Talk:Paper - Development of the human intestine and its position in the adult (1898)

From Embryology
Revision as of 14:37, 16 February 2020 by Z8600021 (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)


Mall FP. Development of the human intestine and its position in the adult. (1898) Johns Hopkins Hospital Bulletin 9: 197-208.

By Franklin P. Mall, Professor of Anatomy, Johns Hopkins University.

Our knowledge of the early development of the human intestine is very complete, and at first thought it seems impossible to contribute anything new to it ; yet, when we consider the topographical anatomy of the adult intestine, we are struck by the fact that there is dispute regarding the position of its various parts, and nothing is known about the development of its convolutions.

The aim in this study has been to follow the successive stages of the development of the human intestine, loop by loop, from the simplest form in the embryo to the adult. As a result, it has been found that the various loops of the adult intestine, as well as their position, are already marked in embryos of five weeks, and that the position of the convolutions in the adult is as definite as the convolutions of the brain.

The present study is closely associated with one recently published upon the development of the human coelom, and the embryos here described were also published in part at that time.f In that paper the shifting of the viscera was emphasized in connection with the development of the coelom, while in this paper only the convolutions of the intestine are considered.

•This paper has appeared in German in the Festschrift fur Professor Wilhelm His zum 22. October, 1897, Archiv fur Anatomie, Supplement-Band, 1897.

f Mall, Journal of Morphology, vols. 12 and 14.

The set of specimens in my possession is fairly complete, as all the important stages are represented. It is self-evident that a subject like this can be studied only by resorting to models, as a simple comparison of sections gives no opportunity to study the loops. A number of important stages were selected aud modeled according to the method of Born. A list of these embryos is given in the accompanying table : Table op Embryos Modeled. Lengtb in mm.


V. B.

N. B.

of Model.

From Whom Obtained.






Ellis, Elkton, Md.






C. O. Miller, Baltimore.






Eyclesheimer, Chicago.






W. S. Miller, Madisou, Wis.





C. O. Miller, Baltimore.






Douglas, Nashville, Tenn.






Ellis, Elkton, Md.






Wilson, Worcester, Mass.

These models were then compared with one another, in order to follow the growth of the loops from stage to stage, using as guides the outline of the intestines in the sections and the blood-vessels, as well as the dissections of other embryos and those of the adult.

The loops which appeared to be homologous in the various models were next painted with the successive colors of the spectrum, beginning with the duodenum, and ending with the



[Nos. 90-91.

csecum. In this way, loops whose position were at first obscure, were finally found to have meaning. It is noteworthy that the successive stages in the development fit into one another accurately, showing that the first loops in the embryo are destined to form certain loops in the adult, and that this primary folding is in no way a haphazard process.


The observations, upon the human embryo, of the first formation of the alimentary canal from the entoderm have now been extended to the very earliest stages.* In Graf Spec's embryo v. H. the entoderm lines the wh ^le of the umbilical vesicle, and is in no way incorporated within the body of the future embryo. In fact, its plane is curved away from the entoderm, and is just the reverse of its direction in a later

The next older stage is found in Graf Spec's embryo Gle, in which there is shown the beginning of the fore-gut. These two stages, given by Graf Spee, are the important ones to make our knowledge of the develojiment of the alimentary canal complete, and from them we can easily follow thi'ough the successive stages until the adult form is reached.

After Graf Spee's embryo Gle, we have next to observe the constriction of the umbilical vesicle from the entoderm. The beginning of this constriction is already well marked in Kollmann'sf embryo Bulle, my embyro No. XII, J and His's§ embryos SR and Lg. Unfortunately, we have no data regarding the extent of the alimentary canal in Kollmanu's embryo Bulle, nor His's embryo SR. My embryo XII, however, is of about the same stage as the other two, and it has been cut into sections which are about perfect. The history of this embryo, as well as its coelom, have been described by me recently, so I need not repeat them at this time.

EMBRYO No. XII. (2.1 mm. LONG).

The Figures 1 and 2, on Plate I, give the external form of the embryo, as well as the extent of the alimentary canal, which was taken from a reconstruction. The entoderm is already divided into fore-gut, mid-gut, and hind-gut. The fore-gut marks the pharynx, from which there are four diverticula on the dorsal side ( Br', Br"), one on the ventral side (T), and two near the mouth ( M and S ). These diverticula mark the first two branchial pouches, thyroid gland, mouth and Seessel's pocket respectively. At the junction of the pharynx and the umbilical vesicle there is a large diverticulum of the entoderm into the septum transversum, L, the beginning of the liver.

