<' is stainabic with greater chromatic affinity and with greater stability than the embryonal cartilage, but that among the different


v.iiitticsuf iitluli carlilum'S, the hyaline variety will take the stainiiij? with Ki<!it*'r clcctivity ami with greator stability than the clastic and lilirous varieties, ami this is readilj' verified in jiractiee.

What we have endeavr)red to explain may be readily demonatrated by the siiriple test of the way in whirh the cartilage of the laniix and trachea behaves as against that of the knee and of the auricle.

Bearing closely to what we have said, we may add that it is necessary to watch with greater attention the decoloratiun of the elastic and fibrous cartilages, where as it is not necessary to ase such care as regards the hyaline cartilage.

Another advice may be given in regard to the embryonal and foetal cartilages, that is, that the conditions being the same, the staining with methylenblue or methylviolet, or with the toluidin blue is much morestable than with the methylgreen.

By the examination of our prejjarations we may conclude that such methods to demonstrate the cartilaginous tissue, always give con.«tant and sure results and exi)lain the jxissibility of a coming re vision of full chapters of .\natomy dealing with the morphologj' of certain and determined cartilaginous organs, which were previously stutlied by the dissecting and macerating methods, followed by the desiccafcion of the cartilaginous pieces.

Our assertion seems to be perfectly justifiable in as nmch as none of these methods arc as satisfactory to the one in which the cartilage is stained.

The revision of the morphology of the cartilaginous organs may give very good results, not only in regard to the largtr hyaline cartilages (lar^Tix, trachea, bronchi, nasal septum, et cetera,) but also as regards the study of the extension and topography of the amorphous ground .substance in the fibrous cartilages, for example. We believe as regards this point of view that with the methods employed by us, combinetl with the methods for histologic research, interesting and new results may be obtained. Certamly, the free of the methods of staining of the cartilaginous pieces in toto and their demonstration through the soft tissues promises to give many good n suits in any systematic stuilies which may be undertaken. On some of these, we expect to contribute very soon with personal and especial researches which we will develop during this j-ear.



.lAMKS H. CASH Aiialoniiral Laburatiini of the Juhiix Uupkiiui UniviTsily

Tliis i)roljlem was approached l)y way of pinhryolojy primarily, because extensive lymphatic injections, in the adult heart, are impossible due to numerous valves in the lymphatic vessels. In embryos imder (iO mm. in lenRth the valves have not formed and retrofi;rade injertion of the cardiac lymphatics is quite possible under the projier conditions. .\lso by observing the orifj;in of the vessels in younger emlu-yos aiul several further stages in their development it api)ears cviileiit that ol)servations made on the adult heart can be more rationally iiitcrpretetl and the solution of the problem will lie nujch nearer complete than if conclusions wore drawn from the adult s]iccimcns alone.

The literature on the nature of the lymphatics of the heart is very scanty. The most pretentious and generally accepted work being that of H. Bock of Munich.

Ranvier ('95) comes to the conclusion that there is considerable free space in the myocardium (that is, space between the muscle cells and blood vessels) and that this space is lymphatic in nature. Tliis assumption is based on the fact that pericardial lymphatics can be injected by plunging the neeiUe at random into the myocardium at any poijit. Thus the intercellular spaces ami the lymi)hatics were sujiposed to be in direct connection. So the entire heart is considered to be a 'lymphatic sponge,' as Ranvier expressed it.

Sahioli injected Berlin blue into the myocardium and made out what he thought were lymphatic vessels whii-h ran in the connective tissue septa between the muscle bundles — but he could find no evidence of their connection with the intercellular spaces as advocated bj' Ranvier.



Alhrecht ('87) in his work on tho heart muscle, made an appar«'ntly \t'ry painstaking in\estigati()n of the subject. He injected tlic livinp iieart. thus using its luitural movement to distribute the injection mass. He came. to the conclusion that the large lymphatic trunks lay in the spaces between the muscle cells. These large trunks he thought, were connected bj' straight Ij'mphatic vessels of smaller size, which run at aliout right angles to the long diameter of the heart nui.-<cle cells and arrange themselves into a capillary network in the interior of the adjoining muscle bundle. 'I"he lyniphatic vessels were observed to be straight anil were tlifferentiated from blood vessels which exhibit multiple and acute branching, and by the observation that blood vessels ran parallel to the long diameter of the muscle cells. Within the muscle bundle the lymphatic capillary network includes a nmsde cell within each of its meshes. .Vlbrecht made the further observation that the intercellular spaces which contain the large lymiihatic trunks were not entirely filled by them, l)ut that the lymphatic vessels lying in these spaces had their own walls, and that when injected, the}' could be seen lying over to one side of the space which was apparently empty. These spaces observed by Albrecht, were probal)ly artefacts due to shrinkage of the tissue. Thus .Vlbrecht concluded that an intercellular lymphatic capillary network existed with its own walls, and that it w;as in connection with the great interfasicular vessels which lay in the large spaces in the heart nmscle (Henle spaces).

Bock ('05) criticized these statements of .Vlbrecht severely. He points out very truly that in the previous work, the great discrepancy in reports lay in mistaking blood vessels for lymphatics. He attempted .solution of the i)roblem by making injections of both blood vessels and lymphatics in the same heart with different colored injectioi\ masses. The heart was placed in jihysiological salt solution and heated to liody temperature, then kneaded thorougiily by hanil to expel as nuich blood as possible. C'annuhis were placed in the coronary arteries and carmine gelatine injecte(l under pressun*. thus filling the blood vessels with red. Herlin blue was then introduced (juite forcibly into the myocardiuui near the apex, by puncture with a hypodermic


n(;t'dle. The lymphatics of that area \von> thus siippos«Mlly irijfrt(>(l, the l)lo()(| v«'ssels licinn alrciuly full of H(lafii\o. The rclatioiis of the two systems of vessels should thus he i lemons! rate<l. The injected areas were cut out and sectioned. From their examination Hock reached the following conclusions:

1) I):is I.NiiipliKefitssiicl/, slelll ciii waiulhahities iiiteriiiii^kiilur verJaufendes Holiroiisyslein dar, gciiau wie die Hintkapillar ( !i-f:'.ss.systpni. nur nopli vielji;e.staltiner und reiciiliaitiniir.

2) Jede Muskelzelle liat ciii Lyiiiphkapillargefiiss vielfach auch dereii zwci und ein BhitkapiliarKefiiss.

3) Lymph- imd Biutkapillaren iiegen meist dicht iichen cinandcr.

4) Lyniphkaiiillareii verlaufen wie die Biutkapillaren parallel den Muskclzollcn und ist die Muskclfihrille hiiufig von einer Anastonios*zur langsverlaufcnden Lyniplikai)illarcn unispannt.

5) Die Lyniphkapillaren saniineln sicii in grosscn L}inphbahn<'n.

6) Eine wirklicho Koniiiiunikation rnit spalten zwischcn Muskeln konnte nirgeiuis naehgewiescn werden.

7) Die Muskelzellen Iiegen in nornialcn Herzen ohne Spalten diclit neheneinander und lasscn nur Rauni fiir die Lymph- und Biutkapillaren frei.

8) Spalten zwischen Muskelzellen sind pathologisehe und steUs mit Bindgewche ausgefiillt.

Thus Bock is of the opinion that the lymjjhaties of the heart» though following closely the pattern of the l)lood vessels, are much more numerous.

The most satisfactory method of approach to these problems in the studj' of lymphatics is through their emhryology. The lymphatic vessels under study are thus followed through several stages of embryonic life.

Such was th(> method of procedure in this work, and it is offered as an additi(ni to the series of emhryological studies of lymphatics which have come from this laboratory. Embryo pigs of three stages were injected. \i that time injection.s were made of both blood vessels and lymphatics of adult pigs' hearts.


Pregnant uteri, recently removed from the body, were obtained from an abattoir near the laboratory. The hearts of the end)rvos were still beating regularly when the injections were made.

4r>4 .lAMKS K. r.VSH

The eiiil)ry<)s were approxiniatoly of three stages: 1) 25 to 35 mm. in length, showin^r tl\e (iowiigrowth of the right and left lymphaiie ducts, l)otii ()l which may be seen gi\ing off branches to tht> trachea and hmgs. Also the right duct may be seen senchng off a branch to the heart forming a superhcial plexus on its surface. 2) Kml)ryos. 'An to 4') mm. in length, showing the formation of the aortic and tracheal plexuses by the right and left ducts. The tracheal plexus may be seen giving off a rather large branch to the heart which is joined near its origin by the cardiac vessel, seen in the earher stage coming from the right lymphatic duct. 3) Embryos, 45 to 65 mm. in length, showing far greater richness of the lymphatic plexuses. The superficial cardiac plexus, seen in even the earliest injections, has increased greatly in area anil its meshes are much smaller. The union of the two vessels from the right duct and the tracheal plexus is very well shown in embryos of this age. Embryos of greater length than 05 nuu. could not be injectetl, due to formation of numerous valves and nodes along the course of the lymphatic ve.s.sels which prevents the injection from flowing.


The embryos were injected with either a saturated atjueous .solution of Berlin blue or India ink. Number 2S hypodermic needles in an ordinary hypodermic syringe were u.sed for injecting.

Injection was made from four sites. In the smallest embryos, the needle was introduced into the right jugular sac. Hy this method the right duct, pa.ssing down the right anterior cardinal vein and giving off a branch to the heart, was filled. In the other specimens (45 to ()5 mm.), the tracheal plexus, giving off the cardiac ves.sel, and the tributary fn)m the right duct w(Te injected by introducing the needle into the region just dorsal to the superior vena cava, which has formed at this age. (Cunningham." Kl, Figs. 1 2.1 Most of the injections of these older en\bryos were made by placing the needle liirectly into the thoracic duct just above the diaphragm. The entire left side of the body wall of the embryo hurriedly dissected away and the needle quickly


placed into the thoriicic duct, which niay he accurately located just hetween the azygos veins and the aorta. When injecting, the pressure' must he sudden and rather j)()\verful. Kven when ihineedle is properly placed in the duct and the |)ressure properly applied, very few einhryos show heart lymphatics. This is quite different from the case of most lymphatics of the body which fill readily, once the injection mass has free access to them. There have been several instances when the thoracic duct has niptured under pressure without filling the canliac lymphatics. Pile explanation of this lies in a mechanical consideration of the origin of the cardiac vessels. The ready filling of the intestinal and mesenteric lymphatics upon injection into the retroperitoneal sac is evidently due to the lumierous and large branches of origin and the rich anastomosis on the vessels ,sj)ringing from them: if one pathway is blocked, the injection can go by another route. Then, too, in this location, the i)oint of injection where the pressure is applied is near the ves.sels to be filled. The sit nation of the heart is entirely different. The thoracic duct reaches the openings of the tracheal jilexus, a very dense and complex mjuss of ves.sels. From this plexus only one vessel goes to the heart, and this is the vessel that must be filled before the cardiac lymi)hatics can be injected from the thoracic duct. The tracheal plexus must be completely injected liefore the canliac branch will till. Injections of the heart are by no means a.ssure(l even when the proper techni<|ue is carried out in making them. This is due. in great part, to the fact that the thoracic duct opens into the veins. Thus, l)ack pressure is reduced, which accordingly greatly les.-^ens the chances of complete retrograde injection of the tracheal plexus from so low a region of (h(> thoracic duct. Such difficulty was elimijmted, to a great extent, by placing a clamp around the great vessels at the of the heart just before injection. So, too. in this way ink was prevented from entering the right chambers of the heart by way of the veins, as it inevitably did in specimens injectetl without the clam]) and so prevented the clearing of the heart for study. Injections were also maih- into the retroperitoneal sjic, and, though fiUing the lymphatics of the lungs, diaphmgm, stomach and intestines verv well, all faile«l to fill anv of the car


(liac vessels, the sac either nipluriiin under pressure or the injection cntcrinR the veins. Inunediately after injection the specimens Wire hardened and l)leached in ( "arnoy's fixing linid.. They were tlun transferred to 70 per rent alcohol and curried up the alcolioN to absolute hy a |)er cent changes and cleared hy the Spalteholz method.

