JAS a ni at EARL Ube SSesee ay 4 ee ae BrEnSs = = OAS SL SSS: i en BEIT HSMN sola ee H Fiat Aventis Md 448 Yad. wil epee zr oi Hatt = a erenats = = aay i Neg tee 8) he hae ce ag Ra a eth THE AMERICAN JOURNAL OF ANATOMY EDITORIAL BOARD CHARLES R. BARDEEN G. CarL HuBER University of Wisconsin University of Michigan Henry H. DoNnaLpson GrEoRGE S. HUNTINGTON The Wistar Institute Columbia University Smmon H. Gace Henry McE. KNowEr, Secretary Cornell University University of Cincinnati . VOLUME 23 1918 FRANKLIN P. Mau Johns Hopkins University J. Puayrarr McMurrica University of Toronto GEORGE A. PIbRSOL University of Pennsylvania THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. CONTENTS No.l. JANUARY RaymMonp Peart AnD Atice M. Borina. Sex studies. X. The corpus luteum in the ovary of the chicken. Six text figures and nine plates............:.........-00002 1 Exior R. Crark. Studies on the growth of blood-vessels in the tail of the frog larva— by observation and experiment on the living animal. Sixteen figures............... 37 J. A. BapertscHer. The fate of the ultimobranchial bodies in the pig (Sus scrofa). [AChE oR) re ore erst Pt. a ee Meeks fo «4 SE AAR DER RMS Coro eee. 89 J. DurssperG. Chondriosomes in the testicle-cells of Fundulus. Twenty-one figures (iwiosmlates) =. 2:5 .cc) occa sh She Fs: << aly hrs SRM Scie aged ere a no acne @ eaeye ale 133 Apvotr H. Scuuurz. The position of the insertion of the pectoralis major and deltoid muscles on the humerus of man. Three figures..... 5-6 bd OOS ES a eMart eae . 155 W. B. Cuapman. The effect of the heart-beat upon the development of the Bene Syscemuom ghechick, sSeventeen. figures)... << .<8..-see eer: bere eck Jetite See ee 175 Epwarp A. BoypEen. Vestigial gill filaments in chick embryos with a note on similar structures in reptiles, Three text figures and four plates............. Orane Me SESE 205 No.2. MARCH Auice Turnc. The formation and structure of the zona pellucida in the ovarian eggs of UIT LES tee aW.G LV Gel OUTE Sian eect otees soisn< 0 cto. souks. o > vast REPO ore, lesen IY RE 237 Avour H. Scuuttz. The fontanella metopica and its remnants in an adult skull. Five 1 Fy PLATS Sco ck gS 058 a A Mae MRD ee 259 FRANKLIN PARADISE JOHNSON. The isolation, shape, size, and number of the lobules of the pig; s liver. ~ Lwelve figures: (two: plates)).... 2. x 0. geemeeicte 2 cbae a crale sie ee oe 273 AsprAM T. Kerr. The brachial plexus of nervesin man. The variations in its formation Ans ran Cheses Mawenty=NIM evil CURES ®, 2) ./)- 0,2. «1: ole Meee ele eines ciel thal tee eee 285 FRANKLIN P. Matyi. On the age of Human Embryos. Two figures..................-. 397 C. R. BARDEEN. Determination of the size of the heart by means of the x-rays. One . 1 Pre? t) oe 4 , y oe ae ae itt Ww - y” ( ? } \ oe Ace i i % ' Cita ee §@ . i, ApPeree es 7 * e P 7) i a ; é ORE Pe a - ‘. ’ . a p a ® r Reyes oe 4 = > « - P _ ’ - 4 = x Menges Geer 3 om pil Py Iai ea uy ‘ ee ov, sta epee. wea ray 1 pte s: AUTHORS’ ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 1) ; - SEX STUDIES X./THE CORPUS LUTEUM IN THE OVARY OF THE DOMESTIC FOWL! ¢ RAYMOND PEARL AND ALICE M. BORING SIX TEXT FIGURES AND NINE PLATES I. INTRODUCTION The corpus luteum is one of the clearly recognized sources of an internal secretion in the mammal. Various functions have been ascribed to it. Its function in connection with secondary sex characters has been discussed by Pearl’ and Surface (15), with one piece of clear cut evidence. The case was that of a cow which developed cystic ovaries and took on male secondary sex characters. The ovaries were compared histologically with those of a normal cow and the two were found to resemble each other in all respects except that the cystic ovaries had no corpora lutea. The interstitial cells were the same in both so that the difference in secondary sex characters could not be at- tributed to them. The implication of the facts is that the cor- pus luteum has an inhibitory influence in the female which prevents maleness from developing and that when no corpus luteum is formed, male characters appear. The chief difficulty with such a view has been that its appli- cation is very limited, as the corpus luteum has been supposed to be a structure occurring only among mammals. The sec- ondary sex characters of birds are particularly pronounced and the results of ovariotomy experiments, such as those of Goodale, (16) show the possibility of changing these characters experi- mentally. Also the many cases of hermaphrodite birds (to be ' Papers from the Biological Laboratory of the Maine Agricultural Experiment Station, No. 115. 1 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 JANUARY, 1918 2 RAYMOND PEARL AND ALICE M. BORING considered in Study XI of this series), with varying degrees of maleness and femaleness indicate the presence of some sex regu- lating substance in birds. Is this substance entirely different from the corpus luteum probably connected with it in mammals, or is there a corpus luteum or its homologue in birds? An in- vestigation of this question has been undertaken in this study. We consider that we have successfully demonstrated the pres- ence of the corpus luteum in the domestic chicken. Further discussion of the bearing of this fact on the whole question of secondary sex characters will be deferred until a later paper of this series, which will unfortunately probably be delayed for some time, as one of the authors (R. P.) has been called upon by the government to turn his attention to practical problems during the war. A eareful examination of the ovary of a bird which has been actively laying shows three kinds of structures: the yolks of various sizes indicating different stages of development, the discharged follicles in various stages of regression, and the atretic follicles or degenerating eggs of different sizes. These are all easy to identify when they are large enough to protrude far from the surface of the ovary, that is, when they are larger than 2 or 3 mm. in diameter. Under this size, it is impossible to distinguish the discharged follicle from the atretic. Both of them show a yellow or orange spot in the center. The ques- tion naturally arises whether these yellow spots are homologous in structure and origin with the mammalian corpus luteum. They never develop into a large mass like the corpus luteum of the mammal. They have the color of the spots on the cow ovary which indicate remains of old corpora lutea. In order to interpret these yellow spots, a study has been undertaken of the progressive .and regressive changes in the cell structure of egg follicles in different conditions, undischarged, discharged and atretic. The material used came chiefly from four birds, an actively laying Bantam, a Barred Plymouth Rock in the same condition, an old Compine past the laying condition, and a guinea-hen with a large ovary containing several large yolks. Material CORPUS LUTEUM IN OVARY OF THE CHICKEN 3 from a number of other birds was used in the study of special points. These are some of the same birds used in Study IX. The ovaries were fixed in Gilson and McClendon. In the Barred Plymouth Rock ovary the different discharged follicles were sectioned separately and arranged in a series, according to size and consequent order of age since ovulation. After the study of this series, it was easy to judge of the condition of various follicles in pieces of the other ovaries cut at random. Various stains were tried, iron haematoxylin and Delafield’s haematoxylin for general histology and Mallory’s and Mann’s stains for secretion granule tests. Il. UNDISCHARGED FOLLICLES OF THE HEN’S OVARY A study of the follicles of large undischarged oocytes shows them to consist of an epithelial layer, the granulosa, and two connective tissue layers, the inner and the outer theca folliculi (fig. 1). In the inner theca are located groups or nests of epi- thelial cells (J, figs. 1 and 2). They have been described by many authors, notably Ganfini, Sonnenbrodt and Poll, but have been called interstitial cells. Poll calls them Kornzellen at first, describes their collection into the internal theca and then implies their function by saying that the biological rdéle of the theca interna in the formation of the corpus luteum still needs to be worked out. That he also confuses them with interstitial cells is shown by his statement that the theca interna fills up the atretic follicle with groups of Kornzellen, which is the same thing as an interstitial gland. These nests of cells in the bird are not anything like the usual glandular interstitial cells of the ovary in structure. They are about three times as large (compare fig. A and C). The nucleus is bigger and plumper, the cytoplasm is usually clear and vacuolated in ap- pearance, only occasionally containing a few acidophile gran- ules which stain with the fuchsin in Mallory’s stain or the eosin in Mann’s stain; while the real interstitial cells are crowded with granules. These large clear cells are seldom found alone, but are usually grouped into nests of various shapes, as already mentioned. The cytoplasm of these cells usually will not take 4 RAYMOND PEARL AND ALICE M. BORING up an acid stain. They remain strikingly clear, when the con- nective tissue all around them is highly colored. So great is the. contrast that they show distinctly even at low magnification in a section such as figure 1. Furthermore, they are found in Fig. A Part of follicle of wall of medium sized oocyte in hen ovary. (X 950.) Compare figure 2. Fig. B- Part of theca interna of sixth discharged follicle in hen ovary, show- ing many vacuolated lutear cells. (x 950.) Compare figure 6. CORPUS LUTEUM IN OVARY OF THE CHICKEN 5 different parts of the ovary, mostly in the theca interna, while the interstitial cells lie in the general stroma, and especially on the periphery. Figure 3 shows several very young oocytes from the same ovary as figure 1. In these, the follicle consists only of a single layer of epithelial or granulosa cells (g). The connective tissue layers are not yet formed. But there are nests of clear cells (1) in the stroma nearby. Presumably these are included with the connective tissue when the theca interna is formed. _ Ii. DISCHARGED FOLLICLES OF THE HEN’S OVARY In the largest follicles before ovulation, the three layers are stretched out very thin by the pressure of the large yolk within them. After ovulation, there is a shrinkage of the follicle walls, probably due to the elasticity of the connective tissue recoil- ing at the sudden release of pressure from inside. On the Barred Plymouth Rock ovary, the ripe yolk measured about 40 mm. in diameter, and the last discharged follicle measured 20 mm. in length from base to tip, while the next to last was 12 mm., and the fourth in the series was 7 mm. As this shrinkage in length takes place, the walls thicken until finally a small oval mass results having no resemblance to a hollow follicle. .The ruptured place through which ovulation took place, becomes gradually closed up, by the growing together of the edges, and the filling of cells into the cavity. Sometimes this mass of cells protrudes slightly from the cavity at the old place of rupture, thus somewhat more resembling a miniature mammalian corpus luteum. Yellow pigment forms in the puckered edges of the follicle and also in the central mass. The microscopic study of sections through discharged follicles of various ages shows that the increase of thickness of walls is due chiefly to a thickening of the theca interna. Figure 4 is a section of the last discharged follicle of the Barred Plymouth Rock ovary. It shows the thickened theca interna (7) and in addition the remnants of the granulosa (g). The latter seems to loosen from the follicle after ovulation, and the cells collect in masses in the cavity and degenerate. 6 RAYMOND PEARL AND ALICE-M. BORING The first subsequent discharged follicle in the series to show any new microscopic features is the sixth (fig. 5), where there appears a marked increase in the number of nests of vacuo- lated cells in the theca interna (l). They are concentrated toward the cavity. The closeness of nests together may be partly due to the shrinkage of the cavity after discharge of the egg. But as this does not seem sufficient to account en- tirely for the increase, the number must be added to either by division or migration. The fact that division plays some part in the process is proven by the observation of several mitotic spindles. The character of these cells shows better in greater magnification, as in figure 6 and figure B. The further progress of the increase of vacuolated cells in the theca interna is shown in figure 7, a section of a discharged fol- licle too small to have been placed in the series as to time of discharge. Here the whole internal theca looks full of holes, due to vacuolated cells (J). The central cavity is nearly oblit- erated, almost as though the edges had been pulled up by a gathering string. There are, however, a few cells in the central cavity (p). These get in there by migration from the internal theca. Figure 13 shows the process in an atretic follicle where it is more conspicuous, but it is true to a more limited extent in the discharged follicles. ‘The cells concerned have a speckled ap- pearance in figure 13 (d). \They are abundant in the follicle wall, some are scattered among the yolk spheres in the central cavity and some are on the border line between the follicle wall and the cavity, indicating that the cells actually migrate into the cavity. Occasionally a very large central plug is formed which protrudes from the spot of rupture. Figure 7 shows a small plug of this kind (p). The cavity usually becomes finally obliterated by the thick- ening of the internal theca and the formation of large masses of vacuolated cells from the original nests. In figure 8, the chief tissue consists of the masses in the internal theca (/). The line between the theca interna and externa is marked by the irregular spaces and blood vessels. The connective tissue in the CORPUS LUTEUM IN OVARY OF THE CHICKEN 7 center (c) shows where the edges of the internal theca have drawn together and obliterated the cavity. We have traced thus far the general histological changes involved in the shrinking and filling up of the discharged follicle. We must consider next in more detail, the cytology of these par- ticular cells involved. Figure 2 and figure A show them in their original condition from a large undischarged follicle. We have earlier in this paper pointed out their especial characteristics in distinction to the interstitial cells. By the time they are close enough together to cause the vacuolated appearance of the whole inner part of the theca interna, the nuclei are somewhat shrunken and pushed to the side of the cell, suggesting active elaboration of secretion material (fig. 6 and fig. B). By the time the closing in of the follicle has neared completion (figs. 8 and 9), the character of the cells is decidedly modified (fig. C). The cell boundaries in any one small mass of cells are indistinguishable. The cells seem to have melted together so that the outlines of the vacuoles are the evidently visible Jines rather than the cell outlines. The vacuoles also are much larger than previously. ‘The nuclei are smaller and less regular in outline, they stain darker, in fact, they look shrunken. These figures show nicely the contrast between the cells which fill up this discharged follicle and the interstitial cells. The interstitial cells lie in the connective tissue of the external theca and of the internal theca in between the masses of transformed epithelial nest cells. They are entirely unchanged from their usual ap- pearance. They show clearly because the granules with which they are packed stain vividly with acid stains. A homologous mass of cells from an older solidly filled follicle (fig. 10) is shown in figure 11 and figure D. Here the nuclei show still further signs of degeneration and the general network of the cytoplasm contains clumps that look like secretion material. These secre- tion particles are yellow in color. They look amorphous in character, and they vary greatly in size (fig. 20). They can not be fatty, for they have not dissolved in the clearing oils. They cannot be of the protein nature of the secretion granules of the interstitial cells, as they retain their distinct yellow color 8 RAYMOND PEARL AND ALICE M. BORING no matter how the preparation may be stained. They make a fine contrast with iron haematoxylin, acid fuchsin, eosin, methyl blue, and still show their own characteristic yellow even with Fig. C Masses of lutear cells from older discharged follicle, with interstitial cells lying in connective tissue between masses. (X 950.) Compare figure 9. ce ee 2 poe £ 0G g,, NG re a FEB 3 os. a She oer ‘ oe ~ Ls ® Fig. D Mass of lutear cells from discharged hen follicle, with pigment par- ticles developed in the network. (x 950.) Compare figure 11 and figure 20. CORPUS LUTEUM IN OVARY OF THE CHICKEN 9 orange G. The cell masses finally become nearly filled with this yellow material, some of it collecting in clumps several times larger than the degenerated nuclei. Further tests of the character of the cell contents in these cell masses were made with Sudan III. Hand sections were made of material in McClendon’s fluid. Although these could not be cut very thin, they showed that the inner lining of the early discharged follicles contains fatty material. In an old follicle with central yellow mass the cells of the yellow mass take the red of the Sudan III, but the yellow amorphous particles show in the midst of the red. They can be squeezed out of broken cells and isolated from the red fatty background, showing they are still yellow, unaffected by the Sudan III, and therefore not of a fatty nature. The fatty substance indicated by the Sudan III reaction in both young and old follicles is probably con- tained in the vacuoles so conspicuous in paraffin sections. The xylol would have dissolved out all the fat leaving the vacuoles in which it had been contained. IV. DEGENERATION OF CORPUS LUTEUM IN COW OVARY In order to show the significance of the yellow mass formed in the center of discharged follicles in the hen ovary, we have made a brief study of the degeneration of the corpus luteum in the cow ovary for comparison. There is an extensive literature on mammalian corpus luteum, but this deals chiefly with the de- velopment and early involution. Now the bird quite evidently has no structure similar to the large corpus luteum which fills up half the ovary of a cow at its full development. The small yellow spot on the bird ovary resembles the small yellow spots on the cow ovary which mark the old remains of former corpora lutea. Ovulation in the cow alternates between the two ovaries. So by studying the two largest corpora lutea on both ovaries we can arrange a series of four involution stages. Beyond that, they all seem equally shrunken and therefore can not be ar- ranged in a further series. Such a series of four involution stages has been studied for two cows, and in addition several older corpus luteum remains. | 10 RAYMOND PEARL AND ALICE M. BORING The last formed corpus luteum is of a salmon pink color, due to a combination of the blood color and the lutein color. Sections show it composed of large plump cells with rounded nuclei, as described by Corner. These luteum cells are scat- tered in the midst of an areolar connective tissue groundwork (fig. 18 and fig. E). In dehydrating for embedding, the abso- lute aleohol and xylol become very yellow, indicating that the cells contain something soluble in these reagents. This is of course one chemical character of lutein. e Fig. E Cells from youngest corpus luteum of cow. (xX 950.) Compare figure 18. Fig. F Cells from older corpus luteum of cow, showing pigment developed in cells. (XX 950.) Compare figure 19 and figure 21. The next to last corpus luteum is much reduced in size. Its color has lost-the pinkish shade and it appears a solid bright yellow. This is also soluble in absolute alcohol and xylol as in the first stage. The cells and nuclei both look a little shrunken. In one cow, this second corpus luteum contained a few amor- phous yellow particles like those described for the hen. CORPUS LUTEUM IN OVARY OF THE CHICKEN ita) In the third oldest corpus luteum, the tissue is shrunken so that a mere speck shows on the surface. This is the stage re- sembling the yellow spots of the hen’s ovary. Dissection shows that it is reduced in all diameters. That part of this decrease in size is due to cell shrinkage is well demonstrated by com- parison of figures FE and F, which are drawn to the same scale. Not only the nucleus but the cell body is at least halved in size. The color now is darker, being a brick red. This is not due to blood vessels, as sections do not show any more than formerly. It is due to the development of a dark yellow pig- ment, the same substance which appeared in small quantity in the younger corpus luteum and in large quantity in the hen ovary. In this stage of involution it is developed in large quan- tities, practically filling up many of the lutear cells (figs. 19 and 21 and fig. F). In unstained sections it gives a yellow color to most of the section. In the fourth oldest corpus luteum of the two series and in the scattered older ones sectioned, the structure is similar to that in the third oldest, the yellow amorphous masses being possibly larger and more distinct. This yellow material certainly looks the same as that in the hen ovary. The chief structural difference is that it is all con- fined within cells with distinct cell walls in the cow, while in the bird, the cells forming it, lose their boundaries and the particles are formed in a vacuolated network with scattered shrunken nuclei (cf. figs. 20 and 21). Sudan III reacts similarly with hand sections of formalin material from both cow and hen ovary. All four stages in the cow series take the red color showing the presence of a fatty substance in the cell. This corroborates the evidence from the solvent action of absolute alcohol and xylol. But in the third and fourth stages, yellow amorphous pigment particles can be seen glistening in the red background. The pigment is not of fatty nature in the cow, any more than it isin the hen. In fact, this substance is so similar in the two animals, that we shall from now on speak of a corpus luteum in the hen, and call the cells forming this pigment, lutear cells. 12 RAYMOND PEARL AND ALICE M. BORING This development of a non-fatty pigment in the mammalian lutear cells has been already described by Mulon‘as occurring in atretic follicles. He speaks of the lipocholesterine as changing over to an indelible pigment. This same substance certainly forms in the involution of the corpus luteum of a discharged follicle as shown in this present work. lt is of especial interest to find that Blair Bell’s deseription of the corpus luteum in Ornithorhynchus, a primitive ovi- parous mammal, shows it very much like that in the hen. It often remains hollow, it never becomes very large. It con- sists chiefly of a thickened theca interna. Sometimes it be- comes a solid fibrous mass. One of Bell’s figures almost exactly resembles figure 4 of this paper. One would like to know whether the yellow pigment is found in Ornithorhynchus thus making its resemblance to the bird even more striking. V. BIOCHEMICAL CHARACTER OF PIGMENT OF CORPUS LUTEUM The identity of this yellow amorphous pigment in the corpus luteum remains in the ovary of the hen and of the cow has been put to chemical tests as well as morphological; first of a micro- chemical nature, as already partially described, and secondly by various special chemical solvents. The work of Escher and of Palmer and Eckles on animal pigments has been consulted in selecting the reagents to use. The microchemical tests have been discussed in previous sections, but will be summarized here. Microscopical technique processes have shown the identical behavior of- the pigment in hen and cow. It does not dissolve in alcohol or oils. It will not stain with basic nuclear stains such as haematoxylin and Kresylviolet, or with acid counterstains, such as eosin, methyl blue, anilin blue, orange G, or with such a stain as iron haema- toxylin. Neither does it stain with the fat stain, Sudan III, although there may be much fatty material in the cell in which it lies. As normal secretion granules of a protein nature take acid stains and secretion granules of a fatty nature take Sudan III, this pigment is neither protein nor fat in composition. CORPUS LUTEUM IN OVARY OF THE CHICKEN 13 A further test of its chemical nature was made by trying some of the various solvents used by Escher and by Palmer. Sections were cut in paraffine and mounted on slides and then the paraffine removed by xylol and the sections treated with different chemicals. This pigment is not the carotin described by Palmer, but we could not reach any conclusion as to its chemical nature, as nothing could be found to dissolve it. But the fact of the identity of this pigment in the hen and cow is proven beyond a doubt. Concentrated HCl, HNO; and H.SO, were tried and had no effect except that the H,SO, turned the particles dark brown and made them even more distinct than before. For an alkali solvent, strong KOH was used; it turned the pigment bright orange but did not dissolve it. In addition to these various other solvents were tried after consultation with the chemistry department, petroleum ether, sulphuric ether, ace- tone, carbon bisulphide, and carbon tetrachloride, but none of these had the slightest solvent effect on the pigment. Acetone cleared the background and this made the particles stand out more sharply. Carbon bilsulphide was allowed to act for sev- eral hours, but the preparations still contained the pigment at the end of that time in undiminished degree. We conclude that any two substances which can withstand the action of as many well known solvents of as many different properties as this list includes must be of very similar chemical nature. This gives us one more proof that the yellow particles in the hen ovary are the same as those in involuted mammalian corpora lutea. VI. CHANGES IN ATRETIC FOLLICLES IN THE HEN’S OVARY Among the developing yolks and discharged follicles of the hen ovary are many degenerating eggs. They can be distin- guish d from developing eggs by the shrunken a»pearance as though the contents did not quite fill out the follicle. Eggs may start to degenerate at different stages. The largest one on the Barred Plymouth Rock ovary was 12 mm. in diameter. Many ot them show dark spots which are masses of coagulated blood. Mostly they are smaller than this when involution begins. The 14 RAYMOND PEARL AND ALICE M. BORING degree of shrinkage shows whether the involution process had recently begun or not. When these degenerating eggs are cut open, the contents is found to be in a more or less fluid state. When these atretic follicles have become reduced in size to 2 or 3 mm., it is no longer possible to distinguish them externally from the discharged follicles; the same kind of a yellow pigment appears in the center. Studied microscopically, the chief difference between atretic and discharged follicles is that the former have a more distinct cavity which becomes obliterated chiefly by migration of lutear cells into it instead of by shrinkage of the walls. The granu- losa is shed similarly. There must frequently be hemorrhage as corpuscles are often found in the cavity. The varying quan- tity of yolk spheres is one indication of the degree of involu- tion, also the number of lutear cells in the cavity. Figure 12 is an atretic follicle with considerable yolk still unabsorbed. A few lutear cells have filled in to the cavity (fig. 13, 7). It is particularly clear here that the cells inside of the inner mar- gin of the theca interna are the same in structure as those of epithelial nature in the interna theca. This is Just as Benthin describes it for the atretic mammalian follicles. Figures 14 and 15 show a later stage where the yolk is almost all absorbed and the cavity is filled with lutear cells. Not until the cavity is filled with lutear cells does the yellow pigment already described in discharged follicles, make its ap- pearance. It forms in the lutear cells of atretic follicles in a similar way to that in the discharged follicle. The cell bound- aries are possibly not obliterated so completely, so that the morphological resemblance to the cow corpus luteum remains is even more striking than in the case of the discharged follicles. Figure 16 is part of an atretic follicle where the cells are filled with pigment. The amorphous character of this material shows in figure 17 a part of figure 16 under higher magnification. It is of interest to notice that the lutear cells in the hen in both discharged and atretic follicles originate entirely from the theca interna. In mammals the origin of the lutear cells is a mooted guestion. Some authors, as Niskoubina, hold that they CORPUS LUTEUM IN OVARY OF THE CHICKEN 15 have a double origin, from granulosa and theca interna, while others such as Benthin and Hegar, claim that they all come from the theca interna. This point is perfectly clear in birds due to the ease with which one can distinguish these peculiar cells in the internal theea of undischarged follicles and follow them to the thickened mass in the center of the discharged follicles, and see them migrating out into the cavity of the atretic follicles. The formation of a corpus luteum in atretic as well as dis- charged follicles makes it possible to identify ovarian tissue in ovaries too abnormal to have ovulated any eggs. Most of the literature of the mammalian ovary considers the involution of the atretic follicle as something distinct from that of the dis- charged follicle. The mass forming in the atretic follicle is called the corpus atreticum or fibrosum in contradistinction to the corpus luteum. However, Hegar says that it is hard to tell one from the other. They are practically identical in the hen. VII. SUMMARY We are now in a position to sum up the points proving the homology of the corpus luteum in the hen and in the cow. There has been much discussion about the origin of the corpus luteum in mammals. In the hen there is no question but that the origin is simply from the theca interna. The course of development in the hen corpus luteum is an abbreviation or fore-shortening of that in the cow. It corre- sponds directly to the late involution stages of the cow corpus luteum. They both contain a yellow fatty substance, as shown by the Sudan III, absolute alcohol and xylol reactions. There develops in both a yellow amorphous pigment in the cells con- taining the fatty substance. This pigment is similar chemically in that it will not stain with basic or acid stains; also in that it will not dissolve in any of the usual solvents, acid alkali or oil. In the hen, a corpus luteum forms in both discharged and atretic follicles. 16 RAYMOND PEARL AND ALICE M. BORING VIII. LITERATURE CITED Breutu, W. Buatrk 1917 The sex complex. Wood and Company, New York. Brentuin, W. 1910 Ueber Follikelatresie in kindlichen Ovarien. Arch. f. Gyniikologie, Bd. 91, p. 2. 1911 Ueber Follikelatresie in Siugetier Ovarien. Arch. f. Gynikol- ogie, Bd. 94, p. 599. Bouin er ANcCEL 1912 Sur la nature lipoidienne, d’une substance active se- cretée par le corps jaune des mammifére. C. R., T. 151, p. 1391. Corner, G. W. 1915 Corpus luteum of pregnancy as it is in swine. Carnegie Inst. Washington, 222, p. 69. Dupatsson, H. 1906 Contribution a l’étude du vitellus. Arch. de Zool. Exp. et Gen. T. 4. Series 5, p. 158. Escuer, H. H. 1913 Ueber den Farbstoff des Corpus Luteum. Zeitschr. Physiol. Chem., Bd. 83, p. 198. FRAENKEL, L. 1910 Neue Experimente zur Function des Corpus Luteum. Arch. f. Gynaikologie, T. 91, p. 705. GaANFINI, C. 1908 Sulla strutturo e sviluppo delle cellule interstiziali dell’ ovajo. Arch. di Anat. e di Emb., 8. 7, p. 378. GoopaueE, H. D. 1916 Gonadectomy in relation to the secondary sex charac- ters of some domestic birds. Carnegie Inst. Washington, 243. Heaar, K. 1910 Studien zur Histogenese des Corpus luteum und seiner Riick- bildungsproducte. Arch. f. Gynikologie, Bd. 91, p. 530. Hennecuy, L. F.. 1894 Recherches sur l’atrésie des follicules de Graaf chez les mammiféres et quelques autres vertébrés. Jour. de |’Anat. et Rhys dia 3O) yp.) lt Mituer, J. W. 1910 Die Riickbildung des Corpus luteum. Arch. f. Gyni- kologie, Bd. 91, p. 263. MuLon AND JonG 1913 Corps jaunes atrésiques de la femme. Leur pigmen- tapions ©. R&Soe Biol) Mie: NiskouBIna 1909 Recherches sur la morphologie et la fonction du corps jaune de la grossesse. Dissert. de la faculté de med. de Nancy. Paumer, L.S. anp Eckurs, C. H. 1914 Carotin. The principal natural yellow pigment of milk fat. I, II, III, IV. Research Bul., nos. 9, 10, 11, 12, Univ. of Mo., Agr. Exp. Stat. PEARL, R. anp Surrace, F. M. -1915 Sex Studies. VII. On the assumption of male secondary characters by a cow with cystic degeneration of the ovaries. Ann. Rept. Me. Agr. Expt. Stat., 1915, p. 65. Pout, H. 1911 Mischlingstudien VI: Eierstock und Ei bei fruchtbaren u. unfruchtbaren Mischlingen. Arch. f. Mikr. Anat., Bd. 78, II, p. 63. SONNENBRODT 1908 Die Wachstumsperiode der Oocyte des Huhnes. Arch. f. Mikr. Anat., Bd. 72, p. 415. Rok NY ; 4 se PLATES: sigh ee eae 2 See * Te DESCRIPTION OF PLATES We wish to take this occasion to acknowledge our indebtedness to Mr. Royden Hammond for all the photomicrographs, and to Mrs. Maud DeWitt Pearl for the paintings on plate 9. PLATE 1 EXPLANATION OF FIGURES 1 Medium sized oocyte in hen ovary (X 40), showing three layers to the follicle, the granulosa (g), theca interna (7), and theca externa (e), with nests of lutear cells in the theca interna (I). 2 Part of follicle wall in figure 1 at greater magnification (X 352). Labels the same as in figure 1. 18 CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 1 RAYMOND PEARL AND ALICE M. BORING PLATE 2 EXPLANATION OF FIGURES 3 Young oocytes in hen ovary (X 352), with follicles consisting of a single layer of granulosa (g). Nests of lutear cells in the stroma nearby (J). 4 Portion of last discharged follicle in hen with thickened theea interna(?) and granulosa (g) being sloughed off into the cavity. (x 40.) 20 CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 2 RAYMOND PEARL AND ALICE M. BORING gr a OSes _~ a dati PLATE 3 EXPLANATION OF FIGURES 5 Portion of sixth from last discharged follicle, showing large number of lutear cells (1) in the theca interna (X 40). 6 Part of figure 5 enlarged (X 176). 7 Small discharged follicle with cavity nearly obliterated. Small plug of cells (p), filling in the cavity. Theea interna filled with masses of lutear cells CG) ACe405) i) bo CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 3 RAYMOND PEARL AND ALICE M. BORING PLATE 4 EXPLANATION OF FIGURES 8 Discharged follicle with cavity completely obliterated. The chief compo- nent is masses of lutear cells (J). The connective tissue center represents original location of cavity (c). X 40. 9 Part of figure 8 at greater magnification (X 176), showing interstitial cells (7.c.) in connective tissue between lutear masses. CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 4 RAYMOND PEARL AND ALICE M. BORING aA LES Pee sats t ; he ES Rey am ME io a ae *r, e PB, ‘ — Bn oot PLATE 5 EXPLANATION OF FIGURES 10 Later stage of solid discharged follicle, showing large development of yellow pigment in lutear masses (X 40). 11 Part of figure 10 at greater magnification (X 176), showing pigment particles. 12 Atretic follicles in hen ovary, with yolk spheres in central cavity (X 40). 13 Part of figure 12 at greater magnification (X 176). Lutear cells (l) show in theea interna and also among yolk spheres in the cavity. 26 CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 5 RAYMOND PEARL AND ALICE M. BORING - NS ae : » ae ee | I b PLATE 6 EXPLANATION OF FIGURES 14 Later stage of atretic follicle (X 40). Only a few yolk spheres remain in cavity. Cavity is practically filled with lutear cells. 15 Part of figure 14 at greater magnification (X 176), showing lutear cells in theea interna (7), as well as the central cavity. bo or PLATE 6 NT vp CHICKE 1 4 u CORPUS LUTEUM IN OVARY OF THI BORING AND ALICE M. RAYMOND PEARL PLATE 7 EXPLANATION OF FIGURES 16 Atretie follicle in which the pigment particles have developed in the lutear cells (X 40). 17 Part of figure 16 (X 176). 30 PLATE 7 T ~ i 4a CHICKI y BORING VARY OF THE O M IN RAYMOND PEARL AND ALICE ) 4 CORPUS LUTI M. dl PLATE 8 EXPLANATION OF FIGURES 18 Section of youngest corpus luteum of cow (x 176). 19 Section of older corpus luteum of cow (X 176), showing cells filled with pigment particles. 32 CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 8 RAYMOND PEARL AND ALICE M. BORING PLATE 9 EXPLANATION OF FIGURES 20 Section of discharged follicle of hen ovary, stained in Mallory’s stain. Connective tissue = blue. Corpuscle = red. Interstitial cells = purple. Lu- tear pigment = yellow. 21 Section of older corpus luteum of cow, stained in Mallory’s stain. Tissues colored as in figure 20. 34 CORPUS LUTEUM IN OVARY OF THE CHICKEN PLATE 9 RAYMOND PEARL AND ALICE M. BORING ce ae “a4 se 3 oe 4 ae 5 seth ‘s f aN “Pee sueatigte oS" AUTHOR'S ABSTRACT OF THIS PAPER [ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 29 STUDIES ON THE GROWTH OF BLOOD-VESSELS IN THE TAIL OF THE FROG LARVA—BY OBSERVATION AND EXPERIMENT ON THE LIVING ANIMAL ELIOT R. CLARK + Department of Anatomy, University of Missouri SIXTEEN FIGURES These studies were begun and part of them were made in the laboratory and under the inspiration of my beloved teacher and master, the late Professor Franklin P. Mall, and it is with a sense of the deepest gratitude and reverence that I acknowledge the immeasurable debt which I owe to him. INTRODUCTION The development of the vascular system falls broadly into two stages: (1) the stage of primary differentiation, or histogen- esis, and (2) the stage of extension and elaboration of arteries, ~ veins, and capillaries. The exact, manner and place, in which the primary differentiation occurs is an unsettled problem, and is, at the present time, the subject of spirited controversy. It has not been satisfactorily decided whether blood-vessel en- dothelium differentiates from entoderm, or mesoderm—and if from mesoderm, whether from mesenchyme generally or from the mesothelial lining of the coelom. Nor has it been deter- mined whether this primary differentiation occurs on the walls of the yolk sac alone, or in the embryo proper, or whether it may take place both on the yolk sac and in the embryo. Another unsettled point is the extent of time over which the primary differentiation takes place. Recent discussions and observa- tions supporting one or another of these views may be found in the following: Minot (12), Evans (12), Riickert and Mollier (06), Schulte (14), Bremer (’14), Stockard (15 A),Reagan (17), Sabin (717). The second stage of vascular development includes the further 37 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO 1 38 ELIOT R. CLARK extension and development of the system after the primary differentiation has taken place and after the circulation has been established, and it is with this second stage that the studies here reported are concerned. In this stage, which continues throughout life, the vascular endothelium spreads through the erowing organism, arteries and veins develop, until the extensive and complicated vascular system of the adult is perfected. It is principally characterized by the formation of new vessels by the sending out of sprouts from the vessels already present, in- stead of by the transformation of mesenchyme, of other undifferen- tiated cells, as in the first stage, and by the action on the vessels of the mechanical and chemical factors concerned with the circu- lation of blood and interchange of substances through the wall. In spite of the abundant evidence in favor of this mode of spreading of the vascular endothelium, after its primary differ- entiation, there: are observers who adhere to the view that at any time throughout life, mesenchyme (or other undifferentiated cells) may be transformed into vascular endothelium. ‘This view is held by Maxinow, Weidenreich, and Mollier (cf. dis- ‘cussion in Schulte, ’14,) who believe that, not only may reticulum cells and leucocytes be transformed into blood-vessel endothe- lium, but that the reverse transformation may take place—in brief, that vascular endothelium is not a specific tissue, but is interchangeable with the other tissues mentioned. The evidence for this view has in no case been conclusive. It is clear, however, that the manifestation of the property of sprouting does not form a sharp boundary line in time of development between stages, for apparently sprouting commences before the differen- tiation of endothelium is everywhere complete (cf. Stockard 15, B and Sabin 717). It is probable that the period of over- lapping is very short. It is also clear, particularly from the studies of Miss Sabin (17), that the development of arteries and veins takes place to some extent before the circulation is established. In chick em- bryos she found that, before circulation starts, part of the aortae, the two vitelline veins next the heart, parts of the cardinal veins and the duct of Cuvier are clearly present as definite vessels. GROWTH OF BLOOD-VESSELS IN FROG LARVAE 39 That there is a secondary stage in the development of the blood-vascular endothelium, in which the endothelium spreads by sprouting, instead of by the transformation of indifferent cells, has been proven by direct observation. In the transparent fin expansion of the tail of the tad-pole, this process has been watched during life by several observers, especially Golubew (69), J. Arnold (71) and Rouget (73), who have seen blood capillaries send out sprouts, which extended until they met and -anastomosed with other sprouts or capillaries and into which a lumen advanced—all without the interposition ‘of outside cells. This view is supported also, among others, by J. Meyer (’53), Bobritzky (85), His (69), Kolliker (86), R. Thoma (’93), Marchand (’01), Ziegler (05) and Evans (09 A). While this mode of growth has not been proven by direct observation for all vessels in all animals, and while the existence of other modes of extension is perhaps not necessarily excluded, it is a fair hy- pothesis that this is the universal mode of spreading of the vas- cular endothelium, once it has differentiated, and cannot be abandoned until more convincing objections are brought than have been produced up to the present time. It is not the primary purpose of the present study to enter either into the problem as to the time, in embryonic development, at which the second stage begins, nor the problem whether growth by sprouting is the universal mode of spreading during this period. It is rather to consider the problem as to how, ina region where, and at a time in development when growth by sprouting has been repeatedly verified, and after the circulation has become established, the capillaries are transformed into ar- teries and veins; to study the modes of action and reaction of | endothelium—the laws which regulate its growth. Such a study is by no means new, for it has been, through many years of fruitful investigations, the object of W. Roux and particularly of R. Thoma and numerous coworkers to dis- cover the factors which regulate the growth of vessels, while many others, including Nothnagel (’84), Mall (06), Evans (09, A and B, 712) have studied the same problem less extensively. 40 ELIOT R. CLARK The initial stimulus to this study was given by W. Roux (’79), in his Inaugural Dissertation, in which he studied the ‘‘angle of branching” in relation to the relative size of the branch, and the shape of vessels in the neighborhood of a branch. He found that this angle, which lies between a line continuing the axis of the main stem and the axis of the branch, varies with the relative size of the branch—that, in general, the larger (rela- tively) the branch, the smaller the angle, and the smaller the branch, the larger the angle. He also found that the lumen of an artery shows a widening with subsequent narrowing imme- diately after branching, and that the opening of the branch is oval rather than round. By experiments with openings made in vessels and in tubes and with the use of malleable substances such as lard placed in such openings and on the interior of tubes, he found that the direction taken and the shape found is, in the case of the artery, practically the same as the shape and direc- tion of the stream of fluid emerging from openings in vessels and tubes. He concluded that the shape and direction of arteries at the place of branching are determined by the action of hemodynamical factors; that the blood-vessel wall responds by taking the shape which allows a minimum of fric- tion. The general and important conclusion was that the size and shape of arteries and veins, in the growing and adult animal, are regulated, not by heredity, but by the action of mechanical factors. Thoma’s conclusions were based mainly on studies made on the extra-embryonic yolk sac vessels of chick embryos. From a series of injections he found that there is formed, first, an indif- ferent plexus of capillaries, interposed between the aorta and the venous end of the heart, and that out of this plexus, those vessels which are so placed as to have the greatest amount of blood flowing through them enlarge to become arteries and veins, while others remain capillaries, or atrophy. The results of these and other studies by Thoma (711) may be briefly summarized. He finds that blood-vessels are regulated in their growth by mechanical factors, which he expresses in the form of ‘laws’ (‘Histomechaniche Principien’), as follows: GROWTH OF BLOOD-VESSELS IN FROG LARVAE 41 1. Das Wachstum des queren Durchmessers, also des Gefisslichtung ist abhangig von der Geschwindigkeit des Blutstromes. Dasselbe beginnt, sowie die Stromgeschwindigkeit der nahe an der Gefisswand stro6menden Blutschichten einen Schwellenwert iiberschreitet, den ich mit U bezeichnen will, und ist innerhalb gewisser Grenzen ein um so rascheres, je so mehr die Stromgeschwindigkeit tiber den Schwellen- wert U, hinaus zunimmt. Dagegen tritt em negatives Wachstum, eine Abname des Gefiissumfanges ein, wenn die Geschwindigkeit der nahe an der Gefiisswand strO6menden Blutschichten klemer wird als der Schwellenwert v. 2. Das Liingenwachstum der Gefisswand ist abhingig von den Zugwirkungen der das Gefiiss umgebenden Gewebe und zwar sowohl von denjenigen Zugwirkungen welche das Lingenwachstum der umge- benden Gewebe erzeugt als von denjenigen Zugwirkungen, welche bei “Anderungen der Gelenkstellungen eintreten,” etc. 3. “Wird das Wachstum der Wanddicke bestimmt durch die Span- nung der Gefisswand.” This is determined by the blood pressure and the size of the vessel. 4. (proposed as an hypothesis, not yet preven). Die Umbildung von Kapillaren ist abhiingig von dem in den Kapillaren herrschenden Blutdruck und stellt such an denjenigen Stellen der Kapillarbezirke ein, an welchen der zwischen dem Kapillarinhalte und der Gewebs- fliissigkeit bestehende Druckunterschied einen gewissen Schwellen- wert p tberschreitet. Dieser Schwellenwert ist jedoch in den ver- schiedenen Kapillarbezirken je nach den Ejigenschaften der die Kapillaren umgebenden Gewebe verschieden gross. Expressed more simply they are: 1. Increase or decrease in the size of a vessel is regulated by the rate of the blood flow.” 2. Increase or decrease in the length of a vessel is governed by the tension exerted on the vessel wall in a longitudinal di- rection by tissues and organs outside of the vessel. 3. Increase or decrease in the thickness of the vessel wall is dependent upon the blood pressure. 