The hind-gut is a sharply defined cavity lodged in the tail of the embryo, communicating on the one hand with the allantois. All, and on the other with the neural tube by means of the neurenteric canal, N. C.

The attachment of the umbilical vesicle to the body indi

Graf Spee, His's Archiv, 1889 and 1896. tKollmann, His's Archiv, 1889 and 1891. t Mall, Journal of Morphology, vol. 12, 1897, p. 417. 'i His, Anatomie mensch. Embryonen, 1885.

cates the extent of the mid-gut from which the future intestine is to arise. The coelom is already beginning to be incorporated into the body to form the body cavity, and in the region of the liver and the omphalo-mesenteric vein the peritoneal cavities of the two sides of the embryo communicate freely, showing that at this early stage there is no complete ventral mesentery as has been described. This communication, marked 0, gradually approaches the communication above the allantois, 0', and ultimately cuts off the umbilical vesicle altogether. A stage just before the umbilical vesicle

Fig. a. — Reconstruction of Embryo No. II. Enlarged 17 times. V and X, fifth and tenth cranial nerves; 1, 2, 3 and 4, cast of the branchial pocliets ; 1 and 8, first and eighth cervical nerves ; 12, twelfth dorsal nerve; A, auricle; V, ventricle; L, lung; S, stomach; P, pancreas; WD, Wolffian duct; K, kidney; M, mesentery; ST, septum transversum; O, openings which communicate with the peritoneal cavity on the opposite side.

is completely separated from the embryo is represented in Fig. A, taken from embryo No. II.* By comparing Figs. 1 and A it will readily be seen that the spaces marked and 0' in the two embryos are the same.

The intestinal canal of embryo II is given in Fig. 3. The drawing was made from the right side and gives the irregularity of the tube more accurately than my previous figures have done. The part between the liver duct and the cjecum, of course, marks the extent of the small intestine, and the part behind this, the large intestine. At this early stage, therefore, the cajcum is distinctly outlined. Attached to the small intestine there is this marked umbilical stem, but the vesicle no longer communicates with the intestinal canal. From the

Mall, Jour, of Morph., vol. 6; and vol. 12, p. 429.


September-October, 1898.1



umbilical stem there haugs down an extensive papilliform process which, from its appearance in section as well as its presence in younger embryos, shows that it is nothing more than an island of vessels and villi from the umbilical vesicle. These seem always to be incorporated within the body at this point and degenerate later on.

Thus far it is very easy to follow the formation of the intestine, when the embryos already described by His are also taken into consideration. My embryos, Nos. XII and II, are from the end of the second and fourth weeks respectively, so it takes about two weeks for the intestine to become outlined after the entoderm is incorporated within the body of the embryo. The intermediate stages have all been described by His in his great monograph. In his Atlas the external form of embryos intermediate to XII and II is given, and in the text the alimentary canal of embryos Lg, BB, Lr and R is again pictured in woodcuts. They all show the gradual constriction of the stem of the umbilical vesicle to form the intestinal tube between it and the liver.

As the umbilical vesicle is being separated from the intestine, all of the viscera are moving from the anterior end of the embryo towards its tail. This is also the case with the diaphragm and the origin of the cceliac axis and the superior mesenteric arteries from the aorta.* A comparison of figures A and 1 shows that the whole stem of the umbilical vesicle in embryo XII must have moved toward the tail through the space of at least ten body segments to have gained the position it holds in embryo II.

At the same time that the intestine is bending towards the ventral median line the loop is also beginning to turn upon itself, so that the aboral end moves towards the left side, and the oral end to the right of the body. This process is already beginning in embryo II, Fig. 3, but rapidly becomes more marked, as is beautifully shown by the His embryos and their models made by Ziegler. By this process the loop is separated into right and left halves, the left half to form the large intestine, and the right half, the small intestine. In a short time, however, as the loof) grows longer and longer, not all of the left half is occupied by the large intestine, as the cascum is now no longer in the middle of the loop.