The gross specimens were prepared by approximately the same method used by Hock. Instead of kneading the hearts by hand to expel the blood, they were perfused througli tlic coronary arteries with Locke's solution. It was found tiiat freshly filled hearts could be revived to contraction in this way and that more blood was expelled by their own movement than could be sfjueezed out by hand. During the perfusion, the heart was suspended in physiological saline heated to 37°C. .Vfter a fewminutes' perfusion, Locke's solution was replaced by a saturated solution of Hcrlin blue which was forced through the coronaries at 40 imn. Hg. pressure. Punctures were then made at various points into the myocardium with a hypodermic .syringe, injecting Intha ink into the muscle in the effort to fill the neighboring lymphatics. No matter how carefully this procedure was carried out, much of the ink flowed into blood vessels and the lymphatics injected at the same time could only be distinguished l)y their morphological characteristics: the color of the injected \essels counting for \'ery little in their identification.

It can readily be seen that this is a very crude method of injecting cardiac lymphatics; however, it is the only one available since retrograde injection, through one of the large lymphatic vessels, lying just under the visceral ])ericardiinn. is impractical. This is due to the stoppage of the injection by the numerous valves in the cardiac vessels which begin to form at about (U) mm. liy injections into the myocardium at ramlom, it is evident that it is only by chance that lymphatic vessels will be struck, and when once entered there is no reason to think that all the lymphatics of a given area will be filled. So, final conclusions should not be drawn from specimens .so injecteii without substantiating the pictures .seen in them by further investigation. In the attempt to obtain such evidence the following experiment was performed: —


The left thoraoif ravity of a cat was entered and several rihs just ahove the dinpliragin resected. The parietal |)ericardiiiiii was split and several drops of ink injected into the myocardium of the heatinp; heart. The ink that entered the l)If)od ves-seis w.iquickly washed through them into the general cin uhition. Part of the ink could l)e seen to enter the pericardial lymphatics inimediately, but a good part of it remained as a circumscribeil in the myocardium. The heart was allowed to heat from two to three hours in order that some of the ink might be absorbed by surrounding lymphatics. By this method no blood vessels were filled with which to confuse lymphatic ves.sels. The animal was then killed and strips of heart niuscle in the regioas of the injecteil drops of ink were fixetl in 10 per cent formalin, cut in paraHin sections, stained in alum carmine for microscopic examination. One hundred mg. of luminal sodium was used an an anesthetic and found to depress the cardio-respiratory centers very little. Since one lung was collapsed, artificial respiration was given throughout the entire experiment by Quinby's method of intra-tracheal insufflation ('()',)) at about 10 nun. llg. pre.s.sure. The animal's body temi)erature was maintainetl by placing the entire preparation in an asbestos lined box heated to 37'^C\ It was found that several drops of ink could be placed into the myocardium at various points without em! arassment of the heart's action.

This work was begun at the suggestion of Dr. R. S. Cunningham of this laboratory. I wish to thank him for the great a.ssistance he has be(Mi to me.


The lymphatics of the heart are found to arise from two sources: (1) the right lymphatic duct, which, in adilition. gives to a lung vessel; (2) the tracheal plexus, which also sends ves.sels to the lungs.

The thoracic duct is formed by the union of ve.s.sels growing down from the left jugular sac and another plexus of vessels which butl off from the metlian niesonephritic vein. Maetjer ('OS). The duct is shown to be a complete plexiform structure



in oinhryo pigs, 25 iiiiu. long (8al)iii i:?, Hiietjer 'OS. Kainpnieier ■12).

The' (icvelopment of tlip trachoal plexus has hccn carefully studied .lud deserihed by ( "uuniiitih;ini in his work <m hiriK lymphatics (U)) from which I (luote the fijllowing:

Almut midway U'twccii the jugular anastomosis antl the arch of the aorta, the thoracic duct leaves its position lat.<'ral to the trachea and hends downward to lie iiciir tlie dorso-lateral border of the esopliaRUs. Ill this position it comes down just bciiind tiie arcii of tlie aorta. Tiiis transition is shown by S;il>in i 1913, figs. r2-l.'i). .lust jis tiic duct begins to bend doi-saliy I lie ctirlicst sprout to the lung is f(»rmc<l. At this point a single large vessel buds off from the thoracic duct and ilown over the arch of the aorta to the hilum of the lung. Tliis vcs.sel usually persists in the adult as one of the drainage trunks from the hilac nodes to the thoracic duct. From the region of the thoracic duct, where this vessel buds off to a point about the level of the aortic arch, a mnnber of other vessels are formed very soon afterwards. These vessels very close together and grow to the lateral wall of the trachea where they and form the primitive left tracheal i)lexus. They lie in the undifferentiated mesenchymal tissue that surrounds the tracheal lumen. These lym|)hatics have formed a plexus i)v tlu! time the emliryo ha.s reached a leiiglli of 'M) nmi. From this plc.xus vessels grow across the trachea to with other vessels from the similar plexus on the opposite side; other lymphatics grow up the trachea and form a coarse meshed plexus around it. This is the aniage of the adult sujiply of that structure, lint the most nnportant of the branches of this plexus, as far as the present work is concerneil, are those from the lower part. These down the trachea, and, being joined 1)V other ve.<.sels that leave the duct near the arch, pass up over the bifurcation into the lung.

The primary lymjjhatics to the heart are evidently those coming from tiu' right lymphatics duct. Sabin (Origin and Development of the Lymphatic Sy.stcm, '1.3) states: "The right lymphatic duct curves ventrahvard and grows to the heart and lungs." Figure ]'\ of that work shows the right duct just anterior to the heart and iocalt'd dorso-laterally to the tracliea in a 2.") mm. pig embryo. Heuer ('09) (figs. 4 to 7) pictures a vessel from the thoracic duct wliich has every appearance of going to the heart. Though apparently leading to the heart, this vessel has never been .seen to actually reach it; always entering the aortic plexus. It has been very constant in all the injections, but has never been seen in anj' way to connect with the cardiac lymphatics.


Tho ri^ht lymphatic duct arises from the riRht juguhir sac and runs posteriorly, foilowiiij; the riglit anterior cardinal vein dnrsaliy. At the level of the right duet of Cuvier it divides into two hranches, one H"iiiK into the liiluni of the linig and the otlnr passing ventrally between the trachea and right duel of Cuvier behind the aortic arch and the pulmonary artery to reach the anterior aspect of the heart, just to the left of the eonus ateriosus. Injections of 40 mm. embryos show a large vessel branching fntm the tracheal plexus at the base of the heart and turning ventrally near where the vessel from the right duct bends down. These two immediately anastomose l)y a single vessel and are seen to run closely parallel to each other. The branch from the right duct is evidently j)rin\ary, preceding the branch from the tracheal plexus, which soon afterwards appears and follows its course to a.ssist in the formation of the cardiac plexus. In some of the speciiuens, these two cardiac l)ranches a|)iieared almost as a single vessel from their point of anastomosis. The vessel to the heart froni the right duct, together with the branch to the lungs, forms the early broncho-medijistinal trunk, the main pathway of cardiac drainage in the adult. However, they do not remain as two single vessels, but become plexiform in nature, receiving the drainage of the broncliial, anterior and posterior mediastinal, and sternal glands. The two small cardiac ves.sels, in the eml)ryo, pass anteriorly under the pulmonary artery, lying between it and the left auricle, to the auriculo- ventricular groove about .") mm. from the root of the pulmonary artery. Here, on the anterior surface of the heart, they form a i)rimary plexus which gives rise to two main branches which in turn branch to form the pericardial plexus. One main l)ranch passes to the left around the of the left ventricle, following the auriculoventricular groove, to the posterior interventricular septum, down which it sends a branch. The other branch of the primary plexus passes around the htxsv of the right ventricle and gives olT a large branch, almost eiiiial to itself in size, down the anterior interventricular septum, then pa.ssing around to the dorsal view of the heart, with the left branch. Thus it is evident that the lymphatic invasion of the heart takes place


along thi' l.'irgor bloinl vcssi-ls. as is cliani'toristic of ;ill lyiuphatir devolopnu'iit. The iiuiiii ncsscIs passing around the \ ciitricuhir bases and down the interventricular septa give ofT numerous anastoiiiiising hranehes along their course, in this way frjnning the ixricardial plexus. I'his plexus lies in the suhinesothelial connective tissue just under the visceral pericardium. I do not feel that any of the injections of this plexus have been complete; however, at 40 mm., the vessels pa.ssing down the intt'rventricular septa have been tilled to the apex and the pericardial plexus extending over at least one-half the surface area has been injected. At tiO mm. the me.'ihes of this plexus are much finer and more of the heart is covered. Whether injection is made through the right duct or the tracheal plexus there are only two vessels by which the heart hinphatics may be filled, so that, mechanically, it is very difficult to obtain good injections. In the cleared specimens of the GO mm. stage, no lymphatic vessels could be seen extending down into the myocardium, but along the larger vessels of the pericardial plexus short, lymphatic buds are visible, starting down into the heart nmscle. .Vt this stage none of the pericardial lymphatics can be filled by injecting into the myocardium, as is so readily done in the adult heart. This experiment has been tried on hearts from jiigs varying in size from tJO to 150 mm., and none of them show myocardial vessels to any great extent. Even in the 150 mm. embryos any of the surface lymphatics are only occasionally filled by injecting into the myocardium. Therefore, it appears that the pericardial plexus is formed almost completely before invasion of the myocardium takes place. Such inference seems logical, for it is evidently easier for the lymphatic growth to take place in the loose connective tissue under the visceral pericaniium than to invade the heart muscle itself.