4. New formation of capillaries depends upon increase of pressure in the capillary area (proposed as an*hypothesis—not yet proven). The ultimate controlling factor Thoma considers to lie in the metabolism of the organs (’93, pp. 49-51). It is this which regulates, primarily, the increase or decrease in capillaries, which, in turn, sets in motion the mechanism which results in the increase or decrease in the size of arteries and veins, the increase in strength of heart beat, ete. 42 ELIOT R. CLARK Roux, in his later writings, discusses, mainly in a theoretical way, the factors involved in the increase in size of vessels, and the new formation of capillaries. His views as to the new forma- tion of capillaries, expressed briefly in 1895, repeated more fully in 1910, and again repeated, in a controversial article in 1911, are perhaps most completely expressed in 1910, p. 88, where he Says: Ist der Verbrauch in dem Parenchym, welches eine Kapillare um- gibt, emige Zeit dauernd derartig gesteigert, dass aus den vorstehend erérterten Griinden mehr Stoff als normal hindurchtritt, so wird wohl die an der Stelle stirksten Durchtritts gelegene Wandungszelle durch die verstirkte Leistung in der Richtung des Austritts zur Sproéssung angeregt. Dasselbe geschieht natiirlich auch an der denselben gris- seren Parenchymtheil von der andern Seite der umschliessenden und ernéhrenden Kapillare. Diese noch nicht als Iapillarenfungierenden sprossen treffen, wohl durch chemotropisch vermittelten Cytotropism, aufeinander, also in &hnlicher Weise wie ich es an von mir isolirten Furchungszellen sah, einerlei ob diese Zellen noch freilagen oder schon wieder an etwas anderem (an der Zellen cder am Boden des Gefisses) hafteten. Der vererbte gestaltende Reaktionsmechanismus der Ka- pillarwand, der zum Hohlwerden und zur weiteren Ausbildung der neuen Kapillaren mit Bildung von Nerven und kontraktilen Elementen fiihrt, wird auf diese Weise aktiviert und so eine neue funktionsfabing Kapillare gebildet. Like Thoma, Roux considers the metabolism of the tissue the primary factor in new growth of capillaries. As for the specific stimulus, however, he disagrees. According to Thoma, increased metabolism causes increase in blood pressure in the capillary area, to which the endothelium is thought to respond by sending out sprouts, while Roux’ view is that the new sprout is sent out as a direct response on the part of the endothelial cell to the passage through it of an increased amount of substances. In criticism of Thoma’s hypothesis, Roux (11, p. 201) calls atten- tion to the absence of any noticeable new formation of capillar- ies in tricuspid or mitral insufficiency, in which conditions there is a rise in blood-pressure in the capillaries. Thoma’s first histomechanical law that the size of the vessel is regulated by the rate of blood flow, is criticized by Roux chiefly because he can see no way in which the moving stream can affect the wall, since, as first shown by Helmholtz, there is GROWTH OF BLOOD-VESSELS IN FROG LARVAE 43 a thin layer of fluid next the wall which is immovable. His explanation for growth in size of vessels is that it is brought about through the agency of the vasomotor nervous mechanism; that, following increased metabolism and formation of new capillaries, there is a reflex widening of the arteries and possibly also of the veins of the affected region. This widening, if con- tinued long enough, results in a permanent adaptation of the vessel wall to the increased volume of blood by growth processes. Roux apparently agrees with Thoma’s law as to the increase in thickness of the vessel wall. Mall (06), in an extensive review and discussion of Thoma’s histomechanical laws, finds support for Thoma’s first law, in his studies on the growth of glands. Like Roux, however, he disagrees with Thoma in his hypothesis that the formation of capillaries is dependent on increase in blood-pressure in the capil- lary area. ‘‘In reality,” he says (p. 250), ‘‘we can only state definitely that with the new formation of tissue new blood-ves- sels may grow into it, for all new tissue does not have blood- vessels.”” The precise stimulus for the formation of capillaries is unknown. Again (p. 251), he says, ‘The first and guiding blood-vessel is the capillary, which grows in all directions, form- ing a plexus. Secondary changes made arteries and veins of them and their laws of growth have been discovered and clearly stated by Thoma.” It has been shown by a series of investigators—among them— Erick Miiller (03, 04), Rabl (07), Bremer (12) and H. Smith (09), and particularly Evans (09, A and B) that many of the larger arteries and veins in the body of the developing embryo are first formed as capillaries, which grow as irregular plexuses, and out of which certain ones are differentiated to form arteries and veins. Evans, who has made the most extensive studies in this field, has described the caudal portion of the aorta, the chief veins, the pulmonary, subclavian and sciatic arteries as developing in this manner. He concludes that the histomechan- cal laws of Thoma are the factors which govern the process. A number of investigators have suggested that new capillaries are formed as the result of the action of specific ‘chemiotactic’ 44 ELIOT R. CLARK (better ‘hemangiotactic’) substances outside the capillaries. Ac- cording to Marchand (’01, p. 148), Leber (’88) first suggested this explanation, to which Marchand is slightly inclined. It was suggested again by J. Loeb (93) as an explanation for the growth of vessels in fish embryos whose heart action was elimi- nated by the action of chemical substances. Evans makes a similar suggestion. In each case it has been proposed merely as a tentative hypothesis and has not been tested. Over against this group of investigators whose studies have gone to show that blood-vessels are regulated in their growth by the action of mechanical and chemical factors, and some of whom have attempted to define this regulation in terms of specific laws of growth, there are others who have supported the view that mechanical factors play little if any part in determining the formation of arteries and veins, and who attribute it rather to the action of hereditary influences. Possibly the strongest ad- herent of this view is Hochstetter, who has made so many im- portant studies on the comparative anatomy of the vascular system. His view is probably most concisely presented by his pupil, Elze (12) in an article criticizing the conclusions of Evans and Thoma. In brief, it is that the primitive form of the vascu- lar system is not a capillary plexus, but a single artery and vein, such as is formed in the limbs and digits of amphibians, and also in the segmental arteries; while capillary plexuses are secondary formations. Now it is interesting that support for this view has come in part from the two men who have been most prominent in advocat- ing the regulating action of mechanical factors, namely, Thoma and Roux. Thoma (93, p. 28) mentions that the aorta is de- veloped as a definite vessel before the heart commences to beat, while Roux emphasizes a first. stage in the development of the vascular system, as of other systems, in which differentiation and growth take place as a result of heredity (preformation)— a stage which includes the formation of ‘the anlage of the typi- cally laid down chief vascular stems’ (’95, pp. 326-7, footnote). Roux bases this conclusion on chance observations made on the area vasculosa of chick embryos, in which the embryo failed to GROWTH OF BLOOD-VESSELS IN FROG LARVAE 45 develop, but in which vessels, including the border vein and some others differentiated in situations corresponding with the normal. That growth of capillaries and larger vessels in embryos is regu- lated not entirely by the metabolism of the tissues, but in part at least by hereditary influences, is shown, he believes, by the richness of the capillary plexus in the lung and the relatively great size of the pulmonary arteries and veins, which, ‘according to Wiener,’ are, before birth, four to six times as large as the weight of the lung tissue justifies, in comparison with other organs. Wiener studied the proportion between size of artery and weight of organ. Support is lent to this view by the results of studies made on embryos whose heart-beat has been eliminated experimentally either by mechanical removal or by chemical inhibition. Dareste (77) J. Loeb (93), Patterson (’09), Knower (’07) and Stockard (15 A) agree in finding certain typical arteries and veins formed in such embryos, in which the mechanical action. of the circulation has been eliminated—in fish, frog and chick embryos. The indications are that the truth hes between the two ex- treme views; that what we are forced to call hereditary factors do play a part, not only in the primary differentiation of blood- vascular endothelium and its capacity for growth by sprouting, but in the formation of some of the main vessels in the embryo (how great a part and how long exerted in embryonic life, has not yet been cleared up, cf, Miss Sabin, ’17, previously referred to) that, on the other hand, the vascular system does become, at an early stage, dependent, at least to a very great extent, upon the regulative action of mechanical and chemical forces. Were it found that arteries and veins in latter stages are com- pletely regulated as regards diameter, length, thickness of wall, and position by the action of mechanical and chemical factors, it would be quite compatible with our knowledge of the develop- ment of other tissues and organs, to find that a crude pattern of such mechanically controlled structures should reappear in the embryo (Thoma, 793, p. 28). As to the precise nature of the mechanical and chemical fac- tors which regulate the growth of the vascular endothelium, ‘46 ELIOT R. CLARK there is, as the foregoing review and discussion shows, differ- ence of opinion sufficient to justify further observation and experiment. ; METHODS USED IN PRESENT STUDIES Since most of the studies referred to were made on successive — stages, usually of injected embryos, in fixed preparations, it seemed that it would be worth while to study the changes in the same vessels of the same living embryo, following certain vessels through the critical stages in their development, keeping rec- ords of the circulatory conditions, and of all changes in the size of the vessels, and the direction of the angle of branching, et cetera. For such a study the transparent fin expansion of the tail of frog larvae is admirably adapted, for a larva can, by the use of chloretone as an anesthetic, be kept under observation over a period of weeks, and careful camera lucida records made as frequently as desired. Since the chloretone interferes but little with the heart beat, records can also be kept of the circu- latory conditions in each of the vessels which is being watched. (For details of the method used see E. R. Clark (’12).) In the most extensive series of studies made on a single tad-pole, the observations were started when the larva (rana sylvatica) first became transparent enough to enable the course of the vessels in the dorsal fin to be made out, and records were made at daily intervals, at first, when new formation of vessels was most rapid, later, when changes were slower, at considerably longer intervals. During the observations the larva increased in length from 10.5 mm. to 29 mm. There was thus procured a record giving the vascular changes, with notes as to the condition of circulation in each vessel for a considerable section of the fin, from a stage at which the entire system consisted of a few capillary loops, to a stage in which a fairly complicated system of arterioles, capil- laries and venules had developed. In addition to this series of studies, numerous shorter studies were made, on larvae of r. sylvatica, r. palustris and r. catesbiana. Brief reference has been made in an earlier paper (E. R. Clark (’09) ) to the blood- vessel changes in the tail of the frog larva, and some of the matter GROWTH OF BLOOD-VESSELS IN FROG LARVAE 47 included in the present study was presented at the meeting of the Am. Ass. of Anat. in 1914 (E. R. Clark (715), ) where draw- ings were shown. DESCRIPTION OF FINDINGS When the blood-vessels in the dorsal fin of r. sylvatica larvae first become clearly visible, owing to the absorption of some of the yolk and pigment present in young larvae, they form a sys- tem of capillary loops, making an irregular meshwork of rather wide vessels, all connected with one another. On the arterial side they are connected with the main caudal artery, and on the venous side with the main caudal vein, which are located ventral to the notocord, and between the two layers of myotomes. The vessels reach the dorsal fin by passing dorsally between the notocord and spinal cord in the center and the layers of myo- tomes on either side. With the low power of the microscope their course may be easily followed from the main caudal vessels to their emergence from between the myotomes. With the higher power this is more difficult, and in most of the studies made, only the vessels in the dorsal fin proper, after their emer- gence from between the myotomes, are drawn. While in many of the studies all the vessels iu the dorsal fin have been followed, a small area is selected for closer study, and for reproduction, because any section illustrates the funda- mental principles involved in blood-vessel development. In the series which is reproduced an area was chosen which included an arteriole and a venule and the region between the two, as well as a part of the regions on either side. This area is sufficiently large to make it possible to follow the changes introduced by the development of new capillaries on the vessels already present. The changes which occur in such a selected area are shown in figures 1 to 8, and will now be taken up in detail and analyzed. There is present in the first record a very simple type of cir- culation. An arteriole, or, perhaps better, an arterial capillary is seen toward the left. Two branches are given off from this vessel on the right, and two on the left, through which blood corpuscles are circulating. In addition there is a third branch 48 ELIOT R. CLARK Figs. 1 to 8 Camera lucida drawings of blood-vessels in a section of the dorsal fin expansion of the tail of a rana sylvatica larva, on April 15, 16, 17, 18, 20, May 12, 20 and 31. Length of larva April 15, 10.5 mm., May 31, 29 mm. Arrows indicate direction of circulation. Relative rate of circulation indicated as FAST, MOD., moderate, SLOW, NO C., no circulation. Corresponding ves- sels are numbered. Vessels present in one drawing which have been retracted in the next are cross-hatched. The positions formerly occupied by vessels which have retracted are indicated by dotted lines. In figure 8, the vessels which were present in figure 1 are stippled. enl. (1: 50). GROWTH OF BLOOD-VESSELS IN FROG LARVAE 49 on the left, with a continuous lumen, but without circulation, and a fourth which ends blindly. The arteriole ends with a bend to the right, from which a long thread extends to another non-circulating vessel. Following the two branches to the right, it is seen that the first pursues a winding course, while the sec- ond passes fairly directly to the main venule or venous capillary, near the right. Between the two branches are three communi- eating vessels, the last of which forms a non-circulating loop. Below the venule there is a rather elaborate plexus of vessels 50 ELIOT R. CLARK containing, for the most part, lumina, but without circulation. . Extending peripherally from them are several blind-ending pro- jections. As regards the rate of circulation, through branch 6 it is relatively ‘fast,’ while through branch 10 it is ‘slow.’ To sum- marize the condition of the vessels for the small area selected, there is a simple irregular plexus of capillaries with wide and _ rather irregular lumen, some of them with and some without circulation. For convenience, the principal afferent and effefent capillaries have been called arteriole, or arterial capillary, and venule, or venous capillary, though they are of capillary size and appearance. In figure 2 (a day later) a ‘slow’ circulation has started in a number of vessels which had been non-circulating on the previ- ous day, and several new sprouts and connections between sprouts have formed. In figure 3 this has continued, and has been accompanied by an increase in the rate of circulation in some of the vessels. On the other hand, there are some vessels in which the circulation has diminished, or ceased altogether. In figures 4 and 5 the same processes have continued—a slight formation of new vessels, with modification of the rates of circu- lation in many of the vessels, an increase in some and a decrease in others. in addition a new change has become marked, namely, the disappearance of certain capillaries, in which the circulation had ceased, or in which the circulation had never started. Owing to the fact that the prolonged use of chloretone had caused a slowing of growth processes, the larva was allowed to develop in fresh water, with observations at less frequent inter- vals, in order to see the fate of the vessels being watched, after a considerable amount of growth had taken place. A record was made on April 22, which showed very little change from the record of April 20. The succeeding records were made May 12, May 20 and May 381. While these later records are not close enough together to show all the growth changes, they show very well the new capillary areas which have developed, their relation to the vessels already present, and the changes which the earlier formed vessels have undergone, in consequence. Thus it will be GROWTH OF BLOOD-VESSELS IN FROG LARVAE 51 seen that in figure 6 and 7 the non-vascular zone toward the mar- gin of the fin has become much reduced in extent by the forma- tion of new vessels, until only a very narrow non-vascular zone is left. It will also be noted that there has been a general expan- sion of the tail, so that the meshes between the vessels are not- \* aes) fl ; yi ae SSS aD) MAY 12- 13 arcs > 52 ELIOT R. CLARK iceably larger, and the vessels longer. The last record shows this enlargement of the tail most markedly. The tail has in- creased not only in length and height, but also in thickness, and with the enlargement there has been a very great development of new capillaries, in the widened spaces of the blood-vessel meshwork. In the half of the fin next the muscle, where the growth in thickness has been most pronounced, many of the new capillaries are in new planes, more superficial than the earlier formed vessels. As regards the fate of the vessels present in the earlier stages, it is seen that there has been a marked dif- ferentiation. In figure 1 the vessels present are nearly all of a uniform diameter. In each successive record there is a progres- sive differentiation, in which certain capillaries increase in size, others remain of the same, or slightly diminished caliber, while others disappear. In the last stage this differentiation is seen at its maximum; definite arterioles and venules have formed, which supply and drain considerable capillary areas. In this elaborated system there are present many of the same vessels and parts of vessels which were present in the first stage recorded. Some have been incorporated as parts of the larger vessels, others are still capillaries, while others have disappeared. Considered as a whole, then, this series shows strikingly that arterioles and venules develop, at least in this region in ‘tad- poles, not by a steady outgrowth of a single vessel, which grows straight ahead into a new region, giving off branches where they are needed, and fulfilling its predetermined destiny to grow in a particular place, but rather by the sending out of numerous capillaries, in various directions, which anastomose, adding new loops of circulating capillaries to those already present. Of these new loops some are so placed that a circulation is never estab- lished through them, and they disappear; others are incorporated | as parts of arteriole or venule or remain as capillaries. The ef- fect of the addition of new capillaries on the system already present depends upon the relation’ which the older parts bear to the new; thus a vessel which is at one stage the chief vessel of the region may entirely disappear, while another vessel, which is small, and has a slow circulation at one stage, may later be a GROWTH OF BLOOD-VESSELS IN FROG LARVAE 53 part of the main arteriole or venule of the region. It is impos- sible to predict at one stage, which way a capillary will go— whether it will increase in size, remain the same, or atrophy and disappear; it all depends upon the relation which it bears to the other vessels in existence at the time, and to those which are developed later. The endothelium is equipotential, then, and its differentiation into arteries, veins and capillaries is determined by factors outside the endothelial wall or in the lumen. Further evidence for this view is found in the following experi- ments on chick embryos. They were performed to test another point, but the results are sufficiently interesting in their bearing on the problem of blood-vessel growth to deserve brief mention. The anterior cardinal view of one side, from a point anterior to the otic vesicle to and including a part of the duct of Cuvier, was dissected out from chicks of two and one-half to three days incubation. The method employed is as follows: Berlin Blue is injected into the vein through a very fine glass cannula. As MAY 20-21 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 54 ELIOT R. CLARK soon as the Berlin Blue mixes with the blood it forms a precipi- tate which plugs the vessel and sticks to the endothelial wall, outlining the position of the vein. Using this as a guide, the vessel is dissected out, with considerable of the surrounding tis- sue, to make certain that all is removed. The egg is then sealed and the chick allowed to develop further. This experiment was performed successfully six times and in every case, there was found to be a large vein in the place of the one removed. In one case, the vein on the operated side was larger than the one on the unoperated side. The chicks were examined four to eight days after the operation. The conclusion seems justified that the secondary develop- ment of a large vein in the neck, in the place of the one removed, indicates that the mechanical conditions of the circulation favor the growth of a large vein in this region. Surely the new vein can hardly be considered as due to inheritance. What are the laws which govern the growth of the endothe- lium, making the differentiation of such an elaborate system pos- sible? What, in other words are the modes of reaction of blood- vascular endothelium? THE FORMATION OF NEW CAPILLARIES The first property to be noted is the capacity of the endothe- lium to send out sprouts. This process has been frequently ob- served in the transparent tails of living frog larvae, and verified by other studies, and is the generally accepted mode of spreading of the vascular system, after its primary differentjation. The sprout consists of an elevation of the endothelium which is sent out, usually starting at right angles to the vessel wall, and with a lumen continuous with the lumen of the parent vessel. The end of the sprout consists of a solid process of varying length, which may be in the form of a single thread, or of a thread with one or more branches. This process usually extends in a straight . line from the parent vessel, for a varying distance, and may then curve. Sooner or later it reaches a similar sprout, or approaches a fully formed capillary, when it shows itself possessed of a prop- - GROWTH OF BLOOD-VESSELS IN FROG LARVAE 55 erty most important for the development of a system of anas- tomosing vessels, namely, that blood-vascular endothelium has an affinity for blood-vascular endothelium, of such a nature that if two processes of blood-vascular endothelium draw near one another in their growth, a union will be formed between them (‘eytotropism,’ Roux). Equally important is the fact, readily ob- servable in the tail of the frog larva, that blood-vessel endothelium avoids, in its growth, the cells of other tissues among which it grows, such as mesenchyme, and lymphatic endothelium. Asa rule, the lumen eventually extends through the entire extent of this new sprout, it widens, and after a varying amount of time 56 ELIOT R. CLARK the circulation of blood cells commences, and a new circulating capillary has been added to the system. ‘This whole process may, however, not be completed, for some sprouts grow out a short distance and are retracted, while some in which the lumen has been formed, never have a circulation, but retrogress—becom- ing solid, and disappearing. ‘Throughout this process the endo- thelium remains complete, the lumen being separated from the tissue fluid outside by a complete investment of endothelium. The facts concerning the morphological changes which take place in the formation of sprouts are clear enough; the question then arises as to why sprouts are sent out, to what sort of stimu- lus the endothelium responds when it sends out a sprout. The answer to this question is not entirely clear, yet certain facts to- gether with certain general considerations justify the proposal of an hypothesis. A study of the positions at which sprouts are formed and of the general direction taken in their growth shows that they are pre- ceded in their formation by the growth of the other tissues and that they extend into regions where the amount of tissue not yet vascularized is greatest in amount. In the tad-pole’s tail, at early stages, vessels develop first along the muscle—the thick- est part of the tail. Later they grow out into the fin expansions, which attain a considerable size before vessels reach them. Growth of new capillaries continues in a general direction toward the dorsal and ventral margins of the fin, until eventually the plexus reaches nearly to the margin. During the growth of this first set of vessels, the fin remains thin, and the capillaries—save for the thickest part next the muscle—are all in a single plane. Later, the fin becomes much thicker and there occurs a corre- sponding new growth of capillaries, from the older parts of the plexus, which pass toward the epidermis, and form plexuses in two new planes. In both cases it is clear that the growth of new blood-capillar- ies has been secondary to the growth of the outside tissue. It has been suggested by Thoma as an hypothesis that the stimu- lus responsible for sprout formation lies in an increase in blood- pressure. If this were so, one would expect to find them growing GROWTH OF BLOOD-VESSELS IN FROG LARVAE ov out from the arterial rather than the venous end of the capillary, since, obviously, the pressure is higher in the arterial end. This, however, is not the case—at least, in the tad-pole’s tail new sprouts grow out as frequently from the venous as from the ar- terial ends. Loeb (’93) suggests that the explanation of the new growth of blood-capillaries must be sought in the stimulus ex- erted by specific chemical substances outside the capillary. 56. 15 From a photograph of a portion of a transverse section through the pos- terior portion of the tripartite complex showing the compactly arranged cell cords of the right ultimobranchial body in which is located a cyst. The area in- side the dotted circle is free from colloid. From an embryo 125mm. long. X 45. 16a and 16b From photographs of portions of a transverse section of ths tripartite complex showing respectively the right and left ultimobranchial bodies. The right one is only partially imbedded in the thyroid gland and contains many cystoid follicles (C.F.) which do not contain colloid and a few small follicles which contain colloid (Co). The black dots in the portion of the figure labeled ‘thyroid’ represent colloid. The left ultimobranchial body is more deeply im- bedded in the thyroid gland. From an embryo 125mm. long. X 38. 17. From a photograph of a portion of a transverse section of the trip2zrtite complex showing the right ultimobranchial body in which are found both small and cystoid follicles that contain colloid. From anembryo 145mm. long. X 38. 18 From a photograph of a portion of a section of the tripartite complex showing the left ultimobranchial body which is represented by an area of small follicles. The black dots in the figure represent colloid. From an embryo 160 mm.long. xX 38. C., cyst T., thyroid gland C.T., cystoid follicles U., ultimobranchial body Co., colloid 128 FATE OF THE ULTIMOBRANCHIAL BODIES PLATE 3 J. A. BADERTSCHER y on PR alr ees Me et Sere BO 129 PLATE 4 EXPLANATION OF FIGURES 19 From a photograph of a portion of a section through the left ultimobran- chial body and a portion of the thyroid gland surrounding it. The ultimobran- chial body is characterized by follicles which contain colloid and which are on an average appreciably smaller than the follicles of the thyroid gland. From an embryo 225 mm. long. X 38. 20 Froma photograph of a portion of a section through the right ultimobran- chial body and a portion of the thyroid gland surrounding it. The ultimobran- chial body contains many cystoid follicles which contain colloid. The colloid dropped out from some oi the follicles during the process of staining. From an embryo 245 mm. long. 38. 21 From a photograph of a portion of a section through the left ultimobran- chial body and a portion of the thyroid gland. The ultimobranchial body is characterized by an area of small follicles in which is located a small area free from colloid. The light dots represent follicles from which the colloid has fallen. This figure represents the ultimobranchial body at C in figure 22 a. From No. 2 of the embryos 270 mm. long (full term). 38. 22 a, 22 b, and 22c¢ These figures show the relative size of the ultimobranchial bodies and the thyroid gland in No. 2 of the embryos 270 mm. long (full term). The extent of the ultimobranchial bodies is outlined by a dotted line. Inside the left ultimobranchial body is a small area (X), also outlined by a dotted line, which is free from colloid (figs. 22 a and 22¢). The portion of the left ultimo- branchial body outside the area \ and all of the right one is characterized by follicles which are on an average appreciably smaller than those of the thyroid gland. Figures 22 b and 22¢ represent cross sections through the tripartite complex at b and ¢ respectively of the structures represented in figure 22a. Uctay. T., thyroid U., ultimobranchial body 130 FATE OF THE ULTIMOBRANCHIAL BODIES PLATE 4 J. A. BADERTSCHER —~ Anterior end Posterior end ’) PAPAS 131 1 ee ea AUTHOR'S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 1 CHONDRIOSOMES IN THE TESTICLE-CELLS OF FUNDULUS J. DUESBERG Carnegie Institution, Department of Embryology, Baltimore, Maryland TWENTY-ONE FIGURES (TWO PLATES) Our knowledge of chondriosomes in the spermatogenesis of fishes is limited, as far as I know, to an incomplete account on Myxynoids by A. and K. E. Schreiner (’05, ’08). In the ripe spermatozoon, however, the same bodies have been studied, es- pecially by Retzius, in quite a large number of species. According to A. and K. E. Schreiner, the chondriosomes are represented in the spermatogonia as well as in the spermatocytes of Myxine glutinosa by very small granules, tightly crowded to- gether in the neighborhood of the ‘Sphire.’ No change in shape is observed during mitosis; furthermore, the behavior of the mass of mitochondria seems to be entirely passive and conse- quently its segregation between the daughter-cells is often un- equal. Concerning the process of spermiogenesis, these authors merely state that the mitochondria build a sheath around the axile filament. It must be added that the preservation of the chondriosomes in the material used by A. and K. E. Schreiner can hardly be considered as satisfactory. Retzius has studied the ripe spermatozoon of Amphioxus (05 b), of several selachians (’09 c; 710 b), of one ganoid (Amia. ealva, ’05 b) and of a number of teleosts (05 b;’10b). As data concerning the process of spermiogenesis In selachians are lacking, in reference to the chondriosomes at least, it is hardly possible to decide what part of the spermatozoon is formed by these bodies. In the other classes however, their identification is easier and the concordant observations of Retzius on Amphioxus, Amia and teleosts can be summarized as follows: the chondriosomes of the ripe spermatozoon are located at the posterior part of the head 133 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 134 J. DUESBERG and surround usually for a short distance the proximal part of the tail. The shape of this sheath varies with the different spe- cies. In Amphioxus the chondriosomes are represented by a rather voluminous body in which, by careful study, one can make out three to five granules. In Amia, such a body appears indistinctly granular and fits the posterior part of the head as the cup fits the acorn. In teleosts similar dispositions are found, for the details of which I refer to Retzius’ papers. I wish to empha- size that in a number of species the granules are very distinct and even constant in number. In Lophius piscatorius for in- stance, Retzius (10 b) invariably found four of them, disposed in regular order around the origin of the tail. It may well be recalled that a similar disposition of the chon- driosomes has been observed in invertebrates. For instance, ac- cording to the observations of Meves (’00, ’03), each spermatid of the apyrene generation of Paludina vivipara contains four chondriosomes. ‘They assume the form of spheres and occupy: the posterior part of the head, where they surround the axile filament. Bonnevie (’07) gives a similar description for Mem- branipora pilosa. In these cases however, this stage is a tran- sitory one, for the shape of the chondriosomes changes during the further evolution of the spermatid, while in other inverte- brates the same arrangement is, according to Retzius, retained in the ripe spermatozoon, namely, in a number of celenterates (04 a and b; 05a; ’09 a), in many echinoderms (’04 a and b; ’03 a; 10 a),! in worms (’04 a and b; ’05 a; ’06b, ¢ and d; ’09 b) and in mofluses (’04 a and b; ’05b; 06a; ’10b). In many species belonging to the two last-named classes the numerical constancy of the chondriosomal spheres and the regularity of their arrange- ment around the axile filament are conspicuous features of the ripe spermatozoon. Especially remarkable is the disposition in 1 Meves (’12) contests the accuracy of Retzius’ description for Parechinus miliaris. He finds that the so-called ‘Mittelstiick’ is not granular, as stated by Retzius, but homogeneous, and that it has the shape of a ring, through which runs the axile filament. I take this opportunity to remind how inadequate is the expression ‘Mittelstiick’ or ‘middlepiece,’ as, between the ‘Mittelstiick’ of the spermatozoon of an echinoderm, of a selachian or a urodele amphibian and of a mammal, there is no homology whatever. TESTICLE-CELLS OF FUNDULUS 135 polychetes: the granules in many species are consistently four in number, their form being exactly spherical, their volume equal and their disposition around the proximal part of the tail per- fectly regular. The origin of these bodies is briefly referred to by Retzius (04 a and b) who states (for molluses) that the large spheres are formed by the confluence of smaller granules. Previously, Pic- tet (91) and Field (95) both had described the formation of the so-called ‘mittelstiick’ in the spermatozoon of echinoderms through fusion of highly refractive granules which, however, they erroneously derived from remnants of the spindle. Quite recently M. R. Lewis (17) has stained the chondriosomes (the so-called ‘middlepiece’) of the spermatozoon of Cerebratulus lacteus and of Echinorachnius parma in vivo, by using a solution of Janus-green in sea-water.? The object of the present investigation is the testicle of Fun- dulus (heteroclitus and majalis), the main purpose being to study the behavior of the chondriosomes during spermatogenesis. The material was collected in Woods Hole, Mass., in June, 1916, and fixed in Regaud’s or in Benda’s fluid, the latter either with or without acetic acid. The sections, 5 u thick, were stained in the first case with iron-haematoxylin or acid fuchsin-methylgreen ; after Benda’s fixation, I resorted to iron-haematoxylin acid, fuchsin-methylgreen or Benda’s stain, the latte? giving, as previ- ously stated for embryos (’17), a very small percentage of good preparations. A number of sections were stained with safranin, in order to study the chromatin. Once more I found that the preservation of the tissue is much better after Benda’s fixation than after Regaud’s. This last reagent has a pronounced tendency to make the seminal cells 2 In the same paper, M. R. Lewis (p. 33) quotes my opinion, as expressed in my review (’12), on the vital staining of chondriosomes and, from this quotation, one might be induced to conclude that, for me, neutral red and methylenblue can be used or have been used to stain the chondriosomes in vivo. To prevent any misunderstanding, I wish to recall that this has never been my opinion, as appears clearly in the quoted place of my article (p. 608), as well as in several others (for example, page 823, in the discussion of Arnold’s plasmosomes). 136 J. DUESBERG swell. The extent to which the ground substance is affected is well illustrated by the difference in size exhibited by the cells represented in figures 10 and 11, both in exactly the same stage of evolution, the first from material fixed in Benda’s fluid, the second from material treated with Regaud’s. Thus, cells which normally float freely in the cystic cavity are made to conglomer- ate and stick together. The chondriosomes are also swollen, and the chromosomes are transformed into an undecipherable clump. In contrast to this, the last-named bodies are well preserved ‘in Benda’s material, an appearance which confirms that swelling in Regaud’s rather than shrinking in Benda’s fluid is responsible for the differences between the two sets of preparations. The testicle of Fundulus is at the time of spawning a rather voluminous organ formed by a considerable number of tubular cysts in which spermatogenesis proceeds from the periphery towards the hilus.? The excretory system of the gland consists of a number of ducts lined with cubic or cylindric epithelium. In the distal part of these ducts the cells (fig. 1) contain, besides secretion-granules, a large number of chondriosomes. These are mostly long chondrioconts running along the nucleus in a direc- tion perpendicular to the basis of the cell and intertwining at both poles of the nucleus. This disposition reminds one some- what of the structure of the cells of the tubuli contorti (Heiden- hain’s rods) or of the salivary ducts (Pfliiger’s rods). The inner - part of the cell is often free of chondriosomes and _ irregularly delimited, an appearance which may be due to the action of the fixing fluid. In the cells lining the proximal part of the excretory ducts, the chondriosomes are all replaced by granules of pigment. This recalls an observation made by Prenant (11) on the skin and cornea of the frog. Prenant found that the cells of both layers in the skin contain mitochondria and pigment-granules. In the upper layer the granules of pigment are located near the surface, the mitochondria in the lower part of the cell, while in the deeper layer mitochondria and pigment-granules are mixed 3 Degenerating cells are, as in other testicles and especially in invertebrates, by no means infrequent in Fundulus. TESTICLE-CELLS OF FUNDULUS eo %/ together. In the cornea no pigment is present. If one studies the point of transition between cornea and skin, one can see how the mitochondria gradually take the place of the pigment-gran- ules. This observation is interpreted by Prenant, apparently not without reason, as indicating the transformation of chondrio- somes into pigment and in the same sense could be interpreted the conditions just described in’ the excretory ducts of the fish- testicle. The seminiferous cysts are reunited by thin sheets of connective tissues containing blood-vessels and cells. Some of these are conspicuous by their large size and by the presence of a great number of bacilli-shaped chondrioconts (fig. 2); others contain also granules which I am inclined to consider as secretion-prod- ucts. In places where the connective tissue is somewhat more abundant, for instance in such stellar spaces as appear between the cross sections of the cysts, they usually build groups of two or more elements. The nearest interpretation of these cells is that they correspond to the interstitial cells of the ma nmalan testicle. Supposing I were right, this would be the first men- tion of them in fishes, ‘or, as far as I know, the literature does not contain any mention of interstitial tissue in this class of vertebrates: in fact Friedmann (’98) and Ganfini (02) state positively that they could not find it. The distal part of the cysts is occupied by cells which are obviously the stem of the whole seminal lineage and as such should be designated as spermatogonia. Since, as we shall see, several generations of spermatogonia can be distinguished, I would call these ‘primary spermatogonia.’ ‘Their size is rela- tively large (fig. 3, two cells on the top row and two cells at the right). Each nucleus contains usually only one large, sharply delimited, spherical block of chromatin. The eventual occur- rence of multiple nucleoli is often accompanied by the presence of indentations (the process is just indicated in figure 3, in the cell of the top row, to the right), which are suggestive of direct division. Mitosis however was repeatedly observed (figs. 4 and 5). It would not be surprising if these indentations were indic- ative of a process described_as occurring in the spermatogonia of 138 J. DUESBERG Salamandra after the period of sexual activity (namely by Meves 91), as once that period over, the testicular conditions are very similar in both the amphibian and the fish. The chondriosomes of the primary spermatogonia deserve special mention. In the resting cell they are numerous, coarse and irregular granules or rods. Most of them are located very close to the nucleus and cover its surface. This disposition might be interpreted in favor of Goldschmidt’s chromidial theory. Such a claim however would be unfounded: Goldschmidt and his pupils basing themselves upon defective observations, ex- pected to demonstrate that the chondriosomes of the germ-cells were formed during the growth-period and they have failed utterly. The continuity of the chondriosomes on the other hand has been demonstrated in a number of animals and is strongly supported for fishes by my observations on the fish- embryo (’17).° It is however far from my mind to deny the 4 For a complete historical and critical account of the chromidial theory, see the third chapter of my review (712). Shaffer, who seems inclined to believe (p. 414) in a nuclear origin of the chondriosomes, gives as an argument that ‘‘in nearly all the growth-stages of the first spermatocytes, there is present a denser and more deeply staining perinuclear zone,’’ formed by the chondriosomes. I should take exception to this statement, for it is characteristic, even if not quite general, that the male auxocytes have their chondriosomes accumulated at one pole of the nucleus, around the idiozome. 5 In a paper on the testicle of opossum, Jordan (711) claims that he has demon- strated the discontinuity of the chondriosomes in the seminal cells. I have been investigating lately the same object and my observations are in direct contradiction with Jordan’s claim: chondriosomes exist in abundance in all the stages of the evolution of the seminal cells. Shaffer (17) enters against the theory of the continuity of the chondriosomes in the following way: ‘“‘(p. 423) the progressive increase in the amount of mito- chondria (during the evolution of the seminal cells) seems to indicate that they are differentiation-products. Hence, if there is any genetic continuity between the mitochondria of successive cell-generations, it is only of a limited sort. The conception that the mitochondria present in the somatic cells are the direct descendants of those of the germ-cells, from which they have arisen, certainly has very little evidence in its favor.’”” I must state that I entirely fail to see an argument against the continuity of the chondriosomes in the fact that their amount may tncrease. Concerning the continuity of the chondriosomes in the somatic cells with those of the germ-cells,Shaffer overlooks apparently the num- erous observations which have shown this continuity, from the egg at least to the embryonic cells. I limit myself to remind of my own observations on the bee, TESTICLE-CELLS OF FUNDULUS 139 existence of nucleocytoplasmic exchanges, as the nucleus is cer- tainly not a sort of impermeable rubber-vesicle enclosed in the cell. But it would be rash to base on the mere existence of such appearances as described above any definite conclusion. The ar- guments for the cytoplasmic nature of the chondriosomes I do not want to repeat here and refer the reader to former papers, limiting myself to state that no indications of a nuclear origin can be found in the staining reactions.® During the mitotic division of the primary spermatogonia the shape of the chondriosomes changes somewhat: they round up and become more regular (figs. 4 and 5). Their location in the cell is also modified: at the stage of metaphase they surround the spindle (fig. 4) and later are found between the daughter- nuclei (fig. 5).7 Next to these cells are others differing but slightly from them. They are somewhat smaller in size and their chondriosomes are not quite so coarse. These cells are assembled in rosettes of the rabbit and quite lately on Ciona, where the chondriosomes form the mate- rial of the yellow crescent, the continuity of which has been demonstrated by Conklin. ° The original colors of the preparations could not be reproduced in the plates; as is well known, they are, in acid fuchsin-methylgreen preparations, red for the chondriosomes and green for the chromatin; in Benda’s preparations, dark pur- ple for the chondriosomes and pale brown for the chromatin. 7 Concerning the fate of the chondriosomes during the mitotic division of the spermatogonia of Passalus, Shaffer expresses himself as follows (p. 410): ‘‘the spermatogonial cysts which are in mitotic activity, stand out very clearly in con- trast with the resting cysts. This is because of their lighter staining capacity; whether this in turn is due to the partial disappearance of the mitochondria could not be ascertained.’’ Shaffer quotes Buchner as having found that in Gryllotalpa vulgaris, the chondriosomes disappear during or just before cell- division and gives three possible explanations “‘for the partial loss of mitochon- drial structure during mitotic activity.” Interesting though they may be, these explanations appear to me for the present useless, as, after my own experience, chondriosomes do not disappear during mitosis, no more in Gryllotalpa, as I have shown (710), than in any other case I know of. Payne (717) quotes both Buchner and me and sees no reason why we should differ so much in our observations: ‘‘In this case, one or the other has certainly made a mistake.’’ Between a negative result, however, and a positive one, there is, In my opinion, no room for hesitation. It must be added that since, Buch- ner has considerably modified his attitude towards the chromidial theory, as appears from a text-book he recently published. - 140 J. DUESBERG three or more (fig. 3 on the left below). So unvarying are these features that I feel justified in considering these cells as a distinet generation of spermatogonia and term them ‘secondary sperma- togonia.’ The primary and secondary spermatogonia are in close contact with each other, the cystic cavity being at these stages only virtual, in contrast with all later stages, when some room, in well fixed material, is left between the celis. The spermatogonia belonging to a third generation are, if any, not much smaller than the secondary spermatogonia. In the nucleus several blocks of chromatin are present. The chondrio- somes are granules, most of them regular, some larger and coarser. Instead of surrounding the nucleus, as in the preceding generations, they are all located at one of its poles (fig. 6). During mitosis a breaking-up into smaller granules appears to take place. Their behavior is the same as described above and is illustrated for the stage of metaphase by figure 7. In fact, the size of the spindle is in proportion to the size of the cellso large that the chondriosomes have to take whatever place, they can in the cell-body, which is practically filled by the karyo- kinetic figure. In the first spermatocytes (fig. 8) the polar location of the chondriosomes persists throughout the whole growth-period until the prophase of the first division and coincides as always with the polar field, while in the nucleus the usual structural changes take place. The chondriosomes are now granules all equal in size and regularly spherical and most of them are very closely heaped together. It must be noted that during this so-called growth- period the spermatocytes of Fundulus actually grow very little and that there is no evidence, as in other spermatocytes, of an increase in the mass of chondriosomes. At the prophase of the first division the mitochondria become scattered all around the nucleus and, when the spindle is formed, they are as previously pushed towards the periphery of the cell- body and very close to it; for here again the spindle is very large in proportion to the cell. I may mention in passing that the centrioles appear very conspicuously at the poles of the spindle (fig. 9). During the anaphase all the mitochondria are found TESTICLE-CELLS OF FUNDULUS 141 between the daughter-nuclei (figs. 10 and 11). The same proc- ess 1s repeated during the second division (fig. 12). Though the cells are very small, it is easy enough to distinguish both mitoses owing to the following characteristics. The first spermatocytes are larger than the second ones. The spindle at the stage of metaphase is more slender in the second division. ‘The number of chondriosomes decreases conspicuously. Finally the size and shape of the chromosomes as observed in Benda’s material pre- sent a most distinctive character: in the first division they are unmistakably heterotypic. The spermatids, which are exceedingly small, very soon form an axile filament. At first the mitochondria are scattered all around the nucleus but only fora short time. In the succeeding stage which is very characteristic and which, judging from its frequent occurrence in the preparations, lasts apparently a con- siderable period, all the mitochondria are found accumulated in one heap at the posterior pole of the nucleus where they surround the proximal part of the axile filament (fig. 13). A glance at these cells readily gives the impression that the number of their mitochondria is constant. When one attempts to count them however, one realizes that to obtain exact figures is almost im- possible, for the granules are very small and not all in the same level. The numbers I found in the most favorable cases came very close to eight. Further stages of spermiogenesis are characterized by changes in the mitochondria (which will be described below), the growth of the tail and the following modifications of the nucleus. First, the posterior side, which is in close contact with the mitochon- dria, becomes flattened or even somewhat concave (fig. 14). Its chromatic content then gradually accumulates at the periph- ery, with the exception of the posterior or flattened side, a proc- ess whose occurrence has been described several times in inver- tebrates and which begins in Fundulus at the stage represented by figure 14. The crust of chromatin thus formed assumes the outline of a horse-shoe, the space existing between the free ends of its branches being occupied by the mitochondria; from the same space emerges the axile filament (fig. 15 et seq.). Later, 142 J. DUESBERG the head becomes somewhat elongated and the branches of the horse-shoe are by the same process brought nearer together (fig. 16, 17 and 18). At the same time the head loses its symmetry inasmuch as it becomes somewhat curved along its antero-pos- terior axis and its posterior facet becomes oblique, instead of being perpendicular, to the same axis. From this time on we can distinguish what I have, arbitrarily of course, termed face-views (figs. 16 and 21) and side-views (figs. 17, 18 and 20) of the spermatozoon. All the modifications of the head are more easily followed on acid fuchsin-methylgreen preparations than on Benda’s, for methylgreen gives a sharper stain for chromatin than sodium- sulfalizarinate. In material fixed with Regaud’s fluid the clear middle-space of the head appears very conspicuous even in the last stages; but curiously enough, as soon as the spermatozoa have reached the excretory ducts, the staining reaction changes and the head takes up acid fuchsin instead of methylgreen. In preparations made from material fixed with Benda’s fluid the ripe spermatozoa, that is, those which have reached the exere- tory ducts, appear somewhat different from those fixed in Re- gaud’s fluid. In a side-view (fig. 20) the clear middle-space ap- pears only indistinctly. In face-views (fig. 21) on the other hand, the same space is very conspicuous and sharply delimited, and has the appearance of a canal running from the posterior to the anterior extremity of the head. During this period changes take place in the mitochondria also. Their number decreases and their size increases: in other words, there is a fusion of granules. This process can be best followed in Regaud’s preparations for the reason that the thin sheet of protoplasm which keeps the mitochondria in place (figs. 14 and 15) and which is hardly visible in Benda’s preparations, swells in Regaud’s fluid as do also the mitochondria themselves. Consequently, the cells and the granules are somewhat larger than in Benda’s preparations and they are more scattered. These differences are well illustrated by figures 15 and 19, which represent approximately the same stage, after Regaud’s and Benda’s fixation respectively. Thus in figure 15 we can count TESTICLE-CELLS OF FUNDULUS 143 exactly six granules while Benda’s preparations of the same stage (fig. 19) show an undecipherable heap of mitochondria. Later when the asymmetry of the head has become conspicuous, we find almost invariably four mitochondria (fig. 16), disposed with remarkable regularity upon the posterior facet of the head. Finally im the ripe spermatozoon the number is still more re- duced, usually to three. Here Benda’s material is more service- able than Regaud’s owing to the change in the staining reactions of the head mentioned above. A comparison of the different stages of this evolution, as they appear after fixation in Regaud fluid, shows that the increase in volume of the mitochondria is not directly proportional to their decrease in number (figs. 13 to 18); and, as there is no evidence of an elimination of mitochon- dria, one would be led to believe in a strong condensation of the chondriosomal substance. This conclusion is however not sup- ported by Benda’s preparations and I am forced to admit that the swelling produced by the formalin-bichromate mixture is greater in the first stages of spermiogenesis than in the later ones. As stated above, the average number of mitochondria in the ripe spermatozoon, as counted in Benda’s preparations, is three. They are especially conspicuous in face-views (fig. 21), where they are found regularly disposed on the posterior facet of the head. Occasionally spermatozoa are found with four, five or even six granules taking the chondriosomal stain. The majority of these granules are undoubtedly mitochondria and in such eases the fusion has, for some unknown reason, apparently not pro- ceeded normally. Whether it is completed later is difficult to say. It is probable also that occasionally the centrioles are stained, for in certain cases it was possible to recognize a rela- tionship between the proximal extremity of the axile filament and a small granule stained like a chondriosome (fig. 20). I cannot give any definite information about the behavior of the centrioles during the spermiogenesis of Fundulus,* but there is no doubt that they are located in that region. 