As the loop of intestine enlarges it extends immediately into the umbilical cord, as was first shown by Meckelf for the human embryo. To what extent this is common to the mammals is not known, but my experience is that it is frequently found in other mammals, and from the examination of many pigs' embryos I can state that in them a portion of the intestine always extends into the cord.

Figures 4, 5 and B are from embryo IX, a specimen about five weeks old. The intestine extends into the cord as a single loop, with the plane of its mesentery horizontal to the long axis of the body. In general its arrangement is much like that

of His's embryos Si, Sch*, KOf and 11M|. It is noticed in the figures that the large intestine lies altogether within the sagittal plane of the body, a position it retains until the intestine is returned to the peritoneal cavity proper. The right half of the loop has a number of small bends in it, which are of great importance in the further development of the intestine. I have marked them with the numbers 1, 2, 3, 4, 5 and 6 in order to follow them with greater ease in the drawings of older embryos.

•Mali, Jour, of Morph., vol. 12, pp. 441 and 442. t Meckel, Meckel's Archiv, 1817.

Fio. B. Reconstruction of Embryo No. IX. Eulars^ed 20 'times.

ST, septum transversum; L, liver; S, stomach; C, caecum; W, Wolffian body; K, kidney; 1 to 13, dorsal f^anglia; O, omphalo-meseuteric artery; SC, suprarenal capsule; X, communication between pleural and peritoneal cavities.

In the middle of the mesentery of the loop and in the median line lie the omphalo-mesenteric vein and artery. At the point where these vessels cross the intestine. Fig. 4, u, we have a landmark which is of use in comparing the intestine of this embryo with that of older embryos. The point of communication between the umbilical vesicle and the intestine also represents the position of the persistent Meckel's diverti

His, A. m. E., Ill, p. 19. fHis, Abhandl. d. sach. Gesch., XIV, Taf. II, Fig. 3.

JHis, ibid.XV, p. 677.



[Nos. 90-91.

culum. If the adult intestine is about six meters long, and if the distance from the cjecum and Meckel's diverticulum is about one meter, then the length of the intestine between the omphalo-mesenteric vessels and the caecum is about one-sixth of the whole intestine. In both embryos IX and X, as well as in His's embryos Si and Sch, the extreme bend of the intestine (Fig. 4, u) marks one-sixth the distance from the crecum to the duodenum.

The blood-vessels to this whole loop within the umbilical cord arise from the omphalo-mesenteric or the future superior mesenteric artery. AVhen this is compared with the arterial supply in the adult intestine it is again found to correspond. In this early stage the omphalo-mesenteric artery supplies the same portions of the intestine that the superior mesenteric artery does in the adult. Not only by the form of the large intestine, but also by its blood supply can we divide it into two portions, that jwrtion which is at right angles to the body, supplied by the superior mesenteric artery, and that parallel with the body and supplied by the inferior mesenteric artery.

The relation of the intestine and liver to the body of the embryo is given in Fig. B.

Fig. C. — Reconstruction of Embryo No. X. Enlarged 8 times. 1 to 13, dorsal ganglia; SC, suprarenal capsule; W, Wolfflan body; K, kidney; L, liver; S, stomach; C, ciecum.

BEGINNING OF THE CONVOLUTIONS. A stage somewhat older than the one just described is given in Figures 6, 7 and C. In comparing Figures B and C it is seen that the liver has descended decidedly ; it has moved

away from the head to the extent of at least three segments. While in embryo IX the septum trausversum is opposite the eighth dorsal nerve, and the lower edge of the liver opposite the first lumbar nerve, in embryo X the septum is opposite the eleventh dorsal, and the lower edge of the liver opposite the second sacral nerve. In other words, the septum has descended three segments and the lower edge of the liver six segments. Not only has the liver descended through its absolute growth, but the whole organ has descended ajso. This movement has had a marked effect upon the form of the large intestine, and the direction of the intestine in general, as the figures will readily show.

While this movement is taking place the convolutions are also becoming more and more distinct. Every loop as outlined in embryo IX is more marked in embryo X. In general, the twisting has become more pronounced as the caecum is approached. The loops 1, 2 and 3 are only slightly more bent in X than in IX, while the loops 4, 5 and 6 have become much more sharply defined. In general, the length of the loops has doubled itself while the diameter of the intestine increased but one-third.