Gross inspection of the adult heart reveals numerous lymphatics on its surface just under the visceral pericardium. They are slender ve-ssels of uniform size and form freiiuent anastomoses. Their course may l)e nearly straight or zig-zag. I'here is a


large lymphatic vossol runiiinp; down each side of the heart over the ventriculiir septum, and this receives a rooi! many tributaries. The other .surface lymphatic.^ eventually drain into the large vessels along the auriculo-ventricular groove. The meshes of the pericardial plexus are much larger, relatively, than those of the ()0 mm. embryo, at which time the plexus is approximate \densest. As the heart increases in size the lymphatics are stretched out, so to speak, to cover it, the plexus thus becoming less dense.

If a needle is ])lunged at random into the myocardium, some of the surface lymphatics can generally be filled. If a section is made of the ventricular wall and injection made from the cut edge near the pericardium, numerous surface lymphatics are always filled. If injection is made at a lower level, say about half the thickness of the ventricle, surface lymphatics will generally be filled, but they are not .so numerous as when injection is made nearer the surface. If injections are made at over two-thirds the distance from pericardium to endocardium, only occasionally will surface lymphatics be filled. Injections made just under the endocardium never fill pericardial lymphatics, and in no case was an endocardial plexus of lymphatics found. In several places the His bundle was entered, and I think that positive statements in regard to an endocardial lymphatic plexus are based upon injections of this system which have been mistaken for lymphatics.

These observations indicate that there is a myocardial plexus draining into the pericardial plexus; and it appears that the vessels of the myocardial plexus become smaller and less immerous as they extend into the heart muscle. The relative ditliculty of injecting them at different levels in the myocardium certainly calls for such an explanation.

The method of preparation of the adult cleared specimens in which both lymphatics and blood vessels were injected, has already l)een discu.ssed. By cutting free hand .sections about 2 nmi. in thickness from the injected regions, small areas can easily be found where l)lood vessels and lympliatics were both injected. The lymphatics are seen to follow the larger blood vessels, but


not as single acconipanyiiiK vessels as described \>y Hock. In the iireas where the injections seem most complete, each of the larger hinod vessels was seen to be surrounded by a scanty lymphatic plexus. The vessels making up these plexuses are all about the same size, and have comparatively few branches in comjiarison to the blood vessels. These lymphatics surrovmding the larger blood vess(>ls are connected by smaller vessels whicli complete the myocardial plexus. In many instances, these smaller lymphatics follow blood vessels, and, in others, can be seen to follow the connective tissue between the layers of the heart muscle. In no case were the lymphatic vessels seen to form a true capillary plexus which followetl the blood capillaries, as has been described. The lymjihatics anastomose freely while the blooil vessels, as is well-known, form few or no anastomoses. All the injected specimens give the impression that the patterns of the blood ves.sels and lymphatics differ \ery much; the lymphatics never forming a capillary bed, but a well-woven plexus of vessels, of which tlic smallest are larger than cajjillaries. The lymphatic vessels become fewer and smaller as they extend deeper into the myocardiun\, their niimlxT being exceedingly small near the endocardium. The largest lymphatics are only al)out onefourth as large as the blood vessels they accompany, and the smallest ones are about the size of venules so that it is evident the cardiac lymphatics are more constant in size than are the blood vessels.

These observations are supported by the injection of living cats' hearts. Pericardial lymphatics were immediately filled in all when puncture was made, and myocardial vessels draining into them were .seen in the sections. None of these myocardial lymphatics were seen to be as small as capillaries and their number was scanty. .Vbsorption of the ink does not take place readily. This indicates the scarcity of lymphatic vessels in the myocardium, since India ink is readily absorbed by the lymphatics in those places where the ves.sels are abundant. Such slow ab.sorption by the luuirt lymphatics would certainly not take place if their number did not justify it. In no case did any lymj)hatic vessels appear on the endocardium, which would have been most likely had they been present.


Most of the (jbsorvcrs (to whom I havi- already rfforred; have doubtless seen i}oth perieardial and myocardial lymphaties, hut have been alto|^ether too sweeping in drawing ri inclusions. Most of these vessels they have described have evidently been blood vessels. Most of Hock's illustrations of lyiui)liatics are perfect pictures of blood vessels, while none of them resemide lymphatics. By this method, it is impo.ssible to get complete injections of the l)lood vessels of the heart, and it appears that he has only described blood vessels injected with different colors. This point can be easily proven by observing a piece <if heart nmscle in which only blood vessels have been injected. In the double injections of the heart, many times the blood capillaries of an area will not be hlled. When lymphatics of the same area are injecte<l by puncture, some of the injection mass flows over into the unfilled i)lood capillaries, thus making a very misleading picture, large lynii)hatics and blood capillaries api)arently belonging to the same system of vessels. .Similar specimens prol)al)ly have been the basis for statements to the effect that the lymphatics of the heart form a capillary plexus accompanying that of the blood vessels, but exceeding the lilood plexus in density.


1. The lymphatics of the heart from two sources: a) the right lymphatic duct; b) the tracheal plexus. Hoth these paths remain as ways of drainage in the adult, the branch from the right duct being the aniage for the bronchomediastinal trunk.

2. There is first fonned a pericardial plexas covering the entire heart. This plexus reaches its greatest density at about IW) mm. and becomes less as the heart increases in size.

3. From the injections it apiiears that after the pericardial plexus is approximately complete, ve-ssels arising from it invade the heart muscle, and, by forming numerous, growing to great extent along the veins, form the myocardial plexus. This invasion of the myocardium takes place rather late, at i.)0 nun., there being relatively few myocardial lymph ve.s.sels demonstrable.


4. The myot-ardiiil plexus l^ecomes less dense as deeper points of the myocardium ;ire reached.

5. The larger blood vessels of the heart are accompanied by a which are connecteti with plexuses around other large blood vessels 1 >\ numerous anastomosing smaller lymphatics. These smaller lymphatic plexus composed of the largest lymphatic vessels ve.s.sels compose most of the myocardial plexus.

(). Cardiac lymphatics, in like manner to all other lymphatics, develop along the blood vessels.

7. Claims for an endocardial lymphatic plexus are evidently founded on partial injections of the His bundle.

8. The pattern of the Iymi)hatics does not correspond to that of the blood vessels nor are the lymphatic vessels so numerous as has been supposed.

It is intended that this description of cardiac lymphatics will be included in a more extensive discussion of lymphatics which will appear later and will include illustrations of all the specimens mentioned in this work.


Albrecht, E. 1887 Greifswald, J. Abel.

Baetjer, W. a. 1908 Am. Jour. Anat., vol. 8.

Bock, H. 1905 Anat. Anz.

BiDOE, A. 1880 Arch. u. Phys. Anat. Abth.

Clark, E. R. 1909 Anat. Rec, vol. 3.

Cu.v.M.VGHAM, R. S. 1916 Contributions to Embryology, vol. 4, .\o. 12, Carnegie Institution of Washington.

Hecer, G. 1909 \m. Jour. .Vnat., vol. 9.

Huntington, G. S. and C. F. W. .McClure 190(>-1 908-1910 Anat. Rec, vol. 1, Anat. Rec. vol. 2, Am. Jour. .\nat., vol. 10.

Kampmeier, O. F. 1912 Xm. Jour. Anat., vol. 12.

QuiNuv lltOO Johns Hopkins Bulletin.

Ra.vvier, L, 1895 C. R. Ac. des sc, vol. 121, p. 1105.

Sabin, F. R. 1913 Johns Hopkins Hospital Reports, Monograph — New Series, vol. 5.



RUTH RAND Department of Histology and Embryology, Cornell Univerxity, Ithaca, N. Y.



Introduction ITu

Historical 466

Material imd methods 469

Description of typical stages -I'O

Discussion 479

Conclusions 480

Literutnro cited 491


The followiiiR study' aroso from an invostigation of the dovelopment of the pharyngeal tonsil. As is well known, the pharyngeal tonsil in man has been found by several ob.servers to be in elose anatomieal assoeiation with a small blind poeket in the roof of the nasal pharynx, known as the bursa pharyngea or the recessus medius pharyngis. In order to diseover if possible the nature and signifieance of the relation between the pharyngeal tonsil and the l)ursa pharyngea, it was decided tr) study the origin and development of eaeh structure .separately. The study of the development of the bursa pharyngea was undertaken first, since it is present in the embryo before the appearance of the tonsil.

' Tlic study wiu* undertaken at the suKKestion of Dr. H. K. Kinitsbury. The writer takes (treat pleasure in ncknowlcdtinK the courtesy of the laboratory, and in exprcssinK lier appreciation of the very tjenerous help of Dr. Kingsbury.


460 KUTll HAM)

A mnnbrr of invostigators of roc«'nt yoars have been intereste<l in tlic detinito relations oxisting (iiiring onibryonic life between the chorda dorsalis and the epithelium of the pharynx. Many indeed have noted actual contacts between the notochord and i)li;irvnf!;eal entoderm and lia\'e regarded such contacts as factors in the origin of the l>ursa pharyngea. In a jiaper dealing with the condition a.s found in the human embryo, Huber ('12) states that of the nianunalian embryos, aside from those of man. which are ordinarily available for laboratory use. only the pig shows notochordal contacts with the epithelium of the i)harynx, and in the pig there is present a i)haryngeal pocket which has been considered homologous to the bursa pliaryngea of man. A study of the relations of the head chorda of this animal was therefore undertaken with the hope of throwing still further light on the relation of these notochordal contacts to the genesis of the bursa pharyngea.


A l)rief historical sketch will be given to outline the general views regarding the origin of the bursa pharyngea. P'or a complete review of the early literature on the subject the reader is referred to both Killian ('88) and Oppel ('00). Reference may also be made to Huber ('12) for the more recent work on the subject.

The bursa pharyngea was first described and named by F. J. C Mayer ('42;, who, according to Opjiel, found it present in many mammals including the pig. Luschka ('68) was the first to study the genesis of the structure. On the basis of an examination of a human monstrosity, this ob.server believed that the bursa pharyngea is a remnant of the primitive union of the mouth cavity with the hypophyseal anlage. I.,uschka held that the pharyngeal pocket f)f the j)ig is not homologous to the bursa l)haryngea f)f man (cf. subse(|. discussion). The observations of fJanghofTner ('79) and Schwabach ('87) did not confirm the view of Lu.schka. investigators could regard the i)ocket merely as a part of the pharyngeal tonsil, being a medial depre.s.sion of

i'liAin vi;kai. kimtiiklilm i\ 'iiii; pk. imukmi 4()7

the iiiucous mciiil)niii(' of tho pliarynx ;it tin- juiictinii of fhinu'clial with the lateral furrows of the tonsil.