8 One thing however is certain: that their behavior is very different from the same in selachians (Suzuki, 798). 144 J. DUESBERG Again as in many and perhaps all cases, the last stages of spermiogenesis bring about a change in the behavior of the chon- driosomes towards reagents. It is well known that, in the mam- malian testicle for example, the chondriosomes become more and more resistant to acetic acid as spermiogenesis progresses.? The test of this resistance was not made here, but it was found that the chondriosomes of the last stages are structures much less labile than the chondriosomes of the early stages and are conse- quently much easier to bring into evidence. The preceding description of the spermatozoon of Fundulus agrees in the main with Retzius’ observations on the spermato- zoon of other teleosts, though differing in the details. It helps at the same time to emphasize the similarity in structure between these spermatozoa and those of a large number of invertebrates, while the spermatozoa of selachians and of the higher verte- brates are widely different. From the same description it also appears very probable that the male chondriosomes, owing to their close contact with the nucleus, are carried into the egg at the time of fertilization. Though this can be ascertained only by the study of the fertiliz- ing process, the evidence accumulated by an imposing number of observations made upon almost all classes of animals, especi- ally in recent years, is certainly very much in favor of the theory according to which the penetration of the male chondriosomes into the egg is a general phenomenon. Shaffer who mentions only Meves’ observations on Ascaris and Vander Stricht’s on the bat, overlooks the largest part of this evidence. That Lillie (12) found in Nereis that the ‘middle-piece’ and the tail of the spermatozoon do not enter the egg does not prove that the chon- driosomes are not carried into it. I still believe, as in 1915, that the real objection to the admis- sion that the male chondriosomes play a role in heredity is to be found in Meves’ observations on the echinoderm-embryo; why their admitted chemical composition should plead against *I found recently that the same changes take place in the spermatids of opossum. TESTICLE-CELLS OF’ FUNDULUS 145 such a réle, as Cowdry (16, p. 437) seems to believe, I fail entirely to see. Concerning the hypothesis of their motile function which, first formulated by Benda, reappears occasionally in the literature, I do not see that any arguments have been brought forward in its favor, nor is there any clear expression of how we should imagine this function. Benda considered his ‘mitochon- dria’ as contractile bodies: how can this conception be applied to the spherical chondriosomes of the spermatozoa of so many inverte- brates and of Fundulus? Furthermore, those who advocate this hypothesis entirely overlook two groups of observations, which we have to accept as long as their inexactitude has not been demonstrated: first, Meves’ experiments on the spermatozoon of Salamandra and second, the observations of a number of au- thors, lately Koltzoff’s, on the spermatozoa of decapods (see Duesberg, 712, p. 687). Finally a few words concerning the occurrence of a constant number of chondriosomes in male germ-cells. The first indication of this was given by Meves (’00) who found that the small spermatocytes (i.e., as the apyrene genera- tion) of Paludina vivipara contain on the average eight loop- shaped chondriosomes. Numerations made on spermatids of the same generation a short. time after the second division likewise revealed an almost unvarying number of chondriosomes,. this time four. Two other cases, much more striking, have been described lately, both in arachnoids, the first one by Sokolov (13), the other by Wilson (’16). In the spermatogonia and in the young spermatocytes of Eus- corpius carpathicus Sokolov describes mitochondria which soon by confluence form filaments. Later rings appear, which are proba- bly formed by fusion of the free ends of the filaments of the pre- ceding stages. The average number of these rings is twenty- four. During mitosis they are not divided as is the case inthe small spermatocytes of Paludina, but are segregated into two equal groups between the daughter-cells. Thus each spermatid contains one quarter of the number of rings, on the average six. 146 J. DUESBERG The result of this process is an obvious and measureable reduc- tion of the chondriosomal mass at the end of the divisions of maturation and Sokolov sees in it a confirmation of the views I have expressed as the result of my study of the behavior of the chondriosomes in the spermatocyte-divisions of the rat (’07). Wilson has studied the chondriosomes in the spermatogenesis of two other species of scorpions, Opisthacanthus elatus (South- ern California) and Centrurus oxilicauda (Southern Arizona). The results obtained from the study of the first named species are very similar to those of Sokolov. Each spermatocyte con- tains about twenty-four hollow spheroidal bodies, which are seg- regated by the spermatocyte-divisions into four approximately equal groups. Each spermatid thus receives as a rule six chon- driosomes (in 73 per cent of the cases, on 200 numerations), sometimes five (in 16 per cent of the cases) or seven. No other numbers were observed. In the Arizona-scorpion, the process is quite different. All the chondriosomal material becomes con- centrated in a single definite body in the form of a ring. This ring divides during mitosis in such a way that each spermatid re- ceives exactly one-fourth of its substance, “the process taking place with a precision that is comparable to that seen in the dis- tribution of the chromosome material.”’ As Wilson points out the body in question represents a hitherto undescribed type of chondriosome. The occurrence of this inter- esting process makes one speculate as to what the field of sper- matogenesis, though so widely explored, still has in store for the investigator. It appears to me that conditions similar to those found in scorpions, at least to those found in Euscorpius and in Opisthacanthus, could be expected in the histogenesis of these spermatozoa in which, as stated above, the chondriosomes are represented by a constant or approximately constant number of well-defined granules. There is some indication of a similar proc- ess in Fundulus, but the small size of the cells unfortunately makes an exact numeration impossible. The same difficulty would certainly be met with in the study of the seminal cells of other teleosts as well as of echinoderms and celenterates;. mol- luses and worms, however, would probably be a favorable material. TESTICLE-CELLS OF FUNDULUS 147 BIBLIOGRAPHY BonnevIE, Kr. 1907 Untersuchungen tiber Keimzellen. 2. Physiologische Polyspermie bei Bryozoen. ITenaische Zeitschr., Bd. 42. Cowpry, E. V. 1916 The general functional significance of mitochondria. Am. Jour. Anat., vol. 19. Duessere, J. 1907 Der Mitochondrialapparat in den Zellen der Wirbeltiere und Wirbellosen. I. Arch. fiir mikr. Anat., Bd. 71. 1910 Nouvelles recherches sur l’appareil mitochondrial des cellules séminales. Arch. fiir Zellf., Bd. 4. 1912 Plastosomen, “‘Apparato reticolare interno’? und Chromidial- apparat. Ergeb. der Anat. und Entwickl., B. 20. 1915 Recherches cytologiques sur la fécondation des Ascidiens et sur leur développement. Contr. to Embr. Carnegie Inst., 223. 1917 Chondriosomes in the cells of fish-embryos. Am. Jour. Anat., vol. 21. Firip, G. W. 1895 On the morphology and physiology of the echinoderm- spermatozoon. Journ. Morph., vol. 2. FRIEDMANN, Fr. 1898 Beitriige zur Kenntniss der Anatomie und Physiologie der miinnlichen Geschlechtsorgane. Arch. fiir. mikr. Anat., Bd. 52. GaNFINI, C. 1902 Struttura e sviluppo delle cellule interstitiali del testicolo. Arch. ital. di Anat. e di Embr. vol. I. Jorpan, H. E. 1911 The spermatogenesis of the opossum (Didelphys virgini- ana), with special reference to the accessory chromosome and the chondriosomes. Arch. fiir Zellforsch., Bd. 7. Lewis, M. R. 1917 The effect of certain vital stains upon the development of the eggs of Cerebratulus lacteus, Echinorachnius parma and Lophius piscatorius. Anat. Rec., vol. 13. Liu, F. R. 1912 Studies of fertilization in Nereis. Jour. Exp. Zool., vol. 12. Meves, Fr. 1891 Ueber amitotische Kernteilung in den Spermatogonien des Salamanders und Verhalten der Attraktionsphire bei derselben. Anat. Anz., Bd. 6. 1900 Ueber den von la Valette St. George entdeckten Nebenkern (Mitochondrienkérper) der Samenzellen. Arch. fiir mikr. Anat., Bd. 56. 1903 Ueber olygopyrene und apyrene Spermien und ihre Entstehung, nach Beobachtungen an Paludina und Pygeara. Arch. fiir mikr. Anat., Bd. 61. 1912 Verfolgung des sogenannten Mittelstiickes des Echiniden- spermiums im befruchteten Ei bis zum Ende der ersten Furchungsteil- ung. Arch. fiir mikr. Anat., Bd. 80. Payne, F. 1917 A study of the germ-cells of Gryllotalpa borealis and Gryl- lotalpa vulgaris. Jour. Morph., vol. 28. PicteT, C. 1891 Recherches sur la spermatogenése chez quelques inverté- brés de la Méditerranée. Mitt. aus der zool. Station zu Neapel, Bd. 10. 148 J. DUESBERG PRENANT, A. 1911 Préparations relatives aux mitochondries. Comptes-Rendus Assoc. Anat. Paris. Rerzius, G. 1904a Zur Kenntnis der Spermien der Evertebraten. I. Biol. Unters., N.F., Bd. 11. 1904 b Zur Kenntnis der Spermien der Evertebraten. Verhdl. Anat. Gesellsch. Jena. 1905 a Zur Kenntnis der Spermien der Evertebraten. 2. Biol. Un- ters, N.F., Bd. 12. 1905 b Zur Kenntnis der Spermien der Leptokardier, Teleostier und Ganoiden. Ibid. 1906 a Die Spermien der Gastropoden. Biol. Unters, N.F., Bd. 18. 1906 b Die Spermien der Enteropneusten und der Nemertinen. Ibid. 1906 ec Die Spermien der Turbellarien. ‘Ibid. 1906 d Die Spermien der Bryozoen. Ibid. 1909 a Die Spermien von Aurelia aurita (L). Biol. Unters., N.F., Bd. 14. 1909 b Die Spermien der Nereiden. Ibid. 1909 ec Zur Kenntnis der Spermien der Elasmobranchier. Ibid. 1910 a Zur Kenntnis der Spermien der Echinodermen. Biol. Unters., N.F., Bd. 15. 1910 b Weitere Beitrige zur Kenntnis der Spermien mit besonderer Beriicksichtigung der Kernsubstanz. Ibid. ScHREINER, A. AND K. E. 1905 Ueber die Entwicklung der ménnlichen Ge- schlechtszellen von Myxine glutinosa. Archives de Biologie., vol. 21. 1908 Zur Spermienbildung der Myxinoiden. Arch. fiir Zellf., Bd. I. Suarrer, HE. L. 1917 Mitochondria and other cytoplasmic structures in the spermatogenesis of Passalus cornutus. Biol. Bull., vol. 32. Soxotov, I. 1913 Untersuchungen iiber die Spermatogenese bei den Arachni- den. I. Uber die Spermatogenese der Skorpione. Arch. fiir Zellf., Bd. 9. Suzuxi, B. 1898 Notiz iiber die Entstehung des Mittelstiickes der Samenfiden von Selachiern. Anat. Anzeiger, Bd. 15. Witson, E. B. 1916 The distribution of the chondriosomes to the spermatozoa. in scorpion. Science, N.S., 43 and Proceedings of Nat. Acad. of Sciences. PLATES 149 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 EXPLANATION OF FIGURES All figures were outlined with a Zeiss camera-lucida, at the level of the stage of the microscope. Lens used: Zeiss apochr. imm. 1 m.m., 5; oeular 12. Arti- ficial light (gas). PLATE 1 EXPLANATION OF FIGURES 1 Fundulus majalis. Fixation: Benda, without acetic acid. Stain: Benda. Epithelium of an excretory duct. 2 Fundulus heteroclitus. Fixation: Benda. Stain: Benda. Supposed in- terstitial cells. 3 Fundulus majalis. Fixation: Benda, without acetic acid. Stain: Benda. Group of primary and secondary spermatogonia. 150 PLATE 1 J. DUESBERG TESTICLE-CELLS OF FUNDULUS ~ ae / PLATE 2 EXPLANATION OF FIGURES 4 and 5 Fundulus heteroclitus. Fixation: Benda. Stain: Benda. Meta- phase and anaphase of the mitotic division of primary spermatogonia. 6 Same material. Tertiary spermatogonium. 7 Same material. Tertiary spermatogonium: metaphase. 8 Fundulus majalis. Fixation: Benda, without acetic acid. Stain: Benda. First spermatocyte. 9 and 10 Same material. Metaphase and anaphase of first division of maturation. 11 Fundulus heteroclitus. Fixation: Regaud. Stain: acid fuchsin-methyl- green. Anaphase of first division of maturation. 12 Same material. Anaphase of second division of maturation. 13 to 18 Same material. Six stages of spermiogenesis; in none is the tail represented in its full length. 16 and 17 are respectively face-view and side-view of approximatively the same stage. 19 Fundulus heteroclitus. Fixation: Benda. Stain: Benda. Group of spermatids in a cyst. 20 and 21. Fundulus majalis. Fixation: Benda, without acetic acid. Stain: Benda. Spermatozoa from the excretory ducts (the tail is not represented in its full length). 20: side-view; 21: face-view. PLATE 2 12 10 TESTICLE-CELLS OF FUNDULUS J. DUESBERG ive) 5 14 his & 153 (Sh: 18 17 AUTHOR’S ABSTRACT OF THIS PAPER {SSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 1 THE POSITION OF THE INSERTION OF THE PECTO- RALIS MAJOR AND DELTOID MUSCLES ON THE HUMERUS OF MAN ADOLF H. SCHULTZ Carnegie Institution of Washington THREE FIGURES The metrical determination of the position of the attachments of muscles to bones is a problem which affords a contribution to topographical anatomy. It is also of importance from the standpoint of musculo-mechanics, because measuring the inser- tions of muscles is analogous to the determination of the lengths of leverage of the body. Since such an investigation promises to give information regarding differences or equalities of race or sex, as well as of the two halves of the body, it is of no less in- terest to the anthropologist. As far as the author can determine from a study of anatomical and anthropological literature, no measurements of the insertions of muscles have as yet been undertaken. In approaching this problem one is working in a new field of osteometry, where it becomes necessary to treat the bone not separately but in conjunction with the associated mus- cles, which bave been so neglected in anthropometry. The following study deals with the insertion of the pectoralis major and deltoid muscles. The measurements were made on the right and left arms of one hundred and five bodies. Forty- six of these bodies were obtained from the University of Mary- land in Baltimore, forty from the Jefferson Medical College in Philadelphia and nineteen from the Johns Hopkins Medical School in Baltimore. The author wishes here to express his ap- preciation to Drs. W. H. Lewis, J. P. Schaeffer, J. Holmes Smith and J. W. Holland for their kindness in permitting the use of this material. All of the subjects measured were adults; juvenile and senile ones were excluded. It is regrettable that the sexes 155 156 ADOLF H. SCHULTZ were very unequally represented, for the females numbered only twenty-seven, as against seventy-eight male subjects. A greater uniformity occurred in race, as there were fifty-one white and fifty-four colored bodies. The author wishes to call attention to the fact that the term race is used in its widest sense in the pres- ent paper, because both the white and colored inbabitants of America have originated from numerous races in a limited sense. In negroes one frequently witnesses a more or less extensive admixture of white blood; in cases where there was evidence of a too great intermingling with the white element the material was discarded. The position of the muscle insertion was compared with the length of the humerus by measuring the distance of the most proximal and the most distal point of attachment from the proximal end of the bone, and further by determining the arith- metical mean of these distances in percentage of the length of the humerus. For this purpose we need first of all six exact points of measurement, a proximal and a distal point on the humerus, two corresponding points on the pectoralis major and two more on the deltoid. The two points on the humerus are found by measuring the length of the bone, choosing the distance of the highest point of the caput humeri from the lowest point of the capitulum and measuring parallel to the axis of the bone (fig. 1, points I and II). The points of measurement for the pectoralis major muscle are the most proximal and the most dis- tal points of its insertion on the crista tuberculi majoris (fig. 1, points III and IV); as a rule they are readily determined. Occa- sionally the distal portion of the insertion is intimately connected with the tendon of the deltoid muscle and the distal point can only be obtained after careful separation of these structures. In a limited number of cases the dorsal reflected portion of the muscle was observed to form a narrow tendinous band in the region where it spreads out proximally to join the tendinous lin- ing of the sulcus intertubercularis (in figure 1 such an instance is indicated at a). In such cases this prolongation was ignored and the point of measurement taken at its distal end. The lower point of measurement of the deltoid is comparatively easy to PECTORALIS MAJOR AND DELTOID INSERTION 157 SS \ MY SE SS Ss xx M PH@ie MAG: Matias ae ESS tiple ele wes SSS, 4% O Le | . SV Fig. 1 Diagram of the points of measurement and distances on a right humerus seen from in front. 158 ADOLF H. SCHULTZ ascertain, namely as the most distal point of the insertion on the tuberositas deltoidea (fig. 1, point VI). The most proximal point of insertion is frequently concealed by the body of the mus- cle and it is necessary therefore to remove it partially. In doing this great care should be exercised as the deltoid is usually at- tached at its uppermost end by very delicate strands (fig. 1, point V). The distance between each of these four points and the highest point of the head of the humerus was measured parallel to the axis of the bone, similar to the Jongitudinal meas- urement of the humerous mentioned above and therefore these measurements are projections. The measuring instrument employed was a modified small anthropometer of Martin (Stangenzirkel). This instrument is composed of a ruled metal bar or beam, possessing two arms at right angles to it, one of which is firmly attached to the end, the other movable in the direction of the bar, while both are movable at right angles to the latter. The modification consists merely in the addition of a third arm from another instrument of the same kind, which ean also be moved both in the same direction and at right angles to the main axis (fig. 2). First one measures the length of the humerus with the two outermost arms of the instrument holding the bar parallel to the axis of the bone, then the middle arm is approximated in turn to the four points of muscle insertion as defined above. This is performed by moving the arm up and down as required, shorten- ing or lengthening it, simultaneously rotating the entire instru- ment around the axis of the humerus if necessary. Readings are taken each time on the ruled bar and correspond with the meas- urements two, three, four and five in figure 1. Indices for the relative position of the middle point of each muscle insertion were obtained by the following formulae: measurement 2-+ measurement 3 Om) aale. —_——_—— > 100 for the pectoralis major measurement 1 measurement 4+ measurement 5 ») — —— —— 100 for the deltoid measurement 1 PECTORALIS MAJOR AND DELTOID INSERTION 159 The greater these indices of position, the more distal, the smaller, the more proximal is the insertion of the muscle. Fol- lowing is a short description of the mathematical treatment of the length of the humerus and the indices which have been used in this paper. A more detailed explanation of these methods, which are absolutely necessary for an understanding of the Fig. 2. Small anthropometer with three parallel movable arms. measurements on a large number of individuals, is to be found in the Textbook of Anthropology by R. Martin, Jena, 1914, pp. 63-103. The average (V/) is the arithmetical mean of the individual values (V) of a group (n = number of individuals): M = —dV. ; n The standard deviation (c) is the square root of the average of 160 ADOLF H. SCHULTZ the squares of the deviations of the individual values from the average of the row and expresses the absolute variability: c= yp2 >(V—M)*. The variation coefficient (v) expresses the standard deviation in percentage of the average, whereby a 10s The M correlation coefficient (7) affords a means of determining the law, according to which two characteristics combine. It is the sum of the products of the deviations of the two characteristics from the corresponding averages taken for each individual, divided by the product of the number of individuals and the two standard deviations: r =~ ro UN A complete correlation No, Cy exists when r = 1. If r = 0, no relation prevails between the two characteristics. A positive correlation coefficient indicates a change of the characteristics in the same direction, a negative one, in the opposite direction. Finally, to test the degree of exactness of the above formulae, the probable error (EK) was de- termined by the following formulae: criterion for the relative variability is obtained: v = E(M) = + 0.6745 —— for the average. vs E (¢) = = 0.6745 —— = for the standard deviation. an Jk (@) === 0. 6745 for the variation coefficient. 2n He eee If » > 10, the last formula must be multiplied by Fi Ce g y N+? \ ipo 7) IOGp) = s= 0.6745 | for the correlation coefficient rn The relation of the insertion of the muscles to the length of the humerus makes a short preliminary discussion of this absolute measurement necessary. Table 1 is a compilation of the aver- ages and the conditions of variability of the length of the two hundred and ten humeri, which were measured. The extremes of these measurements range from 260 to 367mm. The humerus in mate whites is on the average 26mm., in male negroes 31.8 mm. PECTORALIS MAJOR AND DELTOID INSERTION 161 TABLE 1 Averages, standard deviations, variation coefficients, their probable errors and ranges of variation for the length of the humerus RACE SEX pes SIDE M + E (M) o + E(c) »v + E (v) Mins Mee (| o 40 Ge 316.3+1.71) 15.94+1.22) 5.04+£0.38 | 283 | 347 a 40 Ie 316.5+1.82) 17.05+1.29| 5.40+0.41 | 283 | 352 Whites rou 80 | r. 1. | 316.4+1.25) 16.50+0.87) 5.22+0.27 | 283 352 oe Q 11 re 293 .1+=2.58) 12.65+1.81| 4.32+0.62 | 269 | 309 g 11 l. 287 .7+2.02| 9.92+1.42) 3.44+0.50 | 267 | 303 Q 22 | r.l. | 290.4+1.67| 11.69=1.19| 4.03+0.41 | 267 | 309 rou 38 Yr 326.2+1.87| 17.18+1.34| 5.27+0.41 | 290 | 367 ron 38 Ie 323 .5+2.02) 18.50+1.44) 5.71+0.44 | 283 | 365 Negroes... of 76 | r. 1. | 324.8+1.39) 17.89+=0.98) 5.50+0.30 | 283 | 367 : g 16 i 294.6+2.45) 14.51+1.74) 4.92+0.59 | 266 | 321 i) 16 l. 291.5+2.51| 14.84+1.78) 5.10+0.61 | 260 | 312 Q SO hte eee 2950 == leedl lao le woe Os 0242 ae200n noo longer than in females. The averages in negroes exceed in both sexes the corresponding values for whites. The division of table 1 into separate rows for the right and left humerus shows that the variability is greater on the left side except in the group of white females of which the number measured was quite small. Furthermore it shows that the white males, who possess the same average length of the humerus on both sides, form an exception to the rule of the greater length of the humerus on the right side. Table 2, which gives a survey of the absolute and TABLE 2 Absolute and relative numbers of individuals with equal and different lengths of the humert and average differences of the individual asymmetries (mm.) AVERAGE DIFFERENCE IF RACE SEX |BOTH SIDES EQUAL ee ae LEFT SIDE LONGER Right Left side side longer | longer , | 10=25.0% | 17=42.5% | 13=32.5% | 3.88 | 5.62 1 aaa 9 | 0= 0.0% | 9=81.8% | 2=18.2% | 7.67 | 5.00 ‘ {| @ | 8=21.0% | 22=58.0% 8=21.0% | 5.73 | 3.25 NSEIED < eip0 5 \| 9 | 7=43.8% 9=56.2% O= 0.0% | 5.44] 0 162 ADOLF H. SCHULTZ relative number of cases possessing humeri of equal and differ- ent lengths and the average differences of the individual asym- metries, shows that 32.5 per cent of white males have a longer left humerus. It also demonstrates that in white males the differences in favor of the left side are on the average greater than those on the right, which is not the case in the other groups. The greatest absolute asymmetry occurred in a negro whose right humerus exceeded the left in length by 23 mm. TABLE 3 Averages, standard deviations, variation-coefficients, their probable errors and ranges of variation for the position index of the insertion of the pectoralis major muscle RACE SEX pas SIDE M + E (M) a+ E (oc) v = E (v) pe ae | (} | 40} rv. | 28.87+0.17] 1.55+012 | 5.46+0.41| 24.3] 31.3 |} 7 | 40] 1. |°28.50+0.16] 1.49+0.11 | 5.23+0.40} 25.3] 31.8 Whites J] @ | 80] r. 1. | 28.43+0.11) 1.52+0.08 | 5.35+0.28) 24.3] 31.8 VP 9 | id} x. | 26.37+0.35) 1.74025 | 6.59+0.94| 21.9) 28.3 | @ | 11} 1. | 26.35+0.56) 2.73+0.39 | 10.34+1.48) 21.3) 2925 Q | 22} r. 1. | 26.36+0.33) 2.29+0.23 | 8.67+0.88] 21.3] 29.5 (| 0 33. | r. | 28 27+0.19| 1.75+0.14 | 6.18+0.48] 26.1) 35.2 || 38] 1. | 27.99+0.17) 1.57+0.12 | 5.61+0.44| 25.1) 32.6 Nene cts |se7eh one | 28.13+0.13 1.67+0.09 | 5.94+0.33) 25.1] 35.2 2 Ns Obs 16) r. -/26.820:32 1:89-0:23 |” 7050.85) 23 aliea0eo [| 2 | 16) 1. | 26.45+0.31| 1.82+0.22 | 6 89+0.83! 22.3) 30.3 [| @ | 382] r. 1 | 26.63+0.22) 1.86+0.16 | 6.99+0.59) 22.3} 30.9 The averages and the conditions of variability of the index of position for the middle of the insertion of the pectoralis major muscle are given in table 3. This index differs in the entire material between 21.3 and 35.2. Expressing this in terms of the mechanics of levers, one can state that in the adult the lifting arm of the musculus pectoralis major is related to the carrying arm—the length of the humerus—in a ratio varying from 21.3: 100 to 35.2: 100. In other words the relation of the lever arms may differ by almost 14 per cent of the length of the carrying arm, and this expressed in an absolute number equals on the average about 45 mm. “Am vierten und fiinften Tage der Bebriitung kommen auf jeder Seite in der Substanz des Halses drei aufeinander folgende fast linsenformige Hoéhlen zum Vorschein, deren jede nach aussen und innen ge6ffnet ist. Die Aussere Miindung der vordersten Hoéhle wird iibrigens von einem Theile, der Shnlich dem Kiemendeckel der Fische ist, verdeckt.?’ (’25) In a much later publication (’61) the same author carries the analogy still farther. ‘““Von dem zweiten Schlundbogen, in welchem sich ein Zungenbeinhorn ausbilden soll, wichst bald darauf, nachdem sich die vordeste Schlundspalte geschlossen hat, ein klappenartiger Fortsatz hervor, der die zweite Schlundspalte bedeckt und als eine Andeutung der Membrana branchiostega der Griitenfische betrachtet werden darf.’’ * Parker on Apteryx: ‘““The backward extension of the hyoidean fold visible in the previous stages has increased so as to form a true operculum, which com- pletely covers the third cleft, so that it is invisible in an external view. The fourth cleft . . . . lies immediately behind the operculum, and is very probably only exposed by the shrinking of the latter. . . . . The reten- tion of so obviously amphibian a character . . . . appears to be a charac- ter of very considerable morphological interest.” GILL-FILAMENTS IN SAUROPSIDA 209 plate 2 (see figs. 12 to 17 inclusive) show that the mound is an evagination of the thickened, vesiculated ectoderm covering the third arch, and contains a mesodermal core, thus almost repro- ducing the early formation of external gills in the amphibia. The lower end of it has already begun to develop filaments and later the upper end will give rise to tufts of cells (fig. 18). In the same embryo the ectoderm of the fourth arch forms an evagina- tion, which however, is solid, and in this specimen much smaller, tending to fuse with that from the third arch (fig. 15). Owing to the rapidity with which the region behind the third arch is being flattened out, the evagination on the fourth arch, which has been observed in several embryos, has only a transitory existence of its own. As the operculum extends backward all of the third arch except the mound becomes covered over, while the mound itself gradually assumes the shape of a wedge, with filaments at its downward directed point, as indicated in figures 6 and 19. As the hyoid arch continues its backward growth during the fifth day it fuses with the third arch in such a way as to carry with it on its under surface the tufted epithelium at the lower end of the wedge, so that from now on, the filaments of this region of fusion will appear to come from the under surface of the opercu- lum (figs. 19 and 23). By the beginning of the sixth day the whole edge on each side has become differentiated into a ridge coextensive with the lateral margins of each operculum (fig. 7), Serial sections (fig. 24) show that the individual filaments borne by the ridge are solid outgrowths of the epithelium, honey- combed with degeneration vesicles. With the appearance of this pair of ridges the first half of the life-history of the filaments may be said to have been completed. While this differentiation of filaments has been going on at the margins of the hyoid arches, the ventro-medially directed por- tions of the two opercular processes have united to form a single band of tissue slightly overlapping the pectoral body-wall and extending across the ventral surface of the neck from side to side,—the homologue of the membrana branchiostega according to Rathke. From now on, the fused hyoid arches may there- fore be referred to as a single structure, the plica opercularis, pos- 210 EDWARD A. BOYDEN sessing two lateral margins, each fused to the side of the neck and bordered by a line of filaments, and a single pectoral margin whose free edge is directed posteriorly. A notch at its middle point (incisura opercularis) indicates the place where the two arches originally united, and a tubercle on each side of the notch still further accentuates the paired origin (fig. 1). In later stages the free, overhanging, pectoral margin becomes more and more encroached upon by the lateral margins, as its under surface pro- gressively fuses with the surface of the neck, from the sides to the mid-line. As these lateral zones of fusion pass slowly inward they carry the lines of filaments with them so that these likewise come to lie successively nearer the mid-line. Meanwhile the pectoral wall below the opercular fold has developed a pair of surface markings of its own which begin to shift their position from the sides to the mid-line at the same time and rate as the lines of filaments and zones of fusion above. Eventually these markings come to form part of a median ridge extending from the region below the operculum to the umbilicus. In addition to their migration these pectoral markings or ridges exhibit a fur- ther, albeit superficial, resemblance to the filaments in that the cell-proliferations of which they are made often contain scattered degeneration vesicles and pyenotic nuclei, but to a much lesser extent than obtains in the branchial epithelium. Because the pectoral ridges and lines of filaments are thus found to possess these common features it becomes necessary to establish the identity or the disparity of the two sets of structures, the one on the neck and the other on the pectoral wall,—hence the following digression. CHANGES IN THE PECTORAL WALL CORRELATED WITH THE LATER DEVELOPMENT.OF FILAMENTS AND OPERCULUM During the latter part of the sixth day and the beginning of the seventh, four different sets of structures make their appear- ance in the pectoral wall, all of which are represented in figure 1: a pair of pectoral grooves (sulci peciorales) ; 2) a pair of pectoral ridges (cristae pectorales); 3) a pair of mesothelial ridges (cris- tae mesotheliales) and 4) a median epitrichial ridge (crista eprtrichialis). GILL-FILAMENTS IN SAUROPSIDA Pit The pectoral grooves first appear about the middle of the sixth day and at once delimit a roughly triangular area whose base coincides with the intersection of the neck and breast, and whose downward directed apex lies just above the umbilicus. At first the area enclosed by these grooves is transparent throughout, revealing the outlines of the heart beneath. But almost immedi- ately the basal third becomes vascular and much thicker than Fig. 1 Sketch of the markings on the pectoral wall of a seven-day chick (H. KE. C., Ser. 2076; 17.3 mm.; 6 days, 7 hours) together with three transverse sec- tions of the pectoral wall of the same embryo, X 42 diam. 407, section at level of c.p.; 550, section at level of s.p.; 651, section at level of c.e., m.l., and m.p., lateral and pectoral margins of opercular fold; 7.0., opercular notch; f.0., opercu- lar tubercle; f.br., row of branchial filaments; a.0., opaque area; c.p., pectoral ridge; s.p., pectoral proove; c.e., epitrichial ridge; c.m., mesothelial ridge; w. | umbilicus. Compare with frontal section of pectoral wall in figure 20. the apical portion which retains for some time its non-vascular and transparent character. In fresh specimens the upper portion exhibits a semi-opacity somewhat similar to that of ground glass. It is this part which at the beginning of the seventh day gives rise to a series of superficial evaginations which may appear anywhere in this area, but are chiefly ranged along the medial borders of the pectoral grooves. Ultimately those of a side be- come numerous enough to form a pair of pectoral ridges which in their later development, as previously noted, exhibit a super- Bie EDWARD A. BOYDEN ficial resemblance to the filaments when seen in section. About the same time a pair of sub-surface lines may be seen through the translucent wall underlying each groove. A study of serial sections (H. E. C., Ser. 2076; 6 days, 7 hours; 17.3 mm.; fig. 1) proves them to be mesothelial ridges projecting into the peri- cardial cavity. Hardly have these lateral grooves and ridges appeared than they begin to shift their position from the sides of the embryo to the mid-ventral line. This takes place in such a way that the legs of the triangle first approach each other in front of the um- bilicus and thereafter successively forward of that point, thus resulting in an apparent ascent of the apex of the triangle. This progressive movement is recorded on the median line by an eruption of epitrichial cells which follows the retreating apex up the pectoral wall until it reaches the surface ridges described above. Thus a Y-shaped ridge is produced on the ventral sur- face of the embryo the upper arms of which, the two surface ridges, form a broad angle, and the lower arm of which, the epi- trichial ridge, extends to the umbilicus. Micrometer measure- ments of selected embryos indicate the rate at which the two upper arms are coming together. In an embryo of five days, twenty hours (17.5 mm.) the distance between the upper ends of the ridges is 1.64 mm.; while in an older stage (6 days, 3 hours; 17.3 mm.) it has been reduced to 0.91 mm.; and in a still older embryo (6 days, 18 hours; 19.5 mm.) to 0.31 mm. The shifting of these superficial ridges also keeps pace with the shifting of the opercu- lum and filaments to be described later; so that if the ridges were continued upward at any given stage they would strike the tufts of filaments above. In all cases, however, the two structures are separated by an appreciable area of the neck. Eventually the surface ridges become heaped up in the median line thus con- stituting, with the epitrichial proliferations, a continuous median ridge from the umbilicus to the neck. In its upper end the evi- dence of its paired origin is visible for some time and as a whole the ridge persists for a number of days even to the time when it becomes elevated upon the developing feather papillae (H. E. C., Ser. 1967; 11 days, 0 hours; 31.0 mm.). A similarly placed GILL-FILAMENTS IN SAUROPSIDA 213 median ridge has been observed in a human embryo of 45.0 mm. (H. E. C., Ser. 2079) and in a dog embryo of 14.0 mm. (H. E. OFF Ser. 2052), but I have been unable to discover any clue to its origin in these animals. It is possible that in mammals, only the last stages of the process are visible. An examination of the inner surface of the pectoral wall shows that an approximation of the two mesothelial ridges is also taking place (fig. 1). In sectioned embryos each ridge was found to contain one or two small veins although no blood vessels could be detected in the area between the two ridges in their lower ex- tent. This suggested a study of injected embryos which has strikingly substantiated the shifting of tissues in the pectoral wall. The displacement of veins in the pectoral wall. In a five-day chick that portion of the body wall which covers the heart is .entirely free from blood-vessels. On the margins of this roughly triangular area lies a capillary network, continuous with that which fills the rest of the membrana reuniens (Rathke’s term for the thin somatic wall which originally covers the abdominal and thoracic viscera). This appears to be growing into the wall over the heart much as capillary-nets elsewhere invade non-vascular regions. By the middle of the sixth day (injected embryo 5 days, 7 hours; 12.4 mm.) this network of the membrana reuniens has resolved itself into a series of radial veins converging upon the umbilicus from the myotomes. Those immediately adjacent to the non-vascular triangle (that is, under the pectoral grooves) converge upon the umbilical vein of their respective sides at the point where it enters the septum transversum, on either side of the apex of the heart. From now on, the non-vascular area over the heart will become more and more circumscribed, not by the ingrowth of new vessels (except in the uppermost part which is congruous with the opaque area previously described, where a capillary net grows down from the cervical region) but by a shifting of marginal veins already formed. These swing in on two pivotal points, the points referred to above, where the umbilical vein of each side enters the septum transversum. These radial venules are not straight lines but present a convex THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 214 EDWARD A. BOYDEN surface to the non-vascular triangle, so that as they swing in they first meet in the median line under the point where the epitrichial ridge first appears, and thereafter progressively an- terior to this point. In an embryo of 15.5 mm. (fig. 2) they are just reaching the mid-line for the first time. In an embryo of 18.5 mm. (injected specimen 6 days, 0 hours), they have moved Fig. 2. Blood vessels in the pectoral, body-wall of a 15.5 mm. chick. Camera lucida drawing of injected embryo ji (6 days, 0 hours) X 15 diam. Note the non- vascular area in the center and the capillary plexus above it which has grown down from the cervical region. p.o., opercular fold; v.u.s., left umbilical vein. Fig.3 Blood vessels in the pectoral wall of a 20.6 mm. chick. Camera lucida drawing of injected embryo ka (7 days, 2 hours) X 15 diam. Note the disappear- ance of the non-vascular area and the direction of the blood vessels as compared with those in figure 2. w., umbilicus; z., median line, on either side of which are the right and left sets of marginal veins. in upon the mid-ventral line as far up as the point where the epi- trichial ridge meets the surface ridges, (the region of the truncus aorticus). Finally in an embryo of 20.6 mm. (fig. 3), the area over the heart is completely filled with parallel longitudinal veins extending from the neck to the umbilicus. Thus in forty-eight hours the original marginal veins of each side have described an arc of 45°. GILL-FILAMENTS IN SAUROPSIDA Zila As far as one can judge from the meagre description of the region in mammals, it seems probable that the pectoral wall is vascularized by a different method than that which obtains in the chick. Evan’s figure of a fifteen millimeter pig embryo sug- gests that the pectoral wall is invaded by the same capillary net which grows down from the cervical region in the chick, but in this animal continues until it reaches the umbilicus. Kolliker’s figure of a cow embryo of about the same stage suggests that this area is supplied by the capillary net which grows in from the marginal veins. Mall describes the condition in man as follows: I have in my collection a well-preserved human embryo (no. LX XVI) (22 days), in which the membrana reuniens is filled with a plexus of veins much like that in the cow’s embryo . . . . The ventral wall of the heart near the liver contains no vessels, while the membrana reuniens covering the upper end of the heart is filled with a plexus of vessels which communicate with the capillaries of the mandibular arch. The study of the blood vessels thus confirms the testimony of the superficial markings, that there is a considerable shifting of body-wall tissues from the sides of the embryo toward the mid- ventral line,—this in advance of the later invasion of pectoral muscles, nerves, dermal and skeletal parts which enter into the formation of the definitive pectoral wall. Whether there is any- thing of this preliminary movement in other amniotes is difficult to determine from the evidence at hand. But certainly in chicks of the sixth and seventh days it is possible to demonstrate a medial displacement of tributaries of the umbilical veins and a shifting of an internal and external set of ridges. Although I have been unable to find in the literature any record of such ridges or grooves as have just been described, there is some evidence that the older embryologists who discussed this region in birds and reptiles saw something of this early shifting of tissues. Thus Rathke, the first to maintain that the muscle and skeletal ele- ments of the breast wall did not arise in situ, believed that these structures grew in from the sides of the embryo in such a way as to push ahead of them the thin somatic wall which originally covered the thoracic and abdominal viscera, eventually causing the membrana reuniens to disappear completely (allmiilig ver- 216 EDWARD A. BOYDEN schwinden). Remak, following Reichert, believed that the new — elements did not displace the original wall but were merely in- corporated into it as they grew in. (‘‘ Die sekundire Bauchwand entstebt aus der Verschmelzung der unspriinglichen Bauchwand mit der hervorwachsenden Entwickelungsproducten der Urwir- bel.’’). Although my own observations have resulted in the find- ing of new manifestations of a migration of tissues in the body- wall it is still very difficult to analyze the nature of the move- ments. It may be that at a certain stage in the development of the wall that part of the membrana reuniens which lies over the breast is being pushed in toward the median line by the faster growing lateral parts and reduced to a narrow zone,—the slack, so to speak being taken up by the increasing thickness of the membrana and by the protrusion of certain external and internal ridges. Or it may be that this early shifting of structures is more in the nature of a preliminary growth-wave which is passing from vascular to non-vascular territory. The difficulty of ana'yzing these changes, however, does not derogate from the conclusions which we may now form concern- ing the relation of the pectoral ridges and the branchial filaments, —to ascertain which this study of the pectoral region was first undertaken. As has been stated, both sets of structures are evaginations of surface epithelium, both move in toward the median line at certain stages in their development and both exhibit some degree of epithelial degeneration. But beyond these superficial resemblances there is nothing to indicate the existence of any genetic relation between them. At no period in their development are they in continuity nor do they even re- semble each other in histological appearance, unless it be at the end of the seventh day, when the filaments have been crowded in upon the tubercles just prior to their disappearance. The differences which they exhibit may be summarized in the follow- ing paragraph. The epithelial evaginations which constitute the ridges de- velop from a broad zone of thickened ectoderm situated on either side of the pectoral wall. The margins of each zone grade imper- ceptibly into the adjacent ectoderm so that its limits are never ‘GILL-FILAMENTS IN SAUROPSIDA ii sharply defined. Apparently the limits vary considerably in dif- ferent embryos. This variability may be said to characterize the whole development of the ridges,—a marked contrast to the regularity with which the filaments develop. As each zone shifts its position medially, scattered pyenotic nuclei appear between the epitrichial and basal layers of the epithelium and diffuse tufts of cells arise on its surface. The crests of these evagina- tions form a low ridge which becomes the more conspicuous the nearer the ridge approaches the mid-line. The filaments, on the other hand, grow out from a narrow strip of epithelium situated on the lateral margin of each operculum, a germinative zone which represents the fusion of the posterior wall of the hyoid arch with the ectoderm of the third and fourth gill arches. The fila- ments thus arise from a specific epithelium,—a portion of the branchial membrane, the whole of which is characterized at an early stage by the presence of degeneration vesicles. The fila- ments grow out of the region where the cysts occur in greatest numbers and are themselves honeycombed with vesicles. Fur- thermore, they do not differentiate into diffuse clusters of cells but into a row of more or less distinct evaginations. Again, the ‘life cycle’ of the filaments is staged from two td three days earlier than that of the ridges. The branchial evaginations first arise in situ and only later become involved in the movements of the ventral body wall, whereas the pectoral ridges are in process of migration when they first appear. Thus one is lead to con- sider whether the ridges are not more intimately connected with the movements of the body wall than are the filaments, if indeed they are not products of that movement. For as the ridges ap- proximate each other they become heaped up in the midline to form a single structure which persists three days after the fila- ments have disappeared, whereas the latter never meet in the mid-line but maintain their identity as paired rows of individual filaments to the end. It may be that the pectoral ridges repre- sent the survival of some similarly placed outgrowths in a lower vertebrate, but in no other animals, so far as I am aware, do any such structures exist. For the present, then, it seems more reasonable to define the ridges as local manifestations of migra- 218 EDWARD A. BOYDEN tion in the pectoral wall. But whatevcr interpretation they may receive it is evident that they belong in a different category from the filaments. LATER DEVELOPMENT OF FILAMENTS AND OPERCULUM The description of the early development of these structures has been carried to the point where the opercula of the two sides have united to form a swollen band of tissue across the ventral surface of the neck (the plica opercularis) the posterior edge of which (the margo pectoralis) slightly overlaps the pectoral wall. On the lateral margin of the fold (the margo lateralis), a line of filaments has been formed, which is separated from the one on the other side by the whole width of the neck. This is the condi- tion at the middle of the sixth day when the opercular fold has reached its maximum length of two and a half to three milli- meters. From now on, the neck will increase in diameter as the fold undergoes reduction. This process consists in the fusion of the under surface of the margo pectoralis with the ventral surface of the neck, so that its form is changed from an overhanging fold of tissue toa mound, which in turn flattens out and eventually disappears. The striking feature of the whole process is that it proceeds from the sides to the midline, at the exact rate and at the same time that the pectoral ridges and marginal veins are moving across the face of the pectoral wall. The rate of fusion is easily gauged by measuring the decreasing distances between the medial ends of the two rows of filaments as they are borne along on the advancing wave. In all cases they move synchro- nously with the structures below. Thusin an embryo of 5 days and 20 hours (17.5 mm.) the unfused or overhanging portion of the plica, measured by the distance between the filaments of the two sides, is 1.45 mm. In the next seven hours it has been re- duced in length to 0.55 mm. (embryo of 6 days, 3 hours; 17.3 mm.), and in the next fifteen hours to 0.23 mm. (embryo of 6 days, 18 hours; 19.5 mm.). The entire opercular fold including both overhanging and fused portions of the two sides measure as before some two and a half to three millimeters although the fused portion is in process of sinking into the neck and disap- GILL-FILAMENTS IN SAUROPSIDA 219 pearing. By this time the projected width of the neck at this. level measures some three and a half millimeters. In round numbers, during the twenty-four hours following the maximum development of the operculum the unfused portion has been reduced by eighty per cent of its former length while the neck has added twenty per cent to its circumference. By the begin- ning of the eighth day the united opercula have become reduced to a pair of tubercles on either side of the mid-ventral line which are themselves on the point of being incorporated in the neck. While the opercular fold has been undergoing a decline, the filame ts have reached their maximum development and have likewise entered a period of decline which is completed with the disappearance of the tubercles. In the beginning it was stated that the filaments were solid outgrowths of cel’s arising chiefly from the ectoderm covering the third branchial arch; that these filaments first appeared at the lower part of the evagination of that arch; then peripheral to this point as the mound assumed