Froriep f'S2) studicil tho (loV(>lo|)riH'iit of the head chorda in six human crnhryos, iiotinR ospcrially notochordal contacts with thf pharyiiffcal cpilhcliuni. In one of these a hnrsa phar^'n^ea was present, tiie hase of which was in close relation with an accumulation of chorda! cells which had penetrated through a hole in tho chordal sheath. Froriep could not positively state, but was strouKly inclined to the opinion that the strain exerted by tho notochord upon the epithelium of tho pharynx mechanically gives rise to the bursa pharyngoa. Killian ('88) made a careful study of the l)ursa pharyngoa, basing his observations on 05 embryos ranging in age from 'A to S months. 42 of presented a bursa pharyngoa. According to Killian, the bursa pharyngoa ia a structure sui generis, developing always in a definite region just in front of the uppermost fibres of the constrictor pharyngis superior in close relation with tho 'Hbrocartilago basilaris.' Killian criticised Froriep for overlooking tho invariable position of tho bursa. Frorii-p himself had shown that man.v accunuilations of chordal cells come in contact with the pharyngeal epithelium. On Froriep's theory, the definite location of the bursa cannot be explained, since anyone of those retropharyngeal unions could logically give rise to the structure. Killian extended his studios among many of the higher vertebrates including th(> j)ig, and C()ncludead ('(M)) not(>d con

468 Hl'Tll UA.VD

tacts between the iintuchdnl and pharyngeal epithelium in the pig onihryt).

It was not however until recently that Froriep's conception of a (•.lusiil relation ln>tween notochordal contact and the origin of the liursa pharyn^ea was revived. R. Meyer ('!()), basing his ol)servatious on a study of human embryos 2.5 to 40 nun. in length, concluded that the bursa pharj'ngea arises from the persistence of the original connection between the notochord and pharyngeal epithelium. From the fact that Froriep had noted a connection between the anterior end of the notochord and Sessel's pocket in chick embryos, and that Staderini had observed the .same for rabbit and .sheep embryos, Meyer was led to believe that the bursa pharyngea is identical with Sessel's pocket. A. Linck ('10) became interested in the early relations of the head chorda to the pharyngeal epithelium through a remarkable case of a tumor of the epithelium of the nasal pharynx which apparently involved notochordal tissue. Of the 1(5 human embryos examined, ranging from 2 to 25 cm. in length, 9 presented a bursa pharyngea. Although actual contacts between the notochord and bursa pharyngea were not observed, still the apex of the bursa in one case showed a more or less close relation with certain chordal remains in the retropharyngeal connective tissue. Linck believed that the appearance of the bursa may be explained partly through the tension exerted by the notochord on the pharyngeal epithelium, and partly through the active invagination of the pharyngeal mucosa it.self.

Huber's study of the origin of the bursa pharyngea, based on the examination of a series of human embryos ranging from 5 to 145 mm. in length, is the most complete that has yet been done on the subject. The early embryos (5 to S mm.), at the time when the notochord is separating olT from the pharyngeal entoderm, show the notochord maintaining a close contact with the epithelium of the pharynx. This point of contact is at the base of the ventral flexure of the notochord as it bends ventrad on passing into the cranial region, at a level just caudal to that of the thyreoglos.sal pit in the tongue, if there is more than one point of contact between the notochord and the pharnygeal en


t()<linn, tliis special one is always the most cHiKial, beiiiK constant in position with reference to the thyreoglossal pit. Of the embryos ran^inK in size from 15 to 30 mm., all hut one showed a (ievelopinn bursa, the blind end of which was in direct contai t with the nolochord, or bore a close relation to it. Still later, if there were any chordal remains in the retro|ihar>-ngeal region, these were all closely related to the bursa. Since in all cases the embryonic bursae were separated from Rathkr>'s and Se.ssel's pockets ljy practically the whole len^h <jf the vault of the pharynx, Huber was able to controvert the statement of .Meyer that the bursa pharyngea and Sessel's pocket an- identical. The constancy of the relation of the notochord to the pliaryngeal epithelium after the notochord begins to separate from it, and the definite location of the area of contact, lead Huber to l)elieve that this relation of the head chorda to the jiharyngeal epithelium is not accidental, and that thus to a certain extent the bursa pharyngea may be regarded as a structure sui generis. Mowever, the maintenance of the early caudal connection betwe<'n the uotoch(jrd and the epithelium of the pharynx seems necessary for the proper development of the bursa.

J. P. 'rourneux ('12) in an extended comparative study of the relations of the cartilaginous base of the skull in mammals, arrived independently at the same conclusion reganUng the origin of the bursa pharyngea. Tourneux believed that of the mammals examined only the horse anil man present a bursa pharyngea in the proper of the word. Tourneux describes the course of the head chorda in three i)ig embryos, 13, 18, and 5() nun. in length resjjectively ; in none of did the notochonl come in contact with the pharyngeal entoilerm.


The study was based for the most part on the examination of the i)ig embryos belonging to the Cornell collection, which consists of about !()() eml)ryos cut in frontal, .sagittal, and transverse sections, and ranging in length from 2.5 to 54 nun. vertex-breech measurement. Professor S. II. Cage very kindly loaned me a number of additional embryos and for tlxis I wish to expres>s my

47(1 HUTH lUND

!il)prcciatioii. A few eiiil)ry<is iM-loii^iiiK to studoiits in the departnuMit of Histology wen' likewise exaiiiiiied.

Of tlie embryos used, (hose tut in llie sagittal plane (about »)() in all' were found to be the most valuable for the study; but embryo.- cut iu frontal and transverse sections were very useful in cheeking up results and in adding to the data obtained. The drawings were made only from embryos cut in sagittal sections and chosen as being representative of the particular stage under discussion. The drawings consist of reconstructions of the midsagittal plane of that region of the head (pars chordalis) including the hypophysis, the basilar plate and the anterior vertel)ral anlagen. i'or purposes of orientation, however, the lirst figure is extended to include the whole head. The figures were made with the aid of the projection microscope.


Piy embryo 5 nun. i« length. Fi(/. /.

The notochord is seen separating ofT from the epitheliuni of the foregut. Posteriorly the separation is complete, se\eral rows of mesenchymal cells intervening between the notochord and the entoderm of the foregut. .\s the notochord passes forward over the pharyngeal region, it bends slightly downward, presents an enlargement and comes in contact with the i)haryngeal entoderm, which is here slightly thickened. This corresjjonds to the region emphasized by Huber in the human embryo as the seat of development of the binsa pharyngea. Slightly cephalad

Kin. 1 I'iK embryo. I H no. 1.57, Conu'll Collection, .5 mm. Reconstruction of midplane saRitlal Mection of the ph;iryn(;eal region. Xotoehord lilark. X 22.5.

I'"i({. L' Vm eiiiliryo no. 10. ( ornell Collection, 0.5 mm. Ueconslruelion of mlilpliine .saKiltal Mection of the pharyngeal rcfjion. Xotocliord black. X 20.

Kin. ;t I'lR embryo II a no. I.5(i, Cornell Collection, 7 mm. Reconstruction of midplane Ha^ittal section of the pharynKeiil region. N'olochord black. X 20.

Vif. 4 I'iu eMd)ryo II a no. 1.52. ('ornell ('olleetion. S..5 mm. Reconstruction of midplane sagittal section of the pharyngeal region. N'otochord black. X 20.

Kig. .5 I'ig embryo belonging to .Mr. R. .S. Cutsell. ll.."> mm. Reconstruction of midplane .sagittal section of the pharyngeal ri-gion. N'otochord black. X 20.

Kig. (i I'ig embryo III I), Cornell Collection. 11 mm. Reconstruction of midplane sagittal .section of pharyngeal region. Notochord black. X 1.5.


pk; KMiiKYo 471


of this region tho iiotochord presents anotiipr enliirf^ement and again comes in contact uitli tlio pharyngeal epithelium — and still again, anterior to this contact. Just before the dorsal curve of the Iiotochord as it nears the anterior end, the sheath comes in contact with the epithelium of the pharynx. The cephalic portion of the notochord presents a slightly waved course, passes over Sessel's pocket (<S), turns sharply ventrad, terminating in a rhickened notochordal plate in close relation with the base of the forel)rain. Transections of still younger pig embryos show that primitively the notochord is also in contact with Sessel's pocket, in a fasliion .similar to that .shown I)v W. J. Atwell ('15) for the rabbit. The notochordal plate at this stage bears no significant relation however to the anlage of Rathke's pocket (H), which is here present merely as the so-called hypophyseal angle, just anterior to the remnant of the pharyngeal memlirane.

Pig embryo 6.5 mm. in length. Fig. 2.

The mesenchpne between the neural tube and primitive esophagus has become slightly den.ser. The notochord at the end of its curve over the retropharyngeal region presents an enlargement and come-s in contact with the epithelium of the pharynx, which is here slightly invaginated. I'"'onvard the notochord has lost all primary connection with the entoderm, two or three rows of mesenchymal cells separating it from the pharyngeal epithelium. The position of the point of contact is slightly cephalad to the vertex of the angle of the pharynx and just caudal to the level of the lh>TcogIos.sal pit in the tongue. It corresponds in jiosition to that point of contact in human embryos emphasized by Ifiiber as the seat of development of the bursa pharyngea. Forwanl, the notochord bends dorsally over Sessel's pocket, ending in a hooked formation near, but not in contact with, the posterior wall of Rathke's pocket.

Pig embryo 7 nun. in length. Fig. S.

The notochord has lost all primary connection with the pharyngeal epithelium. It pa.sses forward in a straight, even course between the hindlmiin and the roof of the pharynx. Cephalad,


the iiotofhord iiros<Tits the charactoristif' dorsal riirvc over Sessd's pocket. Just hdow the Ix'iid of the iiotoclioril over tho n;tropharj'iigeal region, a condensation of mescnehyme may l)e observed extendiiifi; from the aiihiKe of the vertehral eoluinn to the vertex of the allele of the pharjiix. The pharyngeal opitheliuiii in this region is sUghtly 'pulled out.' The extreme cephalic end of the notochord of this pig is in direct contact with the .superior half of the posterior wall of Kathke's pocket.

Pig embryo 8.5 mm. in length. Fig. 4 The condensation of mesenchyme in the retro|)haryngeal region is now clearly defined and is seen in contact with the vertex of the angle of the pharjTix. At this point the pharyngeal epithelium has been 'pulled out' to form a distinct pocket (A'). The pocket hears no significant relation to the notochord, which courses in a wavy line dorsal to the condensation of mesenchyme. Forward the notochord continues in a wavy line between the hindbrain and the roof of the pharynx, ending in contact with the i)osterior wall of Rathke's pocket. The wavy course of the notochord was found to be characteristic of embryos a little older than the stage represented by figure ^. It can be explained only by supposing that the head chorda begins to grow faster than the surrounding mesoderm in the stage following complete separation from the pharyngeal entoderm. There remains to be noted the condensation of mesenchyme over the e.sophagus, indicating the anlage of the esophageal musculature.

Pig embryo 11.5 mm. in length. Fig. 5.