a. wedge-shape and the wedge became compressed into a filament- bearing ridge, about half a millimeter in length. Such is the condition at the middle of the sixth day, at a time when the plica. opercularis is coextensive with the width of the neck, and when the pectoral grooves and other evidences of the median migration first appear in the wall below (fig. 7). As the zone of fusion between the under side of the opercular fold and the neck moves inward from its original position at the junction of the lateral and pectoral margins of each side, the row of filaments is carried with it. Concurrently each row in- creases its length until it reaches a maximum extent of nearly a millimeter, and numbers some eight to a dozen separate filaments. These are often irregularly arranged and are sometimes grouped into two parallel rows. Starting with the medial.end the small ones with which the line begins pass abruptly into large-size filaments which continue from a third to half way across the line. Lateral to this medial portion there are usually gaps in the line and the different members vary in height, tending how- ever to become somewhat smaller as the lateral end is approached. At exactly what point new filaments are added to give the line 220 EDWARD A. BOYDEN its Maximum extent or how the movement of the line as a whole across the neck is accomplished, is difficult to determine. One is inclined to believe that the medial migration is an apparent rather than a real movement, brought about by the addition of new members to the medial end of the row, this end representing an advancing growth-zone superimposed upon the advancing zone of fusion between the operculum and the neck. In favor of this hypothesis is the fact that the medial half of the row ex- hibits the greatest solidarity, that the large medial filaments are the ones that are usually branched (fig. 20), and that the lateral members are the ones which drop out as the total length of the line diminishes. There is the other possibility, however, that the line as it stands is carried bodily inward, new filaments being added laterally (the order of formation in the earliest stages), or possibly interspersed among the old ones as the line is drawn out. Following the period of maximum development during the sev- enth day, the lateral members of the row gradually flatten out until, as the line approaches the center, only a few of the medial filaments on each side remain (fig. 9). By the beginning of the eighth day both the opercular tubercles and filaments have been absorbed into the neck. During this shifting of the rows of filaments from the sides to the mid-line the under surface of the pectoral margin of the opercular fold between the right and left zones of fusion has given rise to a new line of filaments,—abortive structures which are so small that they cannot be made out with the naked eye. In fresh specimens, however, the margo pectoralis has that pearly lustre characteristic of the marginal filaments. These abortive filaments can just be made out with certainty in sections of seventh-day embryos (H. E. C., Ser. 1950; 6 days, 2 hours; 16.0 mm.) (H..E. C., Ser. 2075; 6 days, 1 hour} 17.0 mm.),, where they appear as low sprouts of cells on the under surface of the opercular fold. The largest of these are to be found on the tu- bercles which lie on either side of the notch. As the marginal filaments move in, they push this secondary line ahead of them and at the end often form with these a confused tuft of cells just prior to the final disappearance of both filaments and operculum GILL-FILAMENTS IN SAUROPSIDA pA In figure 9 the primary and secondary series have maintained their identity to the end. In reviewing the origin of both series it will be seen that the posterior wall of the hyoid arch is poten- tially a filament-bearing surface as well as the walls of the third and fourth branchial arches. This is in accord with what we know of conditions in gill-bearing vertebrates. By way of a summary the history of these vestigial gill fila- ments in the chick may be divided into six stages: 1, the appear- ance of degeneration vesicles in the branchial epithelium; 2, the concentration of these in the ectoderm covering the third arch and, to a lesser extent that covering the fourth arch; 3, the thick- ening of the ectoderm of these two vesiculated areas into tufted epithelial mounds, and, in the case of the third, an apparent evagination of the ectoderm with a mesodermal core; 4, a gradual differentiation of these areas (now crowded into one and fused with the sides of the backward growing operculum) into a trans- verse line of filaments on each side of the neck; 5, a progressively medial displacement of this line, correlated with the medial mi- gration of structures in the pectoral body-wall; 6, a rather rapid reduction of this line and the eventual suppression of both fila- ments and operculum. GILL FILAMENTS IN REPTILES Although the branchial region of reptilian embryos exhibits some measure of transition between the higher amniotes and the gill-bearing vertebrates it is much more nearly akin to that of birds than to that of any other group. Particularly is this true with regard to the development of the hyoid arch, where in tur- tles, lizards, and alligators (as in birds) the opercular processes of the two sides unite to form a conspicuous fold of tissue across the ventral surface of the neck which persists long after all trace of the other gill arches has disappeared. Similarly, vestiges of filamentous structures behind the hyoid arch are to be found in at least three of the main groups of reptiles, although in a more transitory and less conspicuous form than obtains in the chick. That they are not developed to a higher degree in the former may be explained by the fact that living reptiles are themselves a a2? EDWARD A. BOYDEN modern and highly specialized group; and that the degree to which retrograding structures are developed does not necessarily correspond to the rank which the possessor of these structures holds in a graded phylogenetic series. The fact that these struc- tures are present at all in reptilian embryos greatly increases the significance of the better developed filaments in the chick embryo. In discussing the conditions in reptiles it should be borne in mind that considerably less material was available than in the study of the chicks where some seventy embryos between the sixth and ninth days of incubation were examined. Of the four reptile series in the Harvard Collection, which are sufficiently extensive to afford a fairly complete picture of the development of the branchial region, three of them, Lacerta, Kutaenia and Chrysemys, show epithelial outgrowths behind the hyoid arch which are identical with the filamentous structures found in the chick. The first of these, Lacerta muralis, presents a more primi- tive branchial system than is found in birds, five well-spaced ecto- derma! grooves being visible from the outside at an early period. Later in its development the operculum fuses with the region behind the fourth arch in such a way as to form a peribranchial chamber into which portions of the third and fourth arches with their respective aortic trunks freely protrude (H. E. C., Ser. 813; 6.4 mm.). Still later, when the branchial chamber has become obliterated, small epithelial proliferations appear from under- neath the operculum in the region of its fusion with the posterior gill arches (H. E. C., Ser. 811 and 812; 7.4 mm.). Although they have but a transitory existence they occur at the same rela- tive time and place as the filaments in the chick, with the dif- ference that the filaments in the birds appear on the third and fourth arches prior to their fusion with the operculum as well as afterwards. In Aristelliger praesignis this is apparently re- versed; the epithelium of the third arch is very much thickened just prior to fusion with the operculum, but thereafter no fila- ments are to be observed, as if a somewhat premature fusion, as compared with conditions in other embryos, had inhibited the epithelial proliferation which had already started (H. E. C., Ser. — GILL-FILAMENTS IN SAUROPSIDA DIAS 1884; 4.9 mm.). The same is true of Sphenodon punctatum (H. E. C., Ser. 1491; 7.9 mm.). In contrast with these lizards the snake Hutaenia presents a very interesting condition. So great is the lengthening process to which the body as a whole is subjected that the gill arches and clefts are obliterated by being drawn out instead of being crowded together. There is no opportunity for the formation of a peri- branchial chamber nor even for a ventro-medial union of the two hyoid arches to form the plica opercularis, so characteristic of the Sauropsida as a whole. Consquently each hyoid arch is pulled back on the side of the trunk and there undergoes a further development by itself, persisting long after the other arches have lost their identity. Just before these disappear (Eutaenia sirtalis, H. E. C., Ser. 1349; 7.6 mm.; and E. radix, Ser. 1350 7.4 mm.) the epithelium of the operculum (in the first case from the under side, in the second from the outer side) gives rise to a tuft of cells comparable in point of time and position with the filaments of the chick. Again, however, these are rather small and transitory appearances. ‘To see structures in the rep- tile, closely comparable to those in the chick it is necessary to examine turtle embryos (Chrysemys, H. E. C., Ser. 1078; 10.0 mm.; and Ser. 1083; 11.6 mm.). The first of these (fig. 22) has been placed bes:de a chick embryo of exactly the same size (fig. 21, H. E. C., Ser. 2038; 10.0 mm.), which happily was so sec- tioned as to permit a very striking comparison of the two em- bryos, even to such details as the aortic arches, cephalic veins, ete. A glance at the operculum and its underlying filaments in the two specimens shows that at least in the stage at hand we are dealing with almost identical structures. Again, however, these filaments have but an ephemeral existence as compared with the development which the same structures undergo in the chick. DISCUSSION OF LITERATURE Of considerable interest in connection with this paper is the exhaustive work of Ekman on the branchial region of the Anura. He conducted a series of experiments to determine the various factors involved in the production of gill filaments in frog; and 224 EDWARD A. BOYDEN toads. He was able to show by transplantation methods that the ectoderm of the branchial region and immediately adjacent territory has a certain specificity for building gill filaments not possessed by the remaining ectoderm of the embryo; that a polarity of this ectoderm can be demonstrated; and that even when theentodermand mesoderm underlying the future gill region in very young embryos are removed the ectoderm alone will pro- duce abortive filaments devoid of blood vessels. It is the ecto- derm of this same region in reptiles and birds which produces rudimentary filaments and they bear at least a superficial resem- blance to some of the abortive structures thus produced experi- mentally in amphibia by Ekman (ef. figs, 26 and 27). In the case of the higher vertebrates the process never passes beyond the initial stages as evidenced by the early appearance of degenera- tion vesicles and the failure of blood vessels to participate in gill formation. In the light of Ekman’s experiments and the evidence pre- sented in this paper it is doubtful whether the entodermal invagi- nation which Grosser found in the first pharyngeal pouches of young human embryos has been rightly interpreted as an inter- nal rudimentary gill. Grosser recorded his observations as follows: A remarkable observation has been made by the author in all young embryos with the first pharyngeal pouches well developed; these are the embryos R. Meyer 335, Hal, Pfannenstiel III (loaned for this pur- pose), R. Meyer 330, and also a somewhat pathological, young embryo from the collection of R. Meyer. In the revion of the first pouch there projects ventrally (figs. 315 and 316) or caudally (fig. 318) from the closing membrane into the pharyngeal lumen an irregularly knobbed process filled with mesoderm. That it is an accidental structure or due to post-mortem changes seems to be excluded by the regularity of its occurrence (Low has figured, but not described it). It disappears quite early (in the oldest embryo examined, figure 318) it is present only on the left side and is greatly reduced in size; in embryos of 4.25 5.0 and 5.8 mm. and in those still older, it is wanting), and may per- haps be interpreted as a rudimentary internal gill.4 It would not be the first instance of a very ancient rudiment well developed in the human embryo. Similar structures have not yet been observed in other amniote embryos. (Keibel & Mall Human Embryology, 1912). ‘ Italicized by the author of the present paper. GILL-FILAMENTS IN SAUROPSIDA 225 In his description of embryo ‘Robert Meyer No. 335,” the same author (’11) figures a section of the ‘‘gill rudiment”? which he describes as follows: Das Relief der Gegend der ersten Tasche wird hauptsachlich von der erwihnten, in das Lumen des Vorderdarmes vorragenden Einstiil- pung beherrscht; sie mag vorliufig als Kiemenrudiment! bezeichnet wer- den. Das Kiemenrudiment liegt jederseits ventral und zum Teil kranial von der Beritihrungsstelle der Epithelien, kaudal vom ersten Aorten- bogen und ragt zapfenformig dorsal und kaudalwarts in das Lumen der Darmbucht vor. In seinem Inneren findet sich ein mesodermaler Kern, Gefasse sind aber in diesem nicht mit vollen Sicherheit nachzu- weisen. There are at least three difficulties in the way of accepting Grosser’s interpretation, the first of which involves the ento- dermal origin of the structure he presents as a gill filament. Kingsley states in his Comparative Morphology, that the gills of vertebrates ‘‘were long regarded as of entodermal origin but in recent years considerable doubt has been thrown on this; at least for fishes, and there is some evidence for their ectodermal origin.’ Ekman has shown that the ectoderm alone can produce abortive filaments in frogs and toads, while the evidence of the present paper establishes the fact that in the Sauropsida the filamentous structures which have been described as vestigial gills are wholly ectodermal. Another factor unfavorable to Grosser’s interpretation is the position of this structure in the auditory pouch. If gills have persisted at all in so highly de- veloped an animal as man, it is not likely that they would persist on arches which least commonly possess gills in water-breathing vertebrates. For with the exception of a few cyclostomes and fishes, gills are never found in the hyomandibular cleft. Again, the time of development is against Grosser’s interpretation. * The inpocketing which he describes first appears at a time when only the first two entodermal pouches have been formed (Embryo R. M. 335, 1.73 mm., 9 to 10 somites) and when the second has not yet reached the ectoderm. It is last met with in embryos no older than R. M. 300 (2.5 n m., 23 somites) where only the first threé pouches have reached the ectoderm. In fishes the func- tional gills are never formed before all the clefts have broken 226 EDWARD A. BOYDEN through, at a considerably older stage than that represented by the human embryos under discussion. Unless, therefore, some- thing similar can be found in the branchial region of the lower vertebrates, it hardly seems as if the structure described by Grosser could be regarded as an internal rudimentary gill. With more likelihood it may be compared with those outpocketings in the preauditory region of the pharynx of the chick which Kastschenko as doubtfully considered ‘‘vermutliche rudimen- tire Schlundtaschen.”’ Apparently filaments do not occur in mammals. The exten- sive series of mammalian embryos which are available in the Harvard Collection have been searched in vain for traces of fila- ments comparable with those already described for the Saurop- sida. Reviewing the phylogeny of the branchial system of ver- tebrates in the light of these facts, it would seem that the gills of the lowest vertebrates have given place to functionless homo- logues in the Sauropsida and that with the further reduction which the branchial system has undergone in mammals all traces of even vestigial filaments have disappeared. CONCLUSIONS In the Sauropsida the development of the branchial region is characterized by the formation and relatively late persistence of a band of tissue across the ventral surface of the neck, which has been derived from the ventral union of the hyoid arches, and which may be known from its resemblance to the development of the gill cover of certain fishes and amphibians as the opercular fold or plica opercularis (Kiemendeckelwulst of German authors). On the lateral margins of this operculum, after it has grown backward to enclose at least a potential peribranchial chamber, filamentous outgrowths may be observed on its under side, which in reptiles have a very transitory existence but which in the chick undergo a relatively extensive and prolonged development. On account of the filamentous character of these outgrowths, their origin from the branchial arches (the epithelium of which Ekman has shown to possess a certain specificity for gill-forma- tion in the Anura), and their constant relation to the operculum GILL-FILAMENTS IN SAUROPSIDA DD in both reptiles and birds, these structures are adjudged to be true gill filaments, evidently vestigial in character, but none the less comparable in kind to the functional organs of water-breath- ing anamniotes. If this interpretation proves to be correct an unbroken series of gill-bearing vertebrates is thus presented from fishes up to mammals. LITERATURE CITED EKMAN, GunNAR 1913 Experimentelle Untersuchungen iiber die Entwick- lung der Kiemenregion (Kiemenfiiden und Kiemenspalten) einiger anuren Amphibian. Morph. Jahrb., Bd. 47, S. 419-575. Evans, H.E. 1912 The development of the veins of the body wall. Keibel and _ Mall, Human Embryology, vol. 2, p. 686. Greit, A. 1906 Ueber die Homologie der Anammierkiemen. Anat. Hefte., Bd. 28, S. 59-60. Grosser, Orro 1911 Zur. Entwicklung des Vorderdarmes menschlicher Em- bryonen bis 5mm. grosster Linge. Sitzber. K. Akad. Wiss. Wien, vol. 1205 p:.7 1912 The development of the pharynx and of the organs of respira- tion. Kiebel and Mall, Human Embryology, vol. 2, p. 454. KastscHENKo, N. 1887 Das Schlundspaltengebiet des Hiihnchens, Arch. f. Anat. u. Entwickel., p. 258. Kinastey, J. S. 1912 Comparative anatomy of vertebrates. Blakistons 1912. Dp. 237. K6ouiiker, A. 1884 Grundriss der Entwickelungsgeschichte. p. 103. Matt, F. P. 1898 The development of the ventral abdominal walls in man. Jour. Morph., vol. 14, p. 361. Nassonow, N. 1895 Ueber das Operculum der Embryonen des Struthio came- lus, L. Zool. Anz., Bd. 18. Jahrg. N. 492. p. 487. ParKER, T. JEFFRIES 1892 Observations on the anatomy and development of Apteryx. Phil. Trans. Roy. Soc., vol. 182. RatuHke, H. 1825 Isis, H. 6. 1832 Anatomisch-philosophische Untersuchungen tber den Kiemen- apparat und das Zungenbein der Wirbelthiere. Riga und Dortat. 1839 Entwickelungsgeschichte der Natter. Koenigsberg. -p. 60-65. 1861 Entwickelungsgeschichte der Wirbelthiere. Leipzig. p. 71. Remak, R. 1851-55 Untersuchungen iiber die Entwickelung der Wirbelthiere. I. Ueber die Entwicklung des Hiihnchens im Hie. Berlin. 1851. p. 44-49, PLATE 1 EXPLANATION OF FIGURES Development of gill filaments in chick embryos of the fourth to the eighth day of incubation 4 Chick; 3 days, 4 hours; 6.8 mm. X 9 diam. (H. E. C., Ser. 2057) LIV, first four ectodermal grooves (cf. with fig. 10). 5 Chick;3 days, 22 hours;8.1mm. X 9 diam. (ef. with fig. 11). Note hillock on third arch. 6 Chick; 4 days, 23 hours; 13.4 mm. X 9 diam. I, auditory groove; IJ, III, second and third ectodermal grooves limiting the wedge out of which the filaments are differentiating (cf. with figs. 18 and 19). 7 Chick; 6 days, 3 hours; 16.9mm. XX 6diam. IJ, site of the second ecto- dermal groove, occupied by a ridge bearing filaments; S.P., pectoral groove. 8 Chick; 6 days, 5 hours; < 6 diam. (cf. with figs. 1 and 20). 9 Chick; 7 days, 1 hour; 19.7 mm. X 6 diam. 228 PLATE 1 N SAUROPSIDA EDWARD A. BOYDEN ENTS I A THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 1 PLATE 2 EXPLANATION OF FIGURES Sections illustrating early development of filaments 10 Section through the branchial clefts of an embryo of 3 days and 4 hours (H. E. C., Ser. 2057; 6.8 mm.; section 266; * 27 diam.), illustrating the stage when vesicles first appear in the branchial epithelium (cf. with fig. 4). /-/V ectoder- mal grooves; op., operculum. 11 Section of an embryo of 4 days and 4 hours (H. E. C., Ser. 2058; 11.0 mm. ; section 482; 27 diam.), illustrating the stage when vesicles become concen- trated in the ectoderm of the third and fourth arches (ef. with fig. 5). Note the thickened epithelium covering the mound on the third arch. 3, 4, 6, aortie arches. 12 to 17 Consecutive serial sections of an embryo of 4 days and 3 hours (H. E. C., Ser. 1943; 12.0 mm.; sections 215 to 220 respectively, X 27 diam.). The first three, on the left, show the mound which forms on the third arch and its mesodermal core; the last three, on the right, show the thickened epithelium at the lower end of the mound, out of which the first filaments are differentiating. 12 J, 1I, IIT, ectodermal grooves; IJ, IIT, IV, pharyngeal diverticula. 15 E-IIT, E-IV, evaginations of the ectodermal epithelium on the third and fourth arches, respectively. PLATE 2 GILL-FILAMENTS IN SAUROPSIDA EDWARD A. BOYDEN : 231 PLATE 3 EXPLANATION OF FIGURES Sections illustrating later development of filaments 18 and 19 Sections through the hyoid region of an embryo of 5 days; 3.0 mm. (H. E. C., Ser. 1951; sections 245 and 247 respectively) X 67 diam. 3, 4, 6, aortic arches; JJ, second ectodermal groove; ///, diverticulum of third pouch; op., operculum; /;, filaments at the upper end of the ‘wedge;’ fo, filaments at the point of the ‘wedge’ (ef. with fig. 6). 20 Frontal section through the opercular fold and pectoral wall of an embryo of 6 days and 2 hours; 16.0 mm. (H. E. C., Ser. 1950; section 901) X 67 diam. (cf. with fig. 8). op., operculum; f., branched filaments; s.p., pectoral grooves; c.m., mesothelial ridge; b.c., bulbus cordis; p., pericardial cavity; c.p., thickened epi- thelium which gives rise to the pectoral ridges. 21 Filaments of a 10.0 mm. chick (H. E. C., Ser. 2038; 5 days; section 572) < 67 diam. (ef. with fig. 22). 3, 4, 6 aortic arches; III, IV, diverticulum of third and fourth pouch; e., esophagus; t7., trachea; p., pericardial cavity. 22 Filaments of a 10.0 mm. turtle embryo (H. E. C., Ser. 1078 Chrysemys marginata, section 286) * 67 diam. FILAMENTS IN SAUROPSIDA PLATE 3 EDWARD A. BOYDEN SOE Ua esse PT & cart pest ctieim, Hints sen mi 233 PLATE 4 EXPLANATION OF FIGURES 23 Low power sketch of operculum and filament from a chick embryo of 4 days and 23 hours (H. E. C., Ser. 2059; 14.0 mm.; section 1004) X 50 diam. 24 High power sketch from same section as figure 23 showing longitudinal section of an opercular filament. Camera lucida drawing X 750 diam. ect., ecto- derm covering the under surface of the operculum; mes., underlying mesenchyma. 25 Epithelial cyst in process of formation. Camera lucida drawing of sec- tion 207. X 900 diam. (H. E. C., Ser. 1954; 4 days, 3 hours; 9.0 mm.). ect., ecto- derm covering the fourth branchial arch; mes., underlying mesenchyma. The vesicle figured measures 19u in diameter. 26 After Ekman’s figure 29: “Horizontalschnitt durch die Kiemengegend einer Bombinator—Larve 3 Tage nach der Entfernung der entodermalen Mund- hoéhlenwand im I. Stadium. £&g. Blutgefaisse; A/-IJJ 1-3. IWiemenreihe; Op., Operculum.’’? X 200 diam. 27 Section through the hyoid region of a 12.0 mm. chick (5 days, 1 hour, < 100 diam.), for comparison between the normally occurring vestigial filaments and opercular fold of the chick and the abortive filaments and peribranchial chamber experimentally produced in toad embryos by Ekman (ef. with fig. 26). 234 FILAMENTS IN SAUROPSIDA PLATE 4 EDWARD A. BOYDEN & n 16 nS ean , 235 (vat 4) A nga ME a ra: ol 1 ieee : 4g yi 7 ped hee Nig heal tae ay! Feehan Noted hae f abe SAL, Vitra a tijn | Por a , A AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, JANUARY 5 THE FORMATION AND STRUCTURE OF THE ZONA PELLUCIDA IN THE OVARIAN EGGS OF TURTLES ALICE THING Anatomical Department, Western Reserve University, Cleveland, Ohio TWELVE FIGURES CONTENTS lintitoxe (bKern Claas Sante GR Seer es Gene IEE eo. o ono auacbic on gajoo DOO tO DAF 237 bsboniealvs ke tel.) nisi oom evecare oko Rae oe ee ECE LE cee 238 Material and methods (including possible sources of error)..............-. 242 ODSerV ATONE. t.co8os fas ocee hee eee Bes PAE. Mr ss 58 oo conor 244 AN SA BlonneLVSDYb en Co eT Hone Eee Saga cobo,c oes dooma Kd ORS 244 Bae Zomanpel lel dai ccc 2 ccs tcemeesta eee votes CSc cle OCU E OL Tr ara ester ieee 246 USAR Gulls Vast at eyareicrasisaraieis ei aie slings eos ibshe eae eee ae Tee OTe 246 [DEER SUR YE Brat tke eS a EMRE ac cada edd oe ch CMO OO 248 CHRIS EEK UC tks It Caesars Ae Ge eR Geos ool Oooo nas aGe 250 ‘SULIMTENTC CTP 2 a RES ick Oe ere PIERRE Feel 2 ASG hoioc tis oc o.Brh 251 EMO STAY. cain. 2. aire eros Vole sina d Y aylve «Bedok ole eh eee ee 253 INTRODUCTION In the active research upon eggs of all groups of animals, which has been in progress for nearly fifty years, the zona pellucida, a cuticular membrane, formed around the egg in the course of growth at some stage preceding maturation, has not failed to be an object of interest to investigators and to share with other portions of the egg the most painstaking examination. Accord- ing to Waldeyer (’01, ’02, ’03, p. 287) there occurs at least one membrane in all vertebrate eggs whereas among invertebrates some eggs remain naked. The majority of vertebrate eggs which have been subjected to careful study show this membrane to be the zona pellucida. It consists typically of two concentric layers, one of which exhibiting characteristic radiating striations perpendicular to the egg surface is termed the zona radiata. 237 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 MARCH, 1918 238 ALICE THING That in the same species this membrane varies to a considerable degree in thickness is shown by figures given below where in the turtle’s egg its thickness ranges from Iw in initial stages to 17u. The variation of this dimension in the eggs of different groups of vertebrates can be demonstrated by comparing these figures with those given by Prenant, Bouin, and Maillard (Traité d’Histologie, p. 1097) who quote Nagel’s measurements, 1.2u to 1.54 in the mouse and 2.0u to 2.54 in man. The zona pellucida must be distinguished from the yolk or egg membrane, a very thin membrane observed in some vertebrate eggs surrounding the yolk before maturation is completed (Van Beneden, ’80, Fischer ’05, p. 595). In respect to thoroughness and extent of investigation upon the structure of the zona pellucida, mammalian eggs naturally stand first. Less complete and comprehensive study has been given to it in the lower vertebrate groups. HISTORICAL SKETCH Authors who have given detailed contributions upon the mammalian zona pellucida may be grouped into three classes: first, those who regard this membrane as originating from the egg cytoplasm; second, those who maintain that it is derived from processes of the egg epithelium or from an exoplastic or intercellular substance of the cells of this epithelium; third, those who frankly state its origin to be uncertain. ‘The first class includes Van Beneden (’80), Kélliker (98) and Sobotta (’Q2). In the normal Graafian follicles of the bat (Rhinolophus ferrum equinum, Schreb.), Van Beneden observed egg cells in such close contact that the zona pellucida of one touched the zona pellucida of its neighbor to the exclusion of the epithelial layers. He, therefore, concluded that the zona pellucida developed from the surface of the egg cytoplasm in the absence of the egg epi- thelium. Ko6lliker stated that egg cells not yet surrounded by epithelium but already arranged in islets or nests were inclosed in a distinct membrane which he designated as the first anlage of the zona pellucida. ZONA PELLUCIDA IN TURTLE EGGS 239 Flemming (82), Retzius (89), Paladino (90), Von Ebner (00), and Fischer (’05) affirm that the zona pellucida is formed of cytoplasmic prolongations of the epithelial cells. Flemming describes fibers which ‘‘may be protoplasmic connections of the egg cell with its neighboring cells; in the spaces between these bridges the intermediary mass of the zona, gradually becoming firmer, may be laid down.’ Retzius states that in the rabbit cylindrical cells send out branched processes which gradually interlace so that a thick network originates around the egg. A consolidation occurs on the inner belt of this network forming the zona pellucida. In the completely developed zona pellucida the outer zone is also consolidated and between the inner zone and the surface of the egg radiating striations can be recognized. These represent granular filaments, which bore through the sub- stance of the zona and are attached to the egg surface by small conical basee. In the opinion of Paladino, bridges exist in the rabbit between the epithelial cells as a whole, as well as between these and the egg cell. In ripe eggs a fiber net exists between the epithelium and the outer surface of the egg, the inner meshes of which contain finely granular substance. Paladino gave a rather bizarre interpretation of this granular substance, regard- ing it as nutrient material derived from the breaking down of the epithelial cells formerly existing in these areas. A true zona pellucida is evolved from this substance which becomes hyalin in character and strongly refractive. Von Ebner substantiates in general the statements of Flemming and Retzius. The first anlage of the zona shows a network closely attached to the egg surface; this gradually moves back toward the epithelium leaving radiating filaments in connection with the egg surface to give place to the ‘‘secondary zona substance.’’ According to Fischer the zona pellucida arises from unbranched cytoplasmic prolonga- tions of the epithelial cells interwoven and pressed together. Compression occurs to such an extent on the inner portion of the zona as to eliminate the individual outlines and thus form a homogeneous substance. In the completely developed zona pellucida he distinguishes three layers, spongy, radiating and homogeneous, of which the last named is the oldest and firmest. 240 ALICE THING Regaud and Dubreuil (08, p. 152) deny any protoplasmic connections between the egg and the cells of the egg epithelium. In a fully developed ovarian follicle (rabbit) the zona is formed of three concentric layers. The first isa very thin internal layer applied to the surface of the egg but substantially independent of it; to this layer, which is not homogeneous but fenestrated in the manner of a grating (or grill?), we have given the name fenestrated epiovular membrane. The second is an external layer in connection with the prolongations of the cells of the corna radiata: it is formed by a thick felt of filaments running in all directions, the felted layer. The third or middle layer, the zona pellucida properly called comprises two substances, radiating filaments irregularly extending from the felted layer to the periovular membrane and an amorphous or granular substance (following the action of the fixative), laid down in abundance in the spaces between the radiating filaments which it bathes. . . . The felted filaments, the radiating filaments and the epiovular membrane which have been interpreted up to the present time as anastomosing elements are not protoplasmic but an exoplastic production of the follicular cells about the egg. The investigations of Rubaschkin and Waldeyer have left them in doubt as to the exact origin of the homogeneous sub- stance (so-called by them) of the zona pellucida. According to Waldeyer (01, ’02, 03) the zona pellucida is composed of a fiber felt and a homogeneous substance across which protoplasmic connections from the epithelium to the ooplasm make their way. The homogeneous substance is perhaps a product of the ooplasm and the mammalian zona pellucida, derived in part from the epithelium, in part from the ooplasm. Rubaschkin (05, p. 519) describes as the zona pellucida in guinea pigs, a thick homogene- ous layer directly surrounding the egg or yolk membrane. Cen- tral processes from the epithelial cells penetrate this zona sub- stance where they lose their protoplasmic appearance. These processes do not form intercellular bridges because they are pre- vented from actual contact with the ooplasm by the presence of the egg membrane. They do not end with enlargements or knobs as Retzius figured them to do. A number of eggs, however, show a thick layer of coarse fibers, the processes of the epithelial cells which wind about the zona substance but do not penetrate it at any point. This layer corresponds to the perizonal fiber net of Retzius. Waldeyer is inclined to regard the zona pellucida, con- ZONA PELLUCIDA IN TURTLE EGGS 241 trary to his earlier opinion, as a product of the ooplasm, while Rubaschkin favors the view that it is derived from the epithelium. Of those who express an opinion on the zona pellucida of the egg in vertebrate groups below the mammals there may be cited Lams on the European smelt (Osmerus eperlanus), Munson on the turtle (Clemmys marmorata), Waldeyer on_ selachians amphibia, reptiles and birds, and Mlle. Loyez on reptiles in general. In Osmerus eperlanus, Lams (’03, ’04) describes the zona pellucida which he calls a chorion, thick and radially striated. The striated appearance is due to innumerable canalicules running perpendicular to the surface of the egg. Also, in the cytoplasm of the egg directly beneath the yolk membrane, he sees granular striations which ‘‘do not properly, in all probability, belong to the egg cell but correspond to prolongations of the follicular cells which have traversed the canalicules of the chorion and the yolk membrane and become continuous with the cytoplasm of the egg.’’ Munson (’04, p. 331) states that in Clemmys marmorata there occurs an egg membrane which is composed of two layers, the outer homogeneous and the inner striated. In selachians, amphibians, reptiles and birds Waldeyer (’01, ’02, ’03, p. 293) shows that a zona pellucida consisting of an outer homogeneous and an inner striated layer can be seen well only in developing eggs, that it atrophies in mature eggs leaving only a very thin egg membrane. The striated appearance is due to radial canals. According to Mlle. Loyez (’05, ’06, p. 147) three membranes arise in eggs of reptiles. The vitelline membrane which originates directly from the primitive membrane of the oocyte is at first very thin. As it increases in thickness it becomes finally striated and then granular. The heavily striated zona radiata forms on its inner surface early in the course of development. A very transitory third membrane is differentiated from the internal surface of the zona radiata. After its disappearance the inner surface of the zona radiata becomes less and less distinct and finally the striations come to appear in the superficial layer of the egg. Mlle. Loyez’ vitelline membrane and zona radiata together make up the zona pellucida, without doubt and her third transitory membrane is the yolk membrane. 242 ALICE THING This short résumé proves that most authors agree upon the existence of protoplasmic bridges connecting the epithelial cells with the egg cytoplasm. But when they mention the homogene- ous cuticular substance few give satisfactory descriptions and illustrations of the origin of this or of its structure in later phases of development. Only Regaud and Dubreuil go into the subject in detail; they lay emphasis upon the different stages of its development from the exoplastic fibers formed between the epi- thelial cells. The object of the present paper is to show that this membrane in the species studied consists neither of real cytoplasmic structures nor of real exoplastic structures but of intercellular substance and of cytoplasmic prolongations of the epithelial cells combined in a definite manner. ‘The intercellular substance is represented by a series of walls ramified and anasto- mosed in such a way as to create cylinders or canals of which the transverse section appears as a reticular network. Extending down through these cylinders cytoplasmic filaments from the epithelial cells make their way to the yolk substance. MATERIAL AND METHODS (INCLUDING POSSIBLE SOURCES OF ERROR) Twenty-one series have been prepared from the ovarian eggs of the following turtles: Clemmys guttatus (Schneider), Grap- temys geographicus (Lesueur), Emydoidea blandingi (Holbr.), Aromochelys odoratus (Latr.) and Chrysemys picta (Hermann) in various stages of growth. The identification of these species is so simple that I shall not stay to discuss the particular features by which they were identified. The animals were killed as soon as possible after their arrival in the laboratory to reduce errors in observation, the result of any prolonged starvation due to improper feeding, a condition which has marked influence upon the general ovarian structure as shown recently by Walsh (Loeb 17). The time which elapsed between the capture of the turtles and their arrival in the laboratory and the conditions under which they were kept prior to their arrival are not known. The major- ity of the ovaries examined had the appearance of being perfectly ZONA PELLUCIDA IN TURTLE EGGS 243 normal. Ina few of them, however, at least one egg which must already have attained a diameter of 2 to 3 mm. showed processes of degeneration well under way. In several of these pathological eggs I found an object which Dr. Van der Stricht and Dr. Todd identified as a parasite. To the influence of this parasite the pathological condition of the egg was probably due but no one to my knowledge has so far made a study of this subject. All the eggs examined were very much less than the size of the deposited egg. No essential differences are apparent in the structure of the zona pellucida of the various species, hence the stages described below have been chosen as representative of all the material. The following methods of technique were employed. Fixa- tion by the fluids of Hermann, Flemming or Benda, followed by staining with iron haematoxylin-and Congo red or with safranin and picric acid. Fixation by Bouin’s mixture or trichloracetic acid followed by staining either with iron haematoxylin and Congo red or with Mallory’s connective tissue stain. The sections are cut four or five micra thick. All investigators have studied their material in cross section but, judging from their text and illustrations few have seen the importance of examining tangential and oblique sections. Fischer mentions that he could see the fiber work of the spongy layer very beautifully in tangen- tial sections. In tangential sections Lams is able to interpret the structure of the chorion of Osmerus eperlanus. Dr. Van der Stricht called my attention to the significance of this method of study. Because of shrinkage in paraffin and because of flattening from the action of fixatives and from the pressure of the knife in cutting the circumference of the egg almost always becomes ovoid: this necessitates the taking of averages from measurements of the long and short axes. Because also of the method of measuring with the camera lucida, the figures given for the diameters of the egg, taken through the zona pellucida, are only approximate. The figures given for the thickness of the zona are more exact, having been obtained from prints of microphotographs by computing the magnification. 244 ALICE THING The microphotographs, all of which were taken at a magni- fication of 750 diameters, represent the structure of the zona pellucida in eggs ranging in diameter from 0.65 mm. to 2.6 mm. and from younger stages in which the zona pellucida measures only lu up to a stage where it is 17 in thickness. I do not know if this last measurement may approximate the maximum thick- ness of the zona since I have no measurements from larger eggs. Mile. Loyez states that in reptiles the zona is very thin upon completion of development. Since’ it is extremely difficult in microphotography to focus upon an entire field unless that field is perfectly flat in all its parts some portions of the figures are not sharply defined. The endeavor has always been to focus upon the most ‘mportant part of the section. OBSERVATIONS The epithelium When the oocyte has reached the size two or three times, at a rough estimate, that of the oogonium from which it originated it is surrounded by a flattened epithelium which remains of one layer throughout the course of development. With the gradual growth of the oocyte the epithelial cells take on a definite pris- matic shape and increase in height in the axis perpendicular to the surface of the egg until this axis may become as long as the transverse. The transverse axis appears the longer, however, in the majority of cases especially in the later stages herein de- scribed. Upon cross sections through the epithelial layer of oocytes less than 1 mm. in diameter the nuclei of the epithelial cells are seen to be rather widely spaced (fig. 2, ep.) while in older stages, because of reduction in size of the nuclei and in content of the cytoplasm, the arrangement is more compact (figs. 3, 7, 9, 10, 11, ep.). Occasional mitoses prove that to accommodate the increasing volume of the egg the epithelium extends itself by divisions of its constituent cells. In eggs much larger than those figured very numerous mitoses occur. The epithelial cells are sharply marked off from one another by intercellular channels filled with intercellular substance. Unfortunately this does not ZONA PELLUCIDA IN TURTLE EGGS 245 show clearly in the photographs. Some preparations fixed in Bouin and stained by Mallory’s connective tissue method, show this substance very clearly colored by aniline blue. The inter- cellular substance early undergoes a change of constitution and becomes transformed, at the level of the surface of the cells, into the special cement known as the terminal bars (Schiifer ’12, p. 86, Stéhr’98, p. 68). It is well known that sections cut perpendicular to the plane of the surface of the epithelial cells show in well fixed and stained preparatidns a continuous dark line representing the lateral surfaces of the terminal bars sometimes thickened noticeably at points marking the limits of two adjacent cells. In other portions of the sections this line may not be seen but cross sections of the bars appear as dark round spots. The for- mer picture is represented in the turtle’s eggs in figure 2, 0.0. The lateral surfaces of the terminal bars of adjacent cells form a rather thick distinct boundary line between themselves and the oocyte thus marking the beginning of the zona pelucida. Cytoplasmic bridges of various sorts connect the cells with one another (Fischer ’05, Paladino ’90). Filamentous and thin or short and coarse, they traverse the intercellular spaces and retain their identity for considerable distances within the cell cytoplasm where they finally mingle with the denser portions encircling the large nuclei (fig. 1 /.b., s.b.). A dense opaque mass, the attraction sphere, is closely attached to each nucleus usually either on that face which is nearest the surface of the cell or at one side (figs. 2,4, a.s.). Often such clearness is obtained through successful fixation or through the thinness of the section as to determine the character of the sphere.. It is composed of three elements, a small granule (or sometimes two) the central cor- puscle in the center or slightly to the side of an oval or circular clear field, the medullary layer marked off from the mass by a distinctly larger, more opaque zone, the cortical layer (Van Bene- den). Loosely interwoven filaments extend out from the dense attraction sphere to the clear exoplasm at the periphery of the cells thus forming a delicate network. 246 ALICE THING Zona pellucida Solely for purposes of clearness the developmental history of the zona pellucida may be presented in three successive stages. The first stage covers those phases of formation in which the zona pellucida, on cross section, is but a thin one layered cuticle while on oblique and tangential sections the beginnings of a reticular network are found. The second stage includes that period during which the zona pellucida becomes divided into two concentric layers, the inner thin and radially striated, the outer, denser with striations more or less obscured. The third stage is co-extensive with the period of growth during which both layers just mentioned become very much thicker. Stage 1. The terminal bars, as viewed on a cross section, divide the epithelial cells from the oocyte by an apparently continuous line which at first is thin and uniform but later becomes thicker until it is a cuticle of double contour and of rather uneven outline especially on its deep surface where it lies in connection with the epithelial cells. On this front the junctions of the intercellular substance, separating the lateral surfaces of the epithelial cells, make with the bars triangular thickenings. The change in the terminal bars initiates the development of the zona pellucida. From the time when the cuticle reaches an average thickness of ly it may be termed the zona pellucida (fig. 2, 2.z.p.). Filaments of the cytoplasmic network extending from the attraction spheres (a.s.) seem to attach themselves directly to the deeper limit of this cuticle (fig. 2) the actual structure of which is not demon- strable on cross sections. Oblique and tangential sections, how- ever, make it clear that the zona pellucida is of complicated organization even at this early stage. It is perhaps well to ex- plain at once that in an oblique or tangential section of an egg one may see two, three or more irregular rows of epithelial cells, the number depending upon the size of the egg and therefore upon the curve of the epithelial layer. These represent cross sections of the epithelial cells at various heights. These portions in the section furthest away from the yolk show the bases of the cells; ZONA PELLUCIDA IN TURTLE EGGS 247 then appear successively the clear cytoplasm and perhaps the basal segments of the nuclei; next various segments through the nuclei; and nearer the yolk, sections through the central spheres and terminal bars and therefore through the incipient zona pel- lucida. These tangential sections (figs. 3, 4, 5) prove that the cuticle is composed of large polygonal fields (p.f.) marked off from one another by a system of dark lines, the terminal bars (¢.b.). These large polygonal fields are not homogeneous but inclose smaller fields of similar outline formed by a fine pale network, the meshes of which are a little thicker and darker at some points and in close connection with the terminal bars, thus giving the impression of extensions of the bars over the surface of the epi- thelial cells. The meshes of this fine reticulum seem exactly to overlie the deeper cytoplasmic network (fig. 4 ¢.n) of the cell which arises from the interwoven filaments extending from the central spheres. The zona pellucida then takes its origin as a veil-like formation consisting of a mosaic of terminal bars and polygonal fields within which may be recognized the small, pale areas, future canals of the adult membrane separated by pale and dark filaments giving origin to the future fundamental substance of the adult membrane. In older oocytes several changes take place. Those portions of the network, in which the meshes are a little thicker and are stained in the same way as the terminal bars, have become much more numerous (figs. 5,6). - It may render the description clearer at this point to distin- guish the network of darkly stained meshes which follows the pattern of the original terminal bars around the large polygonal fields, calling this the primary network (p.n.) from that which follows the outlines of the original cytoplasmic reticulum, using for this the term secondary network (s.n.) Dr. Van der Stricht observes a similar distinction in structures of the membrana tectoria. The meshes of the primary network appear to send out short extensions to the secondary network and to soften their sharp angles so that these assume circular or oval shapes rather than clear cut polygons (figs. 5, 6, p.n.). So far I have been unable to assure myself definitely of a longitudinal splitting of 248 ALICE THING these meshes and of the development of cuticular bridges con- necting the parts as has been shown to take place in the membrana olfactoria (Van der Stricht) although certain figures do suggest such an interpretation. A superficial and older portion of the veil of the zona pellucida shows the beginnings of the adult structure, regular small round spaces inclosing dark granules, the cross sections of prolongations of the epithelial cells (figs 5, pr.). Stage 2. In more advanced phases of growth the nuclei of the epithelial cells are crowded nearer to one another and lie closely on the zona pellucida. A cross section (fig. 7) shows that the zona pellucida has become thicker and is divided concentrically into two layers, the outer of which (o./.) is more or less homo- geneous and very dark in the figure whereas the inner (7.l.) is less opaque and distinctly striated in a direction perpendicular to the surface of the egg. This layer is separated from the yolk substance by a sharp boundary, the nature of which together with the two layers of the zona must be investigated in tangential sections. The real importance of the study of tangential see- tions is well demonstrated here for the extremely intricate struc- ture of the zona pellucida, of which one could obtain no true con- ception from cross sections, is revealed with remarkable clearness. In many preparations, as portrayed in figure 8, o0.l. the outer denser layer appears separated into three concentric belts, a middle clearer stratum (s’.) between two bordering darker thicker strata (s.s”.). In other preparations stained either more deeply or very slightly this concentric division into belts is not seen. A completely satisfactory explanation for this phenomenon cannot be given. There is a possibility that it may be due to accidental causes, for instance uneven penetration of the fixative or other media used though its explanation is more probably to be found in differences in constitution between the older and the more recently formed parts of the zona pellucida. Far from being homogeneous the outer layer consists of clear spaces, the cross sections of a system of canals (c.) within which are seen filaments, the cytoplasmic prolongations (pr.) from epi- thelial cells. The canals are separated by a meshwork much thicker and larger than in earlier stages, representing the cutic- ZONA PELLUCIDA IN TURTLE EGGS 249 ular part of the zona pellucida (f.s.) already observed in the first stage. Immediately beneath the epithelium in the zona (s.) a series of polygonal or circular fields occupies an area correspond- ing to that originally marked off by the primary network. On the whole one receives the impression that merging occurs be- tween the primary and the secondary networks so that distine- tion between them is no longer possible. The three elements of which the outer layer is composed also make up the clear inner striated layer though in the latter region the network of the fundamental substance of the zona stains far less deeply and appears to be of a much less dense character. The striations (fig. 7, f.s.) are undoubtedly produced by filaments connecting the epithelial cells with the yolk and by walls of the tubes of the fundamental substance of the zona which these filaments traverse. Since tubes, canals and filaments occur in the outer layer it seems at first remarkable that the striations in it are not obvious in cross sections. In favorable and largely decolorized preparations, the outer layer does appear striated but in more darkly stained preparations the fundamental substance obscures the prolonga- tions because of its great affinity for the stain. The striation in the inner layer is quite evident in cross sections because its fundamental substance takes up very little stain. The inner layer is evidently the older part of the zona and must have been originally identical in substance with the outer layer, the later differentiation resulting from a change in properties of the older fundamental substance causing it to become less dense and to have le& affinity for stains. For the site of active proliferation of the fundamental substance is the surface of the epithelial layer which moves back as the epithelial cells withdraw in the cen- trifugal growth of the egg. It is a still more significant fact, I believe, that living eggs show striations in the outer layer also: at least I have lately observed this appearance in preparations of more advanced stages of the living eggs of Aromochelys odo- ratus, the eggs of which differ in no essential manner from those of the species previously mentioned in the general structure of the zona pellucida. In eggs of A. odoratus approximately 1.5 to 2 mm. in diameter examined in normal saline the striations of the 250 ALICE THING outer layer seemed continuous with those of the inner layer yet the line of demarcation between the layers was in no way obliter- ated. The difference in the nature of the layers apparently is one simply of refraction since there is no distinct structure divid- ing them nor indeed any distinguishable boundary line. The presence of an egg or yolk membrane which might have been represented by a sharp line of demarcation between the striated layer and the oocyte in figure 7 cannot be confirmed in tangential sections. The boundary line (fig. 7) seems to be produced by thickenings of the ends of cell prolongations at the points where they reach the yolk. . No trace of an egg membrane can be dis- covered in the living oocytes of A. odoratus. Stage 3. The inner layer of the zona pellucida grows in thick- ness comparatively slowly whereas the outer, increasing more rapidly, becomes two or three times as thick as the former (figs. 9 and 10). The area of proliferation often stains very deeply (fig. 9, a.p.) so that a densely colored belt borders the surface of the outer layer remote from the yolk. At certain points in the outer layer (fig. 9) are seen cross sections of the canals (c.) with their contents (pr.) at other points a real striation, the result of rows of granular filaments in continuity with identical rows of filaments in the striated layer. This confirms the observation made on living eggs in which was noted the presence of striation in the outer layer.