■^Phis eml>ryo presents a remarkaliie picture in which the notochord takes a distinctly wavy course over the roof of the pharynx and comes in contact several times with the pharyngeal epithelium. Of the fifteen embryos examined between the ages 10 to 12 mm., six showed similar though not as many contacts, the usual number l)eing three. There is evitlence for the belief that these contacts are secondary and 'accidental,' due to the e.xces.sive growth of the head chorda at this stjige (cf. sub,se<i. discussion). The pharyngeal wall at the vertex of the angle of the


474 inTii

pharynx has hciMi piillcii nut into a dofinito pookcl. whicli is in contact at its base with tin- coiulcnsjitioii of mcscncliyiiic extending dorsal to the retropharyngeal region in the way shown in figure I. This little out|)ocketing is very characteristic of embryos S to 12 mm. in length. In none of the pigs examined was it ever in contact with the notochord, but was always in relation with the retropharyngeal condensation of mesenchyme. In this enil)rvo the condensation of mesenchyme has extended over and .slightly beyond the pocket. Anteriorl}', the notochord comes in contact with Sessel's pocket and ends in the posterior wall of Rathke's pocket in al)out the middle region. Attention is called to the condensation of mesenchyme in the esophagus, similar to that shown in figure 4.

Pig embryo 14 vim. in length. Fig. 6.

The angle of the i)harynx has become greater as the cervical flexure apparently determining it has 'straightened out.' At the same time the outpocketing at the vertex of the angle of the pharynx has disappeanul. In none of the pig eml^r^-os examined over 12 mm. in length did this pocket persist. The condensation of mcsench^Tne which passes through the retropharyngeal region bears the same relation to the pharyngeal wall as in the preceding stages, the whole region ajiparently having lengthened relatively. The condensation is now readily recognized as the perichondrium of the developing basal plate and includes also the anlage of the fascia pharyngoba.silaris or ligainentum occipitopharJ^lgis, as it is .sometimes called. The notochord courses through the anlage of the basilar plate in a deep ventral curve, the waves shown in figure 4 having decreased markedly in amplitude. This is i)ossil)ly due to the fact that the head region in general has 'caught up' with the excessive growth of the notochord. Anteriorly the notochord has lost contact with Rathke's pocket, ending in a hooked formation near it. Contact with Rathke's pocket is lost usually at any time between the ages 13 to 18 mm. The of the forebrain is evaginated to form the infundil)ular process.

The roof of the i)harynx just at the opening into the esophagus is invaginated to form a distinct pharyngeal pocket, the base of



which is in closo contact with the unla^o of the esophageal nmsculatiire, which we liavo noted hcforc (^fi^s. 4, .")). It is important that this pociii't be clearly distinguished from the hursa shown in figures 4 and 5, the position of which was ce|)halad as indicated by the letter x. This pocket appears regularly after 13 mm. and is constant in occurrence. It is possible that it arises primarily from the tension exerted on the pharyngeal wall at this region by the developing esophageal musculature.

Kig. 7 I'ig embryo III F, Cornell Collection, 17 miu. Ueconstruction of raidplane sagittal section of pharyngeal region. Notochord black. X 15.

Pig embryo 17 mm. in length. Fig. 7.

This embryo shows the notochord extending through the ba.silar plate to come in contact tliree times with the pharyngeal epithelium. These are probably persistences of such secondary contacts as are shown in figUFc 5. The condition is by no means uncommon, three of the ten eml)ryos examineil ranging in length from i;^ to 17 nmi. having shown similar contacts. The epithelium at these points of contact is thickened slightly l)ut not at all evaginated. .Vnteriorly the notochord ends in a sharp, double twist within the basilar plate. Uathke's |)ocket has become constricted off from the pharynx, and Sessel's pocket has disappeared. The condensation of mesenchyme indicating the

476 iMTii m\i)

ptTidiondriinu of the basilar i)liit(> iuid the doveloi^ing fascia pharvnp;<>-l):isilaris has lost contact with the pharviiKal epithelium, ;i coiKUtiou which occurs regularly after the cml)ryo has attained a length of 1(> iimi. ("lose beneath the epithelium there has formed a second condensation of mesenchyiTu* indicating the tunica propria of the pharyngeal mucosa. 'i"he pharyngeal pocket above the entrance into the esojjhagus has become greatly enlarged and still maintains a specific relation to the developing esophageal musculature.

Pig embryo 35 mm. in len^jth. Fig. 8.

The vertebral centra and the basilar plate are present in the form of cartilage. The notochord on pa.ssing out of the ejiistropheus forms a knot, enters obliquely into the dorsal surface of the caudal end of the basilar plate, and through to the ventral surface in a long dorso-ventral curve, thence diagonally dorsal to its termination near the .sella turcica. In certain areas the notochord shows signs of degeneration. It presents also several nodal swellings similar to those ob.served by Killian ('88), Williams ('()8), Mead CO'Jj, and others.

The pharyngeal pocket (C.) at the entrance into the esophagus ha.s developed into a deep cul-de-sac in close relation at its blind end with the developing esophageal muscle. More anteriorly a separate condensation has appeared indicating the anlage of the constrictor muscles of the pharynx. These in the pig appear on the medial plane much later thiin the muscles of the esophagus, being recognizable first usually when the embryo has attained a length of about 23 mm. The angle of the pharynx has quite disappeared as such, as the head has 'lengthened out.' At some distance caudal to the upper limit of the j)harvng('al nniscu l"'ig. 8 I'iK embryo. Serial no. 1-17. Cornell Collection, '.ia mm. Reconstruction of mitlplane aajjittal Hcction of the plmryngeal region. Notochord black. X 10.

Fig. 9 I'ig embryo. Serial no. l.'i!), C'orncll Collection, 54 mm. Reconstruction of niidplanc sagittal section of the pharyngeal region. Notochord black. X 10.

Fig. 10 Pig embryo, special preparation, 170 mm. Reconstruction of midplane sagittal section of the pharyngeal region. X 4.


478 KlTIl HANI)

laturc ill the rcpinn of the limsc rctmpliiin ii^ful cimiicctivo tissue, socoiulary folds have appeared in the pliaryiigeal mucosa. The fascia pharyiigo-basilaris has become distinctly (lifTereiitiated from the perichondrium of the basilar plate, as dense fibrous tissue.

Pig embryo 54 mm. in leif/th. Fig. 9.

The most striking feature in the pharyngeal region of this embryo is the appearance of three distinct pockets in the mucosa of the roof of the nasal pharynx over the soft palate. The blind end of each pocket is in direct contact with strands of dense Hljrous tissue, the developing fascia pharyngo-basilaris. The l)Ocket which is most caudal (-1.) is the largest and corresponds in position to that of the human bursa pharyngea as figiu-ed l)y Ruber. It likewise occurs at the region designated by Killian as the place where one slu)uld look, in the pig, for a bursa pharyngea homologous to that of man, being at the level corresponding to the middle of the basal plate. There is no evidence that these pockets have ever been in relation with the notochord in the way shown by Huber for the bursa pharyngea of man.

The notochord courses through the basilar plate in a manner similar to that described for figure 8. It is interrupted, however, in the anterior portion. The pharyngeal recess is firesent at the entrance into the esophagus, separated from the three pockets by practically the length of the vault of the pharynx. The constrictor muscles of the pharynx and the musculature of the esophagus appear now as a rather continuous sheet in the median plane — an extension of the esophageal musculature having i)a.ssed underneath the pharyngo-esophageal recess (C.) over the entrance of the pharynx into the e-sophagus.

Pig embryo 170 mm. in length. Fig. 10.

The figure .shows the pharyngeal tonsil (T) developing as a series of accumulations of dense lymphatic tissue in close relation with a series of pockets or folds similar to those of figure 9 {A.). Each pocket is in contact at its base with strands of connective ti.ssue, the fascia pharyngo-basilaris. Cssihcation in


the basioocipital and basispherKjitl is now extensive. \o traces of chordal tissue were fdiservcd. The pharyiiuea! refess above t}ie entrance into the esophagus has heconir L'rcatly fnkied and enlarged.


Relation of the notochord to the pharynyeal epithelium. .Attention lias recently l^een directed by numerous obser\'ers (Mrs. Gage, '0(), Mead, 'Oi), Hul)or, "12) to the exi.stence of contacts between the notochord and the pharyngeal epithelium in the pig embryo, which were thought to be similar to those in man. It was therefore to this phase of the study that particular atf(>ntion was paid, since in man according to Froriep, Linck, Tourneu.x. and Huber, notocliordal contact seems intimately bound up with the development of the bursa pharyngea. The results of my study were somewhat negative, in that the relation between the notochord and the pharyngeal epithelium in the pig eml)ryo wa-s found to be quite different from that describe<l for the human embryo. In the first place, the notochord of th(^ pig at an early stage seems to become completely separated from the epithelium of the pharynx, grows more rapidly than the surrounding tissues, and th(>n may acquire secondary contacts with the pharj-ngeal epithelium. It is probable that these secondary contacts are the ones noted by Mrs. Cage ami .Mead for the pig embryo. Notochordal contacts which Linck, Tourneux and Huber described for man are apparently primary contacts which the notochord maintains with the j)har\7igeal epithelium. In the second place, notochordal contacts in the pig ('nil)ryo were never found to bear any specific relation to the pharyngeal pockets.

The early pig embryos — 3 to (i mm. in length — present notochordal contacts with the epithelium of the i)harynx (juite sinular to those figured by Huber for the early human eml)rj'os. These primary contacts which the notochord maintains longest with the pharyngeal entoderm are somewhat posterior in position, usually caudal of the level of the thyreoglossal pit of the tongue (figs. 1 and 2). Embryos of approximjitely 7 mm. in length showed the notochord as a rule completely separated from



the pharyngeal epithelium. Tlu> imtochord of those embryos passes in a straight even course hefweeii fhe hiii(ll)raiM and the roof of the pharynx (fig. •i). At tlie time when tlie embryo has attaineil a length of 8 to 10 min. the notoehonl apparently iindergoes a period of excessive growth. Figure 4 presents a typical picture of this stage, and shows the notoehonl passing in a deeply waved course over the roof of the pharynx. Between the lengths of 8 to 10 nun. the notochord at no time showed true contact with the pharyngeal epithelium. On the otlier hand, in a large percentage of enibrj'os lU to 12 inm. in length, the notochord came in contact several times with the epithelium of the pharynx (fig. 5). The notochord still presents a distinctly wavy course. These contacts ditTer relatively in position from those of the -younger eml)ryos, lying well in front of the level of the thyreoglossal pit in the tongue. A few of the older embryos showed jiersistences of these contacts even after the anlage of the basilar plate was well formed (fig. 7). Thus the statement of Tourneux that the notochord of the pig embryo is always entirely intrabasilar requires modification. In none of the pig embryos at my disposal that were over 17 mm. in length, were notochordal contacts observed, although Meatl figures a condition similar to that of figure 7 for a 30 mm. embryo. This is clearly an exceptional case. At no time were notochordal contacts found in relation to pharyngeal outpocketings.