- When the area of proliferation has chanced to stain less deeply one can see that the filaments are actually prolongations extending down from the scanty cytoplasm sur- rounding the epithelial nuclei into the substance of the zona (figs. 10, 11, pr.) There are no indications that these filaments branch as Retzius has reported in the case of the rabbit oocyte but the small conical or knob-like bases described by him appear (figs. 9, 10, k.e.) as enlargements of the prolongations. Among the granular filaments within the striated layer there appear more homogeneous elements in continuity with the meshes of the fundamental substance of the outer layer (figs. 9, 10, 12, f.s.). In this stage the constituents of the zona are shown to be the same as in stage 2: asystem of clear openings, cross sections of cylinders (c.) with their contents the prolongations (pr.) of the epithelial ZONA PELLUCIDA IN TURTLE EGGS eS k cells and the mesh work of the cuticular fundamental substance (f.s., fig. 12, 0.1., 7.l.). In the inner layer the meshes of the funda- mental substance stand out more clearly than in figure 8 since they are more deeply stained. Here the tubes and filaments have increased greatly in length and the fundamental substance in _ amount. In the series of stages showing these elements in vari- ous phases of development it can be noticed that whereas in numerous openings the prolongations are very well seen in other openings no contents are perceptible. The absence of prolonga- tions from some spaces may be due first to imperfect fixation and staining, secondly to the real lack of systems of cavities corre- sponding to and overlying the intercellular spaces and conse- quently the primitive terminal bars in the first stages of develop- ment. A very thin discontinuous line between the knob-like enlargements of the ends of the granular filaments and the yolk substance in figure 9 may represent a real egg membrane. This appearance is very rare and further investigation with more refined methods is required to explain it. In tangential sections one sees nothing convincing of the presence of an egg membrane. It may be as Van Beneden asserts regarding the eggs of the rabbit that it can never be isolated in ovarian eggs until a short time before impregnation. In that case it could not be seen in turtles, eggs as small as these which are at present being investigated. SUMMARY 1. The epithelium surrounding the ovarian egg in all turtles herein reported is represented by one layer of prismatic cells between the sides of which short and long bridges extend. The intercellular spaces at the surface of these cells are closed by a special cement, the terminal bars. The cell is formed by a nucleus and by cytoplasm consisting of an attraction sphere composed of a central corpuscle, a medullary and a cortical layer. These spheres form a dense endoplasm around the nucleus from which filaments extend to a clear layer near the periphery, the exoplasm in a delicate network. 252 ALICE THING The zona pellucida varies in thickness from lu to 17y ac- cording to the stage of development of the egg. Beginning with a stage where it is on an average 3u thick two different layers appear, the outer denser and thicker and the inner narrower, clearer and striated. In the course of development the outer layer differentiates, grows and extends to a greater degree than — the inner. 3.,The zona pellucida during its growth is always formed by two or three different elements: . The fundamental homogeneous substance filling up Ls spaces between b. A system of numerous canals or tubules which inclose c. Filaments or prolongations of the epithelial cells which are connected with the surface of the yolk. The fundamental sub- stance of the zona pellucida is more abundant and dense 1 in the outer layer than in the inner. 4. The fundamental substance of the zona pellucida is de- veloped as a cuticular element by the terminal bars or primary network, that is by a definite special intercellular cement possess- ing the property of extension over the free surface of the epithelial cells and forming connections there with the delicate secondary network apparently produced directly by the superficial cyto- plasm of the epithelial cells. The secondary network seems able to give rise at its surface to a cement similar to that resulting from the activity of the terminal bars. This superficial cuticular network gradually becomes thicker and by the development of fresh cuticular material builds up the entire fundamental sub- stance of the zona pellucida. The prolongations of the epithelial cells, at first short, traverse the zona pellucida and become longer as this increases in thickness. Enclosed in canals, the pro- longations reach the surface of the yolk to end in knob-like enlargements. The structure of the zona pellucida just described presents a condition most favorable for the conveyance of nutritive mate- rial from the epithelial area in contact with the maternal capillaries to the actively growing and extending yolk. ZONA PELLUCIDA IN TURTLE EGGS 235 In conclusion I wish to acknowledge my indebtedness for constant advice and criticism to Dr. Van der Stricht under whose direction this work has been carried out and to Dr. Todd who obtained and identified the material. BIBLIOGRAPHY CarTraneo, Donato 1913-14 Richerche sulla struttura dell’ovario del mammi- feri. Arch. Ital. di Anat. edi Embriol., vol. 12, pp. 1-34. Fiscuer, A. 1905 Zur Kenntnis der Struktur des Oolemmas der Siugetierei- zellen. Anat. Hefte, Bd. 29, pp. 557-689. FLemMMING, W. 1882 Zellsubstanz, Kern, und Zellteilung. Leipzig, p. 35. Kor.uiker, A. von 1898 Uber die Entwicklung der Graafschen Follikel und Eier. Sitzungsber. d. physiol. med. Ges. Wurzburg, pp. 1-7. Lams, H. 1903-04 Contribution & l’étude de la genése du vitellus dans l’ovule des Téléstéens. Arch. d’Anat. micr., T. 6, pp. 633-652. Lors, L. 1917 Factors in the growth and sterility of the mammalian ovary. Science, N.S8., vol. 45, pp. 591-592. Loyrez, M. 1905-06 Recherches sur le développement ovarien des oeufs méro- blastiques 4 vitellus nutritif abondant. Arch. d’Anat. micr., T. 8, pp. 69-237. Munson, Joun P. 1904 Researches on the oogenesis of the tortoise, Clemmys marmorata. Am. Jour. Anat., vol. 3, pp. 311-347. Parapino, G. 1890 I ponti intercellulari tra l'uovo ovarico e le cellule folli- colari, e la formazione della zona pellucida. Anat. Anz., Bd. 5, pp. 254-259. Recaup ET Dusrevit 1908 Sur les productions exoplastiques des cellules folliculeuses de l’ovaire chez la lapine. Verh. d. Anat. Ges. Berlin, pp. 152-156. Rerzius, G. 1889 Die Interzellularbriicken des Eierstockeies und der Folli- kelzellen sowie tiber die Entwickelung der Zona pellucida. Verh. d. Anat. Ges. Berlin, pp. 10-11. Rupascukin, W 1905 Uber die Reifungs- und Befruchtungsprozesse des Meerschweineneies. Anat. Hefte, Bd. 29, pp. 509-548. Scuirer, E. A. 1912 Text-book of Microscopic Anatomy. London, p. 86. Soporra, J. 1902 Atlas und Grundriss der Histologie, p. 89. Stéur, P. 1898 Text-book of Histology. 2nd Amer. from 7th German ed., Philadelphia, p. 68. VAN BENEDEN, E. 1880 Contributionsala connaissance de l’ovaire des mammi- féres. Arch. de Biol., T. 1, pp. 475-551. VAN DER Srricut, O. 1909 Le Neuro-epithélium olfactif et sa Membrane Lim tante Interne. Mém. cour. del’Acad. roy. de Méd. de la Belgique, AD 2D ite Be Waupeyrer, W. 1906 Die Geschlechtszellen. Handbuch der vergl. u. exper. Entwickelungslehre der Wirbeltiere. Hertwig, pp. 287-293. THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 PLATE 1 EXPLANATION OF FIGURES For abbreviations see page 256 Fig. 1 Tangential section of the epithelium of an egg 0.75 mm. in diameter from the ovary of Chrysemys picta. Benda. Safraninandpicricacid. 5u. Note the long filamentous (l.b.) and short thick (s.b.) cytoplasmic bridges connecting the adjacent cells. The intercellular substance is not stained. Fig. 2. Transverse section of an egg 0.69 mm. in diameter from the ovary of Chrysemys picta. Bouin. Mallory’s stain. 4u. The epithelial cells one of which shows an attraction sphere (a.s.) very well are widely spaced. The terminal bars (¢.b.) form the anlage of the zona pellucida (7.z.p.) whichis lw in thickness. Fig. 3 Oblique section of the egg represented in figure 2._ The zona pellucida (z.p.) is seen to develop from a system of large polygonal fields (p.f.) marked off by the terminal bars (¢.0.). Fig. 4 Tangential section of the egg represented in figure 2. A number of epithelial cells are cut through their bases, others at various heights through the nucleus, a third group through the attraction sphere, a fourth through the cyto- plasmic network and terminal bars at their surfaces. Central corpuscles can be seen in some of the spheres. The polygonal fields (p.f.) are sharply outlined by the terminal bars (¢.b.). Fig. 5 Tangential section of an egg 0.99 mm. in diameter from the ovary of Chrysemys picta. Bouin. Heidenhain’s haematoxylin, Congo red. 4u. The primary network (p.n.) of the zona pellucida follows the outlines of the original terminal bars and the secondary (s.n.) the outlines of the superficial cytoplasmic network of the epithelial cells. Fig. 6 Tangential section of an egg 0.74 mm. in diameter from the ovary of Chrysemys picta. Bouin. Heidenhain’s haematoxylin, Congo red. 4u. The details are similar to those of figures 4 and 5. Fig. 7 Transverse section of an egg 1.1 mm. in diameter from the ovary of Graptemys geographicus. Trichloracetic acid. Heidenhain’s haematoxylin, Congo red. 5yu. The zona pellucida has divided concentrically into two layers, the inner of which shows radiating striations very clearly. It measures 3.6u in thickness. Fig. 8 Tangential section of the egg represented in figure 7. Same fixation andstain. 5u. Bothlayersof the zona pellucida7.l. ando.l. are seen to be formed by three elements: 1. Asystem of canals (c.) separated by 2. Meshes of the fundamental substance (f.s.) which enclose 3. Prolongations of the epithelial cells (pr.) PLATE 1 ZONA PELLUCIDA IN TURTLE EGGS ALICE THING ” > na rete Py oh ABBREVIATIONS a.p., area of proliferation a.s., attraction sphere c., canals c.c., central corpuscle c.n., eytoplasmic network ep., epithelium f.s., fundamental substance 7.l., inner layer 1.2.p., incipient zona pellucida k.e., knob-like enlargements l.b., long bridges o.l., outer layer p.f., polygonal fields p.n., primary network pr., prolongations r.s., radiating striations ., outer stratum ’, middle stratum "inner stratum s.b., short bridges s.n., secondary network t.b., terminal bars y., yolk z.p., zona pellucida $ s Ss The figures are not reduced in reproduction. They are microphotographs taken at a magnification of 750 diameters. Leitz microscope. Obj. 7. Oc. 1. PLATE 2 EXPLANATION OF FIGURES Fig. 9 Transverse section of an egg 1.42 mm. in diameter from the ovary of Clemmys guttatus. Benda. Safranin and picric acid. 5u. The outer layer (o.l.) of the zona has thickened to a greater extent than the inner (7./.)._ The pro- longations show knob-like enlargements at their tips (k.e.) The area of pro- liferation (a.p.) isdeeply stained. Zone measurement 12u. Fig. 10 Transverse section of an egg 2.6 mm. in diameter from the ovary of Chrysemys picta. Bouin. Heidenhain’s haematoxylin. Congored. 4u. The prolongations (pr.) are clearly seen extending from the scant cytoplasm of the epithelial cells down into the outer layer. The zona measures 17y in thickness. Fig. 11 Oblique section of the egg represented in figure 10. Same fixation and staining. 4u. Note the canals and the prolongations of the epithelial cells. Fig. 12 Tangential section of the egg represented in figures 10 and 11. Bouin. Mallory’s stain. 4y. With figures 10 and 11 this shows the great increase in thickness in both layers of the zona (ef. with figure 8). PLATE 2 EGGS nl v ZONA PELLUCIDA IN TURTLE ALICE THING 57 Ki a9 is P ; * ‘ E ; ; vat Ne oii ase AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 15 THE FONTANELLA METOPICA AND ITS REMNANTS IN AN ADULT SKULL ADOLF H. SCHULTZ Carnegie Institution of Washington FIVE FIGURES It is not uncommon to find in the skull of a newborn infant a small fontanelle between the two frontalia in their nasal third. This is usually called fontanella metopica (f. medio-frontalis, fonticulus interfrontalis inferior) (fig. 1). A considerable num- ber of skulls, both of children and of adults, showing short, irregular, transverse or V shaped sutures or fissures in the mid- line of the frontal bone above the level of the superciliary ridges have been described in the literature and interpreted as remnants of the fontanella metopica. The author has found in the skull of an adult an abnormal suture, which is comparable to those above mentioned, but which is more extensive than in any of the cases previously described; accordingly its publication ap- pears justifiable. This specimen (fig. 2) belongs to the Anatomical Department of the Johns Hopkins Medical School and was kindly placed at my disposal by Dr. W. H. Lewis. The skull is that of an American negro, fifty-five years of age. It might be mentioned that the skin over the frontal region was absolutely normal, therefore any external factor, whether acci- dental or surgical (trepanation) can be excluded as the cause of the anomaly. The greatest length of the skull is 193 mm., the greatest breadth 148 mm., the basion-bregma height 128 mm. and the horizontal circumference 553 mm. The weight of the skull, including the mandible, is 985 grams, a figure, which is close to the upper limit of variation of weight for the human skull. This is an indication of the thickness of the bones of the skull, which is characteristic of the negro. Most of the sutures are 259 260 ADOLF H. SCHULTZ Fig. 1 Frontal view of the skull of a male negro fetus with a fontanella metopica. Fig. 2 Frontal view of the skull of a negro with an abnormal suture on the frontal bone. FONTANELLA METOPICA IN AN ADULT SKULL 261 obliterated, on the inner surface more than on the outer. This is also true of the internasal suture, although its course can still be recognized (the suture was retouched in fig. 2), and therefore the right nasal bone is found at its upper end to extend far into the left. On each side, at the incisura parietalis, there is a Wormian bone. The lambdoid suture is rich in Wormian bones. It is noteworthy that there is present on both sides of the man- dible a well pronounced processus anguli mandibulae (apophysis lemurica), which points downward and outward and shows rough outlines for muscle-insertion. The latter are likewise present on the thick zygomatic arch. Attention may also be called at this point to the prominent processus marginalis on the posterior border of the malar bone. The processus anguli mandibulae assumes in our case special interest, in as much as Herpin (’07) reported that this anomaly is rare in the negro and when present is poorly developed. The abnormal suture on the frontal bone, which is situated not exactly median but somewhat to the right, consists on its outer surface of a transverse, irregular, dentate part, 15 mm. long, and of two lateral, ascending limbs, which diverge upward and have a length of 9 and of 13 mm. on the right and on the left respectively. The distance between the upper ends of these diverging limbs is 23mm. The middle of the transverse part is situated 25 mm. above the nasion and 15 mm. below the line con- necting the two tubera frontalia. If, as according to Schwalbe (01) the length of the frontal are is represented as 100, then the transverse portion of the suture lies 20.3 above the na- sion. It is of interest to compare the position of the abnormal suture on the frontal bone in the author’s case with those re- ported by Schwalbe (01), Fischer (’02) and Davida (14). Table 1 isa compilation of the tables of the two first mentioned authors with the corresponding measurements of Davida’s case and of that herein described. The figures show that the suture or fis- sure is always situated below the level of the tubera frontalia, and with only two exceptions always in the upper half of the nasal third of the nasion-bregma are. In the twelve European skulls the average relative distance between the nasion and the 262 ADOLF H. SCHULTZ TABLE 1 Position of the transverse abnormal suture on the frontal bone of adults DISTANCE POSITION OF THE SUTURE|OF THE SUTURE THE FRONTAL LINE IN ARC MILLIMETERS (| European | 32y.| o 1825 17:5 European ad. a 1733 11.0 Schwalbe: .....s. 45.2%... +)|/ Huropean, *| 31 yz1|Fo3 ilgfes 19.0 European | 58 y.| of 20.3 19.5 European |4ly.| o& 13.8 20.0 (| European ad. || o 18.0 18.0 European | 64y.| o& 23.9 19.0 European |40y.| o& 22.2 20.0 European ad. fot 18.5 16.0 LENSING Pagene carci n oro ne re: European ad. 2 18.2 15.0 European | 41 y. Q 20.0 29.0 European | 19 y. Q 17.4 16.0 Negro ad. fot 27.5 12.0 Negro ad. ot 21.8 11.0 DD) BVP Bh, acts, aoe alee ee ? 50 y. t 15.4 Schultze... Skee Negro DOMVe lames 20.3 15.0 suture is 18.8 mm., and the distance between the intertuberal line and the suture 18.3 mm. The average of the two negro skulls of Fischer and that of the author’s case is, for the corre- sponding measurements, 23.2 and 12.6 mm. respectively. There- fore the suture in the negro seems to be relatively higher above the nasion and closer to the intertuberal line than in the white. The exact determination of the position of this abnormal suture is also of importance in the explanation of its origin, as will be seen later. Upon examining the inner surface of the skull, the suture is like- wise found to beextensive (fig.3). Incidentally it might be stated that the crista frontalis interna is only moderately developed, as was the case in the skull with the same anomaly described by Rauber (’03). In two of Schwalbe’s cases the crista frontalis was examined, and found in both to be broad and blunt. For the most part the abnormal suture on theinner surface communi- cates with that on the outer surface, often allowing the passage FONTANELLA METOPICA IN AN ADULT SKULL 263 of a fine bristle. The transverse portion presents itself on the inner surface as eight short perpendicular adjacent fissures, with a total width of 9 mm. and located 19 mm. above the foramen coecum. ‘The lateral limbs of the suture, which also diverge upward, are straight regular fissures, in contrast to those on the tabula externa. The right limb is 13, the left 19 mm. in length and they are 16 mm. apart at their upper ends. The bony part included by the suture is narrower but higher on its inner side than on the outer. On a horizontal section through the frontal Fig. 3 Frontal bone of the skull in figure 2 seen from the inside (upper part sawed off). bone, at a level somewhat above the transverse portion of the abnormal suture, a trapezium is formed by the lateral limbs, with its shorter base directed inward. This wedge-shaped piece of bone is plainly shown in Rauber’s section of a similar case. Schwalbe directed attention to the fact that adult skulls showing remnants of a fontanella metopica present an unusually large interorbital breadth. Table 2 is a compilation of tables by Schwalbe and Fischer with Rauber’s case and that of the author, in which the interorbital breadths and the interorbital indices are given. ‘The interorbital index represents the relation between the interorbital breadth and the internal biorbital 264 ADOLF H. SCHULTZ TABLE 2 Interorbital breadth and interorbital index on adult skulls with the abnormal suture or fissure on the frontal bone AUTHOR RACE SEX ty aa ae mm. European of 28.5 26.6 European of 32.0 30.2 Schiwalibesast.t nck ck ccecte a meee European i! 27.5 PA lB European of 28.0 26.4 _| European Ci 31.0 29.8 European of 31.0 30.1 European a 29.0 26.9 European of 29.0 26.6 ischer... hese: ss eee European Q 24.0 25.8 European Q 30.0 29.4 Negro of 26.0 26.3 Negro a 37.0 33.3 Rae rs Ss ib eeaee tay ic ene 2 ae es European of 30.5 Solas voc tats) sleiecg oe eet tee Negro fot 32.0 29.1 breadth; the technique of these measurements may be found in Schwalbe’s studies on Pithecanthropus erectus (99). In this same work are published similar measurements of a considerable number of normal skulls of most heterogeneous races. Accord- ing to these measurements, the interorbital breadth varies be- tween 18 and 31 mm. with an average of 24.2 mm., the inter- orbital index lies between 20 and 30.1 with an average of 24.3. A comparison of these figures with table 2 shows that both the absolute and relative interorbital breadths of skulls showing remnants of a metopical fontanelle are much above the average. For the determination of the relative frequency of the anomaly in the two sexes, the material at our disposal has been much en- larged through the published cases of transverse fissures in the frontal bone of children and newborns. In Schwalbe’s cases sex is stated in 9 juvenile and in 5 adult; all were male except one newborn. Fischer found in 1 newborn and in 7 adults, in which sex was known, that the female sex was represented twice. FONTANELLA METOPICA IN AN ADULT SKULL 265 Adding to these the case of Rauber and that of the author, both of which were males, the total of males is 21, of females 3. The anomaly, therefore, would appear to be of much greater frequency in males. This same preponderance has been found by the author (16) in another anomaly, namely, the persistent canalis cranio pharyngeus, and this relatively greater frequency has been likewise shown in respect to other anomalies. From this it would seem probable that anomalies are more common in the male, but whether this is a rule for progressive or for atavistic anomalies, or for both, can only be determined when care is taken by investigators to always mention the sex in reporting anomalies. Short transverse sutures or fissures occurring in the lower third of the frontal are in adults have always been interpreted by the various authors as remnants of the fontanella metopica, but the origin of the latter has been explained in widely different ways. The metopic fontanelle was first described by Gerdy in 1837. He was followed by Hamy and the Italian scientists Maggi, Riccardi, Staderini and Zanotti. Of these, Hamy (’72) sees in the metopical fontanelle a divergence of the lines of ossification of the tubera frontalia. Maggi (’94,’98, 799) inter- prets the fontanella metopica as a product of the approximation of the four frontalia media. These assumptions are based upon his isolated comparative anatomical observations. Zanotti (01) explains the medio-frontal fontanelle as the last trace of the foramen, which corresponds to the location of the paraphysis in primitive vertebrates; in other words, a foramen frontale for the paraphysis similar to the foramen parietale for the epiphysis. Both Maggi and Zanotti to a certain extent place atavistic in- terpretations upon the fontanelle, but these must be considered as extremely hypothetical. Bolk (11) was led to believe that the fontanella metopica arises at the site of the primitive or primary nasofrontal suture. This opinion was based upon observations on monkeys, in which the nasal bones have become shortened, that is the supramaxil- lary portion of the nasalia is displaced by a medial growth of the frontalia, by which process a secondary naso-frontal suture, situated closer to the apertura nasalis, is formed. This theory 266 ADOLF H. SCHULTZ does not explain in a satisfactory manner the extremely rare oc- currence of a true metopic fontanelle in monkeys, together with the relatively frequent appearance of incomplete nasal reduction. On the other hand, the relative frequency of the metopic fon- tanelle in man according to Schwalbe is 15.2 per cent in children up to 14 years, whereas high reaching nasal bones, such as are found in monkeys, have never been described in the human skull. Moreover, it must be borne in mind that the remnants of the fontanella metopica are often situated in the adult high above the nasion. Asshown in table 1, the lowest point of the remnants of the fontanelle is located as much as 27.5 per cent of the frontal are above the nasion, its middle point, being even higher. If the fontanelle really corresponds to the original uppermost end of the nasalia, then the latter must have extended between the frontalia high above the orbits and the superciliary ridges. Bolk assumes that the supranasal portion of the frontal suture (supra- nasal field or triangle)—a frequent finding in adults—is the result of the reduction of the nasalia. However, this supranasal suture reaches as arule only slightly above the glabella and not, as Bolk supposes, to the level at which the fontanella metopica occurs. Rauber (’06) describes the skull of a child with two fontanelles at the frontal suture (fonticulus interfrontalis superior et inferior) which in his opinion had become separated from the frontal arm of the anterior fontanelle. The fonticulus interfrontalis inferior corresponds to the metopic fontanelle, and as a factor in its remaining patent Rauber considers it possible that the site of the anterior neuropore of the medullary canal of vertebrates exerts its influence under special circumstances, even to the ossification of the skull. Schwalbe (01) in contrast to the explanations offered by pre- vious authors, considers it possible that the metopic fontanelle is to be conceived as a progressive variation, which bears a rela- tion to the greater development of the frontal lobe of the cere- brum. The adult skull described in this paper would seem to support this theory inasmuch as its capacity was 1520 cc. and its smallest frontal width was 109 mm. Both these measurements are rather large for the negro; on the other hand Fischer’s cases FONTANELLA METOPICA IN AN ADULT SKULL 267 showed the metopic fontanelle to be present in two idiots, one of them a microcephalus with a skull capacity of only 704 ce. Schwalbe in his explanation makes use of the hypothetical sup- position that the tubera frontalia might consist of two adjacent ossification centers, which usually join immediately, but in ex- ceptional cases remain separate, later forming two independent systems of lines of ossification. The divergence of these lines forms the metopic fontanelle, which in children is situated on a plane with the tubera frontalia. Schwalbe emphasizes the fact that the metopic fontanelle and its derivatives are always found at a definite location, while the fontanelles and fontanelle bones which are found at times in the upper portion of the frontal suture have a more variable situation and are to be included in the great fontanelle. Schwalbe cites among other the cases described by Staderini, in which the fontanella metopica is connected with the great fontanelle by a wide space. In spite of this, however, he makes a distinction between the two above mentioned fontanelles, which rests purely upon the situation of the metopic fontanelle. According to Schwalbe in children up to 138 months the latter varies in respect to the lower end of the fontanelle from 5.6 to 17.8, in respect to its middle point from 11.2 to 22 per cent of the frontal arc above the nasion. Fischer described the skulls of two children in which interfrontal fontanelle bones are divided in two and in three part’ respectively. In one of these the mid- dle point of the fontanelle bone was situated 30.6 in the other 50 per cent of the frontal are above the nasion. It is evident that the position of the metopic fontanelle is not as definite as claimed by Schwalbe, who makes the following statement: In the rare cases in which two or even three groups of Wormian bones occur in the frontal suture, only the lowest corresponds to the normal medio-frontal fontanelle; those situated near the parietal bones, how- ever, are to be considered as Wormian bones in an abnormally wide suture (hydrocephalus). The latter may even represent the anterior end of the large fontanelle, which has extended abnormally far into the frontal region. It sometimes occurs that the anterior end remains open for a longer period than that portion lying directly posteriorly; there- fore the anterior end may become separated as a secondary fontanelle. 268 ADOLF H. SCHULTZ This distinction of Schwalbe seems somewhat arbitrary, in- asmuch as all transitions can be observed in juvenile skulls. On the basis of original observations the author is convinced that the metopic fontanelle is derived from the bregmatic fontanelle, and at some time has become separated from it. Figure 4 gives the best proof. An interfrontal suture wide at its upper part, that is, a very long arm of the great fontanelle, as shown in num- bers 1, 2, 3 and 4 in figure 4 is not of rare occurrence. Among 35 skulls of infants up to a few months’ old the frontal arm of the great fontanelle was found to extend six times to within 10 to 17 T.White 7m. intrauter. 6. Negro aa 8. White 4m. A. Fig. 4 Normae frontales of frontal bones of juvenile skulls with a long arm of the bregmatic fontanelle, which has been constricted in the lower cases to form a metopic fontanelle. FONTANELLA METOPICA IN AN ADULT SKULL 269 mm. of the nasion. In three other cases the great fontanelle reached within 22 mm. of the nasion. This prolonged arm of the great fontanelle is an extreme variation, and is not necessarily a result of hydrocephalus. In the skull of a year old hydrocephalic negro, the author found the great fontanelle reaching to within 16 min. of the nasion; in contrast to the cases in figure 4, however, it was, even at its lower end, 17 mm. wide; in the middle of the frontal are 29 mm. and at its upper end 35 mm. _ It is striking that the lowest portions of the frontal bones always approximate Fig. 5 Norma verticalis of a skull of erethizon dorsatus with fontanelle bones. each other and indeed to a height which is considered typical for the position of the metopic fontanelle, that is, to a point to which the frontal arm of the great fontanella may extend un- interrupted or constricted. As a designation for the lowest portion of such long bregmatic fontanelles extending into the nasal third of the frontal arc, the name fontanella metopica may well be retained. However, no fundamental difference is to be made between the two mentioned fontanelles. It is more fre- quent for the lower end alone to remain patent in children and THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 270 ADOLF H. SCHULTZ to be recognizable in adults. The constriction of different por- tions of the frontal arm of the great fontanelle results from locally decreased or increased growth of the lines of ossification, and may occur in any situation, but appears to be most common between the two tubera frontalia. Double constriction to form secondary fontanelles has also been described (Rauber ’06). This identity of the metopic and the great fontanelle is also demonstrated by the position of the fontanelle bones, which occur anywhere from the bregma to the upper portion of the nasal third of the frontal are (Hartmann 1869, Barclay-Smith ’09 and 10, Gulliver 1890). Whether the above described case of partial persistence of the metopic fontanelle in an adult was associated with a fontanelle bone can not be determined with certainty, but seems probable, especially upon examining the inner surface. Before any definite statements can be made as to the cause of the occurrence and partial persistence of a long frontal arm of the great fontanelle, more material must be available, and at- tention must be paid to correlations, especially in the frontal region. The author hopes by this contribution to stimulate interest in this anomaly in order that further cases may be re- ported. Observations on the occurrence of fontanelle structures in the frontal bones of mammals have been reported in a limited number, and further cases would be of great value. Among 10 skulls of erethizon dorsatus, which the author collected recently, 3 cases presented paired symmetrical fontanelle bones extending far between the frontalia. Figure 5 shows one of these cases. FONTANELLA METOPICA IN AN ADULT SKULL Zi BIBLIOGRAPHY Barcuay-SmitTH, E. 1909 __ supra. clav. suprascap. lat. ant. thor. i . thor. dors. subscap! ol vaws = 9 a ; med. br. cuta. ‘ eS \ x med. ant. thor. | . ' ax. subscap. . med. antibr. cuta. Fig. 3 From the right side of a white male, age about 35 years, reversed. Group 3, Type F. Fig. 4 From the right side of a white female, age 95 years, reversed. Group1l, Type A. 382 to phren. N Gs ie : suprascap. C.6 fat. ant. thor. Cay Gs Th.2 med. br. cuta:-~” te med. ant. thor.” med. antibr. cuta. subscap. ‘ y thor. dors: a ax. subscap. = supra. Clav. C.4 = — — ~= \ S . > to phren. C.s — subclav LNs ITS subscap- “Ss SS =‘ suprascap SS SS LFA ’ , 0 AS =~ ==! (ss. Sa lat. ant. thor. med. ait. thor. = 8 thor. dors ax. subscap! Fig. 5 From the left side of a colored female, age 14 years. Fig. 6 From the left side of a white male, age about 40 years. Type C 383 : med. br. cuta. med. antibr. cuta. mus. cuta. subscap oa a SS Sao Se. un. Group 1, Type A. Group 1, _—— C.7 a= WU! // hE WAC thor. dors. Ce Gabe _fad. ’ WS WN SS ‘~ . cuta \ mus. cut AS \N £ i : SS KS med. antibr. cuta. SN SA y. Y \ Sy \S ‘ AAS NEN S SA, eM Sina ax. subscap. SA ve; INS A \ ~ ~-.uln. ___ supra. clav. subclav. _..Suprascap. lat. ant. thor. med. ant. thor: thor. dors.” med. antibr. cuta. Fig. 7 From the left side of a white male. Group 2, Type D. Fig. 8 From the left side of a colored male, age 25 to 30 years. Group 3, Type G. 384 BRACHIAL PLEXUS OF NERVES IN MAN 389 CG. () 9 tosubscapular = — median Fig.9 Showing the connections and interlacings of the funiculi of a plexus as they ap- peared after the connective tissue was removed by maceration. _ 386 ABRAM T. KERR C.4 C.s suprascap. Gx ‘ lat. ant. thor. C1 Cus e mus. cuta. That med. br. cuta. ~S med. antibr. cutay subscap.” 1 ~-med. thor. dors: ax. subscap. C4 to phren_ Cs : subclav, suprascap. lat. ant. thor. Cs 2 thor. dors. Ca re subscap. Cs 7 ‘: Tha subscap. ~ 2nd intercosto. br. ( med. br. cuta..!- de 3rd intercosto, br. -- f: med. antibr. cuta. Fig. 10 From the left side of a white male, age 35 years. Group 2, Type F. Fig. 11 From the left side of a colored female, Group 1, Type B. BRACHIAL PLEXUS OF NERVES IN MAN 387 C.4 12 suprascap. CNG tophren.. - - -- lat. ant. thor. C.7 "lat. head uln. C.3 Th.q aise subscap. - -~~ _-- tmed-br. cuta. med. ant. thor-~ ig ale 2nd intercosto. br! e thor. dors..’ med. antibr. cuta, C.4 4 t — supra. Clav. SS > aes \ Ne subclav. 1 3 iN Boe C.s suprascap. at lat. ant. thor. C.6¢ Zz : == lat. head uln. C.7 : C.38 eS Th. 2 ‘Ist intercosto. br. to phren 2nd intercosto. br.-.-f = ; med. br. cuta.” med, ant. thor: | subscap: -- - ~~ ax. subseap. --*\ Suln, Rhoridors4 med. antibr. cuta. Fig. 12 From the right side of a colored female, age 25 to 30 years, reversed. Group 1, Type A. Fig. 13 From the right side of a colored female, age 30 years, reversed. Group 1, Type B. 388 ABRAM T. KERR C.s suprascap. lat. ant. thor. 14 C7 C.s cor. br. Th.z é : med. ant. thor. \ or Sg ait : % . tus. cuta. med. antibr. cuta. ; Nes a P: med. thor. dors: subscap: ns Ni: = ~ ulin: ax. subscap_ C.s ¢ suprascap. ‘ <= lat. ant. thor. : 7 SS, phe és subscap. 15 mS med. ant. thor: subscap: thor. dors: rad.” med. br. cuta. med. antibr. cuta. ~— uln. Fig. 14 From the left side of a white female, age 61 years. Group 2, Type E. Fig. 15 From the right side of a white male, age 60 years, reversed. Group 2, Type E. BRACHIAL PLEXUS OF NERVES IN MAN 389 cs subclav. : suprascap. 20 lat. ant. thor. C.7 cor. br. ax. C.8 Lb ; mus. cuta. med. ant. thor. S es subscap. { , med. br. cuta: thor./dors. -med. ax. subscap: med. antibr. cuta. -y--.uln. a C.+ ‘ \ to phren. CxS we suprascap. c. 6 ! ___tat. ant. thor. rad. lat. head uln. cor. br. if cal “mus. cuta. Th. 2 subscap.-” cuta 1st intercosto. br...._.__ med. br. cuta..---- >> 5:~ med. 2nd intercosto. bra.+~.--! : med. antibr. cuta..../ ~~uln. ax. subscap. thor. dors: Fig. 16 From the right side of a white male, age 55 years, reversed. Group 2, Type E. Fig. 17 From the left side of a colored male, age 20 to 25 years. Group 1, Type B. 390 ABRAM T. KERR Gs to phren. 1 8 < suprascap. = = ___ Jat. ant. thor. C.6 a ax. J ! rad. ? C.7 ATEN S) SS 1.2 cor, br. =a — ‘ . eae SY > : ae es lat. head uln. wa ¥ . oe C.8 = : . —_—__ a mus. cuta. Th.t med. ant. thor. ~ subscap.-— NN thor. dors. 2nd intercosto. br... .. e / med. ms : ax. subscap: ln! C.4 cap. Sunes P: lat. ant. thor. Cis ra 19 C.e Cait Cig Th. 2 med. ant. thor,” med. br. cuta: subscap:’ 2nd intercosto. br... SX : thor. dors.” s ; ax. siiserap med. antibr. cuta. Fig. 18 From the right side of a colored male, age 20 to 25 years, reversed. Group 1, Type B. Fig. 19 From the right side of a white male, age 55 to 60 years, reversed. Group 1, Type B. BRACHIAL PLEXUS OF NERVES IN MAN 391 4 e to phren. aN 20 suprascap. C.6 __lat. ant. thor. C.7 i C.8 Th. Bae rh : No mus. cuta. SS mi oS . . subscap..-’ NS \ thor. dors..-~ x 5 2 -45--. med. ax. subscap.- * 35) Bede br cutlass = Neg : -uln. med. antibr. cuta-~ aN Ci4 1G _—_ Supra, Clay. r \ to phren. C.5 SS a5 subclav. 1 suprascap. Gee lat. ant. thor. Caz rad. C.é mus. cuta. Th. + NS 2 med. br. cuta.. .- --- -—- Xe subscap- thor. dors..° ax. subscap. LSS med. med. ant. thor. uln. med. antibr. cuta. Fig. 20 From the right side of a white male, age about 40 years, reversed. Group 1, Type B. Fig. 21 From the left side of a white male, age about 70 years. Group 2, Type F. 392 ABRAM T. KERR (hs suprascap. . , lat. ant. thor. 22 ES x ax. at g rad. C.s . —— cuta. Th. ~ NS Cor. br. \ SS med. ant. thor. \ N med. br. cuta..-” SS thor. dors..." x ~,med. ‘ ax. subscap. fea uln. med. antibr. cuta. C.¢ es S suprascap. ae lat. ant. thor. 23 C.6 C.1 C.a mus. cuta. Th. + ax. subscap. uln. Fig. 22 From the right side of a white male, age 38 years, reversed. Group 2, Type E. Fig. 23 From the right side of a colored male, age 45 years, reversed. Group 1, Type B. BRACHIAL PLEXUS OF NERVES IN MAN 393 to phren. q suprascap. Subclav. ; s } lat. ant. thor. 24 mus. cuta. med. ant. thor..” NSS NE oD SS SS subscap.*2~ NN RWS med. br. cuta.-~~ thor. dors. |. i? a » med. med. antibr. cuta. ax. subscap. ve ~uln. to phren. “J Subctav. ; “ suprascap. Sr lat. ant. thor. ; ax. \ \ Pe irad: > --=yCOF, br. med. ant. thor.‘ : e mus. cula. subscap.’ ‘ a ax. subscap.” ‘ thor. dors.‘ ~-med. subscap!: 2nd intercosto. br’ med. br. cuta.. med. antibr. ‘cuta, Fig. 24 From the right side of a colored male, age about 25 years, reversed. Group 1, Type B. Fig. 25 From the right side of a colored male, age 35 years, reversed. Group 2, Type E. 394 ABRAM T. KERR ‘ to phren. C.s Sees 26° suprascap. a lat. ant. thor. C.7 rad. “= aX: eae < mus. cuta. Th. 2 med. ant. thor. y subscap....../ ~ YN ¢ 2nd intercosto. br. eS = thor. dors.” D med. med. br. cuta: x * ‘2. + S ~Juln. med. antibr. cuta. ax. subscap. C.4 subclav. to phren. . : o suprascap. 2h Gis = --....lat. ant. thor. Cc. . Cir C.s Th. + med. ant. thor. ....\ subscap- mus. cuta. thor. dors.-” med. med. antibr. cuta. ~uln. Fig. 26 From the right side of a colored male, reversed. Group 2, Type E. Fig. 27 From the right side of a colored male, age 40 years, reversed. Group 1, Type C. Fig. 28 Fig. 29 BRACHIAL PLEXUS OF NERVES IN MAN 395 C.4 : to phren. 28 C: suprascap. C6 Cay C.s mus. cuta. Th.: rE é subscap. 2 med. ax. subscap: ; thor. dors. ny ula. med. br. cuta. A med. antibr. cuta. e~--) \ subclav. Ste ceph. Le / suprascap. 29 lat. ant. thor. "Jat. fase. "post. fase. LF = / to med, ant. thor. SSeS : subscap. -- ~~ med, br. cuta. -- SN SS Nat head uln. 2nd intercosto. br.» = Se: : 5). thor. dors:-- med. fasc. N Sy -- med ax. subscap. ' Sait med. antibr. cuta. From the left side of a colored male, age 40 years. Group 1, Type C. A composite, typical plexus. in) 9) . ioe AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY » THE BIBLIOGRAPHIC SERVICE, DECEMBER 29 ON THE AGE OF HUMAN EMBRYOS FRANKLIN P. MALL Johns Hopkins Medical School, Baltimore, Md. TWO FIGURES AND EIGHT TABLES In the Manual of Human Embryology, published seven years ago, I presented the evidence by which we may determine the age of an embryo or fetus, in my chapter dealing with this sub- ject.!. It was there pointed out that the best check in arranging embryos in time sequence is obtained from our knowledge of comparative embryology; also, that the only factor which can be depended upon in every case is what I then termed the ‘menstrual age;’ that is, the age of the embryo as computed by the time elapsing between the beginning of the last menstrual period and the date of the abortion. In order to procure a satisfactory curve of growth for the whole period of gestation, I succeeded in collecting about 1000 specimens from the different months of pregnancy with the data given concerning them; namely, the measurements of the embryos and the dates of menstruation and of abortion. It was also necessary to establish standard measurements for the embryos, chief of which are sitting-height and standing-height; 1 Mall, Franklin P. 1910 Determination of the age of human embryos and fetuses. Manual of Human Embryology, Chap. 8. Edited by Franz Keibel and Franklin P. Mall, Philadelphia; German edition, Leipzig, 1910. 1903 See also Note on the collection of human embryos in the Anatomical Laboratory of the Johns Hopkins University. Johns Hopkins Hospital Bulletin, vol. 14. In the second paper I gave a formula by which the age of embryos up to 100 mm. long could be determined. That is, to multiply the CR length in millimeters by 100 and extract the square root; the product is the ageindays. I wish to state that this formula gives the age according to the His convention, which I now believe to be incorrect, as demonstrated in my chapter in the Manual. This con- clusion was also reached independently by Bryce and Teacher. A fairly complete bibliography is to be found in the papers by Mall (1), Bryce and Teacher (3), Triepel (2) and Grosser (6). : 397 THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 398 FRANKLIN P. MALL these are known respectively as crown-rump (CR) and crown- heel (CH). Tables were prepared by which the average measure- ments of the’ one for a given stage could be converted into the average measurements of the other; for it is well known that embryologists are given to using the crown-rump measurement for smaller specimens, while anthropologists and obstetricians generally use the crown-heel measurement for larger specimens. MENSTRUATION AGE My tabulation of the menstrual age was made as follows:. All the measurements of the embryos and fetuses were converted into crown-rump or sitting-height measurements. These were then used as ordinates, while the menstrual ages were used as abscissae; in other words, each specimen was entered upon a chart in which the menstrual age and the sitting-height together made a co-ordinate. In this way the 1000 specimens were spread over a millimeter chart, 500 mm. high and 350 mm. wide. It was found that the individual records arranged themselves along a path about 20 mm. wide at the base line, and about 40 mm. wide toward the upper margin of the chart. In addition to this central zone containing most of the records there were numerous scattered entries far out of line. These were especially numerous at the bottom of the chart, which would indicate that in early abortions there is an undue number of poor records; or, at least, records showing greater irregularity in the menstrual periods. In order to determine a mean menstrual age the chart was marked square by square in such a way that exactly one- half of the records were circumscribed by two lines enclosing the usual or normal cases, leaving one quarter of the scattered records to the right of one line, the other quarter to the left of the other line. ‘The first group includes those specimens which grew very slowly and may have been pathological; the second, those cases in which menstruation continued after pregnancy. The two lines which include the middle group are practically parallel, beginning about 20 mm. apart, around the records of the early specimens, and ending about 40 or 50 mm. apart around the specimens from the latter part of pregnancy. The distance AGE OF HUMAN EMBRYOS 399 between these two lines was then divided exactly, and it is this line which marks the mean menstrual age of embryos throughout pregnancy. In a general way it is reproduced as the line CH in figure 145 in the Manual of Human Embryology (p. 200). I have spoken of the age thus determined at different times as the menstrual age, or more properly speaking the mean men- strual age, because there is a very large probable variation. For instance, a number selected at random from the table on page 199 of the Manual, with a mean menstrual age of 51 days, would also show a probable deviation of from 40 to 62 days. Such embryos have a height of 11 mm.; therefore, when we obtain embryos of this length, we may expect that one-half of them have a menstrual age of from 40 to 62 days; in other words, in small specimens there is a probable variation of three weeks. Viewed from another angle, one-half of the embryos with a menstrual age of 51 days would range from 4 to 25 mm., while the average would be 11 mm.; hence, we are probably dealing with a pretty large error which cannot be definitely located. At present it would appear that pregnancy may begin at any time during the intermenstrual period, but it is difficult to determine the most probable time. What I published in the Manual has received careful criticism from Triepel,? but he nevertheless also accepts the term menstrual age, and recommends that we use it in the future. . COPULATION AGE After constructing the curve and table referred to above, showing the mean menstrual age, I entered as the probable or true age a line in the curve and a column in the table which fall in a position exactly ten days earlier than the mean menstrual age. This was done for the following reasons: According to the more recent statistics of Issmer, that writer found the average duration of pregnancy in 1220 cases to be 280 days, when esti- mated from the first day of the last menstrual period; and in 628 cases, 269 days when estimated from fruitful copulation. In general these figures correspond with those of Ahlfeld, Hecker 2 Triepel, A. 1915 Alterbestimmung bei menschlichen Embryonen. Anat. Anz., Bd. 46, 1914. Also Bd. 48. 400 FRANKLIN P. MALL and Hasler, who collected about 500 cases in which the date of fruitful copulation was given. Therefore, in a group of 1200 eases the duration of pregnancy. when reckoned from the last menstrual period, was fully ten days longer than when computed from the time of copulation; and it seems to me that in order to determine the true age it is necessary to deduct these ten days from the menstrual age. Even then I believe we should be care- ful not to use the word true, since the time of copulation does not necessarily record the time of fertilization. For this reason it might be well if we introduced the term, copulation age, to distinguish it from menstrual age, and from two other ages I am about to give. These could be termed ovulation age and fertilization age, the latter being the only true age, since we must always figure the beginning of development from the time of fertilization. The curve in the chart, which I gave in my publication, as representing the true age, but which I now will speak of as the copulation age, was constructed from cases of newborn children, and is probably the more valuable because it eliminates all of the irregularities of early pregnancy which accompany natural abortion. After the curve was completed, however, we received into our collection a few embryos, measuring less than 25 mm., the accompanying records of which gave the time of copulation as well as of menstruation. The copulation ages of these specimens were then entered upon the chart shown in the Manual with stars (fig. 147), and curiously enough nearly all of them fall exactly upon the line of the curve, showing that what was as- sumed to be a difference at the end of pregnancy is also indicated again in specimens from the beginning of pregnancy. In both cases the difference between the menstrual age and the copula- tion age is about 10 days. When this chart was made it con- tained seven stars, but after it had been sent to the printer I found another case in the literature. Also, about the same time I received a copy of a book published by Bryce and Teacher,’ which gave a second case, and these were added to the curve. - 3 Bryce and Teacher 1908 Contributions to the study of the early develop- ment and imbedding of the human ovum. Glasgow. AGE OF HUMAN EMBRYOS AO] To my great surprise and pleasure I found that these authors had reached a conclusion similar to mine; namely, that the age of young embryos is no longer to be computed according to the convention of His. They not only give a detailed and excellent account of their own specimen, but also reconsider all other cases of young specimens in relation to their age, which have been published by well-known writers. They assume that the copula- tion age is probably very nearly the true age of embryos, and that henceforth we will have to consider the question from this standpoint. According to Bryce and Teacher, it is now generally admitted that the menstrual cycle in man and monkeys is homologous with the oestrus cycle of the lower mammals. The oestrus cycle is divided by Heape into pro-oestrum, oestrus and dioestrum, and this division has been confirmed for many mammals by his own researches and those of F. H. A. Marshall. During pro- oestrum the generative organs of the female show signs of special activity, such as swelling of the vulva, coloration or flushing of the surroundings, and a discharge of blood or mucus from the vagina. This is immediately followed by the ‘oestrus,’ or ‘period of desire,’ during which alone the female is capable of impregna- tion and will receive the male. If pregnancy does not occur, oestrus, after a brief space in which desire subsides (metoestrum), is succeeded by a period of quiescence or dioestrum, which lasts till pro-oestrum again sets in. In polyoestrous mammals several cycles of this kind may follow one another. Menstruation in the human female is homologous with pro-oestrum, as first pointed out by Heape. Though there is no fixed ‘period. of desire’ there is an indication that a vestige of this persists, in the fact that a phase of more pronounced oestrus commonly succeeds the cessation of menstruation. ‘This view is confirmed by our records, for we frequently hear from a patient that a fruitful copulation occurred shortly after the menstrual period; and it may be that this opinion records also the rupture of the Graafian 402 FRANKLIN P. MALL follicle. According to J. G. Clark‘ this is accompanied by vascu- lar hyperemia of the ovary, and the possibility of a spasm of the ovary is not to be excluded, for there is an abundance of muscle in this organ which no doubt has a function to perform. The following histories include all cases from our collection in which the copulation history is given. I have also added the ~ Watt® case because it is the only one I have been able to find in the literature since the publication of the Manual. I have included all cases because I think it is best not to select those which suit my convenience in making a curve, but to give the poor material together with the good. A few of the records are sufficiently complete to be unimpeachable; the remainder are given for what they are worth. No. 1399 (Dr. H. N. Mateer, Wooster, Ohio.) Embryo, GL 1mm. Chorion 10x9mm. From a hysterectomy. Copulation September 19 and September 27. Operation, October 19. (I have been unable to find out date of last period, but it is probably recorded.) Copulation age 22 or 30 days. If the former is taken, it matches the curve exactly. No. 779 (Dr.—————, Baltimore.) Embryo, GL 2.75 mm. The specimen though otherwise normal was later found to have spina bifida. It came from the physician’s wife. She is 37 years old, and is the mother of one child and this is her first abortion. She is very anxious to have children. Last period, August 29 to September 2. Abortion, October 12. Fruitful copula- tion, in the woman’s opinion, September 25 and later. She does not state specifically that copulation occurred between September 2 and September 25. Menstrual age, 44 days. Copulation age, 17 days or less. Doubtful case. 4Clark, J. G. 1899 The origin, growth and fate of the corpus luteum as observed in the ovary of the pig and man. Johns Hopkins Hospital Reports, vol. 7. 1900 The origin, development and degeneration of the blood-vessels of the human ovary. Johns Hopkins Hospital Reports, vol. 9. ’ Watt. J. B. 1915 Description of two twin human embryos with 17 to 19 paired somites. Contributions to Embryology, vol. 2, Carnegie Institution of Washington, Publication No. 222, AGE OF HUMAN EMBRYOS 403 Watt’s case (Dr. Watt, Toronto.) Twin embryos, GL 2.75 and 3.55 mm. Mother, a German Jewess, 30 years old, robust and healthy, four children and this one abortion. Last period, December 3 to 6, 1907; first copulation December 20. Slight flow January 3, with similar flow on January 11, 12 and 13, abortion following on January 14. Menstruation age, 42 days. Copulation age, 25 days or less. No. 1182 b (Dr. C. E. Caswell, Wichita, Kansas.) Woman aged 27, four living children and one abortion. Husband has syphilis and so has one child. Mother seems to have escaped. Last period, March 25 to April 4. Abortion, May 10. Mother is sure that conception took place April 14. Menstrual age 46 days. Copulation age, 26 days. Doubtful case. No. 470 (Dr. H. C. Ellis, Elkton, Md.) Embryo, CR 4mm. Chorion, 20 x 13 mm. Mother, 24 years old, two healthy children. Abortion during an attack of mumps with very high fever. Last period October 5, 1910, and copulation about October 15. Abortion November 9. Menstrual age, 35 days; copulation age not over 25 days. No. 588 (Dr. G. L. Wilkins, Baltimore.) Embryo, CR 4mm. Last period January 26 to February 3. Had no intercourse with husband for several weeks prior to this and only three or four days after period but not later. Abortion March 16, 1912. She has had two healthy children, 14 and 20 years old respectively, and not less than eleven abortions. Dr. Wilkins believes that all the abortions were induced. Menstrual age, 50 days. Copulation age, 38 or 39 days. No 1507 (Dr. C. B. Ingraham, Denver, Colorado.) Macerated embryo, GL 4 mm. A Jewess who last menstruated May 7 to 11; abortion June 22. Woman had opportunity to become pregnant shortly after this period or again just before the next. Menstrual age, 46 days; copulation age, 40 days or 17 days. Record not satisfactory, especially since specimen s also pathological. 404 FRANKLIN P. MALL No. 208 (Dr. J. Y. Dale, Lamont, Pa.) Normal embryo, CR 7mm.,GL 8mm. The specimen was enclosed in an almond-shaped ovum, measuring 22 x 11 x 11 mm., and there, was considerable magma within the exocoelom. The specimen came from a married woman whose last period began on December 28, 1901, and who had coitus only twice, January 5 and January 7, between this period and the time of abortion, February 15, 1902. Dr. Dale informs me that the data are entirely reliable, as both the woman and her husband are thoroughly trustworthy. The specimen was secured for me by Prof. John G. Clark of the University of Pennsylvania, who thought that its unique history gave it greater value. Menstrual age, 49 days; copulation age, 39 to 41 days. No. 1461 (Dr. H. A. Wright, Seattle, Washington.) Embryo, CR 9. 8 mm. Mensirual age, 28 days; copulation, 27 days. Data inaccurate and incomplete. Ne ot a reliable case. ’ No. 443 (Dr. William Grant, Baltimore.) Embryo about 10.5 mm. long. The specimen was sent at the suggestion of Prof. T. S. Cullen on account of its interesting history. Menstrual age, 27 days and copula- tion age, 28 days. On account of the manner in which the history was given, and because of the degree of development of the embryo, the data can hardly be admitted as correct. The husband had been away from home for four months prior to the time of coitus, which was on the last day of menstruation. The woman is the mother of four healthy children and menstruates regularly every 28 days. The patient was reluctant to show the specimen to the physician, and both she and her family defended her character, a fact which would seem still further to convict her. For this reason the record is not to be considered reliable so far as the age of the embryo is concerned. No. 167 (Dr. A. H. Ritter, Brooklyn, NoYes) Embryo, CR 14. 5 mm., NL 13.5 mm. The normal embryo w as sent in a beautiful normal ovum measuring 30 x 30 x 20 mm. The specimen came from a multipara whose last period was from November 26 to December 2, 1899. First copulation ry AGE OF HUMAN EMBRYOS A405 after the period on December 15. Due toa surgical operation on Janu- ary 24 there was continuous hemorrhage until January 3 when the ovum was passed. In the event that conception took place after the last period, this specimen could not be more than 46 days old. Men- strual age, 65 days; copulation age, 46 days. No. 1390 Dr. G. N. J. Sommer, Trenton, N. J.) Embryo CR 18 mm. Last period, December 18 to 22. Operation for tuba! pregnancy, February 10. Conception took place on December 25, as the woman was in the habit of using preventive means and same were not used on Christmas eve. Menstrual age, 54 days; copulation age, 47 days. Reliable case. No. 1584 (Dr. F. H. Church, Bonnville, N. Y.) Embryo,CR18mm. Chorion 35 x 31 x 25 mm. Unmarried woman, age 21, first pregnancy. Last period August 15; menstruation regular, every 28 days. Criminal abortion, October 10. Coitus, September 13 and 14. Menstrual age, 56 days; copulation age, 26 and 27 days. Not reliable. No. 26. (Dr. C. E. Simon, Baltimore.) Fetus, CR 25 mm., CH 30 mm. This specimen, which was some- what injured and therefore difficult to measure with precision, was brought to me by Dr. Simon, February 25 (?) 1894 with the following history. The mother, an unmarried woman 27 years old, was a serv- ant in Dr. Simon’s family, and had but recently come from Germany. She remained at home continually until New Year’s eve, when she went to a ball and remained out all night. Her last -period took place on December 12 and lasted six days. During the night of December 31 she was with her lover and the abortion followed on February 25. On January 16, after missing her January period, she took a cupful of mustard powder with the hope that it would produce abortion, but . instead it yearly killed her. On January 21, she recovered, and re- sumed her household duties. (See record of the case by Simon, N. Y. Med. Jour., March 17, 1894.) Later she fell into the hands of an abortionist and the embryo came away during the night of February 24. Dr. Simon assured me at the time of the abortion that it was absolutely impossible for the pregnancy to have taken place at any time excepting the night of December 31. Menstrual age, 75 days; copulation age, 56 days. Reliable case. 406 FRANKLIN P. MALL No. 616 (Dr. S. P. Warren, Portland, Me.) Embryo, CR 26 mm. Unmarried woman, 18 years old; last period began August 9 and continued 7 days. Coitus three times within 9 days after the last period. Curetted after 24 hours of pain, October 13, 1912. | ; Menstrual age, 65 days; copulation age, 56 days (?). Not reliable. No. 1535 (Dr. Philip F. Williams, Philadelphia). Embryo, CR 28 mm. Chorion, 50 x 45 x 15 mm. Unmarried woman, 20 years old; first pregnancy. Last period, May 3 to 7. Abortion, July 6. Last coitus, May 10 (?). Menstrual age, 62 days; copulation age, 55 days. Doubtful age, but it falls close to the time of the probable age. No. 373 (Specimen loaned by Prof. Simon H. Gage, Ithaca, N. Y., Cornell Collection, Homo No. 11.) Embryo, CR 31 mm. Last period, May 9. Conception, May 21. Natural miscarriage July 17, after 2 to 3 days bleeding. No other data. According to these records the menstrual age is 69 days and the copulation age 57 days. No. 849 (Dr. Shipley, Baltimore.) Embryo, CR 52.5 mm. Mother, white, unmarried, age 20. Last period, December 11 to 15, 1913. Abortion, March 3. Coitus from which mother dates preg- nancy occurred just after the cessation of the last menstrual period in December, but she admits that she had frequent intercourse previous to this period. Menstrual age, 82 days; copulation age, 77 days (?) Doubtful case. (Dr. G. C. McCormick, Sparrows Point, Md.) Embryo, CR 62mm. _ End of last period, January 1, 1912. Coitus, January 7; abortion, March 29. Menstrual age, 93 days; copulation age, not over 82 days. No. 1635 (Dr. Henry Leaman, Philadelphia.) Embryo, CR 70.5 mm. Mother, 40 years old, 9 children. Last AGE OF HUMAN EMBRYOS 407 period, August 26 to 30; abortion, November 29. Copulation but ‘once about September 3 to 15. Self induced abortion. Menstrual age, 95 days; copulation age, 75 to 87 days. Records contradictory. No. 322 (Dr. West, Bellaire, Ohio.) Embryo, CR 85-90 mm. The specimen is probably from an induced abortion. The mother says that fruitful coitus took place on June 17 and the abortion on September 20, copulation age 95 days. No. 1295 ¢ (Dr. L. J. Commiskey, Brooklyn, N. Y.) Embryo, CR 87. Woman, 43 years old, mother of one child and this is her second abortion. Last period, April 24 to 28; abortion, July 27. Woman states with great certainty that the productive coitus took place either May 4 or 6. Menstrual age, 94 days; copulation age, 84 or 82 days. Doubtful case. No. 1310 (Dr. B. G. Pool, Washington, D. C.) Embryo, CR 95 mm. First pregnancy of unmarried woman, 18 years old. Last period, July 25 to 30, 1915; abortion, November 6. Said to be from a single coitus on August 9. Menstrual age, 104 days; copulation age, 89 days. Doubtful case. No. 894 (Dr. E. L. Mortimer, Baltimore, Md.) Embryo, CR 121 mm. White mother, age 29, three children and two abortions. Last period, July 24 to 28; criminal abortion, November 21. Husband works on a boat and returned home August 1. Menstrual age, 120 days; copulation age, not over 112 days. No. 142 (Dr. G. H. Hocking, Govans, Md.) Embryo, CR 142 mm. Mother, age 43, has five children. Menstruated May 29 to June 9; abortion, October 5 after several weeks’ flow. Woman says pregnancy could not have taken place before June 18. Menstrual age, 129 days; copulation age, 109 days. Doubtful case. 408 FRANKLIN P. MALL The summary of these cases together with all others of the same kind which I have been able to gather from the literature, is given in table 1. This is an elaboration of the table given in the Manual. The data are sufficiently complete so that those who choose may look up the original records. Most of them, however, will be found in abstract form in the articles by Triepel and by Grosser. The specimens in the Carnegie Collection are recorded above. All of the specimens just given are entered upon figure 1. The mean menstrual age and the mean copulation age are taken from the data given in the table and in the curve published in the Manual. For the specimens here considered the menstrual ages are indicated by means of dots, the copulation ages by large solid circles. The time is calculated by days, and the measure- ments of the embryos are crown-rump. The numbers of several specimens for which the copulation age is given are marked in figure I; for instance, No. 448 and No. 1310. One of the Rabl cases is also indicated. I am of the opinion that all these marked records should really be excluded from the figure as they do not appear to be very reliable. However, I have included them for the sake of completion. Six of the copulation cases are crossed in the figure with an X, and are given again in table 2. It will be noticed that these six records fall almost exactly upon the curve given. They are, I believe, the only ones which are entirely reliable; that is, they record embryos which are the product of single copulations, and for this reason their maximum ages are established. A word regarding specimen No. 26, which is recorded i in the literature as representing an embryo 30 mm. | long. As we are at present dealing with CR measurements, this should be 25 mm. It appears on the chart in the Manual as 30 mm., for the reason that the curve was constructed on the basis of the standing height, or CH length of the embryos. The tables given by Triepel and by Grosser should, therefore, have this measurement corrected accordingly. I have also entered upon my figure the ages of the embryos 6 Grosser, O. 1914 Alterbestimmung junger menschlichen Embryonen; Ovulations und Menstruationstermin. Anat. Anz., Bd. 47. TABLE 1 MEN- POSSIBLE TIME OF COPULATION IN EE Sa Sey | DAYS BEFORE ABORTION OO mm. days Embryo 0.15 38 | Exactly 16 days Bryce-Teacher, 1908 Ovum DEOESORS 42 | 20 days before and earlier Reichert, 1873 Embryo 1.0 22 and 30 days before No. 1399 2s (2) 38 | 19 days (Delaporte) See Grosser _ Anat. Anz. xlvii, 1914 183 34.| Exactly 21 days Eternod, Anat. Anz. xv, 1899 1.5(?) 35 | 14 days Fetzer, Anat. Anz. Bree, Hit. xxxvil, 1910 eG 44 | 17 days and later No. 779 2.75 42 | 25 days and later (Watt, Carnegie Con- tributions to Em- 3.33 42 | (Twin) bryology, ii, 1915 3 46} 26 days (?) No. 1182b. 3.0% 48 | 40 days and later His, AME., vol. 2, 1882 4.0 35 |) 25 days (?) No. 470 4.0 50 | 38 days No. 588 4.0 46 | 40 days and 17 days (?) No. 1507 6.0 50 | 40 days and later Kollmann’s Atlas, 1907 @ 49 | 39 and 41 days No. 208 7.75 57 | 45 days and later His 8.8 42 | Exactly 38 days Tandler, Anat. Anz., xxi, 1907 9.8 28 | 27 days (?) No. 1461 10 60 | 49 days and earlier His 10.5 27_| 22 days (?) No. 4438 11 55 | 31 days (?) Rahl, Entwickl d. Gesicht 13.6 63 | 53 days and later His 14 65 | Exactly 47 days Rahl 14.5 65 | 46 days and later No. 167 18 ‘54 | Exactly 47 days No. 1390 18 56 | 26 or 27 days (?) No. 1584 25 75 | Exactly 56 days No. 26 26 65 | 56 days No. 616 28 62 | 55 days (?) No. 1535 31 69 | 57 days No. 373 5205 82 | 77 days (?) No. 849 62 93.| Not over 82 days No. 591 70.5 957| 75 or 87 days (?) | No. 1635 85 (2?) 95 days No. 322 87 94 | 82 or 84 days (?) No. 1295¢ 95 104 | 89 days (?) No. 1310 121 120 | Not over 112 days No. 984 142 129 | Not over 109 days No. 1284 409 410 FRANKLIN P. MALL according to their degree of development as given by Triepel in order to show that he has practically adopted the curve of de- velopment given by Bryce and Teacher and also by myself. He has really taken what I have designated as the copulation age, minus about two days for each stage, assuming, as do also pees beet seeeaet att i im T “ips - ‘i 1310 et 90 : 1295¢ | 80 : 70 | +H COPULATION AGENGHt EEA PELE i i HH 60 +E HEH t - | : iH i ue 50 Pe Hnaaicoatet i HEH { gpebeaG tee 1 HT ETT srr 40 CTF) FERTILIZATION AGE fi E4111 UHHUG BRYCE-TEACHER Hii t Eee AND TH Act : x t ee + t th ~ = : hae i} | 77 MENSTRUATION AGE HEE - i E E E a 20 HET aro ep Zr py dca eeseuguce geegGE 444 }. iy i : : : Aeeesecacahad ope ated edcabapan ataatetest ataset cL HH a HH if i TE 30 | 40/50) ei) je 70.60") (90) tea) | nas 4 5 6 7 8 9 10 i 12 13 14 I5 WEEKS Fig. 1 Menstruation age and copulation age taken from the curve constructed by me and published in the Manual of Human Embryology. All embryos are entered with CR length. IT have also added for the sake of comparison the curve giving the convention of His. NL and CR give the neck-rump and crown-rump lengths respectively, according to His. The fertilization age is according to Bryce and Teacher for smaller embryos, and according to Triepel for larger ones. The dots record the menstrual age of the embryos under consideration, and the squares the copulation age. The crossed squares mark the best records, as men- tioned in the text. It may be noted again that the curves are not constructed from.these records, but the records are entered to test the curve. ,AGE OF HUMAN EMBRYOS 411 TABLE 2 AUTHOR ~ ES ie sels COPULATION AGE mm. days 1 BY NAS a0 (0 bal DCH) 01) Sa A A 0.15 16 TEs NSU AYOYG | OA Sve oe AO ee ES Se eM OAS 1.3 21 TMD TO eet Sabie: cee nein: Etec ot olor a Ee ReItaT 8.8 38 LIONS? WAG copies fe Se a a Ue a ce 18 s a 14.0 44 LOOKS ters cs Givers erm cere iaae wateiwi alae & ace 18.0 47 IN Ds: AD ete cits Gye RENCE EEC? Coos CECA ate eee NER 25.0 56 Bryce and Teacher, that there is this interval of two days between copulation and fertilization. For the sake of completion the curve giving the His convention is also included in the figure. I wish again to emphasize the fact that the curves given in the figure are not constructed from the records of the specimens in question, and it is quite clear, I think, that the new cases give no reason for materially altering the mean copulation curve as given by me in the Manual seven years ago. The relation of these curves to the ovulation age and to the fertilization age remains to be established, and as far as the evidence will permit this will be done in the following paragraphs. OVULATION AGE The question of the time of ovulation in relation-to menstrua- tion or to copulation is by no means answered, although the literature upon the subject is extensive. If the time of ovulation could be definitely determined we would then be able to ascertain the ages of embryos with very fair precision. Wherever possible we have collected ovaries with our specimens, but so far have obtained only one accompanying a young ovum. This speci- men, No. 970 in our collection, is from a Filipino girl, 16 years old, who died four days after taking hydrochloric acid with sui- cidal intent on account of her condition. The ovum, which measures 5 x 3 mm., is not quite normal in appearance but is well implanted. The corpus luteum is well formed, and solid, with no remnant of blood within it. The Herzog’ specimen, which 7Herzog. 1909 Acontribution to our knowledge of the earliest known stages of placentation and embryonic development inman. Am. Jour. Anat., vol. 9. 412 FRANKLIN P. MALL is also from a Filipino woman who was killed in an accident, likewise had a small ovum measuring 2.3: x 1.2 mm., well im- planted in the uterus. In this case the corpus luteum was ‘fresh but closed.’ The well known Reichert specimen which measured 5.5 x 3.3 mm., with a copulation age of 16 days or more, has in one ovary a well-formed corpus luteum, 20 x 17.5 mm. which has within it a small cavity containing some blood. Finally, Johnstone’ describes and pictures the corpus luteum of an ovum almost the size of Peters’ specimen, which measures 13 x 10 mm. Its center is occupied by a large mass of pale, finely granular material which stained pink with eosin. The periphery is com- posed of a layer of lutein cells bordered on the inside by a layer of red blood corpuscles. The lutein layer, which is 8 x 10 cells deep, is crinkled, owing to papillary ingrowths of connective tissue. There is a great deal of vacuolation of the lutein cells and the whole layer is quite vascular. The specimen came from a woman, aged 29, who died suddenly, not having missed a period, nor was it suspected that she was pregnant. A step in advance on the study of the structure of the corpus luteum was made by R. Meyer? in his excellent paper on the subject. He classifies its development into four stages as follows: 1. Proliferation or early hyperemic stage of the Graafian follicle with transformation of the granular cells into lutein cells. 2. Early hyperemic stage of the corpus luteum with beginning transformation into the second stage of granular metamor- phosis. The blood-vessels now permeate the layer of lutein cells. 3. Mature or blossoming stage of the corpus luteum. 4. Stage of involution. Sometimes when the follicle ruptures it simply collapses, and hemorrhage does not always take place within it. The speci- mens studied by Meyer were increased in number and reported 8 Johnstone, R.W. 1914 Contribution to the study of the early human ovum. Journal of Obstetrics and Gynaecology of the British Empire. ® Meyer, R. 1911 Ueber corpus luteum-Bildung beim Menschen. Archiv fiir Gynaekologie, Bd. 93, 1911. AGE OF HUMAN EMBRYOS 413 in relation to the menstrual cycle by Ruge II!° who gives the following data: TABLE 3 TIME OF OCCURRENCE nA TAPeCTEHE | og NON on IPTG SERGIO, ats Ge NI cea ee hl AO ea ee 10 1 to 14th day Wisigrensl Ee) 8 eodcotue cite o Aah oben Oo Otte Te eae 10 10 to 16th day RAL RU SRE, 5 ee de des eens RENN enc cccheMO cid ott CR Sree 44 16 to 28th day ra POUionNGT, Sleek cen pieces Gee Oe se Teneo 18 1 to 13th day TABLE 4 ie wenn or | THE OF cecumnenee SEBCUTENS TO MENSTRUATION PRION, Bs acemde aud Gee e abou docue ome: % 10 1 to 14th day WiIS CUI Laren een aint tre tengo Birt tor ge REM a mr Me hy, 10 10 to 16th day IAT RUOT EES Vogts II ee ape en 8 Say te ek tas Ee ee aS Ne a Oe 44 16 to 28th day ADirreyie CLG UO MME es aise eee rece IRN he SIRs oe ee 18 1 to 13th day Ovulation occurred in the stage of proliferation, and always during the first 14 days of the period. However, this stage does not form a regular sequence of development during the first two weeks, but the specimens were of unequal development and could not be arranged in the order of time. It is impossible to determine the time elapsing between ovulation and the formation of the third stage of mature corpus luteum, but Meyer and Ruge believe that always a number of days must intervene. Finally, the stage of involution overlaps that of proliferation. At any rate the work of Meyer and Ruge demonstrates that the fresh corpus luteum as described by Fraenkel'! appeared a number of days before he thought it did, thus completely overthrowing Triepel’s assumption that the probable time of ovulation is on the 19th day. According to Ruge, it occurs sometime during the first 14 days of the menstrual month which supports the theory I am advocating. These are all the reliable data I have been able to collect 10 Ruge II, Carl 1913 Ueber ovulation, corpus luteum and menstruation. Archiv fiir Gynaekologie, Bd. 100. 11 Fraenkel: Archiv fiir Gynaekologie, Bd. 91. THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 414 FRANKLIN P. MALL regarding the time of development of the corpora. lutea in human beings. I had thought that it would be possible to extend the subject somewhat further if the corpus luteum in the pig could be standardized in relation to the size of the embryo found in the uterus. This work was carried through by Corner, but unfor- tunately does not include the earlier stages of the corpus luteum, and it is just these data that we need if we are to determine accurately the age of freshly ruptured Graafian vesicles. Corner!” made a careful study of the histological changes in the corpus luteum of the sow for all but the earliest stages of pregnancy. He finds that the corpus luteum is already solid at 20 days, this stage being reached earlier, he believes, than in human beings where this central cavity remains longer. By the aid of refined cytological methods he recognizes seven distinct stages during pregnancy as follows: TABLE 5 STAGES } LENGTH OF EMBRYOS APPROXIMATE AGE days le breparatory periodte. 7.. eeee cers Less than 20 mm.|} (?) 25 2. Exoplasmic development.............. (1) 20-380 _ 25-30 3. Exoplasmic development............... (II) 30-55 30-40 AS AUrANSULOLY: DCTLO Cee es aaa eae eeion 55-140 40-75 5. Endoplasmic development............. (I) 140-170 75-105 6. Endoplasmic development............. (II) 170-220 105-110 (i MRETLOLTESSION : sae ey ee ee en eee 220-290 110 to term Although this study cannot be transferred to the human directly, it at any rate suggests that the latter may be standard- ized. It is hoped to establish at least a relation between the early stages of the corpus luteum and the size of the ovum and embryo; and that in the course of time the age of this body may be estimated with precision. — It may be noted that Corner showed definitely that the size of the embryo found in the uterus could be estimated with con- 12 Corner, George W. 1915 The corpus luteum of pregnancy; as it is in swine. Contributions to Embryology, vol. 2, Publication No. 223, Carnegie Institution of Washington. AGE OF HUMAN EMBRYOS 415 siderable accuracy by the cytological condition of the lutein cells; however, all his specimens were from corpora lutea pre- sumably a little older than the human ones mentioned above. In a measure we may fill in the gap in the earlier stages from the report by Sobotta! on the development of the corpus luteum in the mouse. He found in this study that during the first 24 hours after ovulation the cavity of the follicle fills with serous fluid or blood, at the time the lutein cells become cut up into com- partments by the formation of connective tissue septi. This process continues during the following day or two, and finally the central cavity is nearly obliterated, containing, however, a central mucoid nucleus at the middle of the third day after ovulation. The irregular summary from the several species is about as follows: (1) In the mouse the central cavity of the corpus luteum is obliterated about the middle of the third day after ovulation; (2) it is obliterated in human specimens accompanying ova about the size of those studied by Bryce and Teacher, and by Peters; and (3), it is obliterated in the pig considerably before the 25th day. It may also be noted that Corner states that the corpus luteum of menstruation is of irregular shape in its develop- ment, while that of pregnancy is uniform and even. He speaks of the former as if the cells were arranged like a mob, and the latter as if organized like an army. Finally, a few words regarding Fraenkel’s studies, out of which Triepel has made so much capital. According to Fraenkel, Villemin in 39 operations found no freshly ruptured follicles in the first two weeks after the menstrual period, but observed many from 12 to 14 days before it. Fraenkel himself describes hemor- rhagic follicles as follows: Very fresh, fresh, quite fresh and not very fresh, showing that his average of 19 days after the last menstrual period is not the average time of ovulation, but the average of older corpora lutea in several stages of development. From a study of Fraenkel’s papers it may be seen quite clearly that what he reports as fresh corpora lutea are by no means 13 Sobotta. 1896 Ueber die Bildung des corpus luteum beider Maus. Archiv fiir Mik. Anat., Bd. 47. 416 FRANKLIN P. MALL necessarily fresh, but may possibly vary in age fully a week. In fact he intimates that they are not all fresh, and Triepel makes a slight allowance for this reason. These papers have been care- fully analyzed by Grosser, who finally reached the- conclusion that ovulation does not take place on the 19th, but at the latest on the 16th day after the beginning of menstruation. This figure is not so very far from the average given in my curve; in fact it is a little more than the average age accepted by Triepel as the normal according to the degree of development of the embryo. ‘Triepel has attempted to force a curve which runs exactly 12 days after the average menstrual age of specimens, into one which should be exactly 19 days after this curve, in order to fit Fraenkel’s opinion regarding the proper time of ovula- tion. This of course is an impossible feat. The conclusion to be drawn, therefore, is that we cannot possibly establish a satisfactory ovulation age of embryos from the data now at our disposal; but I believe that we have material within our reach whereby we may eventually be able to determine with greater certainty the probable time of ovulation. Before this can be done with the human, however, it will be necessary to study anew the degree of development of the corpus luteum for various days after menstruation, with new material selected from cases which are otherwise normal. This can be done in any large gynecological clinic. FERTILIZATION AGE According to Bryce and Teacher, the comparative infrequency of pregnancy during continuous cohabitation points to some special circumstance connected with successful impregnation. This circumstance would appear to be simultaneous ovulation and limited power of fertilization on the part of the spermatozoa. As regards the former the work of J. G. Clark is of interest. According to this writer ovulation is accompanied with hypere- mia of the ovary, and, he believes, is hastened by it. He in- jected the blood vessels of an ovary in which there were fresh corpora lutea, as well as swollen Graafian follicles, and found that AGE OF HUMAN EMBRYOS 417 the injected fluid immediately ran out of the ruptured follicle. In a few instances the fluid entered mature follicles, causing them to become dense and finally to rupture when the vascular pressure was continued for a sufficiently long period. This suggests at least that a factor in fertilization is the rupture of a Graafian vesicle, due to orgastic reaction in the uterus, tubes and ovaries when copulation takes place immediately after menstruation. At this time ovulation is most likely to oceur in lower animals, and all the facts indicate that the same is true in human beings. It is known that in the rabbit, dog and pig there must be repeated copulation in order to insure impregnation. A single mating rarely suffices. Thus, for instance, according to Weysse,!* only three out of the nine sows became pregnant after being covered but a single time. This would indicate that the fertilization power of the sperm was of short duration, as Bryce and Teacher seem to think is the case in human beings. According to Waldeyer! live spermatozoa were found in the bitch eight days after copulation, and dead cells, that is motion- less cells at the end of 17 days. Living moving spermatozoa were found in a woman three days after death. Living sperm cells were found in the Fallopian tube of a patient 9 days after admission to the hospital and 33 weeks after copulation. On the other hand, spermatozoa have been found upon the surface of the ovary of the rabbit and sow two hours after copulation. In Waldeyer’s opinion the power to fertilize remains as long as the sperm cells retain normal motility, and there are no facts to deny that human sperm has the power to fertilize over a week after copulation. Spermatic cells of animals that emit them into water die in a very short time if they are greatly diluted, and have a much longer life if only a little water is added. Thus in fertilizing trout eggs ‘dry’ sperm is used, while if the sperm is added to water containing the eggs but few eggs are fertilized. This 14 Weysse, Arthur Wisswald 1894 The blastodermic vesicle sus scrofa domes- ticus. Proc. Amer. Acad. Arts and Sciences, vol. 30. 15 Waldeyer, W. 1906 Hertwig’s Handbuch der Vergleich. und Exper. Entwickelungslehre der Wirbeltiere. Bd. 1, Tl. 1, Erste Halfte. 418 FRANKLIN P. MALL question has been tested recently in Arbacia by F. R. Lillie,1 who makes the following interesting statements: The sperma- tozoa are absolutely immobile while they are in the body of the male, but become intensely active when suspended in sea-water. They then become relatively inactive, but can be restored again by the addition of fresh sea-water. When greatly diluted they lose their fertilizing power completely in about an hour, and when diluted by 250,000 times their volume in water this power lasts but a few minutes. The loss of fertilizing power cannot be due to a loss of motility, for long after the former occurs no loss of vitality or motion is observed. In man the secretion of the prostate gland maintains the motility of spermatozoa much more effectively than does normal saline solution, and it is said that the secretions of the mucous membrane of the uterus and tubes have a similar influence. Thus it would seem that when motility is accelerated it does not indicate that the power to fertilize is prolonged, as asserted by Waldeyer. Lillie’s experiments certainly do not favor such a view, and Bryce and Teacher infer the same when they state that were the spermatozoa to retain for a long time their power of fertilization, no ovum could escape fertilization. For the sake of argument Bryce and Teacher deduct 24 hours from the copulation age of their specimen (163 days) .and esti- mate that it would have been 154 days old had it lived up to the time of abortion. This seems to me to be reasonable, as are the other statements in their admirable paper. In view of the difference between the fertilization power of spermatozoa and their motility, as expressed in Lillie’s report, we may admit with considerable safety that the fertilization power of sperm is of shorter duration than is the power on the part of the egg to be fertilized. Furthermore, the theory that a fruitful copulation should be accompanied by ovulation at about the same time is a necessary one, in order to account for all of the combinations which are encountered in human beings. Nor is the assumption of Bryce and Teacher of an oestrus following 16 Lillie, F. R. 1915 Analysis of variation in the fertilizing power of sperm suspensions of Arbacia. Biol. Bull., vol. 28. AGE OF HUMAN EMBRYOS 419 menstruation untenable, and the possibility of a relation between orgastic reaction and ovulation is not to be overlooked. An interesting study in this connection has been made by Siegel,!7 using the wives of German soldiers as his subjects. These women, who became pregnant during their husband’s furlough, came to the maternity hospital to be confined, and DAYS % t 2 8 4 OS 78 8) WO a ie = 19) 16) 17, 1& eee | WEEE hese fa a INTERMENSTRUUM aeccuon tite Si meR 2 MENSTRUATION] POSTMENSTRUUM| PoE Ee HEH + PE Se ee ele 4 rae ali Heer ale) ane Sea aaa ees Pe Ere Vie WAN SPEC EEC EEE ES cae Sa Gennes See a er isn Alcea ate NTNU Soa a a [2 Palade. a as MINTS [Te me a He | nee as SS ee a HEH | HAS EE |S pa] (al a (ica a bo eat ;, «|e a | | te ee 2 ree if | wer tele 2 A ee ae | Set} }4-}-+4 Jide Beebe ee ae VE Awe SS Fig. 2 Cohabitation curve according to Siegel. The main division of the menstrual month and the probable time of ovulation are given. One hundred cases of pregnancy, occurring in the wives of soldiers after their husbands’ fur- lough of one week. Each day of the furlough is entered as a possible day of con- ception. In all probability the 1st to the 4th day and the 18th to the 21st day belong to the sterile portion of the month. it was easy to obtain records of the menstrual history as well as the times of furlough, which in each case was of about a week’s duration. Figure 2, taken from Siegel’s paper, gives the result of the tabulation of 100 of such cases. Each day of the furlough is entered in the curve. Thus, if the furlough lasted from the 17 Siegel, P. W. 1915 Warum ist der Beischlaf befruchtend? Deut. Med. Woch., 41. 420 FRANKLIN P. MALL Sth to 16th day of the menstrual cycle it was entered for each of these days. It is noteworthy that there was no entry for the last seven days of the menstrual month, indicating that pregnancy did not take place either a week before this nor within the week following; that is, there is a sterile period of about 18 days and a fertile period from the end of menstruation to the 15th day, which includes the probable time of ovulation. Of course only those cases which came to the maternity hospital could be recorded, and it is interesting to note that none of the 100 pregnancies dated from a furlough during the last week of the menstrual month. Such did not end in conception. Siegel was able to gather 10 cases in which the husband was on furlough a few days before the menstrual period, and in none did pregnancy follow. He cites further cases gathered by Wohler from the records of the same maternity hospital for the past ten years. These included 160 pregnancies among newly married women, in whom con- ception had occurred during the first five weeks after marriage. Among this group there were 65 cases in which marriage took place within the eight days preceding the menstrual period, and in each of them one more menstruation followed, which fact alone would indicate a sterile condition during the week pre- ceding it. The records of Siegel, although not entirely satis- factory, demonstrate quite conclusively that the most probable time for conception is during the week or ten days after the period of menstruation. From what has been written above we may, for the sake of argument, accept one day as the average time between copulation and fertilization. The time at which this is most likely to occur is between the 4th and 13th day after the first day of menstrua- tion, as shown by the following table. This table is compiled from the records of our own eases, given above, each datum being obtained by subtracting the copulation age from the menstruation . age. 7acn ooo eee aioe 429 AREA mleswAw andere rls to seth AB ARC AUS ES 0 es Be Re eee 431 aeHeartisilhouetteranea and: body welghtia..5.---aoscade eee eee te 431 b. Silhouette area and transverse diameter..............:%.:...-.....- 445 Cupeleartawelehiteamd so Odivsswiele Meme sete nts 2 ee ea a 449 Cap bleantevolumes sje RLS pn aabs CREE ces eo): 2 ok ete ee eee 465 enaVienbtricularyout puter. hiss oka cence sels Se esclelclts eh ete eee Ok ee 476 f. Relation of size of heart to height, weight and sex.................. 481 The x-rays are of value in the study of the relations, the shape, the action and the size of the heart. We shall treat here of methods of determining the size of the heart and the relation of the size of the heart to the size of the body. Of all the organs the heart is probably normally the most closely related in size to the size of the body as a whole. It is well known that a noticeably enlarged heart usually means some lesion either of the heart itself or of the blood vessels. Under- sized hearts have been less studied but the more accurate methods of studying now being developed in x-ray technique show that it is of clinical importance to know when a heart is disproportion- ately small as well as when it is disproportionately large. 1. THE HEART SILHOUETTE Of the various methods which have been devised for the study of the size of the heart those which have proved of greatest value are the orthodiagraphic and the teleroentgenographic. Ortho- 423 424 Cc. R. BARDEEN diagraphy has the advantage of giving a graphic outline which theoretically at least, corresponds exactly in size to the contour of the object casting the shadow and it is economical in material, but it takes much time and skill to exercise and is subject to errors when a moving organ like the heart is studied. Tele- roentgenography with our modern machines is quick and accurate but demands that a proper allowance be made for enlargement of the heart silhouette due to divergence of rays. Fortunately this is relatively simple when the distance from the target to the plate is the usual two meters and the patient faces the plate. ‘The average distance of the heart contour from the front of the chest is approximately one-third of the distance from the front to the back of the thorax measured at the lower part of the sternum during expiration. Albers-Schénberg (08) has shown that the greatest transverse diameter of the heart lies in a plane parallel to the front of the thorax and at a distance of about one-third of the distance from the front to the back of the thorax at the level of the 6th thoracic vertebra. I have been’ able to confirm this observation by studies on numerous cadavers and on cross sections of the trunk and also to show that the average distance of the contour of the heart which casts the outline of the heart silhouette in parallel dorso-ventral rays is about the same distance from a plane parallel to the front of the chest. The contour of the apex is of course nearer the plate than the contours of the right atrium and the left atrium (W. Guttmann, 06) but we are concerned with the average distance of the heart contour from the plate. I have substantiated these studies on the cadaver by means of stereoscopic methods and half-distance methods of study of the distance of the heart contour from the plate in the living. Knowing the average distance of the heart contour from the plate it is possible to calculate the percentage of reduction which one must make of the heart silhouette in order to get the true size of the heart contour. In round numbers the silhouette area must be decreased one per cent for each three centimeters of distance from the front to back of the chest and a given diameter one per cent for each six centimeters. As a DETERMINATION OF SIZE OF HEART BY X-RAYS 425 routine for an adult of average size six per cent reduction of the silhouette area or three per cent reduction of a given diameter will give the actual size of the heart contour with sufficient ac- curacy to obviate the necessity of measuring the antero-posterior diameter of the chest and making a special calculation. But for very large or very small individuals and for children the simple formula given above should be followed.! 1 When a shorter distance from the target to the plate than two meters is used, allowance must be made not only for this variation in distance but also for a variation in distance of the heart contour from the plate which enters in as the target is brought nearer to the heart. This factor also enters in at the two meter distance in ventro-dorsal pictures, the heart contour being further from the front of the chest in the dorso-ventral than in the ventro-dorsal position. For determining the average distance of the heart contour from the plate the stereoscopic method gives the best results. This is based on the distance of cor- responding points in two silhouettes from a fixed line on the plate perpendicular to the line of shift of the tube. The average distance of a series of such points on the contour gives the average distance of the contour. Knowing the distance of the target from the plate, the length of the shift of the tube and the distance of the shift of a given point in the two silhouettes it is easy to calculate the distance from the plate of the point on the heart which casts this Ae B where 4 x = distance of point on heart contour from the plate A = distance of shift of given point in the two silhouettes B = distance of shift of tube C = distance from the target to the plate. The formula for determining the relation of the size of the area of the sil- houette to that of the area enclosed by the heart contour is as follows: pe 100 B? A? x = area enclosed by the heart contour 100 = area of silhouette A? = square of the distance from the target to the plate 2 = square of the distance from the target to the heart contour The formula for determining the relation of a given diameter of the heart silhouette to a given diameter of the heart is as follows: a 100 B A x = length of diameter of heart contour 100 = length of diameter of silhouette A = distance from the target to the plate B = distance from the target to the heart contour point of shadow. The formula is x = ll 426 Cc. R. BARDEEN 2. POSITION OF THE BODY IN RADIOGRAPHY OF THE HEART In the orthodiagraphie studies the position of the patient is determined by the convenience of the operator and patient and general physiological considerations. It makes no particular difference whether the tube is behind or in front of the patient. In teleroentgenography it is important to have the heart as near the plate as possible and hence as a rule the patient should face the plate. It is difficult to place the tube two meters from the plate when the patient is in the supine position. The prone position is inconvenient and somewhat unnatural. A sitting or standing position as a rule is more convenient and comfortable. I have found the best position for routine work to be the sitting position in which the patient leans slightly forward against a plate holder with an inclination of 20° from the vertical. The ventral surface of the gladiolus of the sternum should be approxi- mately parallel with the surface of the x-ray plate. This position In the two-distances method of estimating the size of the heart two pictures are taken one with the target a shorter distance from the plate than the other. The most convenient distances are one and two meters. When these distances are chosen the following formulae may be used: (1) _ 200 A — 2004 < 2A-—a x = distance of plane of the heart contour from the plate A = square root of area of heart silhouette in picture taken at one meter a = square root of area of heart silhouette in picture taken at two meters. If desired a given diameter may be used instead of the square root of the area. (2) A.a x= —— 2A-—a x = diameter of the heart A = diameter of the heart silhouette at one meter a = diameter of the heart silhouette at two meters (3) a? (200-x)? 200? = area enclosed by heart contour = area of silhouette at two meters 200—x = distance from the target to the plane of the heart contour. oo. x= te 8 i) =] DETERMINATION OF SIZE OF HEART BY X-RAYS ADT is comfortable and throws the heart forward toward the plate. The tube is raised high enough to direct the central rays per- pendicular to the center of the plate. These central rays pass approximately through the tenth thoracie vertebra. Asa routine the pictures are taken during deep but not forced inspiration and with two half second exposures’ with an ‘intervening half second so as to insure a diastolic outline. For special studies we have taken pictures in the prone and standing positions as well as in the sitting position, during inspiration as well as during expira- tion. We have also taken instantaneous pictures timed by special electrical devices at any desired period of the cardiac and re- spiratory cycles. The studies of Moritz, Dietlen (09), and others have shown that as a rule the heart is larger in the supine than in the sitting position and in the sitting than in the standing position. Ac- cording to Dietlen the difference as a rule is more marked in young healthy individuals than in those with less healthy hearts. For normal individuals he gives the average percentage difference in area of heart silhouette between the supine and standing positions as 20 per cent of the supine area with the extremes at 30.4 per cent and 10.6 per cent. For those with slight lesions he found an average difference of 12.8 per cent, those with marked lesions a difference of 9.5 per cent and in cases of recent decompensated hearts a difference of only 5.6 per cent. On the other hand in some cases of acute dilatation he found variations of from 23.3 per cent to 34.6 per cent. Dietlen’s figures for normal individuals seem somewhat high. (Cf. Otten, ’11-"12.) In eight normal individuals, taking the pictures during deep inspiration, I found an average difference in area between the prone position and the sitting position of 6.7 per cent of the prone area (extremes 2.8 per cent and 12.1 per cent) and between the prone and standing positions of 15.3 per cent (extremes 2.4 per cent and 17.4 per cent). In nine other individuals of whom ‘instantanedus’ pictures were taken in the prone and sitting positions during quiet respiration at the height of diastole I found an average difference of 4.7 per cent of the prone area (extremes 0 per cent to 10.9 per cent). Veith’s 428 Cc. R. BARDEEN studies of twenty-three boys in the supine and sitting positions (08) show an average difference of 7.6 per cent between the su- pine and the sitting positions (extremes 2.6 per cent to 22.7 per cent). For the average normal individual we may therefore take 5 to 7 per cent as a conservative estimate of the reduction in area which we may expect in the heart shadow area between pictures taken in the prone and sitting positions when the patient leans slightly forward in the latter position. These changes in size of the heart associated with change in posture are due chiefly to changes in hydrostatic pressure in the inferior vena cava. To a large extent they may be overcome by binding the lower extremities. As a rule but not always the pulse rate is faster in postures in which the heart is relatively small. The effects of the respiration on the size of the heart seem to be less constant. According to F. M. Groedel (711) the heart does not as a rule change in size during quiet respiration al- though there is a slight fall of blood pressure during inspiration. In forced inspiration there is a marked fall of blood pressure which may be followed by a passive rising of the diaphragm and a rise of blood pressure. At the height of deep inspiration the heart, or at least the heart shadow, may be smaller than normal owing to the pull of the pericardium. In three experiments in the sitting position I found a decrease in the area of the heart silhouette in expiration as compared with inspiration in two instances (—2 per cent, —2.5 per cent), an increase in one in- stance (+5 per cent), average +0.2 per cent. In the prone position I found a decrease in the cardiac area during expiration in two instances (—5 per cent, +6 per cent) and an increase in one instance, +4 per cent), average —2.3 per cent. In the standing position I found an increase during expiration of 6.6 per cent and 9 per cent in two instances, average +7.8 per cent. From these few experiments it would appear that in the prone position, when the heart is relatively large in size, it tends to be smaller during expiration, while in the standing position, in which the heart is relatively small, the heart tends to be larger during expiration than during deep inspiration. We need a much more extended series of observations before this can be DETERMINATION OF SIZE OF HEART BY X-RAYS 429 considered definitely determined. In the position we have chosen as a standard it is probable that a moderately deep in- spiration increases the diastolic filling to a slight extent over that in expiration. The negative pressure produced in the thorax by the inspiration tends to fill the heart during diastole, while no marked restraint is exercised by the pericardium. 