'1 he following data gives in tabular form the evidence for regarding as secondary those notochordal contacts with the pharyngeal e[)ithelium which are present in jiig emt)ryos 10 to IS mm. in length:

LC!>ram or























10 24 7 15 10 66

8 6

6 3

80 25

40 33

4 3



3 1

3 2

Caudal Caudal

Cephalic Cephalic

Even Even Wavy Wavy Deep curve Deep curve


Xotiichonl (iiul In/pophijsis. In iiiakiiin tlic al)OVc ohsrrvations the writer has hud occasion to study aiso the relation of the notochord to the hypophyseal anlage. A ii umber of early investigators, M. Miller ('(kS) and Dursy ('08), have held the viewthat the notochord acts mechanically in drawing out the infundibular process, basing their belief on the fact that in early stages the notochord is attached to the base of the forebrain. Observations on the pig embryos at my disposal do not permit me to confirm this view of the relation of the chorda and the infundibular process. The early embryos, 2.5 to 3 mm. in length, at a time when the anterior end of the notochord was not yet folde<J off from Se.ssel's pocket, showed the notochord in contact for a short distance along its dorsal surface with the biuse of the ff)rebrain. Several of the embryos, 4 to 5 mm. in length, at a time when the .separation of the notochord from Scs-sel's pocket is complete, also showed a point of contact between the notochord and the of the forebrain, similar to that shown in figure I. These however were not true contacts according to the criterion of Woerdeman, being more juxtaposition of parts, since the notochordal sheath and margin of the brain wall could be clearly seen between the two structures. Later the notochord and of the forebrain are separated by the ingrowth of mesenchyme (fig. 2), which occurs ordinarily when the embryo hiis attained the length of 6 mm. The infundibular process, however, does not begin to appear until the 11 to 12 mm. stage, at a time when the base of the forcl)rain is free from notochordal contact. .V cau.sal significance, therefore, cannot be attached to the relation of the notochord to the base of the forebrain in the appearance of the infundibular process.

Other investigators have hold that the notochord is respon.sible in a similar fashion for the origin of Rathke's pocket. Among these may be mentioned Reichert ('40) and His ('68). Mihalkovics (74) denied that the chorda is resonsible for the drawing out of the hypophy.soal .sac in the rabbit, on the grounds that its attachment is to the inferior portion of the posterior wall of Rathke's pocket instead of to the apex. Woerdeman ('13) however has pointed out that this argument is not conclusive, since


in the pin rnihryo the notochonl is in contact with flic superior portion of the posterior wiill of Kathkc's pocket; nevertheless Woenlenian does not consider this a positive arRiinierit for the existence of a causal relation between the chorda and Hathke's pocket. In HH5 \V. J. Alwcll published observations on the relation l>etween the notochonl and Hathke's pocket in the rabbit and chick embryos. Contacts between the chorda and Rathke's pocket were constant in the chick, but existed in a very small jjcrcentage of cases in the rabbit. .Vtwell explains notochordal contacts with Hathke's pocket as secondary, being due to the growth of the forebrain and sharpness of the cervical flexure. In no case could he find that the entoderinal contribution to the hypophjsoal sac was anything more than an accidental union of parts. Recently the old idea of the notoehord acting as a causative factor in the origin of Hathke's pocket has been revived bj' M. M. Miller ('15). In a preliminary note on the development of the hypophysis in the pig embryo, he a.s.signs a mechanical function to the notoehord in the gene.sis of the structure, and .states that the entoderm as well as the ectoderm enters into its structure. According to Miller's observation, the notoehord pulLs away with it from the foregut a mass of entodermal cells, which at first comes in relation with the apex of the hypophyseal angle. Later the contact shifts from the apex down to the constricted stalk of the hypophyseal sac. I am unable to confirm these observations of Miller on the pig embrj'o. In none of the pig embryos examined under (i nun. was the notoehord in contact with Hathke's jjocket. All of the emiiryos 4 and 5 mm. in length and a few of the (i mm. embryos .showed the developing Hathke's pocket wholly free from notochordal contact. .After the embryo has attained a length of (5 or 7 mm. the notoehord regularly comes in contact with the superior half of the posterior wall of Rathke's pocket. The contact is brought about probal)ly through the excessive growth of the notoehord over the rest of the head, which begins at this time. The contact between the chorda and Rathke's pocket was found to be entirely similar to that described by Woerdeman, being a true contact without any intcr\ening membrana propria. There was no


cN'itlpiice that the ilionhi conlrihutcd any eshcntiul CKiiiponcnt to the structure of the hy|)oi)hysis liowever. Any entodermal tissue that might he found later in the develoiHri^ hypophysis may he exphiinod as the result of an areidental union with the not*iehord. In later stages the notoehordal eontaii usually shifts down relatively from the superior to approximately the middle region of the posterior wall of Rathke's pocket. ( 'ontact between the notoehord and the hypophysis is usually iiiainlained until the embryo attains a length of I'.i to 15 mm., wIhh it is lost through the ingrowth of mesenchyme.

The pharyngeal pockeUi. The material pre.sented shows very clearly that it is necessary to distinguish among three different pharyngeal pockets in pig embryos, —each one of which has appeared in the literature under the name bursa pharyngea or reccssus medius {)haryngis. It has been the main object of this study to determine which, if any, of these pocket.s is the true homologue of the bursa pharyngea of man.

With the exception of Minot, no one to my kn(jwledge has ever noted the pocket which has been shown as the first to appe&r, and wiiich at the \ertex of the angle of the pharynx (figs. 4,X, 5, and G,X). In an outline drawing of a model of the" pharynx in a 12 mm. pig embryo as seen from the dorsal side, Minot ('11) figures this pocket at the summit of the pharynx at the level of the third gill i)ouch. and calls it the bursa pharyngis (fig. 28, p. 26). In a drawing of the profile view of the phar>-nx from a wax reconstruction of another 12 mm. pig embryo, Minot again figures tliis pocket, but gives it no name. Farther down in the pharynx above the entrance into the esophagus and on a level with the fourth gill pouch, he figures the pocket which has been shown is the .second to appear, digs. (> and 7i and calls this the pharyngeal bursa ((ig. 173, p. '2'.i7). Thus .Minot failed to distinguish between the two pockets. In none of the pig embryos at my <iisposal were both pockets present at once in the fashion .shown by Minot, though the condition is probai)Iy not uncommon.

.\lthough the pocket at the vertex of the angle of the |)harynx which is fh<> first to d(>veIop, is very cunstant in appearance


jiiiKiiin S to I'J HUM. pin cinljryos, I iicvortheloss Ix-lii-vo that no spi'cial niorplKildgical sinnilicancp can 1)0 attached to it. It«  close approximation to the point described by Huber as the seat of (le\rlnpnu'nt of the bursa pharyngea of man, sugKosts the possiliility that this j^ocket in the pig is homologous with the human iiursa. Careful study shows, however, that the region where the notochord longest maintains its primary attachment witli the pharyngeal entoderm was some distance anterior to the point where the pocket develops. The pocket persists only up to the time when the angle of the pharynx increases as the 'head i)ending' which seeminglj- determines it 'unbends.' In none of the seventy-six embryos over 12 mm. in length that were examined did this pocket persist. I therefore believe that it should be regarded merely as a temporary out-pocketing of the pharyngeal epithelium arising mechanically in growth.

The pharyngeal pocket (figs. 8 and 9, C) which has been shown as the second to develop, appearing usually when the embryo has attained a length of 12 mm. and which is situated in the roof of the pharynx over the entrance into the esophagus, has likewise been called the bursa pharyngea or the recessus medius pharyngis. Luschka' ('68) was the first to point out that this pocket in the pig is not homologous to the bursa pharyngea of man, since the structure in the pig is constant and well-developed, whereas the bursa of man is inconstant and rudimentary. Later Killian ('88) also denied that this pocket was homologous to the bursa pharyngea of man. He figures the pocket in a 65 mm. pig embryo as Ijing in the roof of the pharynx at the level of the l)oundary between the second and third vertebrae and opposite the entrance into the larynx, at the .same time designating the place where one must look for a pharyngea comparable to

' "Mit jencm rudimrntiircn, iiberdics nicht rcgcltnitssigc vorkommenden beutolformiKpn .\nlian(je <lpr Pars n!itiali.s hat die constanlc. cine geselzmassige Art von Pharynx sociindaiiu.s darstellendc .Vussackung der Dorsalwand des Pharynx boitii .Schwcin nichts Koinein. Hei diescm Cifschcppfe stiilpt sich niimlich in der Hichtung gegcn die .Speisemhrc die Schleimhaut der IJorsalwand zu einer 4 cm. langer cylindri.Mchcn fingcrdickcn Taschc aus, dercn abgerundete, freiesKnde den Hand des Constrictor pharyngis inferior ftberragt, und also schon am unzerlegIcn Urgane sichtbar ist."


that of man, as farther up in tho vault of tlio pharynx at a level eorresponding to the middle of the hasal plate. Since Killian, however, this poeket has been called the l)ur>a pharyngea or median pharyngeal recess by writers among whom may be mentioned Sisson Cll), Stilling I'll), and Minot '11 . The present study has convinced the writer that Luschka and Killian were right in their statement that this pocket is not homologous to the bursa pharyngea of man, since it differs from it both in mode of development and in its anatomical position. The bursa pharyngea in man has been shown to develop in relation with the notochord and the oceipito-phar3'ngeal fa.scia: the pocket under discus.sion arises in close relation with the developing esophageal mu.sculature, (figs. 5, 6). The bunsji ])haryngea of man is constant in position, being situated always anterior to the upper boundary of the constrictor muscles of the pharjiix at a level corresponding to the middle of the basal plate, and bears a definite relation to the fa.scia pharyngo-ba.silaris and the pharyngeal tonsil. This pocket of the pig, however, is situated in the posterior part of the pharyn.x just over the entrance into the esophagus at the level of the boundary between the second and third vertebrae, and bears a close relation to the esophageal musculature. It would therefore be advisable to call the pocket by some other name, as, for example, the pharynpo-esophageal recess. The recess is constant in appearance after the pig embryo has attained a length of 13 mm., and is therefore of morphological interest. It is not to ray knowledge present in any other mammal with the possible exception of the horse. .\s was shown in the figures, it develops in close contact with the developing esophageal musculature, possibly arising from the tension which muscles exert on the pharyngeal wall of this region during the growtli of the head. Dissection of older pig embryos showed that the laryngeal apices fit into the recess thereby probably enlarging it. It may be that the laryngeal apices act in conjunction with the to shut off the (>sophagus. thus affording a more direct channel for the pas.sage of air. In the adult the recess is described as 4 centimeters deep, and often the seat of |)ath()logic;i! disturbance due to the fact particles of food become lodged within it.


Of tho pliarviiKcal pockets which appear in the development of the pin ciiihryos. I hcMeve that the series whirh was shown to arise hist (tig. 9, ^4.) in dose rehition with the (leveioping fascia pharyiiiro-basilaris, is most nearly homologous to the bursa phar\ iijica of man. .Viiatomically, the most caudal of the series of ]iii(kcts in the pig emluyo bears the .same relations to the surrounding structures as does the bursa pharyngea of the human embryo, — being situated just in front of the upper limit of the superior constrictor pharyngis at the level corresponding to the middle of the, in close relation with the fascia |)harnygo-ba.silaris and the pharyngeal tonsil. On the other liand, it was found irnp().ssii)le to compare step by step the de\ elopment of the structures in the two embryos, since the relations of the head region differ greatly in growth. Until the (» or 7 mm. stage however, the general relations of the head region are the same in the two embryos. Huber has shown that the human bursa pliaryngea develops at the most caudal j^oint of contact which the notochord maintains longest with the pharyngeal epithelium. In the human embryo the attachment is maintained until the develoi)ment of the bursa pharyngea. Pig embryos 5 to 7 mm. in length usually show the notochord maintaining its connection with the pharyngeal epithelium longest in the region corresponding in position to the point of contact emphasized by lluljer in the luiman embrjo. \\'ith the growth of the pig embryo however, this connection is early broken, and thus the landmark is lost which would enal)le one to locate the bursa pharyngea of the pig embryo as arising definitely at this point. On the other hand, the older pig embryos show the most caudal of the series of pockets under discussion in the same general anatomical relations as the bursa pharyngea of man at a similar stage of development, -being situated in the roof of the anterior portion of the pharynx just in front of the upper limit of the nun. constrictores j)haryngis at the level corresponding to the middle of the l)asilar plate, and in close relation with the developing fascia pharyngo-ijiusilaris. The fact that in both the human and pig embryos of younger stages the point of attachment between the notochord and the pharyngeal epithelium, which is

i'II.\kvv(;kal kpithhijim in iiik i'k; kmbrvo 487

most caudal and which is inaiiitaiiicd longest, is in th<' sumie position; the fact that the bursa pharyngea of the human ('ml)ryo arises at this |)oint of attachment: as well as the fact that the bursa pharynRca of the human embryo and the most caudal of this series of pockets in the older pig embryos possess the same anatomical relations, incline one stronKb' to the belief that this pocket in the pig embryo arises at the point where the notochord maintained its attachment louKest with the i)harynKeal epitheUum, and therefore may be considered comparalile to the bursa pharyngea of man. The early loss of the relation of the notochord to the pharyngeal epithelium in the pig embryo complicates a direct comparison.

Concerning the part played l)y the notochord in the development of the human bursa pharyngea, Froriep, Linck, Tourneux, and Huber have published observations, which indicate that the tension of the notochord on the pharngeal epithelium is the essential factor iji the origin of the bursa. Froriep believes that the occipito-phiuyngeal fascia is another possible factor. Linck believes that to some extent the pharyngeal mucosa invaginates of itself. Tourneux considers the occi|)ito-pharyngeal fascia an accessory factor in the genesis of the bursa. From his observations Tourneux' is led to distinguish between two kinds of pharyngeal pockets, the simi^le pharyngeal recess and the bursa pharyngea, which in the horse may l>oth be present at the same time, and often not separated, the recessus opening into the bursa directly. In the pig a series of pockets (fig. !)) corresponding in general position to the bursa pharyngea of man has been shown to aj)pear independently of the action of the notochord, ^the develop ' "II convieni do ilitTorencier le recessus nu'di.'in ilu phiirynx d'avec la bourse pli:iryngicnc . . . . Le roccssus sans rehition ilirocteavec la cliorde paralt ri'.sultcr de I'onflcxioii copliuliquc di'tonninant I'anislo dii pharynx, et surtout dea adhiTi-nces que lo liK:iment <iccipito-|iKaryiiKienni> contraole avoc la muqueuse du pliarynx. (Juant A la riiniiation ile la bourse pharynRii-nne. olio est provoqude ainsi tpie la inontric Krorie)). par une adluTonce locale (|ue la cliorde a conservtfe avec rendodcrnie dans I'epaisscur duquelle elle elait priinitiveinent enclavec le lonKc de la ligne niedianc. .V cette adherence vient s'ajouter coinine cause adjuvantc le ligament occipito-pharyni;icnno. dont riiusertion sup.^rieurc n'po nd i la parlie infi'-rieure de la bourse."


iiiR o(Tipito-pharynK<\il f;iscia ai)i):ir('iitly actinp; as an essential factor ill their origin. HuIxt discussos the possihle action of the occipito-pharynReal fascia in the development of the hinnan bursa pharyn^ca. In one of the embryos at his disposal there was present a small bursa in dose relation with a well-developed ociipito-pharyngcal liRament and free from notochordal contact. The fact that there was present only an imperfectly develo|)ed bursa independent of notochordal contact on the one hand, and a well-developed fascia pharyngo-basilaris on the other hand, seemed to llul)er to argue against regarding the fascia as a factor in the origin of the bursa. Nevertheless there was present a bursa pharyngea. Moreover, examination of Huber's detailed drawings shows that at the time when the pharyngeal epithelium becomes distinctly invaginated to form the bursa, the mesench>TTie about the notochord at its point of contact is thickening to fomi a sheath, later incorporated with the fascia pharyugo-l)asilaris. Hul)er figures also a (50 mm. human embryo in which there is present a well-developed bursa pharyngea in direct contact with the developing fascia pharyngobasilaris, which involved at some distance from the bursa some degenerating chordal remains. It may be suggested that while the human bursa pharyngea regularly develops in contact with the notochord, it arises not so much through tension exerted on the pharyngeal epithelium by the notochord itself, as by the sheath of (developing) connective tissue surrounding the notochord. However, maintenance of the notochordal connection with the pharyngeal epithelium may influence the mesenchyme to condense earlier, since the bursa together with the fa.scia begins to develop in the human embryo at a relatively earlier stage than ill the pig. If this be the, Tourneux's distinction Iwtween the bursa pharyngea and sim|)le pharyngeal recess is not valid. Indeed the fact that both kinds of pockets may be found at once, the one in contact with the notochord, the other with the occipito-pharyngeal fascia, together with the fact that sometimes these pockets are inseparal)le. indicates that there is no essential difference between them. Killian's distinction between the bursa pharyngea and the simple pharyngeal recess is


interesting in roniparison with that of Tournoux: "Dua Wrhitltnis zur Filirocartilage basilaris bietet ein wichtiges Kriteriiim zur 10ntschri(hing zwisohen cinoin einfachos Rocossus und einer wirkUchcr liursa. Krstcrer gohort stets zu tier Schloimhaut an." The appearance of a series of bursae in embryos of the pig and of the horse (Tourneux), instead of a single bursa as in the human embryo, may be explained by the fact that tlic tension exerted by the fascia pharyngo-liasihiris extends over a longer area, corresponding to the longer Une of the crista phjiryngca of the ba-sioccipital, which, as Skoda ('12) points out, corresponds in long skulled animals to the tuberculum pharyngeuin of man.

As is shown in figure 10, the pharyngeal tonsil (T) develops at the site of this series of pockets (figs. 9 and 10). I believe that the name bursa pharyngea is preferable to that of recessus medius pharyngis. The latter tcnii was originally proposed by (langhofner (79) who, with Schwabach ('87), believed that the pocket is merely a part of the pharyngeal tonsil, — being the junction of the medial with the lateral furrows. By way of comparison with the rece.ssus laterales, (fossae of Rosenmuller) he suggested therefore the name recessus medius pharyngis, but oddly enough proposed that the term bursa pharyngea be retained for the pathologically over-developed form. The study of the genesis of the pocket, however, in both the human and pig embryo has shown it to be an independent anatomical structure. For this reason bursa pharyngea is the preferable term.


1. The relations of the notochord to the pharyngeal epithelium in the pig embryo differs essentially from that described for the human embryo, in that

a) The notochord early becomes separated otT from the pharyngeal entoderm, and then may acquire .secondary contacts through excessive growth and 'bendings.'

b) The notochord is never in contact with pharyngeal outjjocketings.

2. It cannot be .said that in the pig embryo the notochord either acts mechanically in the drawing-out of the infundibular



process ami hyi)opliysoal siif, or (■iintril)ute.s any essential i-innpoiieut til the structure of the hypophysis.

{. In the pip; eiiihryo (listinrtii)ii must he made anidhfi three

(lilTerent jtharyngeal pockets:

a) The first pocket, which can he regarded merely as a temporary invafiination of the pharyngeal wall at the summit of the pharynx, arising from the ten.sion exerted on the wall by. the flexion of the head in its growth.

Ill Tlie pharj-ngo-esophageal recess above the entrance into the esophagus, a constant structure in the pig, possibly arising from the ten.sion exerted on the pharyngeal wall liy the developing csojihageal musculature.

c) A series of pockets which develop in the anterior portion <if the roof of the pharynx in front of the constrictor nm.scles and which are comparable to the bunsa pharyngea of man.

•1. In the pig embryo, the series of pockets corresponding to the bursa pharyngea of man arises independently of notochordal contact — the developing fascia pharyngo-basilaris acting as the essential causative factor.

5. The pharyngeal tonsil in the pig embryo develops in close connection with the series of pharyngeal pockets corresponding to the bursa pharyngea of man.



Atwkm.. \V. J. lOI.'i The rt'lution of tho chi>rijii ilor.xjilJH to the entodcrmal

comporienl uf the hyi)oi>hy«i». Anal. Kcc, vnl. 10. pp. Ift-.'Mi. FitoKiEP, A. 1882 Kopfthcil der Chonia Doreali.i bei ineiuchlichen pjinbryonen.

Als Fe8t(?iibf fiir Jiicob HeuU;, DoDn, pp. 2ft-l(l. (tAOE, ScsANNA PiiKi.i's ISKMi Thc notochorcl of the hciid in human embryos of

the third to the twelfth week, and cotnpari.son with the other viTtc brates. Science. New Serici. vol. 24, pp. 2!).VL'iKi. (Ia.vgiiufnek, I'"lilKi)F<iiii IST'.t Cber die Torusilla und Btirxa pharyogeii .Sit zungsbcr. d. k. .Vkad. d. wiss. in Wien. Hd. 7S, pp. 1S2-2I3. Hi;iiKi<, (r. ('. 1912 On the relation of the chorda (li>r-:ili~ tn thc aniagc of the

pharyngeal bursa or the median pharynRi-al n ri«K Anat. Kec, vol.