3. MEASUREMENT OF THE HEART SILHOUETTE Methods of measuring the heart silhouette vary. As a rule the right and left margins of the heart silhouette are clearly defined while above the heart silhouette merges with that of the great vessels and vertibral column and below with that of the diaphragm, liver and stomach. The apex of the heart is most clearly defined when there is a well marked gas accumulation in the stomach. Sometimes a Seidlitz powder is given a patient in order to insure a well marked gas bubble in the stomach. The pressure of an excessive amount of gas may, however, to some extent distort the picture. As a rule if the patient is given a glass or two of water immediately before the picture is taken and is requested to swallow as much air as possible with the water a gas bubble of sufficient size will be formed in the stomach to aid in outlining the heart. But gas in the stomach does not serve to make a clear demarcation between the shadow of the heart and that of the liver. To complete the lower margin of the heart shadow it is necessary to draw a line to connect the outline of the left margin with that of the right margin of the heart. With practice it becomes possible to draw this line with fair approximation to its true position. As one gets used to visualiz- ing the heart one learns to continue the swing of the line from the right side into that from the left side. I have drawn many hearts in position in the dissecting room using an apparatus which enables me to draw a line perpendicularly above the margin of the heart. By comparing these drawings with those made from x-ray plates it becomes evident that the approximately correct completion of the lower margin of the heart outline is less difficult than one might expect. It is more difficult in fat than in thin individuals. THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO 2 430 Cc. R. BARDEEN At the base of the heart a purely arbitrary line must be drawn since there is no simple line of demarcation between the heart and the great vessels. If, however, the right and left margins of the heart silhouette be connected by a line which curves gracefully from the curve of the right margin into that of the left we have a line which will include within the territory of the heart the right and left atria and the cardiac extremity of the pulmonary artery and of the aorta. A small portion of the left auricle may be cut off by the line that curves toward the right from the left border but as a rule this is insignificant (fig. 1). By practice in employing this method of outlining the heart silhouette, which is essentially that of Mortiz and Dietlen, one - may acquire sufficient skill to make practically identical esti- mates of the heart silhouette area when plates are studied at widely different intervals. This is perhaps the best test of one’s own consistency with the method. Different observers may establish slightly different methods of completing the outline of the heart silhouette which will lead to slightly different estimates of the heart silhouette area but these differences should not be serious when careful studies are made of the anatomy of the heart in the dead body in conjunction with the heart shadow in the living. The extent of the area included within the outline of the heart shadow may be quickly estimated with a planimeter. If the outline is that of a teleroentgenograph the appropriate reduction for ray divergence should then be made. The chief objection to the method of estimating the size of the heart from the heart silhouette area as outlined above is that it is not sufficiently objective. For this reason it has not been used by a number of foreign and American investigators who have made x-ray studies of the heart. Among these may be mentioned Otten (12), Groedel (08), Claytor and Merrill (09), Williamson (15) and Shattuck (16). The most objective measurement that can be made of the heart silhouette is that of the greatest transverse diameter. This is probably the measurement most frequently made. For the study of comparative size the transverse diameter of the heart shadow is compared with the transverse diameter of the DETERMINATION OF SIZE OF HEART BY X-RAYS 431 thorax according to some such formula as that suggested by Kreuzfuchs (712). Some investigators add to the measurement of the transverse diameter of the heart the measurement of the long diameter from the point where the curve of the right border of the heart is broken by the line of the aorta or of the superior vena cava to the apex of the heart silhouette. Since, however, an accurate outline of the apex of the heart is the chief difficulty that con- fronts one when he attempts to complete the line of the lower border of the heart the measurement of the long diameter of the heart is subject to the chief error that may arise from measuring the area of the silhouette and the long diameter gives a far less satisfactory standard on which to base an estimate of volume. This is true of the numerous other diameters that may be measured on the heart silhouette. The area, which combines them all, _gives the best standard from which to estimate the volume of the heart. For study of variations in the shape of the heart however, some of these various diameters may be of value. 4. TABLES A AND B Chosing then the area of the heart silhouette, reduced in case of radiographs to conform in size to the contour of the heart, (see p. 424) as the standard from which to estimate the size of the heart we have tabulated in tables A and B the normal relations of a silhouette area of a given size to transverse diameter, to body weight, to heart weight, to heart volume and to height in either sex at various ages. The data on which these estimates are based may be summarized as follows: a. Heart silhouette area and body weight The relations of silhouette area to body weight are based primarily on the study of radiographs of 188 men, 42 women and 9 children, all healthy and normal from the clinical standpoint which we have studied at the Wisconsin Clinic according to the teleroentgenographic method outlined above. With the data obtained from these studies have been compared the orthodia- 432 Cc. R. BARDEEN TABLE A Table showing relations of a heart silhouette area of a given size to approximate trans- verse diameter, body weight, heart weight, heart volume in diastole, and height for either sex, at a given age during childhood. Individual at rest, sitting WEIGHT OF BODY APPROXI- siumov- | gute na | sq. cm. cm. kilos 16 4.7 554 17 4.9 3.0 18 5.0 3.8 19 Hol 4.1 20 5.3 4.5 21 5.4 4.8 22 5.9 5S 0 23 Hot 6) 24 5.8 5.9 25 5.9 6.3 26 6.0 6.6 27 6.1 7.0 28 6.2 7.4 29 6.4 7.8 30 6.5 8.2 31 6.6 8.6. 32 (0), ¢/ 9.0 33 6.8 9.5 34 6.9 9.9 35 7.0 10.4 36 eal. 10.8 37 7.2 11.2 38 7.3 iLL 2 39 7.4 12.2 40 eo 12.6 41 Hole 13.0 42 13.5 43 Fol 14.0 44 ats 14.5 45 7.9 15.0 46 8.0 15.5 47 8.1 16.0 48 8.2 16.5 49 8.3 17.0 50 il 33 51 8.4 18.0 52 8.5 18.5 pounds to ee EE See ae me ke ene ee : Sq qoge So Soe e oS oS OES) Tho) TS en en te) Bey Tos fh ter) a) ba 8 Sb WEIGHT HEAT EMPTY grams alia on on bo bo oo — COON nN ND OD NOonrrfrnw ON oOo or (ore) (0%) b pt aed ga an ted oe Cocco OoOmoonmonmnononmoenonocoooorrarnannodadcsd vo) part 94. I ~ 100.0 103.0 VOLUME HEART DIASTOLE SEX AGE YEARS ce. 34 37 40 44 47 51 59 59 62 66 70 74 79 83 87 92 96 101 105 110 115 119 124 129 134 139 144 149 155 160 165 171 176 182 187 193 198 M BIRTH Cl leo ESTIMATED HEIGHT cm. inches 51 20 Sy 53 il 54 56 22 ayy 58 23 60 61 24 62 64 25 65 66 26 67 69 27 70 28 72 74 29 76 30 78 80-3) tee 81 32 83 33 85 87 34 89 35 91 36 93 95 3d 97 38 99 39 101 40 103 41 104 41 106 42 108 43 110 DETERMINATION OF SIZE OF HEART BY X-RAYS 433 TABLE A—Continued APPROXI- HEART MATE WEIGHT | VOLUME AGE Pee TRANS- | WEIGHT OF BODY | HEART HEART SEX YEARS |ESTIMATED HEIGHT AR VERSE EMPTY | DIASTOLE BIRTH DIAMETER sq. cm. cm. kilos pounds grams ce. cm. inches 53 8.6 19.0 | 43.0; 106.0 204 M 6 12 tet 54 8.7 20.0 | 44.0} 109.0 210 114 45 55 8.8 20.5 | 45.0 25) 216 F a 116 56 ZO PAGE OR eae 520 222 M 7 117 46 57 8.9 21.5| 47.0| 118.0 228 119 47 58 9.0 22hOF 4 OR On| l2220 234 F 8 121 59 9.1 D745) |) BO)o0) EO) 240 M 8 122 48 60 PaO) || GLO) |) 12320) 246 124 49 61 9:2 ZAOR Pb 320 132.0 253 F 9 126 62 9.3 | 24.5) 54.0 135.0 259 M 9 127 50 63 25,0) 50L0 138.0 265 129 64 9 34 Pr) |) EXorl0 141.0 Dif 130 51 65 9.5 26.0 | 58.0 144.0 278 F 10 131 66 9.6 27.0 | 59.0 147.0 284 M 10 132 52 67 20) |) 6120) 15020 290 133 68 9). 0 280 ON 6220) 5420 297 134 53 69 9.8 28,5) ||| 163.0) 158.0 304 F 11 135 70 29.5 | 65.0) 161.0 311 M 11 136 54 71 9.9 30.0 | 66.0 165.0 317 137 12 10.0 30.5 | 67.0 168.0 324 138 73 10.1 SIO) N69F Os 7220 331 139 55 74 3220) 7050 175.0 337 140 75 10.2 ayy Ih isc) | al7)a00) 344 F 12 141 56 76 10.3 305 OFio5Ok | aetSs20 351 M 12 142 56 77 34.0] 75.0] 186.0 358 143 78 10.4 34.5 | 76.0] 189.0 365 144 we 10.5 30-00 ai Onl 19320 372 M 125 145 57 80 36.0 | 79.0] 197.0 379 146 81 10.6 36,0) |) +802 07)" 2010 386 147 82 10.7 37.0} 82.0] 204.0 393 M 13 148 58 83 38.0 | 83.0] 208.0 401 149 84 10.8 38.5 | 85.0) 212.0 408 F 13 150 59 85 SOLON SOLOn |e 2t5 80 415 151 86 10.9 40.0} 88.0} 219.0 423 : 152 87 11.0 40.5 89.0 | - 223.0 430 M 14 153 60 88 Alo, |) 9120) | 22720 438 154 89 Hi a 42.0 93.0 Dail O 445 155 90 12 42.5 94.0 235.0 453 91 43.5 96.0 239.0 460 F 14 156 61 92 11.3 44.0 | 97,0} 243.0 468 93 45.0 99.0 247 .0 475 157 434 Cc. R. BARDEEN “TABLE A—Concluded HEART APPROXI- . 2 MATE WEIGHT | VOLUME AGE SILHOU- | orans- | WEIGHT OF BODY | HEART HEART SEX YEARS |ESTIMATED HEIGHT aa VERSE EMPTY | DIASTOLE BIRTH DIAMETER sq. em. cm. lilos pounds grams cc. cm. inches 94 11.4 AD eos LOOZOR eZole0 483 95 15 46.5 | 102.0 255.0 490 158 96 47.0 | 104.0 | 259.0 498 F 15 159 62 97 11.6 48.0 | 105.0 | 263.0 506° M 15 160 63 Sse’; | 48.53 107.0) 267Onlaenl4 ||) 99 49.5 | 109.0 271.0 522 161 100 11.8 S020) T1000 27520 530 M 153 162 64 101 SLSOR 2 OR e227 980 538 F 16 160 63 102 11.9 51.5 | 114.0 | 283.0 545 163 103 §2.9 | 115.0 | 92870 554 164. 104 12.0 HonO NOM B292a0) 562 M 16 165 65 105 12.1 54.0 | 119.0 | 296.0 570 166 106 54.5 | 120.0 | 300.0 578 F 17 163 64 107 p22: aE) || 1PP7e(0) 304.0 586 M 163 167 66 108 12.3 56.0 | 124.0 | 309.0 595 168 109 57/0) || WAS) || BBO) 603 169 110 12.4 58.0 | 127.0 S70 612 M M7 170 67 111 58.5 | 129.0 322.0 620 171 112 12.5 59.0 | 130.0 326.0 628 113 60.0 | 1382.0 | 330.0 637 172 114 12.6 61.0 | 134.0 335.0 645 115 = 615i) 1386.0) 338920 653 M 18 173 68 116 Nai 62.5 | 138.0} 343.0 662 DETERMINATION OF SIZE OF HEART BY X-RAYS TABLE B 435 Table showing relations of a heart silhouette area of a given size to approximate transverse diameter, body weight, heart weight, heart volume in diastole, and height for either sex, at a given age. Individual at rest, sitting HEART SILHOU- ETTE AREA APPROXIMATE TRANSVERSE DIAMETER WEIGHT OF BODY . WEIGHT HEART EMPTY VOLUME HEART DIAS- TOLE sq. cm. 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 11.8 11.9 13.3 13.4 kilos | pounds 99 100 102 104 105 107 49.5} 109 50.0} 110 51.0) 112 51.5} 114 2.5) 115 53.0} 117 54.0) 119 54.5} 120 55.5} 122 56.0} 124 57.0) 125 98.0) 127 58.5| 129 59.0} 130 60.0} 132 134 136 138 140 141 143 145 147 149 150 152 154 156 158 160 162 we Or Ce eS) PS ~J SS) Sx _ va Or loner) eS Re (SS) > Ou aod © on ~] wWNrere oO Sx Suen ey ey ey) Cn “ISI NI ws grams cc. 475 483 490 498 506 514 522 530 538 546 554 562 570 578 586 595 603 612 620 628 637 645 653 662 671 679 688 696 705 714 723 732 741 750 759 768 Se dladard (di HEIGHT AT GIVEN AGE FOR GIVEN WEIGHT 20 years 30 years 50 years Sex Height Sex Height Sex Height cm. | in. cm. | in. cm. | in. F | 140) 55 F | 142) 56 | F | 130) 41 M | 145) 57 | F | 135) 53 1D aye eles || iisye| Ge! M | 147) 58 | M | 137| 54 | F | 124/49.0 F | 150) 59 | F | 142) 56) F | 132/52.0 M | 150} 59 | M | 152) 56 | M | 135/53.0 F | 152) 60 | M | 145] 57 | M | 137/54.0 M | 152} 60 | M | 147} 58] F | 140/55.0 F | 155) 61 | F | 150) 59 | M | 142/56.0 M | 155} 61 | M | 150) 59 | M | 145157.0 F | 157| 62.) F | 152) 60°) F | 145|57.0: M | 157} 62 | M | 152) 60 | M | 147/58.0 F | 160) 63 | F | 155] 61 | F | 147/58.0 M | 160) 63 | M | 155} 61 | M | 150/59.0 F | 163} 64| F | 157] 62 | F | 150/59.0 M | 163} 64 | M | 157) 62 | M | 152/60.0 M | 160) 63 | M | 155/61.0 F | 168) 66 | F | 163) 64) F | 155/61.0 M | 165) 65 | M | 163) 64 | M | 157/62.0 M | 168} 66 | F | 165) 65 | F | 157/62.0 F | 170} 67 M | 170] 67 | M | 165) 65 | M | 160/63.0 F | 160/63.0 F | 173} 69 | M | 168) 66 | M | 163/64.0 M | 173) 68 | F | 170) 67 | F | 163/64.0 By) 178) 270) | ES sis M | 175} 69 | M | 170] 67 | M | 165/65.0 F | 180) 71 | F | 175) 69 M | 178} 70 | M | 173) 68 | M | 168/66.0 F | 168/66.0 F | 183) 72 | F | 178 70 | F | 170/67.0 M | 180) 71 | M | 175] 69 | M | 170/67.0 Fels wi Ee |) 171/67 .5 M | 183} 72 | M | 178! 70 | M | 173/68.0 436 HEART SILHOU- ETTE AREA sq. cm. 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 APPROXIMATE TRANSVERSE DIAMETER cm. C. R. BARDEEN TABLE B—Continued WEIGHT WEIGHT VOLUME HEART HEIGHT AT GIVEN AGE FOR GIVEN WEIGHT OF BODY eee pas 20 years Sex | Height kilos | pounds | grams ce. cm. | tn. 74.0) 163 408 786 (520) 165 Wes G9 dt |e Mia| LS5leve 76.0; 167 | 417] 804 76.5) 169} 422} 8138) M | 188) 74 Tils}| ~ lef 427 | 823 78:5) 173°) 431) 832 79.5| 175 | 436) 841 | M | 191) 75 80.0} 177 | 441 850, 81.0} 179 | 446] 860] M | 193] 76 82.0) 181 451 869 83.0, 183} 456] 878 83.5) 185 | 460] 887 84.5) 187 | 465 | 897 85.5) 189] 470 | 906 86.5) 191 475 | 916 7.5} 193 | 480] 925 88.0} 195 | 485 | 934 89.0; 197] 490 | 944 90.0) 199} 495] 954 91.0) 201 500 | 964 92.0} 2038 | 505 | 974 93.0) 205} 510} 984 93.5) 207] 515 | 994 94.5} 209 | 521 | 1004 9525 eT 526 | 1013 96.5} 213] 531 | 1023 97.5) 215 | 536 | 1033 98.5} - 217 541 | 1043 99.5; 219] 546 | 1053 100.0) 221 551 | 1063 101.0) 223 557 | 1073 102.0} 225 562 | 1083 103.0) 227 567 | 1093 104.0) 280 572 | 1104 105.0} 232 57 1114 106.0} 234 583 | 1124 107.0) 236 588 | 1134 108.0} 238 593 | 1144 109.0 599 | 1154 Sex F M M M M , Z 4, —=O281> 18) | —OFN i203 AB“EGlbs...:.:. BS ar aha elite ia SIE AQT 8.2 : —2.1) 7 |—1.1] 12 |—4.2) 1 |+7.5 6788's. -3.. eo ate ANDO Scat —2.1 —7.8 89-110 lbs...... 2 - Motalse....-.-.06 80 |+10.8) 25 |4+2.1) 25 |—0.8, 25 |—5.1} 6/41.1) 3 00 The average divergence from the standard of the 188 normal men studied by me is —0.1 per cent (table 2). The most note- worthy feature of the divergencies of the sub-groups is that shown by the heavier sub-groups, a decreasing size of the heart silhouette relative to the body weight. This is also shown in the athlete column, the lighter sub-groups with one exception showing - relatively large silhouette areas while the heavier sub-groups show smaller relative areas. Among the athletes here tabulated are included men who have taken an active part in strenuous inter- collegiate athletics but we have excluded from the table athletes whose hearts gave clinical evidence of abnormality. The forty- DETERMINATION OF SIZE OF HEART BY X-RAYS 44] two women with clinically normal hearts studied at the Wis- consin Clinie show hearts relatively slightly small for the lighter groups, more markedly small for the heavier groups. While it appears that the standard of heart silhouette area in relation to TABLE 2 Average percentage of divergence from the standard silhouette area corresponding to a given body weight in youths and adults OBSERVER Bardeen, sitting position Dietlen, supine position Se EG EIT, Men Athletes Women Men Women Men o Bho 3 ELo ® Bu a) Ee (te sae ® | fhe Ss |sge| 3s | som] 3 | poh] 5 | pom) 3s | see] Ss | eee Z| Z| As Z| Z| Z| Ai _4 |e 40-50 K.... — 3. 2 PAY, — 17 20|/— 1.6 FASTET ATUL ul cee te oor er Mii ol ke mec) | ae S| 2 4 — 4. 3 |— 1. 82 : 388)}— 8.1) 55 . eees ibe ole lie Os iamee : +8.6 61-70 eA land : 4.44 9 |/— 8. 2s 16;—10.5| 56 ; 134-155 lbs. 90 |+ 1.4) 14 |+ Bosh 7 16 6 +3.2 71-80 K... a — i. D2. 1 |—28. t2|/—0» .10)4-2. elisha ilo Ma ee vee. oe oe a 81-90 K... — 4, 2 |— 2. 4\—9. —l. Ieecootbe. (0° 2 Bee =e iy 91-100 K... : aieoinibe. flu: lac” OORKG Sear 22D Mosse a le 12-0 Totals ......|/188 |— 0.1) 30 |4+ 3.4) 42 |— 2.5) 187|—0.01) 74/— 6.6) 123)/+5.5 body weight given in tables A and B is slightly large for the average women the difference between the size of the silhouette for men and women of a given weight is so little, probably not over 2.5 per cent for individuals of average size, that it does not seem worth while to try to establish a separate curve for women. 44? Cc. R. BARDEEN The relatively smaller heart silhouettes of fat individuals appears in groups of lighter weight for women than for men because heavy women average less in height. See below for a discussion of the effects of height. ye The figures in table 2 based on Dietlen’s data show on the average a close correspondence for the 187 men, a divergence of only —0.01 per cent. We should, however, expect in this group an average plus divergence of over 5 per cent since Dietlen’s observations were made on individuals: in the supine position ‘while the standard table is based on individuals in the sitting position. The relatively small size of the silhouette area in Dietlen’s figures may be due in part to differences in method and in part to racial differences. The relatively small size of the heart silhouette in Dietlen’s studies of women is more marked than in those studied by me. Schieffer’s studies of the hearts of individuals engaged in strenu- ous muscular work show an average increase of 5.5 per cent of the size of the heart shadow above the normal but since his studies were made on individuals in the prone position, we should expect about this difference from a standard based on the sitting position. The lighter groups of individuals show relatively large heart shadows. Giegel (14) suggested as a method of determining the heart quotient, the division of the 3/2 power of the area of the heart silhouette by the body weight in kilograms. He showed that the heart quotient thus obtained varied in Dietlen’s cases from fifteen to twenty-three in 93 per cent of the cases. The extremes were twenty-seven (two cases) and fourteen (three cases). Expressed in terms of divergence from the normal standard this would mean + 21 per cent to —9 per cent for the 93 per cent of cases, —18 per cent and +27 per cent for the extremes. The amount of diver- gence from the standard found in the normal individuals in the groups studied by me will be discussed below in connection with other factors which must be considered, age and height. The area of the cardiac silhouette estimated as 70 per cent of the long diameter of the heart times the transverse diameter gives for Claytor and Merrill an average divergence of —5.7 DETERMINATION OF SIZE OF HEART BY X-RAYS 443 per cent, for males and —15.3 per cent for females. The Dietlen males, on the other hand, show by this method of estimation an average divergence of +12 per cent. By direct measurement of the area enclosed by the completed cardiac outline, the average divergence in 187 men studied by Dietlen is —0.01 per cent. The average divergence of the cardiac shadow of the men studied by Otten is +4.9 per cent. Making allowance for the prone TABLE 3 Average percentage of divergence from the standard silhouette area corresponding to a given body weight in youths and adults Area estimated as 70 per cent long diameter times transverse diameter. OBSERVER Claytor and Merrill Dietlen Otten WEIGHT Men Women Men Men 2 Pick 3 Paks 2 poh 3 Pek Ae KGS dee tig a : CETNI Ly eee me ae ls 128) CBN ace) PI GO. see et oh r » fe 2 ; 22189 Ibs... see. BSED) te) elec Ce are BE fp Sees GIETOR Kh. Vc hte. ee ee An Ria ee 5 PAP ES bs.) ae Dost iE eee parce) | Se eascae HAE SMEG late = eee? 2 = 119 MSG a17Sillbe.. ese. : EU es a fas <2 : PTET, MIG OO ee ee 179-200 bse eek J eRe Sle ERG tallsy:: 2-00 oe ra 7a 57 51 |= 15e3ly aaGmieede 100 | +4.9 position makes the average for the Otten figures correspond closely with the standard based on the sitting position. The heavier individuals show relative small cardiac shadow areas as in the case of the individuals tabulated in table 2. In table 4 are given dissection room data obtained according to the method outlined above, p. 4838. The bodies studied for this tabulation were all embalmed with equal parts of carbolic acid, 444 Cc. R. BARDEEN alcohol and glycerine. They were measured and weighed before being dissected and the heart was measured and weighed as soon as the thorax was opened. As a rule the embalming fluid was injected into the femoral veins under a pressure of six pounds. In many bodies this leaves the chambers of the heart moderately TABLE 4 Average percentage of divergence from the* standard silhouette area and standard heart weight corresponding to a given body weight from dissectin) room data AREA HEART OUTLINE HEART WEIGHT Male Female Male Female AGE AND WEIGHT OF BODY 5 ese) 2 pese| 2 ase) eae Ee less! € |es8| & | ass] € | ose = ee oD 2 a Se 5 2 Cc 5 2 Cc Foetus; 2.06 K., 4.5 lbs... 1 |—23.0 1 |—23.0 New born; 2.9K., 6.4lbs...| 1 |—46.0 1 |—37.3 New born; 4.3 K., 93 Ibs... 1 Serr 1 |+20.5 Child 23 years; 5.9 K., 13 ID Stewv.is53 .08 Saeco ae 1 |+25.0 PS een Child 4 years; 7.3 K., 16 IID SER Paco. et ope oes ee eee 7 \—<5e8 1 |—17.8 Child 2 years; 8.6 K., 19 IDS: hese See 1 |+ 9.7 1. alco Child 9 years; 20 K., 44 | 05: eae Ue perl Sy oy ee 1 |— 9.3 1 | ee3 Motal- childrens sss ss: 6 |— 8.2 IE et of 6 |—12.0 1 |+20.5 Adults: 21-30 K., 45-66 Ibs...... 1 |-- 9.4). 2) te 927) Ll 83) 62 eae 31-40 K., 67-87 Ibs...... s |+ 3.0). 6 |+9:9) 8 1= 22 6 eeu 41-50 K., 88-111 lbs......| 10 |-+ 5.1 1 Siem Oera 51-60 K., 112-133 lbs...... 3 |+ 0.5) 1 |—138.9) 3 |— 9.6) 1 |—25.2 61-70 K., 134-155 lbs...... 71-80 K., 156-178 lbs...... 2 I+ 3.2 2 |— 6.1 81-90 K., 157-198 Ibs. .... otalvadults) seek 24 |+ 3.9) 9 |4+ 7.2) 25 |— 2.0) 9 |— 1.1 distended. Such bodies were selected for this study. Bodies showing cardiac lesions or in which from one cause or another the heart was distorted were not utilized for the data here tabu- lated although frequently utilized for other data in the study of the heart. The weight of the heart was controlled and in some DETERMINATION OF SIZE OF HEART BY X-RAYS 445 instances estimated from the displacement of water or oil by the empty heart. While data of this kind are necessarily crude, especially since the bodies utilized were the regular material used in teaching medical students, it is believed the data obtained have some value, especially in connection with x-ray studies on the living. Twenty-four adult male bodies show an average divergence from the normal standard of +3.9 per cent, nine females an average divergence of +7.2 per cent. The individual variations are to be attributed in part to variations in the extent of dis- tention of the heart after embalming. The relatively few female bodies studied do not lend support to what seems to be on other evidence fairly well established that the female heart relative to body size is slightly smaller than the male. The foetus and one of the new born infants studied show small hearts, the other new born, a large heart. One of the young children shows a small heart, the other two large hearts. The nine year old child shows a heart small from the standpoint of area and from the standpoint of weight. While the data here given for children are scanty they tend on the whole to lend support to the belief that the standards of tables A and B are approximately correct for young children. b. Silhouette area and transverse diameter The transverse diameter varies in size in relation to the area of the heart silhouette and the volume of the heart according to the position of the heart. If the long axis of the heart is trans- versely placed, as during childhood and in fat adult individuals, the transverse diameter of the silhouette is relatively large. If the long axis is more nearly vertical as is usual during youth and in thin adult individuals, the transverse diameter of the silhouette is relatively small. I have found far wider variations in the transverse diameter than in the area of the heart silhouette in relation to the size of the body. To compare the extent of variations of transverse diameter with those of area the former should be squared. If this be done the variations in transverse THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 446 Cc. R. BARDEEN diameter are from five to fifteen times as great. Dietlen’s tables show the same thing. We have therefore discarded the transverse diameter of the heart silhouette in favor of the area as a means of estimating relative heart size. In order, however, to make possible a com- parison between the results of studies based on measurements of the transverse diameter and the data in our standard tables we have introduced a column showing the approximate transverse diameter corresponding to the silhouette areas given in tables A and B. To determine the approximate transverse diameter correspond- ing to a heart silhouette area of a given size we have tabulated the various transverse diameters corresponding to a given area, reported by Dietlen in his study of the hearts of adult men and women, those reported by Veith in his study of the hearts of children, and the data obtained in our own x-ray studies of the heart in the living and those obtained from a study of cadavers. While the variations in the size of the transverse diameter cor- responding to an area of a given size are considerable the formula 1.18 V/area =transverse diameter gives a fair general standard as the following examples may show (table 5). In slender, youthful adults, especially during deep inspiration the long axis of the heart tends toward the vertical and hence the transverse diameter of the heart becomes relatively small. The large number of such individuals included in the series studied roentgenographically by me tends to make the transverse diam- eter of hearts with a silhouette area of from 90 to 125 square centimeters average below the figures called for by the formula given above and utilized in tables A and B. The average is also low in several of the older groups of children studied by Veith and in the cadavers of small slender individuals studied by me. In order to test out the values of the transverse diameter given in tables A and B from the standpoint of body weight the fol- lowing table has been prepared (table 6). For the lighter weights, the number of observations are re- latively few. Veith’s supine individuals number 80. The male individuals sitting are of two groups; (1) is composed of TABLE 5 Observed transverse diameter for a given area compared with he standard based on the formula transversed diameter = 1.18 area a & AREA H2ea OBSERVED TRANSVERSE DIAMETER B23 Rea SHA sg. cm. cm. cm. 10 3.7 | 3.4 (foetus) 21 5.4 | 5.3 (infant cadaver) 30 6.5 | 6.4 (infant cadaver) 34 6.9 | 7.0 (infant cadaver) 49 8.3 | 7.7 (infant cadaver) 59 9.1 | 9.3 (child, C.R.B.) 8.4, 9.3 (children, Veith) 70 9.9 | 8.8, 9.3, 9.8 (cadavers) 9.0, 9.30, 9.45, 9.55, 9.65, 9.75, 10.2 (boys, Veith) 9.5, (girl, Veith) 80! 10.5 | 10.3 (cadaver) 10.9 (male, Dietlen) 9.35, 9.5, 10.15, 11.00 (boys, Veith) 85 | 10.8 | 10.7, 10.9 (cadavers) 11.1 (woman, Dietlen) 90 11.2 | 11.1 (cadaver) 11.2 (boy, Veith) 95 11.5 | 10.8, 11.4 (cadavers) 10.6 (woman, C.R.B.) iL 11.9 (men, Dietlen) 11.1, 11.3, 11.3, 11.7, 12.3 @vomen, Dietlen) LOOPS ies) | MSs Gniane CARB) | 11.5 (woman, C.R.B.) 11.5 (man, Dietlen) 11.3, 11.7, 12.0, 12.2, 12.4, 12.8 (women, Dietlen) 105 12.1 | 12.4 (cadaver) 12.3 (man, C.R.B.) 10.9 (woman, C.R.B.) 12.1, 12.6, 13.0 (women, Dietlen) 11.6, 11.6, 12.3, 12.3, 12.4, 12.4, 12.8, 12.9 (men, Dietlen) 110 Aa Ue Celie Os 2a Gnen, iCoRB:) 1273501284) women, ©. Re.) 1S) Mes 126, 12.8. Ws. 1 13255 1358) GnensDretlen) Plone 2 ton mul owe petite 2 12.2) 12 ole ele semenk ©.Reb.) 12.0 (woman, C.R.B.) 12.7, 12.8, 18.1, 14.0, 14.2, 14.38 (men Dietlen) ZONA laa leo alae es ee (menns©sEvelss) 1a ee oles, 12.9) 130) tole isnoetsese 1306. 1358, 1338; 14 Gy wAR 30), .8, .2 men, Dietlen) 125) || TBD MPG PG, AO) Nes PH Wei) (Ceaveray, (Oy 18%,,183,) 12 13.7, 13.9 (men, Dietlen) USO) || TRG) | 8), TB.74 sia, SO) (Coaveroy (Gr18 4183, ) 13 13.9, 13.9, 14.0 (men, Dietlen) 140'| 14.0 | 13.6, 14.8 (men, C.R.B.) 150 14.4 | 14.8, 15.1 Gmen, C-R.B.) 14.7 (man, Dietlen) 447 448 Cc. R. BARDEEN 25 individuals (orphans), 23 of whom are also in the supine list; (2) is composed of 25 healthy school boys. The female sitting group is composed of 25 individuals. The children in the groups studied by me number only 7 boys and 13 girls below 44 K. weight. The individuals studied by me above 44 K. weight number 188 men and 42 women with apparently normal hearts. TABLE 6 Table showing the average transverse diameter corresponding to a given body weight as reported by various observers compared to the standard given in tables A and B AVERAGE TRANSVERSE DIAMETER Veith Bardeen, sitting eee WEIGHT eee Supine Sitting Pee Otten, ard supine supine : Male Femaie Male |Female} Male Male | Female 1 2 lilos cm. cm. cm. cm. cm. cm. cm. cm. cm. cm. cm. 15-19| 8.3*| 8.9 8.9 S05) Sol 8.7 8.1 20-24) 9.0] 10.2 8.6 9.42) 9.1 9.3 8.6 25-29| 9.6 9.4 OO | Qo Web Bee! 30-34] 10.2 | 10.8 10). 5) |) O68.) OES) O83 30-39) 1027 | 10225) 1054) 1Ok5 7 928) |) ORG 98 40-44} 11.1 10.7 Wks} | al cal OFZ 45-49] 11.5 | 10.8 | 11.3 11.4 | 11.4 10.2 90=54) 11.9) 12.0 | 11-7 1 DEAS ste 10.7 95-59 | 12.3 | 12.4 | 12.2 1259) LAE Sa OS OS iO 60-64] 12.6 | 12.7 | 12.2 Hee UA | ibs || al © 65-69 | 13.0 | 18.0 | 14.2 MBoPI WO abs |) GL 1 70-74. 13.3 | 12.9 | 12.8 13249) 13/0) 1253) |G 75-79 | 18.6 | 13.5 14.3.| 13.2 | 12.4 | 11.9 80-84; 138.9 | 13.7 14.4 12.9 * For 17 K. The Dietlen list is based on a study of 156 adult men, that of Otten on a study of 100 adults, those of Claytor and Merrill on a study of 37 men and 54 women. Veith’s boys sitting on the whole correspond well with the standard adopted. Boys supine show in general a larger trans- verse diameter, the girls a smaller transverse diameter than the standard. Considering the relatively small number involved DETERMINATION OF ‘SIZE OF HEART BY X-RAYS 44y and the great individual variations found in the transverse di- ameter the averages in my own cases correspond fairly well with the standard. Dietlen’s cases studied in the supine position show a greater transverse diameter than the standard which is based on the sitting position. On the other hand, the individuals studied by Otten show an average difference of 0.43 em. below the standard for each group. This is probably due in the main to the fact that the individuals were relatively slender, averaging 3.2 per cent above the normal height for the average weight at age thirty, as given in tables A and B. Yet Otten gives the position of the long axis of the heart in the individuals studied by him in the supine position as obliquely placed in 35 per cent of the cases, perpendicularly placed in 13 per cent and trans- versely placed in 52 per cent so that in over half the cases the transverse diameter should be relatively large. The individuals studied by Claytor and Merrill have unusually narrow hearts. The upright position chosen by these investigators may have tended to bring the heart into the vertical position but with due regard for this the hearts seem to average abnormally narrow. In general it may be said that for the supine or prone position about 7 per cent should be added to the transverse diameter over the figures given in the standard table; for the standing position, about 4 per cent should be subtracted. I have found in seven cases at the end of deep inspiration in the prone, sitting and standing positions that the transverse diameter averaged respectively 14.1, 13.1 and 12.6 em. In these the average for the prone position was 7.6 per cent greater than for the sitting, 3.8 per cent less for the standing than the sit- ting position. In eight cases studied during normal respiration but at the height of inspiration in the sitting and prone positions, the average transverse diameter sitting was 13.3, prone, 14.3, or 7.5 per cent greater in the prone than in the sitting position. In deep inspiration the change in the transverse diameter corresponds with the change jn the area of the heart silhouette but in quiet inspiration the heart is relatively broader in the prone than in the sitting position. 450 Cc. R. BARDEEN c. Heart weight and body weight In dealing with the relation of heart silhouette area to body weight we are dealing with factors which can be objectively studied on a large number of individuals. The determination of the relation of the size of the heart silhouette to the volume and weight of the heart is not open to so direct a study in human beings. After death the heart can be weighed and the weight of the heart may be compared with the weight of the body but unless death has occurred from accident we are not likely to be dealing with normal conditions. Conclusions as to what obtains in the living must be cautiously applied from study of the dead. The relations of the size of the heart to the size of the body in cadavers has been studied from various points of view. Among the chief contributions to the subject are those of Boyd (’61), who made an extensive study of the average weight of various organs, including the heart in relation to body weight and age; of Thoma (’82), who utilized mathematical theories of probability in a valuable analysis of his own data and that of other investi- gators In a study of the relation of the weight of the heart to body weight; of W. Miller (’83) who utilized extensive data in a study of the relation of heart muscle weight to the weight of the heart as a whole and of the relation of heart muscle weight to body weight, height and age; of Beneke (’78) who studied the volume of the heart substance from the standpoint of body length; of H. Vierordt (90) who has summarized the work of previous investigators and added data of his own; E. Kress (’02) who studied the weight of organs in children; and of Greenwood and Brown (718) who have applied modern mathematical methods to a study of a small but carefully selected material. The studies of these and of numerous other investigators have shown that there is a close correlation between the size of the heart and the size of the body, due probably to the need of a given mass of heart muscle to pump the blood to a given mass of tissue. Greenwood and Brown conclude that the correlation between the weight of the heart and that of the whole body is not much less than 0.5 and that the weight of the heart can be DETERMINATION OF SIZE OF HEART BY X-RAYS 451 deducted from the weight of the body and kidneys by means of a linear equation with an average error of about 8 per cent. If the mean heart weight of the cases studied by these authors be di- vided by the mean body weight, the heart weight is found to be approximately 0.575 per cent of the body weight. Individuals show a variation from this mean per cent of body weight of about 25 per cent in either direction that is, from about 0.45 per cent of body weight to 0.70 per cent of body weight. These figures probably very nearly express the average relative weight of the heart. Other investigators who have studied a greater number of individuals have furnished data which differ more or less widely according to the material studied and meth- ods used. Asarule the data have been presented from the stand- point of average body weight and average heart weight for a given age. By dividing the one by the other one obtains a rough estimate of the proportion between heart weight and body weight for a given age. Data obtained in this way have led to somewhat divergent results as may be seen in table 7. Boyd studied a large number of individuals at the Marylebone Infirmary and a smaller number at the Insane Asylum at Som- merset. The figures for the latter are placed immediately below those for the former for age groups above 30. It will be noted that in general the hearts studied by Boyd are heavy in relation to body weight, running from 60 to 80 per cent of the body weight instead of less than 60. There is no great difference between the relative size of male and female hearts but the hearts studied at Sommerset are notable relatively smaller in size than those studied at Marylebone. The figures from Thoma are based on a mathematical study of the average heart weight found by Caspar-Liman, Blosfeld, Reid, Peacock and Boyd for a given age combined with Thoma’s, study of the average body weight for corresponding ages. They show a low relative weight of the heart and indicate that the high relative weight shown in Boyd’s figures is due largely to body emaciation. The hearts studied by W. Miiller are from a more carefully selected material and average relatively smaller in size than those of Boyd, 0.604 per cent of the body weight in males; .594 per cent in females. ‘The very 452 C. R. BARDEEN Proportional heart weight at various ages as reported by several investigators 7-9 mo....... Boyd Thoma E. Kress Male | Female | & Male Male | Female so) ——~ % > | a8 2 > | |e Es = > S | Boles ie. | 01S | asi cee S|ge|sl|ee| 88 ge |8/82|/5|/22|8)/22|3/ 82/8] 25 SSS || el || Re we oF | 5) G7 | 5) 8215) 52] 5) 52 | Ss) a Zila | 4 | ca) oe AI oey Iz yeu Cea feme Weazie | | 44) 648] 42|.572| 0.625 |23/0.62)14/0.63/62 0.76/59 3/0.66, 610.89 16| 594) 21 0.64/47|0.63) 7|0.51)12 20.83 0.58 52, 30 6.0.68) 70.65 15|.605] 24). 0.444 28) 8/0. 67|11|0.64 29:0 70.64) 90.65 46.516} 40). 6 6.0.54) 5.0.67 0.412 15 40.59] 70.58 34|.736| 32 0.422 45 120.60 4.0.60 27| .662| 29 0.417 31 100.58 70.61 0.40 32 100.64 80.61 0.382 24 8 0.54) 70.60 | 27) .676| 20}.592| 0.383 7 70.65 70.64 0.392 18 7\0.61) 5.0.64 0.402 3 5|0.80| | 0.417 6 5.0.65, 20.66 0.433 10 5.0.60; 20.58 0.452 12) 20.49} 2.0.60 0.474 1 21| .634| 17]. 0.481 8 2.0.63) 3/0.54 0.490 8 40.61) 10.60 0.500 9 5/0.62/35|0.52 0.481 11 0.471 17 18} .699} 15). 0.474 20 0.474 23 0.476 15 DETERMINATION OF SIZE OF HEART BY X-RAYS 453 TABLE 7—Continued OBSERVER Boyd Thoma Miller H. Vierordt E. Kress Male Female 5 Male | Female! Male | Female} Male | Female AGE and By aot > > ag > > > > = > z Zz as 3 Z 3 as 3 as } fe) 2s } } } } } } , cee 4 rate 1D 3 ay ae » 25 Fa fs Baty by ats ny se a BE = cto ease |asect fa | 8 re S =) 2\85|3|sm| 22 |8| fm|3| sh) 38| $m| 3] s5| 3) $m|/ 2] so et) Se el Sa) ce PENS Ee ie) SO | el) 2S | El) Sones 2) Ge |) | Bue Ae S| SEs] SE Ss ee Ss se oo Se | os) ee A |e Z| A ea Palio | Zales Zee meee alae ella PA NA ee ease : 0.481 33/0.49)'22,0.48 D2 VES ese 0.483 7\0.50,21\0.48 DSBYS ose a 0.490 24'0.48)22:0.49 ARV TS Ses eos: 0.495 30,0.46)22;0.49 25-30 yrs....| 58].675| 74] .654| 0.502 |73)0.58'45'0.50)30,0.46)27|0.49 46) .597| 29) .608 or bo 31-40 yrs... .{118].720| 87].682 70'0.56/59 0. 59|.609| 49|.470 41-50 yrs... .|187| .705) 106) . 706 84.0.59)69)0.56 76| .690| 49) .676 51-60 yrs... .|119}.719) 106} . 764 87\0.62/61/0.59 42) .637| 39) .710 61-70 yrs... .| 126). 763) 149) .758) 88 0.64/83)0.64 39) .728) 41).716 71-80 yrs... .|100} 774150) . 786 64/0.64/61)0.67 21) .693) 20) .612 81-90 yrs....| 24|.840| 76).806 11/0. 75)12|0.69 7|.740| 5) .655 young and the older age groups show relatively larger hearts than the others. The Vierordt figures show relatively smaller hearts than the Miller groups but larger than the Thoma groups. In this connection it should be noted that like Thoma, Vierordt takes a theoretical weight for a given age and divides this by the average heart weight of hearts of individuals of that age. Vierordt bases his estimates of body weight on data , 454 Cc. R. BARDEEN from Quetelet and Lorey. In Vierordt’s tables of the weight of organs in the adults selected from various investigators to illustrate normal build (’06, p. 34,35) the percentage of body weight given for the heart varies from .477 per cent to .633 per cent with an average (not weighted) of .558 per cent. The figures of Kress for children which are based upon the average body weight divided by the average heart weight for each age show a considerably higher relative heart weight than that given in Vierordt’s tables. In part this is due to emaciation in the children studied by Kress but it is not due entirely to emaciation because in case of several of the Kress groups the average body weight is to be looked upon as normal. If one bases his estimate of the relative size of the heart in the living upon the relative size of the heart in the dead, as done by Boyd, Miller and Kress one is apt to get too high a relative heart weight owing to the relatively large degree: of emaciation in the dead. If one bases his estimate upon a di- vision of the average heart weight of a group of individuals by a body weight assumed to be normal for such a group like Thoma and Vierordt he is apt to get too low a proportional heart weight because the heart weight compared with the ‘normal’ body weight is not the average of a normal group but of a more or less emaci- ated group. The mean between the two estimates will proba- bly more nearly approach the normal relative heart weight than an estimate based on either method alone. The following table (table 8) shows the relative weight of the heart for each age group as reported by W. Mueller, the relative weight of the heart based on normal body weight as given in Tables A and B and the mean between the two. Owing to the fact that the average height and average age for each group is not given in the Miller tables the estimates of normal weight for each group are necessarily somewhat rough. This table shows that the heart is relatively much smaller for a given age group if the normal weight for the group be taken instead of the average weight of the group and that the differ- ence in general is greater during childhood than in adult life. The mean between relative weight based on average weight DETERMINATION OF SIZE OF HEART BY X-RAYS 455 TABLE 8 Body weight, average weight of heart at various ages and relative heart weight after W. Miller, body weight normal for a given age, relative weight of hearts studied by Miiller based on normal body weight and mean between relative body weight based on observed and that based on normal body weights 58 AVER- nie |) Seer sor 6 | gtia| waome | Meuaners [Nom | gWEIONE, | nauscivs pe HEART | WEIGHT aa 1. Males grams | kilos Premature births....... 42 1.15} 7.06) 0.00615 | Mature births.......... 23 3.35), 20.79 0.00620) 3.2| 0.0065 | 0.0064 Mem Ontbs ieee ee 45 2.52) 16.19 0.00643} 3.8 0.00426 | 0.0054 2-6 months........... 50 3.49} 20.13 0.00576 | 5.9) 0.0034 | 0.0046 (AD months. s.:.s< 2: 34 5.13] 30.64 0.00597 | 8.2) 0.0037 | 0.0047 DEO RVCATS YI te tA 34 8.57) 52.7 0.00615 | 12.2 | 0.0043 | 0.0052 AE ROVY.CATS oy) see a. 16 | 11.26) 65.2 0.00580 | 15.5 0.0042 | 0.0050 G-10tyears: 2)... 1.220. 15 | 16.63) 103.6 0.00623 | 21.5 0.0048 | 0.00551 N= yey ears 5 ssa. 9 | 27.3 | 163.8 0.00600 | 37.0 0.0044 0.0052 16-20 yyeaTs....... 25... 23 | 43.2 | 236.9 0.00548 | 61.5 | 0.0038 | 0.0057 21-30 years.............| 7% | 51.3 | 297.4 | 0.00580 |-62.5 | 0.0048. | 0.00525 SiAOnvears: ke ta. 70 | 51.6 | 289.6 0.00561 | 65.0 | 0.0046 | 0.0051 AI=5Osyears..:....5.2. 84 | 52.0 | 304.2 0.00585 | 70.0 | 0.0043 | 0.0051 HOO veatses)-. 25.5002 87 | 55.3 | 340.8 | 0.00615 | 70.0 | 0.0049 | 0.0055 Gl=nUmycalsnena ples ae 88 | 54.0 | 345.9 | 0.00640 | 70.0 | 0.0049 | 0.0057 Me SOny canst. esc. 64 | 52.7 | 335.5 | 0.00637 | 70.0 | 0.0048 | 0.0056 Rieo0ryenrs..\..2.00 2). 11 | 42.3 | 315.7 | 0.00746 70.0 | 0.0047 0.0061 2. Females Premature births....... 48 1.24) 7.29) 0.00587 | | Mature births.......... 14, | 3.06] 19.24! 0.00629 | 3.2 0.0060 | 0.0062 ET ONIE ee eget eee 47 2.27; 14.36) 0.00632 | 3.8 | 0.0038 | 0.0051 oe Gamonths: ene GIA | aes 20.18 0.00610 | 5.9 0.0034 | 0.0048 (AA) NOM soso cose Fall Ge 5.34) 32.14) 0.00602 | 8.2 | 0.0039 | 0.0050 DORVAL Ste yey eo 42 7.34, 43.2! 0.00616 | 11.7 | 0.0039 | 0.0050 AON VCALS s(n 5 19 | 11.67) 69.0 0.00591 | 15.5 0.0044 | 0.0052 GalOhyvearsess: 84a 18 | 14.7 | 82.0} 0.00561 | 21.5 | 0.0038 | 0.0047 P=aT5ryea4rs. sao .....a62 H 32.2 | 177.4 | 0.00551 | 38.0 | 0.0047 | 0.0051 NG= PON VeAtSiae ys aes snes 13 | 43.5 | 205.2 | 0.00495 | 54.5 | 0.0039 | 0.0044 Di SO MyCATSeeces.. + ese 45 | 46.2 | 220.6 | 0.00499 | 55.5 | 0.0040 | 0.0045 SI=40h years: --e.....-. a2 59 | 44.9 | 234.7 | 0.00523 | 58.5 | 0.0040 | 0.0046 Al -SObyeatssn.. 2s..2085. 69 | 47.1 | 264.1 | 0.00561 | 61.0 | 0.0043 | 0.0050 LE GONVC aIEStas rsa sR 61 | 43.4 | 256.9 | 0.00592 | 61.0 | 0.0042 | 0.0051 Tl SOkyeatshere ne sia a 61 | 44.1 | 294.3 | 0.00667 | 61.0 | 0.0048 | 0.0051 SIIEC D) Avetishs une oebues 12 | 36.7 | 253.0 | 0.00689 | 61.0 | 0.0041 | 0.0055 * 456 Cc. R. BARDEEN and that based on normal weight runs between 0.5 and 0.55 per cent of the body weight for most groups in both the males and females but averages higher in the males. At and immediately following birth most investigators have found that the heart is relatively large. If, however, the weight of the membranes at birth are included the proportional weight given above holds approximately true. Thus Thoma estimates the normal average weight of the new bornat 3.96 K. including membranes but not including the amniotic fluid and at 3.35 K. including the membranes. The heart he estimates at 0.532 per cent of the body weight if the membranes are included, at 0.625 per cent if they are not included. The following table (table 9) from Miiller indicates that similar relations prevail during the latter part of foetal life. TABLE 9 Data from W. Miiller on the relative weight of the heart in foetuses (a) (b) (¢) (d) (e) 2 Rare “yl camara) (fede Fy aa eras etd mm. grams 25 212 201 120 1.15 0.00354 6 S300 © al 783 232 4.44 0.00436 5 325) | hen. l20G 315 8.08 0.00528 22 423 1727 394 10.74 0.00507 456 2252 487 13.81 0.00504 3 492 2756 501 18.68 0.00574 24 522 3448 572 21.36 0.00531 From the data given above we conclude that fifty-five hun- dredths per cent of the body weight approximates closely the normal relative proportions of the heart weight in males at all ages except at and immediately following birth and that in fe- males the heart is slightly lighter, about fifty-three hundredths per cent of body weight. In our tables we have not, however, . attempted to plot separate curves for males and females. The estimates of heart weight given in tables A and B are based upon the assumption that the weight of the heart is 0.55 per cent of the body weight. In estimating the relative weight at birth the weight of the foetal membranes is included in the body weight. DETERMINATION OF SIZE OF HEART BY X-RAYS 457 On testing this estimate on dissecting room material we have obtained the data shown at the right in table 4, p. 444. The bodies studied were all embalmed at the time of weighing but it is assumed that the ratio between heart weight and body weight was not thus markedly altered. The percentages of divergence given are the percentages above or below the standard adopted for the weight of a heart belonging to a body of a given weight. Thus for a body weighing 50 K. we should expect a heart weight of 275 gr. Ifthe heart weighs 286 gr. we designate it. + 4per cent; if 264 gr. — 4 per cent. The average weight for twenty-five adult male hearts was 2.0 per cent below the stand- ard, that of nine adult female hearts 1.1 per cent below the standard. A foetus and one new born child had hearts below the standard. One new born child had a heart considerably above the standard. Of four young children two had hearts below the standard; two had hearts heavier than the standard. Into the weight of the heart there enter four main factors, (1) the heart muscle tissue, (2) the connective tissues of the valves and supporting structures, (3) the intrinsic blood vessels of the heart and the great vessels near their attachment to the heart, and (4) the fat deposited beneath the pericardium and elsewhere in the heart. Of these factors the heart muscle tissue is dynamically the most important and varies in amount with the dynamic demands on the heart. These demands to a large extent are determined by the weight of the body and hence the heat varies in size with body weight. The mass of the intrinsic blood vessels of the heart probably varies normally directly with the mass of the heart muscle. The great vessels near the heart are relatively heavier than one might estimate so that the relative weight of the heart found by different observers varies to some extent with the amount of great vessel tissue included with the heart. The fat likewise constitutes no inconsiderable part of the heart mass. To some extent it varies directly with the relative amount of fat in the body as a whole. The two observers who have studied most carefully the rela- tion of heart weight to body weight, Thoma and W. Miiller, have excluded in some of their tables so far as possible the great 458 Cc. R. BARDEEN vessels and cardiac fat from the heart weight so as to get the proportion between the heart muscle tissue and the body weight. With the heart muscle tissue the connective tissue framework and the intrinsic blood vessels of the heart are, however, included. The following tables (tables 10 and 11) show the relation of the weight of the heart muscle tissue to body weight found by W. Miller: From these tables it may be seen that the average percentage -of body weight made by the heart muscle tissue in the males studied by W. Miller was 0.534 per cent; 52.8 per cent, in those above one year of age. In the females studied by W. Miiller the average is slightly lower, 0.523 per cent; 0.50 per cent for those above one year of age. The range in the various weight groups is for males from 0.590 per cent for the 1-10 K. weight group to 0.391 per cent in the 100-110 K. group; for females from 0.584 per cent in the 1-10 K. group to 0.302 per cent in the 100-110 K. group. For each weight group Miiller gives the average height and average age. In a subsequent section I give a brief description of statistical data relating to the normal weight for a given height and age. From the average height and average age for each weight group in the Miiller table we may estimate a normal weight in contrast to the actual body weight. By dividing the average heart weight of each group by the estimated normal body weight we get the percentage of estimated normal body weight. This percentage indicates what the ratio of heart weight to body weight would be if the body weight were normal for height and age and the heart weight that actually found. This per cent of normal! body weight is smaller than the per cent of actual body weight for the lighter groups, greater in the heavier groups, indicating that the majority of individuals in the former groups were underweight from the standpoint of height and age; those of the fatter groups over- weight from the standpoint of height and age. The heart mus- culature is heavy compared with the body weight in thin indi- viduals, hght compared with the body weight in fat individuals. By taking the mean between the heart weight-body weight ratio actually found and that which would have been found had TABLE 10 Relative heart-muscle weight in males as reported by W. Miller compared with Weighted average the relative weight when the body weight is normal for height and age a fs 1 4 Z BODY WEIGHT a P is PI F a Ac fe Zs E E E = 3 a | ag | «@ = : oa Ve ee nae lee, Wee < a Z < < 4 my Ay kilos years | kilos 1 10 28.58} 0.590) 158 62.2 1 6.3 | 0.454) 0.522 10.001— 20 [ 81.2 | 0.535) 40] 104.4 7 17.1 | 0.473) 0.504 20.001— 30 137.5 | 0.545 13 | 149.5) 18 | 42.0 | 0.327) 0.436 30.001— 40 199.6 | 0.562 98 | 160.6} 48 | 62.5 | 0.319) 0.441 40.001— 50 235.8 | 0.522) 165 | 165.3) 51 66.5 | 0.355} 0.439 50.001— 60 276.0 | 0.505) 127 | 167.2) 50 | 67.5 | 0.409) 0.457 60.001— 70 323.7 | 0.470} 59] 170.0} 50 | 71.0 | 0.456} 0.464 70.001— 80 360.8 | 0.488} 23] 169.8) 53 | 71.0 | 0.508} 0.488 80.001— 90 424.6 | 0.502 7 | 169.3) 55 | 70.0 | 0.606) 0.554 90.001—100 382.6 | 0.401 3) PLA OW 252 74.0 | 0.578) 0.459 100.001-110 400.4 | 0.391 3 | 176.6) 63 | 76.0 | 0.527] 0.459 Rt ant 0.534 0.404; 0.469 Weighted average above 1 eats tence Pear ahs ysrekets 0.528 0.390) 0.459 TABLE 11 Relative heart muscle weight in females as reported by W. Miiller compared with the relative weight when the body weight is normal for height and age a ry t =] - a é 3 SI 5 = 5 a BODY WEIGHT fe 5 & A 2 : 2 3 & e & Seis eel aS g Bbilese lige as a ad ms a 2 as) OB Bm ee Bp S eS a Fae le pee < a Z < < 4 Ay yi kilos years kilos 1- 10 29.2) 0.584) 171! | 6325) 13) 1656: |) 04443) 0.514 10.001— 20 74.8) 0.504, 41) 107.0' 8 | 18.0 | 0.416! 