6, pj). ;}7.i-tOI. KiMjAX, G. 1888 Cber die Bursa und Tonsilla I'haryngeu. Morph. Jahrb.,

Bd. 14, pp. 618-711. KoEi.LicKEB, A. 1879 Kntwicklungsgeschichte des Menschen und dcr hoheren

Thierc, pp. 426, 441, 4.5!), ,S2.S-,S.J2. Ll.NiK, .V. 1911 Beitrag zur Konntnis der inenschlichen ("horda dorsalis im

Hal.s und Kopfskelet. Anat. Heftc, Bd. 42, pp. H(),>-7:i6. Lu.sciiKA, H. V. 18t)8 Dcr .Schlundkopf des .Menschen. Tubingen, pp. 24-27. .Mead, C .S. 1909 The chondrocraniuiii of an embryo pig, sua .scrofa. .\m.

Jour. .Vnat., vol. 9, pp. 167-20S. Mevek, H. 1910 l^er die Bildung des Uecessus pharyngcus mcdius. Bursa

pharyngcH, in Zusanimcnhang mit der Chorda des menschlichen Gm bryonens. Anat. Anz., Bd. .37, pp. 449-4.>3. MlLHALKoyiES, \'. 1874 Wirbel.saite und Hirnanhang. Arkiv. f. Mikr. .\naf.,

Bd. 11, pp. .389-431. .Mii.LBR, .M. .M. 191.i .\ study of the hypophysis of thc pig. .Vnat. Rcc, vol. 10,

pp. 226-228. Mi.voT, CiiA.s. S. 1911 -V laboratory text book of embryologj". Second Edition, revised. P. Blackiston's .Sons & Co., Phila.. pp. 62, 2.37. Oppbl, .\. 1900 Lehrbuch der vergleichendcn mikroscopischen Anatomic der

VVirbelthiere. Dritter Theil, Fischer, .Jena 1900, pp. 10;{-117. ScHWAnAtiT 1.S.S7 Zur Kntwickelung tier Ftachen-tonsille. .Vrch. f. .\Iikr.

.Viial.. Bd. t)2, pp. 187-213. Si.ssoN. S. 1911 \'eterinary anatomy. Saunders i^t Co.. Phila.. pp. 415-416. Skoda. K. 1912 Die sogennanten Tubercula pharyngea der flaussiiugetiere und

(lie .\nsatzverhaltnisse der Kopfbeugemuskel an der Sclmdelbasis.

-Vnat. .\nz.. vol. 42, pp. 33-37. Stillixo, Cieo. 1911 Die Ilachenhohle. die Hiirlrompete. und dcr l.uftsack

des Pfcrdcs. Kllcnberger's Handbuch dcr Vergl. mikr. .\nat. der

Haustiere, Paul I'arey, Berlin, Bd. 3, pp. 114-124. TimuNEii.x, J. P. 1911 BAse Cartilagineusc du CrAne et Organcs .Vnnexea.

Toulouse Th^se. vol. 71. Williams, L. \V. 190S The later development of the notochord in mammals

.VnL Jour. .Vnat., vol. S, pp. 201-2.S."). Woerdemann, M. VV. 1913 Cber einen Zu.sammcnhnng der Chorda Dorsalis

mit der llypopliysen .Vnlagc. .Vnat. Vnz.. Bd 43. pp. 37.'v-38S.

AirrnoHA AitsTftACT or thir papbr lasccD bt



SAMUEL RANDALL DETWILER Otborn Zoological Laboralory, Yale Univernly


Nile blue sulphate as a vital stain has been successfully used for determining the relation of various parts of the egg to the future embtyo. The first appUcation of this method was made by Goodale ('12) in a study of the development of Spelerpes. After removing the outer jelly from the egg, the inner capsule was punctureti, thus allowing the perivitelline fluid to escape. Dry stain which was then applied with a needle to desired positions on the jelly worked its way through the jelly and produced stained areas on the egg. These stained areas could be followed during the shifting of the egg substances, thus enabling a determination to be made of the relation of these parts to the resulting embryo.

Except for slight modifications in application, this same method was used by Smith ('14) in an experimental study of concrescence in the embryo of Cryptobranchus allegheniensis. In this case an aqueous solution of proper strength was applied by means of a fine pipette after the gelatinous capsule has been removed from the egg. Distinct blue spots were thus produced which did not wash out and which were not sufficiently toxic to interfere with normal development. It was observed that the stain did not spread to any extent by diffusion and that stained areas were carried from one position to another only by an actual movement of egg material. Thus, the direction and amount of movemeiit of the tells could be determined.




Many fiiiltryoiiic rudiniciits can bo located only with difficult}in early cinbryos because there are no surface markings to indicate tln'ir position; conseeiueiitly successful extirpation or transplantation of such rudiments cannot be effectually carried out unless some method is employed whereby their exact location can be Msccrtained. By producing a color contrast between the tissue of the two components, this object may often l)e attained, for then the grafted tissue can be marked out inunediately with reference to the outline of the whole embryo, and can also be correlated with local landmarks in older stages. In grafting together parts of eml)ryos of two species difTcrenth' pigmented, we have a means of iloing this, as in the lateral line»experiiiients of Harrison ('03). It was suggested by Dr. Harrison that, in homoi)lastic operations, this same purpose might be accomplished by staining one of the components with some vital dye before transplanting. Accordinglj', with the use of Nile blue sulphate, a method was devised whereby the anterior limb rudiment of Amblystonia could be located in eml)rvos with open medullary folds. At this stage the somites and the pronephros, which in later stages, indicate the position of the liml) rudiment, (Harrison, '15) are not yet formed, and the liml) rudiment is present merely as a region of mesoderm without \isil)le local characteristics.

The method employed was as follows: Embryos, which were either removed from the cap.sule or allowed to remain within it, were placed in an aqueous .solution of the stain until they had acquired a deep blue color, after which they were removed to tap water. An unstained emljryo, which is of a brownish yellow color, and a blue stained embryo were placed side by side in a watch glass filled with 0.4 per cent salt solution. A circular cut was made in the unstained embryo just posterior to an imaginary line pa.ssing dorso-ventraily midway tlirough the embryo. The ectoilerm and mesoderm were removed and discarded and the wound was cleaned of all free me.soderm cells. The wound was then covered over by a circular piece of ectoderm


taken from the stairiod embryo and cut to fit the excision. In about one-half hour the stained piece had healed in. Its size and position cinilii then be indicated by a camera firawjng. The stained disc now occupied in the embryo the place of the original extirpated tissue and the jiosition of this inserted piece of ectoderm, when dcve!oi)ment had proceeded .so that the somites could be seen, served to indicate whether or not the limb rudiment had been removed.

^^'hen the stained disc eventually lay just ventral to the third, fourth and fifth somites, it could be a.s.sumed with a high degree of certainty that the limb rudiment had been removed. If the extiri)ated tissue was taken from a more posterior position than that indicated above, .so that the stained ectoderm which covered the wound came to lie ventral to the fourth, fifth and sixth or the fifth, sixth and seventh .somites, it was then clear that only a jiortion of the limb cells had been removed and regeneration could be predicted.

The position of the stained disc is not always an absolute indicator of the region of the excised mesoderm, since it may be changed somewhat bj' ectodermal shifting. There is, however, very little .shifting in the inunediate limb region. The more pronounced ectodermal shifting, which does jiot usually occur until a somewhat later stage, takes place (Kirsal to the limb rt>gion and proceeds in a dorso-posterior direction ; the general character and direction of the movement being similar to that obser\-ed in frog embryos by Harrison ('0.3) in his experiments on the development of the lateral line.

The migrating ectoderm, however, occasionally involves the dorsal portion of the stained area which then loses its circular shape and lengthens out dorso-posteriorally.

\'arious concentrations of Nile blue .sulphate have lieen used ranging from 1 part of the dry stain in 100.000 parts of water to 1 part of the stain in 500,000 parts of water. The optimum concentration was found to lie between 1: I.'jO.OOO and 1 : 2tM),(K)0. The length of time required to produce the desired intensity in a solution of 1: KiO.OOO depends on whether or not the capsule was removed before the embryo was stained. Embryos remove<l


from the capsule wero properly stained in about three hours, but when the cai)sule was left intact from ten to twelve hours were required. Staining in the capsule, though a longer process, produces a more uniform color and for this reason is the more desirable method.

When ('ml)ryos were stained in the capsule for twelve hours in stronger concentrations, the ectoderm took on a reddish color with a distinct luster, a condition which served to indicate that the embryo had been overstained. The vitality of overstained ectoderm is decreased, its healing properties, when transplanted, are more or less impaired and some diffusion of the stain takes


Ectoderm taken frcim an embryo which has been properly stained has no reddish color, but presents a deep blue appearance; and, when transplanted to an unstained embryo, it heals as readily as normal ectoderm. ' There is no diffusion of the dye into the ectoderm of the host and the original color of the stained tissue is preserved for several days. As a result of continued cell division, however, the size of the stained piece becomes larger and the stain gradually loses its intensity.

Embryos stained in the capsule for twelve hours in various concentrations ranging from 1: 150,000 to 1:200,000 have been kept for twenty-five days after the application of the stain, during which time development proceeded normally, and at the end of which period all reactions were normal. The larvae at this stage were still light greenish blue in color.

Neutral red was also used in a few experiments with satisfactory results. \Mien this stain is used the optimum concentration for a fixed staining period of twelve hours lies between 1 : 400,000 and 1 : 500,000. Embryos stained in the capsule for twelve hours in solutions of such strengths stain a uniform red and such embryos were also kept for twenty-five days during which period development wa.s normal. M the end of this time practically all of the stain had disappeared. In so far as the effects on the embryo are concerned one stain can be used equally a.s well as the other; yet the Nile blue sulphate is the more desirable because it produces a better contrast.


(ioodale states that in Spolorpos the stain has a strong affinity for yolk griiiiulcs and that the protoplasm is loft uastaiiiod. On the other h;inil. Smith states tluit in C"ryf)toliraiichus the niicn)nieres stain ili>tiii(tly and keep the stain while the more heavily laden yolk parts of the egg stain with difficulty. A study of stained ectoderm taken from Amblystoma embryos shows that some of the yolk Kranules take up the stain while others are left entirely uastaincd.

The application of Nile blue sulphate has not been restricted to limb experiments, but has been used this season by Dr. Harrison in experiments on the ear vesicle and on the gills. In as much as the method has given satisfaction in these several types of experiments, it is hoped that it may be applied to still other experiments dealing with embryonic tissue transplantation.

The details of the limb experiment will appear in a later publication.


GooDALE, H. D. 1912 The early development of Spelerjws bilineatu.s. Am.

Jour. .\nat., vol. 12. Smith, B. C. 1914 .\n experimental study of concrescence in the embryo of

Cryptobranchus allcghenicnsis. Biol. Bull., vol. 26. Hakriso.n, R. G. 190;5 Experimentclle Untersuchungen (ib;-r die Entwicklung

der Sinnesorgane der Seitenlinic bei den .Vmphibien. .\rchiv fur

Mikros. .Vnat., Bd. 63.

191.5 Experiments on the development of the limbs in .\mphibin.

Proc. Nat. .\cad., vol. I.