0.460 20.001— 30 139.6] 0.554; 20) 144.3) 18 | 44.0] 0.317) 0.436 30.001— 40 187.0) 0.532) 144 | 153.0) 54 | 58.0 | 0.323} 0.428 40.001— 50 224.5) 0.499} 137 | 156.3} 51 | 60.0) 0.374! 0.437 50.001— 60 252.5] 0.457| 55 | 159.1) 49 | 62.5 0.404 0.431 60.001— 70 270).3) 0.420) - 28 |} 161.1) 51} 64.7 | 0.418) 0.419 70.001— 80 283.9) 0.386 6 | 160.3) 66 | 63.5 0.447 0.417 80.001— 90 227.8] 0.280 1 | 162.0; 23 | 56.6 | 0.402) 0.341 _ 90.001-100 363.6) 0.400 1} 166.0) 64 | 66.5 0.547 0.473 100.001-110 316.6) 0.302 1 | 163.7; 46 | 65.5 | 0.483) 0.393 Weighted average.......... 0.523 0.388] 0.455 Weighted average above 1 MOAN ae encyaiceate repeals aust es stejos 0.500 0.367) 0.434 460 Cc. R. BARDEEN the body weight been normal for height and age we get the re- sults shown in the column at the right in tables 10 and 11. The weighted average for males is 0.469 per cent, for males above one year of age 0.459 percent. The weighted average for females is 0.455 per cent, for females above one year of age 0.434 per cent. In six of the weight groups of the males the per cent of mean body weight is less than 0.46 per cent, in five greater than 0.46 per cent. The highest percentage 0.554 per cent is found in the 80-90 K. group. The lowest 0.436 per cent in the 20-30 K. group. In eight of the weight groups of the females the per- centage is below 0.46 per cent, in three 0.46 per cent.or higher. The highest percentage, 0.514 per cent is in the 1-10 K. group, the lowest, 0.341 per cent, in the 80-90 K. group. Thoma from a mathematical study of a less extensive material than that of Miller but carefully selected, found an average heart-muscle body-weight ratio of 0.463 per cent. This figure lies midway between the averages given above for the mean body weight in men and that in women. In round numbers we may take the heart muscle weight to be approximately 0.46 per cent of the body weight in individuals of normal build, slightly higher in men, slightly lower in women, higher in thin individuals, lower in fat individuals. . The Miller data arranged according to height (table 12) show the relative heart muscle weight ranging from 0.501 per cent to 0.532 per cent, in men, but this relatively high figure is due as pointed out above to the inclusion of a large proportion of relatively thin individuals. According to Miller there is an increase in the relative weight of the heart with age irrespective of size of body as illus- trated in table 13. However, it must be remembered that there is normally an increase in weight during adult life until old age comes on. The heart enlarges to meet this increase in weight. Sickness reduces the weight of the body usually more than the weight of the heart musculature so that after death the heart may appear relatively large in proportion to body weight. The relative distribution of the musculature within the heart has been studied by several investigators among whom the work DETERMINATION OF SIZE OF HEART BY X-RAYS 461 TABLE 12 Relative heart muscle weight in individuals grouped according to height after W. Miiller nove reno ee ee Men 1501-1550 27 228.5 0.00532 1551-1600 62 219.1 0.00502 1601-1650 115 227.6 0.00528 1651-1700 89 216.9 0.00501 1701-1750 62 222.0 0.00519 1751-1800 29 224.9 0.00522 384 Women 1401-1450 21 213.3 0.00495 1451-1500 55 213.4 0.00488 1501-1550 98 210.1 0.00487 1551-1600 93 200.1 0.00455 1601-1650 51 204.5 0.00498 1651-1700 25 204.5 0.00465 1701-1750 9 208.9 0.00497 352 TABLE 13 Relative heart muscle weight in adults grouped according to age after W. Miiller MALES FEMALES aaa Absolute heart Relative heart Absolute heart Relative heart weight weight weight weight 20-40 243.3 0.00493 190.6 0.00432 40-60 265.0 0.00499 216.6 0.00473 60-80 274.9 0.00513 240.9 0.00523 Over 80 258.9 0.00606 203.9 0.00539 of W. Miller seems to have been the most extensive. He found the musculature of the left ventricle to weigh approxi- mately twice that of the right ventricle and the musculature of auricles to weigh approximately one-fourth that of the ventri- cles. The weight of the values of the heart is about two per cent of the weight of the heart as a whole. THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO. 2 462 Cc. R. BARDEEN Taking the relative weight of the heart musculature to be 0.46 per cent of the body weight we may next consider the weight of the non-muscular structures. Of the non-muscular structures fat constitutes the chief mass after early infancy. The most careful study of the amount of fat in the heart is that of W. Miiller. He has shown that amount of fat in the heart increases with age. In the new born there is relatively little but later in life it constitutes no inconsiderable part of the weight of the heart. Part of the fat is as a rule easily removed with the peri- cardium but about 8 per cent can be removed only when special methods are used. The average amount of fat may be illustrated by the following table based on data from Miller (table 14). The body weight is estimated from data given in table 8. The relative amount of fat may be judged by comparing the weight of the heart as a whole at various ages with the weight of the heart muscle as shown in the following table (table 15). The data concerning the weight of cardiac fat relative to body weight given in table 14 do not quite correspond with the data of table 15, owing probably to some variation in the individuals composing the various age groups but the differences in the two tables are not serious. To some extent, at least, the greater amount of fat found by Miieller in the older age groups is due to the greater emaciation of the younger as compared with the older individuals studied by him. A rough estimate of the de- gree of emaciation may be made by comparing the average body weight for each of the Miller age groups with the body- weight estimated as normal for a given age and height in tables A and B. This comparison shows that. while the new born in Miiller table are of about normal weight male infants during the first year of life are nearly 50 per cent underweight and during the second and third years about 33 per cent underweight. After this period the average body weight for each age group appears to be only from 25 to 30 per cent under the weight of healthy living individuals during youth, 20 to 25 per cent during adult life until the oldest age period when the average body weight seems to be 40 per cent underweight. The female in- fants appear to be 4 to 50 per cent underweight, girls up to ten Amount of fat in the heart at various ages after W. Miiller a | & z F 5 PERICARDIAL FAT IN : a ee = re PER CENT Aca aa sities ‘rom | mum | OF Beant | pg (=e <5 aS S| Bs zs a < & grams Premature births ...:... 42) | 706 | 00 0 0 0 Mature births |23 | 20.79) 0 0.1 0) 0 imionthiv.. 5-|45 16.19} 0O 0 0 0 0 Prmonths .-) 14] 0 0.510} 0.000 3- 4 months.. .|22}| 20.13] 0 0. 0.520) 0.000} 1.6 5- 6 months.. .|14] 0. 0. 2°10) |FOHOOR es 7-12 months...|34 | 30.64) 0. .079 iL BalOWOROOR Rona DoWSayeatse a) 134) | 52070 | a2: Pos) ens: 8.10] .00| 5.9 4- 5 years..... 16) 6225, |) 5281010242, 16.23) 9. 465\e2.57 11080 6-10 years...../15 |103.6 | 9.2 |0.7 OO Bae Wale || Ba 11-15 years..... OU GAS 47a aile2= sihe9) 6.8) |SeOunleoed 16-20 years..... 23 |236.9 sSpel2ee 109 72-4 5. Semloed 21-30 years ....|73 |297.4 4 12.5 Ni OO WL 31-40 years..... 70 |289.6 oe |259 LOS eGR 72 NOkGm Niges 41-50 years.....|84 [304.2 Ox 9 |240.4 | 9.7 |16.4 51-60 years.....|87 [340.8 2 14.4 ‘6 1266.2 113.0) izes 61-70 vears.....|88 [345.9 De yeni AE a) age) NTIS 5 71-80 years...../64 {335.5 oe a 169.4 | 9.4 |20.8 81-90 years..... I isan 0 |4.6 146.7 |47.9 |19.8 Premature Dirths.4.<) 6.8 ASE a7e 200 0 0 0 0 Mature births. .|14 19.241 O 0 0 0.3 0 0 0 Gmronths 5... 47 | 14.36, 0 0 0 0 0 0 2 months...... 14| 0.1050. 0. 1.10: | 0} Worse 3- 4 months.. .|28} | 20.18] 0.138/0. 0. 1.60 al ie i 5- 6 months.. .|10] 0.6340. 0. 1.63 | 0 3.4 7-12 months.. ./32 | 32.14] 1.37 |0. ily 3.70 | 0 4.1 0 2- 3 years.....|42 | 45.2 | 3.09 |0. 3. 7.03 | 0 73 0 4_ 5 years...../19 | 69.0] 5.86 (0. 6. 17 NO 9.4 0 GAOpvears. ulm S258) oelsalO.73 19.91 1723) ROR TONG 0 11-15 years..... 10) 7A etGe Sy 3 = 17 6s | 15e3h Poe sen One 0 16-20 years..... 13° |215s20B23R2ee eS: - (25.2) ADoser OM KG 0. 21-30 years..... AD 220RGHIESOn20 124 (32.60 | 74a 2ealeonOen its 2 0 31-40 years.....|59 |234.7 | 38.2 |3.0 [41.2 | 85.0 |10.9 /17.6 0 41-50 years.....|69 |264.1 | 45.9 |3.7 56h HOLE Sa ees ILS 0 51-60 years..... 61 |256.9 | 44.2 |3.5 Tila Ne Se nle Oh te. (19%3 0 61-70 years..... 83 |285.1| 52.2 |4.2 |56.4 |192.0 |19.4 |19.8 0 71-80 years...../61 |294.3| 56.4 |4.5 |60.9 [179.1 |14.4 [20.7 0 81-90 years.....|12 [253.0] 49.1 [3.9 |53.0 | 79.7 |25.8 |20.9 0. TABLE 15 Relative weight of the heart, of the heart musculature, and of cardiac fat after data from W. Miiller NUMBER RELATIVE RELATIVE RELATIVE = poprouase | THIGHD OF | orsicun or, | WatGan oF Males per cent per cent per cent Bivthiae ee eee aoe oe 23 0.620 © .620 0.000 te weeks epee oh: ams ot ob 18 0.645 DWT CCIE eae Ws cee. sin abe 13 0.627 0.016 BU CISC ARES nic, bre prea ee 10 Wee 0.655 A ECan: -5,...5 site 5) 0.645 Dutik OMG GRAY rts 6 sie ae te. 15 0.590 SUE OMG Speer <<. sacs oid creo 14 0.576 0.563 0.013 AAOBIMOMENS s. 2. . 1. a Gere 24 J 0.557 (MRM ONGOS': - - es... 6 wee 34 0.597 0.580 0.017 Do GEES ae RS BE SA ae 17 0.557 0.058 BMYICRES! Foe 5 sec sas toke eeateae 13 ue 0.522 A PORVGAT S:, o'= ws. ctacwarpeeks sane tnt 16 0.580 0.498 0.087 GOLORVEATS ..: sc centers wale ake 16 0.623 0.542 0.081 Sey Cats':. «of Waa anne 8 0.600 0.514 0.086 1GS20 VOATS: ..45. 22 eee gee 23 0.548 0.491 0.057 PSE VOCALS... a aaeeNe sas ss 69 0.580 0.500 - 0.080 AQ VORTS:. «oss cs Bae oe 69 0.561 0.486 0.075 gol ONY CATS 3c eee oa ae . 84 0.585 0.494 0.091 MA GHONVCALS: Saree Bas pode Sere: 87 0.615 0.504 0.111 Gil AO nViCalseer oneness ane 87 0.640 0.522 0.118 l= SOny CATSHs oh ee oma ce 63 0.637 0.504 0.183 Sil Inv EarSiere case eee 11 0.746 0.606 0.140 Females aire eee. 3. .c ee ee ee 14 0.629 lRweeks.. «(Aer ane 18 0.624 DEW COKS «cee et TOE 13 0.652 Siete CMS beeen alae - 10 Ube 0.578 ATWGOKS: . Hees eked eee 5) i 0.649 e) STIG HS che sein ae eee 15 0.613 DdaKou late wake oss eb ges eae 14 0.610 0.583 0.027 Ha OMMONtOS

ls]olels R Zz B a ve SE ee aE a) | | Volume underestimated M 2 Tl 1 490-990 lok +20 1 |1(31.2) 510 M 4 |1(48.6) a) alt 210-525 i) F +15 1 1 410 2| M +10 2 a a 600-685 M 6 D1 all a tlk 287-1000 ile ee 1 a es 1 1 55 22 |18M 17 |3 a} 3) ai ai alal | 4 | 3F 1G Volume as estimated gs | M 6 |1(37.8) | 1 9 | 2 350-790 Bal 8 Fy EA) 5 Dol | et 290-400 TaB to — 2! 3 9} | 1(41.5) 15-105 16 | 8M 14 1 1 2| 216 | 2 | | 5F | | | | 3B | | DETERMINATION OF SIZE OF HEART BY X-RAYS 469 TABLE 17—Continued Volume overestimated M 6 1 ili 1| 2(1-30) 300-810 1 F il 1 510 I B — 5 1 1 16 6 M 6 |1(63.0) wy) aut a 2 320-700 1 F —10 475 4 M 4 1} 3(25, 31, 33.5) 390-1155 1 B —15 1 1(36.7) 75 2 M 2 2(21.8, 34.2) 560-600 1 F —20 al 1 290 24 |19M eee Sal eS 31 e2, 5} 9 3F 2B Weight underestimated! Weight overestimated 53| 5 |3j4l3 |e 7 5 12 ___. _—_— — SS = 15 30 Weight correctly estimated 8 determined for each heart. The hearts were then grouped according to the extent of divergence of observed from calculated volume. Hearts showing a divergence of 23 per cent or less are classed together. The others are classed to the nearest 5, 10, 15 or 20 per cent + or —. Markedly distorted hearts have been excluded but no attempt has been made to select hearts that conform to theory. For the sake of comparison the divergence of observed heart weight from the heart weight calculated from silhouette area is likewise given. This table shows that of the 62 hearts included in the study the volume was correctly calculated from the area to within 23 per cent in 16 cases, to within the nearest 5 per cent + or — in 36 cases and to within 10 per cent (the nearest 10 per cent + or —) in 45 cases. It is probable that greater accuracy can be obtained in estimating heart volume from silhouette area in 470 Cc. R. BARDEEN the living than in the dead since the condition of the heart with relation to the distention of its chambers is more uniform in the living. We believe that the formula given above enables one to calculate diastolic volume from silhouette area to within 5 per cent of the volume in diastole in the majority of instances in the living. In cadavers there is a tendency to underestimate volume from silhouette area when the heart is contracted; to overesti- mate volume when the heart is more distended than is normal in diastole. Whether or not this is true in the living we have no means of ascertaining at present. If we take the heart weight considered standard for a person of a given body weight as de- scribed in Section C, p. 449—and given in tables A and B, and the silhouette area considered standard for a given body weight as described in Section A, p. 431, and given in tables A and B we see that there is a constant relation between silhouette area and heart weight if each is assumed to bear a constant relation to body weight. We may express this relation by the formula: sy area” < 0.0055 = heart weight. The area is here assumed to be the area in square centimeters of the heart silhouette in diastole while sitting at rest and the heart weight that of the whole heart in grams. If the heart is more contracted than is normal in diastole when the body is sitting at rest the weight of the heart in relation to the silhouette area is increased. If the heart is more dilated than is normal for this position the weight of the heart in relation to the sil- houette area is decreased. We thus have a method of deter- mining in a more or less rough way whether or not a heart in the cadaver is more or is less dilated than is normal in diastole when the body is at rest in the living. In table 17 the percentage of divergence of the observed from the heart weight estimated from silhouette area is given for 53 of the 62 bodies in which the relation of silhouette area to volume was studied. From this table it may be seen that of the 17 hearts whose volume was un- derestimated from the shadow area, ten were underestimated from the standpoint of weight and four overestimated. This DETERMINATION OF SIZE OF HEART BY X-RAYS A471 would indicate that there is a tendency to underestimate volume from silhouette area when the heart is contracted. On the other hand of the 22 bodies in which the volume was overestimated from the shadow area, the weight was overestimated in 16 and under- estimated in 3 indicating that these hearts were more dilated than is normal for diastole at rest. This same condition is, however, also true of the hearts in which the observed volume fairly closely corresponded with the estimated volume. Of the fourteen hearts in this group in 10 the weight was overestimated, in 2 underestimated. Of the total of 53 hearts studied in 15 the weight was under- estimated from the silhouette area, in 30 overestimated. We may therefore assume that the method used in preparing the bodies tended in the main to cause a somewhat greater disten- tion than is normal in diastole in the living under the conditions described above. We may likewise estimate heart volume from heart-weight, which we have assumed to be 0.55 per cent of the body weight. To determine a formula to express the relation of heart weight to diastolic heart volume we need to know the relation of heart weight to heart tissue volume and the relation of the volume of heart tissue to the volume of the heart and its contents in diastole. In order to estimate tissue volume from heart weight we have to determine the specific gravity of the heart. Vierordt, quot- ing Davy, gives 1049 as the specific gravity of the left ventricle. I have estimated the specific gravity of a considerable number of fresh dog hearts, of one unembalmed human heart and of numerous embalmed human hearts. The method used was to measure the displacement of the heart in oil and to estimate the specific gravity from this. The heart was in each case freed from extraneous substances but the subepicardial fat was left in place. The chief difficulty met with was to get rid of air bubbles. To aid in this the heart was cut into sections. For exact work the displacement should be measured in a vacuum but this was deemed unnecessary for the purpose in view. While there were individual variations, due chiefly to differences in the amount of subepicardial fact, the figure 1050 was selected as a 472 Cc. R. BARDEEN round number which expressed with fair accuracy the specific gravity of the heart as a whole. The volume of the empty heart in centimeters may therefore be taken as equal to the weight of the heart in grams divided by 1050. The ratio between the volume of the empty heart and that of the heart in diastole can be estimated from cadavers and from experimental work on animals. The chief difficulty lies in the determination of the volume of the heart in diastole. In order to determine the ratio between the volume of the empty heart in dogs and the volume of the heart in diastole, I have made a number of experiments in coéperation with Dr. J. A. E. Eyster and other members of the department of physiology at the University of Wisconsin. The dog was weighed and its pulse at rest under morphine was counted before beginning the experiment. The animal was then anaesthetized, the thorax opened and ligatures were placed about each of the vessels en- tering the heart. With the help of several assistants these liga- tures were tightened simultaneously at a given signal so as to close off the vessels during diastole. The heart was now re- moved from the body and its volume estimated. It was emp- tied and the volume of the cardiac tissue was measured and its weight determined. The ratio of the volume of the empty heart to that of the heart in diastole could then be ascertained. In order to make the pulse correspond approximately with the normal pulse as determined before the experiment, or some- what slower, the vagus nerve was stimulated during the experi- ment to the requisite amount. The chief difficulty in the ex- periment is that of tying off all the vessels simultaneously at the height of diastole. 7 Table 18 shows the result of six experiments. The percentage of the diastolic heart volume occupied by the blood in the heart chambers varied from 26 to 46 with an average of 40.6. It is probable that the smaller percentage represents a heart in which we did not succeed in tying off all the vessels in diastole. If we omit this heart the average becomes 43.5. The average empty heart volume in these dogs was therefore 59.4 per cent if ex- ‘periment 5 is included, 56.5 per cent if this experiment is not DETERMINATION OF SIZE OF HEART BY X-RAYS 473 included. It is of interest to note that the heart of the dog weighs more in relation to body weight than the human heart does and is subject to wider variations. ‘This is in accord with the observations of Joseph (08). In the human heart the percentage of the diastolic heart vol- ume occupied by the blood in the cavities appears to be greater than in the dog, the percentage occupied by the heart muscle less. In the study of embalmed cadavers the empty heart vol- ume was found to vary from 33.8 per cent to 80 per cent of the volume of the heart as a whole. The hearts, the outline of which seemed most closely to correspond with radiographic outlines of TABLE 18 Relation of diastolic volume to the volume of the empty heart in the dog WEIGH | porgx | WEIGHT |2ERCENT|pr.rorzc|| voruun | or | “UME | on Dog RATE HEART WEIGHT VOLUME EMPTY pepe: BLOOD Beare kilos grams 1 | 10.4 SOR Oeil 10 7G on) 102.64)" so8ek 73.9 41.9 Zr We 12).0 96 84.0 | 0.7 135.0 80.0 | 59.2 55.0 40.8 3 1) 10.2) 110 57.1] 0.56 106.0 60.0 | 56.6 46.0 43.4 4 | 13.0 for W250 0286 195.0 | 106.7 | 54.7 88.3 45.3 5 8.36} 120 70.0 | 0.84 90.0 66.7 | 74.0 26.0 26.0 6 | 16:0.) 120) . 135.5), 0.847 | 238.9) 129.0 | 54.0 | 10929 46.0 PANY GIG O Clstcess, o-0b8, od eee 0.801 59.4 40.6 the living diastolic heart, showed an average percentage of about 49.4 per cent heart tissue, 50.6 per cent heart chamber space. If the specific gravity of the heart be taken as 1050 and the per- centage of diastolic heart volume occupied by heart tissue be taken as 49.4 per cent we may estimate diastolic heart volume from heart weight by dividing the latter by 1.050 x 49.4 or 0.5187. The results thus obtained may be compared with the estimates given in tables A and B, in which the heart weight is estimated from the body weight and this in turn from the sil- houette area while the volume is estimated directly from the silhouette area. The use of round numbers in the tables gives rise to slight divergencies but otherwise the estimates of heart volume based on silhouette area and those based on heart weight correspond. ATA Cc. R. BARDEEN The ratio between heart volume calculated from heart weight in bodies studied in the anatomical laboratory and the meas- ured heart volume is shown in table 19. The hearts are grouped according to the extent of divergence of the observed from the calculated volume. Those showing less than 23 per cent of divergence are grouped together. The rest are grouped accord- ing to the nearest 5, 10, 15, 20, 25, 30, 35 and 45 per cent of TABLE 19 Hearts from cadavers grouped according to percentage of divergence from the assumed ratio between empty heart volume and diastolic volume AVERAGE ¥ EXTREMES OF PER CENT OF SEX HEART VOLUME DIVERGENCE PERCENTAGE OF | cupen OF cases ee Num- | HEART WEIGHT Male |Female ber BODY WEIGHT cases RATIO ‘ cc. +45 1 1 562 1 — 5.2 +35 1 1 470 il =13).2 +30 2 2 950-990 1 S11! 25 1 1 365 1 =i) 0 +20 7 (1 child) 6 i 75-687 6 ae (et!) +15 4 (1 child) 3 1 105-1000 4 + 3.2 +10 6 (1 child) 4 2 70-790 6 = 1 ap ty) 5 (1 foetus) 2 3 290-453 4 = 0.5 + 23 to —21 | 10 (1 child) 8 201" 167-810 8 es — 5 5) 5 0 210-700 7 — 026 —10 3 3 0 360-1155 2 + 4!5 —il5 D, 2 0 350-420 2 — 3.9 —20 4 (1 infant) 3 1 15-520 3 — 9.4 = 75) 1 , 1 0 320 otal. c. 2264| 52) 41 11 46 positive or negative divergence. Separate columns show the number of males and females in each group, the extremes of heart volume (including contents of chambers) and the average per cent of divergence from the ‘normal’ heart weight-body weight ratio of those of each group for which records were preserved. From this table it may be seen that while the greatest number (10) of hearts fall within the group assumed to show normal DETERMINATION OF SIZE OF HEART BY X-RAYS A475 diastolic volumes (+ 24 to — 24 per cent divergence), the number (27) of those which show a volume above what is assumed to be the normal diastolic volume is greater than the number (15) which show a volume smaller than the normal diastolic volume. This is what we should expect from the conditions of the hearts studied. The post-mortem condition of the heart has been studied by several investigators including MacWilliam (01) and Roth- berger (’04). At the time of death the heart is in diastole. The amount of blood in the heart depends on the general circulatory conditions at this time. After death there is a tonic contraction of the heart followed by a rigor mortis contraction. The post- mortem contraction of the heart is usually much greater in in- dividuals in whom the respiration stops before the circulation than in those in whom heart failure is a primary cause of death. The postmortem contraction is followed by a subsequent dila- tation but the extent of this depends to a large extent on the amount of fluid blood under pressure when the dilatation occurs. The bodies received at the Anatomical Laboratory at the University of Wisconsin have usually been dead at least a week. As a rule they are embalmed by injecting equal parts of alcohol, glycerine and carbolic acid into the femoral arteries and the thorax is not opened until the body is dissected. In some in- stances we have opened the thorax in order to study the condi- tion of the heart before embalming. As a rule the right atrium is fairly well distended with blood and frequently there is con- siderable blood in the right ventricle. While there is usually some blood in the left atrium this is less apt to be distended than the right atrium. The left ventricle is usually practically empty. When the embalming fluid is injected under a pressure of five or six pounds into the femoral arteries it usually enters the chambers on the left side of the heart and distends them to a moderate degree. The right side of the heart is less affected by the injection than the left side. The embalming fluid is usually followed by a shellac and Prussian blue arterial injection mass which also usually partially fills the chambers in the left side of the heart but not those on the right side. We have not meas- 476 Cc. R. BARDEEN ured the pressure of the fluid in the heart at the time of em- balming but it is probably considerably higher than the pressure in the heart during life at the beginning of systole. When the injection is completed both the right and left sides of the heart are probably as a rule more distended than is normal during life; the right as a result of natural factors active just before and follow- ing death, the left as a result of the pressure of the embalming fluid. The embalming fluid causes some shrinkage. The end result appears to be in many cases a heart having approximately the size of the living heart in diastole during bodily rest. The dilatation of the various chambers is probably seldom quite the same in the cadaver heart as in the living but the heart as a whole frequently appears not dissimilar in outline. If there has been an antemortem acute dilatation of the heart or if the em- balming fluid causes unusual distention we may have a heart large in proportion to the weight of its component tissue. If less blood than usual is sent into the right side of the heart before death or if the distention of the heart by the embalming fluid is less than usual or the shrinkage greater the size of the heart in relation to the weight of its component tissue is relatively small. The table shows that no clear relation exists between the weight of the heart compared to the weight of the body and the cadaver size of the heart compared with the weight of the empty heart. The best estimate which we can make of the ratio of heart substance to heart: content is on the one hand from the heart- weight-body weight ratio based on post mortem studies, on the other hand from the heart-silhouette area-body-weight ratio based on x-ray studies of the living. But it is of interest to see how closely the estimates thus made are approached by direct studies on the hearts of embalmed cadavers as shown in table 19. e. Ventricular output The chief interest in arriving at an approximate knowledge of heart content in diastole is in relation to the systolic output of the heart. Various methods have been used to determine the DETERMINATION OF SIZE OF HEART BY X-RAYS AT7Z amount of blood discharged from the heart at each systole. While the results have been far from uniform the results of the more recent work including that of Krogh and Lindhard (12) and Lindhard (’15) appear to indicate that the output of the human adult heart at rest is not far from 1 ee. per kilo of body weight per beat. Since the weight of the heart substance may be esti- mated at 5.5 gr. per kilo, its specific gravity as 1050 and its vol- ume at about 49.4 per cent of the volume of the heart in diastole the volume of heart content in diastole may be estimated as 5.365 ce. per kilo. About 20 per cent of the blood in the heart in diastole is thus sent into the aorta at each systole during rest. If we estimate one-third of the blood in the heart during diastole to be contained in each ventricle and one-third in the two atria we have 60 per cent of the contents of the left ventri- cle sent into the aorta at each systole during bodily rest. In the upright position the diastolic heart is smaller than in the sitting position and in the sitting position than in the prone position. It appears that to the lessened hydrostatic pressure in the inferior vena cava and to the moderate exertion accom- panying sitting and standing the heart accommodates itself by beating faster, contracting more completely during systole, and expanding less during diastole. Nicolai and Zuntz have shown, however, (14) that during severe exercise the heart expands more during diastole than when at rest. Muscular action acts as a pump to force blood toward the heart. In all probability the heart also contracts more completely so that the output of the heart is increased by pulse volume as well as by pulse rate. The experimental work of Henderson and Barringer on the dog which has led these investigators to opposite deductions does not seem to me at all conclusive. In order to test the estimate of heart content in diastole given above and to estimate the reduction in size of the heart during systole we have devised with the collaboration of Dr. J. A. E. Eyster, an apparatus for taking ‘instantaneous’ radiographs of the heart at any desired period of the cardiac cycle. The mechanism is adjusted to the carotid pulse. As a rule two successive radiographs are taken on the same plate, one at the THE AMERICAN JOURNAL OF ANATOMY, VOL. 23, NO 2 478 Cc. R. BARDEEN height of systole, one in diastole and the outlines of the two superimposed shadows are compared. The pictures are taken at the usual distance of two meters. Two intensifying screens are used, one on each side of a photographic film. Drum tracings of the respiration, carotid pulse and of the period of exposure are made while the pictures are taken. The estimates of change of heart volume from diastole to systole based on these plates correspond well with the data given above as the following table will show (table 20). In the sitting position observations were made on sixteen in- dividuals. For one individual two: sets of observations are recorded in the table. During the change in heart volume from diastole to systole blood from the ventricles is forced into the pulmonary artery and aorta. Since the systolic picture was taken as nearly as possible at the height of ventricular systole it is possible that in most cases diastole had already begun in the atria and some new blood had entered these chambers. The actual output of the heart may therefore have been some- what greater than that estimated from the change in the size of the silhouette area fron diastole to systole. We have however shown above from studies on cadavers that there is a tendency to underestimate volume from silhouette area when the heart is contracted so that to some extent the error due to diastole filling of the atria is offset by the error due to underestimation of volume from silhouette area. The average output per beat in the sitting position was esti- mated as 37.8 per cent of the cardiac contents or 18.9 per cent from each ventricle. This corresponds closely with the 20 per cent estimate based on the work of Lindhard, as outlined above. The lowest output was 27.4 per cent of the cardiac content or 13.7 per cent from each ventricle. The largest was 58.2 per cent or 29.1 per cent from each ventricle. If we estimate the ventric- ular content as 334 per cent of the blood in the heart in the lat- ter case the ventricle was nearly completely emptied at each contraction while in the former case it was less than half emp- tied. In eight out of the sixteen cases the per cent of cardiac blood expelled varied from 39.2 per cent to 41.8 per cent or close DETERMINATION OF SIZE OF HEART BY TABLE 20 X-RAYS 479 Volume of heart estimated from diastolic and systolic silhouette areas, difference im volume, percentage of reduction in heart volume and percentage of heart blood expelled during systole a 3 } = S S a as Be SUBJECT S 5 4 Semoats REMARKS 4 = a a5 % 5 8 S oy 58 | 2 D E eo | aS 5 a a Caan ee GCs cc. Gee A. Sitting position J. F.S., 5’ 10”, 135 Ibs.......|691..2/595.8) 95.4) 13.8 | 27.4 E. J. V., 5’ 63’, 148 Ibs..... .{861.8/741 .0|120.8|140.3 | 7.8) Enlarged heart E. F. S., 5’ 9’’, 155 lbs....... ./723.9/619.2)104.7) 14.46] 28.7 rare f{|675 .0|580.0| 95.0) 14.1 | 27.9) LE AN a | 699 .0|590.0}109.0| 15.6 | 31.0 JOG, 5 3% 115) bss. + 2.-|620701522/0) 98.0) 15.8 | 31.3 R. W. T., 5’ 81’, 143 lbs..... .1667. 4/555 .6/111 .8) 16.75) 33.3) pes DAG 4 GO bse 3th) ats: 862 .0/692.0)170.0) 19.72) 39.2 (CC OVE OVO, UGS Lbsiea 5. 732 . 91587 .8|145.1| 19.8 | 39.3 Crores Oe lO ola bse asce 2 612.0/490.0|122.0} 19.93) 39.5 C. E. G., 5’ 114”, 164 lbs... . ./775.0,619.0)156.0) 20.13} 40.0 Av Weeote Ae 229) Noses. 3d cacti 657 .8]522.0)135.8) 20.64} 41.0 S. A. M., 5’ 8’’, 171.5 lbs... .../658.41522.0/186.4| 20.72) 41.2 P. M. D., 5’ 113’’, 146.3 lbs... |676.0/535 0/141 .0| 20.85} 41.3 F. C. K., 5’ 8’, 154 lbs......../750.0/586 .0/164.0] 21.87) 41.8| Rather large heart Ee Wes: We 142Ibs ee 699 .6)/514.0/185.6| 26.25} 56.1 Crys a? dO"; 145 lbsxseske 715 .8|506 .0/209 .8} 29.3 | 58.2 PAR CT EVE C ner serie. cia weve Micra Sua ORT oye | eres een 19.04| 37.8 B. Prone position E. J. V., 5’ 64’, 148 lbs..... . .|887.0/792.3] 94.7) 10.68) 19.2 R. W. T., 5’ 8}’’, 143 lbs.... . .|658.9|579.6| 79.3] 12.04) 23.9) W. E. G., 5’ 43’’, 130 Ibs... . .|603.0|522.0] 81.0] 13.43] 25.4! J. F.S., 5’ 10’’, 135 lbs....... . 741.9635 .2/106.7| 14.38) 28.5 E. F.S., 5’ 9’’, 155 Ibs...... . .|723.9)619.2|104.7) 14.46) 28.7 H. A., 5’ 92’’, 152 lbs........./699.0}590.0)109.0). 15.60) 31.0 M. D. W., 5’ 9’’, 1387 lbs...... .|851.8/715.8/136.0} 15.96) 31.7 C. E. G., 5’ 114’, 164 lbs..... .|835.6/698.7|136.9) 16.39) 32.5 NSERC NORE Bie Ova, 5 dry Gi vcore paces | Okcee Oke | RR Nee renee 14.1 | 27.6) 480 Cc. R. BARDEEN to the estimate given above of 20 per cent from each ventricle, 60 per cent of the ventricular content. In the prone position we have observations on eight individ- uals. The average output was 27.6 per cent of the cardiac con- tent or 13.8 per cent for each ventricle; 41.4 per cent of the ventricular content. The extremes are 19.2 per cent of the car- diac content, 9.6 per cent for each ventricle, 29 per cent of the ventricular content; and 32.5 per cent of the cardiac output, 16.8 per cent for each ventricle, or 50.4 per cent of the ventricu- lar content. The estimates of percentage output of ventricular content are based upon the assumption that one-third of the blood in the heart in diastole is to be found in each ventricle, one-third in the two atria. It is probable however that in the prone position a greater proportion of the blood in the heart in diastole is to be found in the atria and that the percentage output from each ventricle is greater. On the assumption that in the prone position there is an equal amount of blood in each chamber of the heart in diastole the percentage output from each ventricle would average 55.2 per cent, with variations from 38.4 per cent to 62 per cent. The ‘relation of cardiac output as determined by the method given above, to various factors has been studied in our labora- tories by Mr. E. J. Van Liere. He found no correlation be- tween body weight, height, or build and the proportional amount of blood expelled at each contraction of the heart. Hearts whose diastolic volume was 5 per cent or more above the normal as compared with body weight showed less proportional cardiac output than normal and small hearts. High pulse pressure was accompanied by large relative output in a given position, al- though in the prone position the pulse pressure was higher than in the sitting position while the relative output was smaller. No definite correlation between systolic pressure or pulse rate and output was found under the conditions of the experiment. The average ieft ventricular output sitting was 80.8 cc. per kilo per minute, the average pulse rate 82 making the output approximately 1 cc. per kilo per beat. The average output DETERMINATION OF SIZE OF HEART BY X-RAYS 481 lying was 56.9 ec. per kilo per’ minute with a pulse rate of 67 or 0.85 ee. per kilo per beat. f. Relation of size of heart to heiqht, age and sex In the preceding sections we have considered heart size chiefly from the standpoint of body weight with which it 1s most closely correlated. In case of a given individual, however, other fac- tors than merely body weight must be taken into consideration before we can form an accurate judgment as to whether or not the size of the heart is normal for that individual. Of these other factors the chief are height, age and sex. The size of the heart for a given body weight is estimated in the tables on the assumption of normal height for that weight for a given sex and age. The chief studies on the relations of weight, height, sex and age in the adult have been made by the insurance actuaries. The most important of these studies is the Medico-Actuarial Mortality Investigation vol. 1 published by the Association of Life Insurance Medical Directors and the Actuarial Society of America. For the period of childhood and youth we have a large number of studies made on school children of which special mention may be made of those of Roberts C7883). He. Pe Bowditeh! (755479, 9), > Burk. (98) a Key. (89), F. Boas (96-97), Hastings, W. W. (’02), Baldwin (14), W. T. Porter (94) and Ethel M. Elderton (14-15). The pio- neer work in this general field is that of Quetelet (32, ’48). These studies have shown that a closer correlation between height and weight is found if age and sex be taken into considera- tion than if these are ignored. The figures given in tables A and B are based upon an analysis of the data available in the literature together with studies made in the Clinical Department at the University of Wisconsin. A full account of these studies is reserved for publication in a subsequent paper. The figures for height are those which these studies have led us to believe represent a fair normal average for a given weight, for the age and sex indicated in healthy Americans. Weight means weight without clothes; height, height without shoes; age, age at nearest birthday. 482 Cc. R. BARDEEN Table A gives figures for childhood and youth. Table B gives figures for adults at three ages, 20, 30, and 50. In making use of these tables one compares the parellel ray silhouette area of the heart of the individual under consideration (a) with the silhouette area given by the table as normal for a person of the individual’s weight and (b) with the silhouette area normal for a person of the individual’s height, sex and age being taken into consideration. If the silhouette area is normal for weight or for height or is intermediate between the two we consider that the heart is one of normal size. If the heart volume correspond- ing to the silhouette area is more than 10 per cent too large both from the standpoint of height and of weight we consider that it is disproportionately large. If it is correspondingly small both from the standpoint of height and of weight we con- sider it disproportionately small. Our own practice is thus to estimate cardiac size in percentage of variation of volume from that assumed as normal for height and from that assumed as normal for weight. For instance we will suppose that a man 30 years of age 5’ 10” tall and weighing 150 pounds shows a heart silhouette area (reduced about 6 per cent to allow for divergence of rays if a radiograph is used) of 120 sq. em. From table B we find that an area of 120 sq. em. corresponds to a volume of 696 cu.em. Fora weight of 150 Ibs. we should expect a volume of 723 cu.em. Inaman 5’ 10” tall at 30 years of age we should expect a volume of 768 cu. em. The heart of the individual under consideration is therefore 27 cu. em. or 3.7 per cent below the standard from the standpoint of weight, 72 cu. cm. or 9.1 per cent below the standard for height. A slight variation of this kind is within the limits of error of the method used and the heart would be considered of normal size. In conclusion I desire to thank the members of the staffs of the departments of anatomy, physiology and clinical medicine and Prof. Max. Mason of the department of physics for valuable aid in carrying out the investigations described in this paper. DETERMINATION OF SIZE OF HEART BY X-RAYS 483 LITERATURE CITED ALBERS-SCHONBERG 1908 Die Bestimmung der Herzgrésse mit besonderer Beriicksichtigung der Orthophotographie. Fortschritte aus der Gebiete der Roentgenstrahlen, Bd. 12, p. 38. Batpwin, B. T. 1914 Physical growth and school progress. United States Bureau of Education Bulletin no. 10. BarpDEEN, C. R. 1916 A standard of measurement in determining the relative size of the heart. Anat. Rec., vol. 10, p. 176. BENEKE, F. W. 1878 Die anatomischen Grundlagen der Constitutions-Anoma- lieen des Menschen. Boas, F. 1896-1897 Growth of children, U. S. Bureau of Education, Report, vol. 2, p. 1541-1599. Bowpitca, H. P. 1875, 1879, 1891 The growth of children. Mass. Board of Health Reports. Boyp 1861 Phil. Trans. Royal Soec., London, vol. 151, Pt. 1. Burk, F. 1898 The growth of children in weight and height. American Journal of Psychology, vol. 9, p. 253-326. Cuaytor, T. A. anp Merriti, W. H. 1909 Orthodiagraphy in the study of the heart and great vessels. Amer. Jr. of the Med. Sc., vol. 138, p. 549. DererRMANN, W. 1900 Die Beweglichkeit des Herzens bei Lageveranderungen des Kérpes. Zeitsch. f. klin. Med., Bd. 40, p. 24. Dretien, H. 1906-1907 Uber Grosse und Lage des Herzens und ihre Abhingig- keit von physiologischen Bedingungen. Deut. Archiv f. klin. Med., Bd. 88, p. 55. DintiteNn, H. 1909 Klinische Bedeutung der Veriinderungen am Zirculations- apparet bei wechselnder Kdérperstellungen (Liegen and Stehen). Deut. Arch. f. klin. Med., Bd. 97, p. 132. Experton, Erser M. 1914-1915 Height and weight of school children in Glasgow. Biometrika, vol. 10, p. 288. GinceL, R. 1914 Die Klinische Verwertung der Herzsilhouette. Miinch. Med. Wochenschrift Bd. 61, p. 220. GroepEL, F. M. 1910 Beobachtungen iiber den Einfluss der Respiration auf Blutdruch auf Herzgroésse. Zeitschr. f. klin. Med., Bd. 70, p. 47. GREENWOOD AND Brown 1913 , cee 175 DvuesserG, J. Chondriosomes in the testi- ele-cellsyon vem Uls es afeicisccieteie icles eteress/elere 133 | ere of turtles. The formation and struc- ture of the zona pellucida in the ovarian. 237 Embryos. On the age of human.......... Bo Gill Embryos with a note on similar structures in reptiles. Vestigial gill filaments in chick. 205 ILAMENTS in chick embryos with a note on similar structures in reptiles. Ves- (ateaeail, fori lS ae aie Sad Ota leceairs 205 Fontanella metopica and its remnants in an Gye hvilbie irl LAS hoy eenee heads canwaoohoe 259 Formation and structure of the zona pellucida in the ovarian eggs of turtles. The...... 237 Frog larva—observation and experiment on the living animal. Studies on the growth of blood-vessels in the tail of the.......... 37 Fundulus. Chondriosomes in the testicle- Cells) Of his rare reclaniaten croaiiennsteeces 133 ILL filaments in chick embryos with a note on similar structures in reptiles. WWiEStiei al map eertmeney ely ateeee ai cstetars ucuctoielere! Ve 205 Hee by means of the x-rays. Deter- mination of the size of the.. .......... 423 Heart-beat upon the development of the vascular system in the chick. The effect Of thet s ere te eroe bone etag doen 175 Human embryos. On the age of............ 397 Humerus of man. The position of the inser- tion of the pectoralis major and deltoid muscles! OM theseaeseeerenitcsrasnecsee cee 155 OHNSON, FRANKLIN PARADISE. The isolation, shape, size, and number of the lobules of the pig’s liver................ 273 | eee Asram T. The branchial plexus of nerves in man, the variations in its formation and branches 316.605 Son 285 ARVA—by observation and experiment on the living animal. Studies on the growth of blood-vessels in the tail of the 1080) SR Ren DESO nonce dociyc)s.5 ool eee 37 Liver. The isolation, shape, size, and num- ber of the lobules of the pig’s............ 273 Lobules of the pig’s liver. The isolation, shape, size, and number of the........... 273 Luteum in the ovary of the chicken. Sex Studies oX. Lhe(corpussepeeeeee ees cc 1 ALL, FRANKLIN P. On the age of human EMbDEVOSSs sic eee O eee 397 Man, the variations in its formation and branches. The brachial plexus of nerves TD 0s (9s web aichay aks: caw noc ata e e is 285 Metopica and its remnants in an adult skull. whe) fontanella..o ae ane. reOo Muscles on the humerus of man. The posi- tion of the insertion of the pectoralis major and! deltoid: «..<\<:..0 gee oe eee es tes 155 ERVES in man, the variations in its for- mation and branches. The brachial plexus) Of... sos siacsotenciooemeiehioe seis'elc 285 @* ARIAN eggs of turtles. The formation gad structure of the zona pellucida in Ona ae the chicken. , Sex studies. X. The Corpus) luteum) inythereeeeeeereaeee ccc EARL, RAYMOND AND Borine, Auice M. Sex studies. X. The corpus luteum in the ovary of the chicken... ............. Pectoralis major and deltoid muscles on the humerus of man. The position of the in- sertion/of thes preemie ce esc ccceeee « 155 Pellucida in the ovarian eggs of turtles. The formation and structure of the zona...... 237 Pig (Sus serofa). The fate of the ultimobran- chial bodies in the.....-. 89 Pig’s liver. The isolation, shape, size, “and number of the lobules of the... 273 Plexus of nerves in man, the variations in ‘its formation and branches. The branchial.. 285 EPTILES. Vestigial gill filaments in chick embryos with a note on similar AULICHITES PLL el tdlepal-lese sis'e.eloie elelstets;= alain ioe 205 489 490 ° CHULTZ, Avotr H. The fontanella me- S topica and its remnants in an adult skull. Scuutrz, Apotr H. The position of the in- sertion of the pectoralis major and deltoid muscles on the humerus of man.......... Sex studies. X. The corpus luteum in the Ovary OL thevehicken ees. - ose eee ie Size of the heart by means of the x-rays. termination voteube: 6. 6.3.0 seceiaaeeloeens Skull. The fontanella metopica and its rem- MANS ID ANUAAUIG | ok oe eee eee eee Structures of the zona pellucida in the ovarian eggs of turtles. The formation and...... Te of the frog larva—by observation and Stud- INDEX 259 Turina, Aticr. The formation and structure of the zona pellucida in the ovarian eggs Of turtles? ves-ctcn wns eu eee eee ore ; Turtles. The formation and structure of the 237 zona pellucida in the ovarian eggs of...... 237 Le bodies in the pig (Sus scrofa). The fate of the.......... YASCULAR system in the chick. The effect of the heart-beat upon the de- velopment) Of the poe. --o-h eee ey. Vestigial gill filaments in chick embryos with 175 a note on similar structures in reptiles.... 205 -RAYS. Determination of the size of the heart by means of the...........-++-+.<- 4 Y Aarts pellucida in the ovarian eggs of tur- tles. The formation and structure of es om | | | ui | 1] TCO Cave ecru ey Et ay SITET het an PEC CED LODO Tar CoO Cnet ISHS HEV ECU CHOOT Oto in Waiter? Ht { yeeteidceyes petite a LOCC LECH) TUR TC) AACA CASES D he Teetnyeedter WCU LUT NU Vestereaye eeepc rerce ree (1G Ter COUPER ERE COREEE DEERE PTeCP CPN Peedey, cried re cue es VEC ete sperenarant COTE OTE eed ul Hritperi fer { HEU i rt ft i ; , est ? $ t : PAU Pattr eta taa aedreattnagey i byte tbe ‘ rite rf iff. j eyedat TUT: Hedin canbe et oH Steet rt stint i polert male +EESELE 1s Sit : { i uth i i { t rf f rey { ‘ CECT ¢ COT Cea IE Caircgaers (sey Vad teecedneed ieeceeedayeryy Vetere atppereee(reesas faut , : SEPTALGE Hintited facies Weer iy SURGE etter tet Seen oe hese LS peas SASS at) t iti its ait Hee eet} Hate : Mt Tact 4 : i $ Peetteetineany hy 4 4 aritetire ater ate 1 a ? : fi Fitts ECE, it iayattreti fl i ] Wt TEITE Hiern iii Aiba tits ts, , y if 3 ECTS CUE EGE a 4 2 / H Tefitittiel Sa Fee LGTEL TEP a a . 4 ‘ ii (elites ey tr] Piet Hi iTstratts ita Hs} i f fi eatin (ibeabnedfeectendsligg re Hatt SHARE HEA A Sea a +f PYet FE i i ENE TIREI ECT Een ALES Puss rset TAMEHRE NEP net AN MHPE ate AITE i j j ity Gbatiiaisstnatty x ==" tae iasead pie Ht yeep } renee tins Hate ates COLES UNAS LER EE Pestle asin { felis seacencebelebetd ee Oe freerriti titre r { | APA tad ae seetaunaye i ih FUERTE ERR \ CSERREAE CADE LOSE ESTEE UREA LEE LARACBEA TASCA SE ASTRO TACT ES REELED RLT ih cranE petra ciscatny (HbeUbbasbaseadeacsagl REE i : Ai : Hit : [ j PEECUEUEeL ey eal teh sst ys AA ie 4 aaties TH ; WGh ; i Peres Crating HIT StHt eiieeH ti) RAE str} Hy : : { tiff AO : Vebitanecteteedncayny ee eit basicseashot ts STTTEFTPTE nt i Hittite t fsdatisds reriise eeMaabestiah THY rots oe rants eS thOACLAL LE ; ‘ Fiteeitrieiits ut ; ny a (HARE ER ELL ik Ade ohn bing We EA Scent ert ; RAbectiectiuyes A ie UE i { (ie th I ists treet Itt tits ) ant ‘ VECHUUA AISLES f iL : f \ Miya tet bib hy Atta Alea 44 tite A a ih Latest i | } { ie ee ee ety HOAs OCU se ----- < a speseee Hay SSS CECT ELAS IONE nonce en nem ean ye ae pe