Pjo d. — Reconstruction of Embryo No. VI. Enlarged 8 times. S, stomach; SC, suprarenal capsule; C, ca'cum ; K, kidney; W, Wolffian body.

The next embryo (VI) I have modeled is only slightly larger than No. X. I give the same views for this embryo as I gave for Nos. IX and X. Fig. D compared with Fig. C gives the general relation of the intestine to the body. The large intestine has not changed its position much ; it has

Septembee-Octobee, 1898.]



elongated, but has uot increased its diameter. There is now added to the cisecum a marked vermiform appendix. The stomach has descended more than before, the great omentum forming a sac and extending well over the large intestine.

A comparisou of Figs. 6 and 7 with Figs. 8 and 9 shows the growth of the small intestine. The duodenum is still bulged at its stomach end and only the lower portion of it is as small as the rest of the intestine. This enlarged duodenum is so decided that, at first sight, one might think it belongs to the stomach, but since the liver and pancreatic ducts open into it in all the specimens, there is no doubt but that it belongs to the intestine. Of course there is the possibility of these ducts shifting, but this seems to me very improbable.

The second portion of the intestine, 2, is now curved towards the dorsal side of the embryo, and, as in embryo X, this is also the case with its mesenteric attachment. We are all familiar with this portion of the intestine in section, as it has this hooked mesentery showing that the intestine has bent backward. The next portion, 3, is bent upon itself to such an extent that it rolls around on the dorsal side of the omphalomesenteric artery, to project to the left side of the clump, as shown in Fig. 8. It also has the hooked mesentery in section, as the mesentery is very much bent upon itself.

It has been customary for embryologists in discussing sections of the intestine with me to call this portion of the intestine the duodenum, on account of its position, as well as for its very characteristic mesentery. At first I was strongly inclined towards this view, but more mature consideration of models convinced me that both the loops 3 and 3 are finally transferred to the left side of the body to form the upper part of the jejunum.

The fourth portion of the intestine. Fig. 8, 4, has its beginning in the earlier embry.j on the left side of the mesentery as shown in Fig. 6. It is readily seen by the comparison of the two figures that the loop 4, in Fig. 8, is only an exaggeration of the same in Fig. 6. While this loop begins on the left side it ends on the right. In all the figures the extreme bend of the loop 4 is marked a, and a comparison of the figures will readily show that this is always the homologous loop.

Following the loop 4 there is the loop 5, which is altogether on the right side in embryo X, Fig. 6, and about equally distributed on both sides in embryo VI. In both Figs. 6 and 8 the end of the loop 5, b, approaches the loop 3, and this relation is also present in Fig. 10. In the figures this point is marked b, and the similarity of this loop in them is very apparent. Loops b and 3 are just touching in Fig. 6, while in Fig. 8, through the elaboration of loop 4 and its gliding to the right side of the mesentery, the loop 5 has been brought nearly in contact with loop 3. At any rate the relations of loops 5 and 6 to the umbilical vein in both figures show that the numbering of the two loops is not far amiss.

The loop b in Figs. 4, 6, 8 and 10 holds the same relation to the ca3cum and umbilical vessels in all four embryos. This point seems to be the fixed point for the loop on the right side of the mesentery as the point marked a is for the left. Between b and the ctecum the intestine is thinner and the loops are smaller than in the upper part of the intestine. |

For this reason as well as for the fact that there is no sharp landmark other than the umbilical vessels between loop b and the cascum, I have classed this whole region as one group and marked it with the single number 6.

The convolutions in embryo XLV, Figs. 10 and 11, are only an exaggeration of those of embryo VI. The loops 1, 2 and 3 are much the same as before. The loop 1 is again defined by the extent of the head of the pancreas. The loops 2 and 3 together are now S-shaped instead of a simple curve as in embryo VI. My interpretation of this is that the loop 3 is held in place by the opening of the umbilical cord, as at this point the intestine leaves the body, while the loop 2 is beginning to rotate to the left side of the body with the rest of the intestine. The loop 4 on the right side has enlarged, however, and has pushed its way in between the loops 3 and 5. On the left side the loop 4 has made for itself another twist, so it now appears as several loops. The loop 5 is much as it was before, only it has increased its length somewhat. It is easy to see the loop 6 of embryo VI converted into that of embryo XLV by imagining x in Fig. 9 to be drawn over to the opposite side to form x in Fig. 10. In so doing x' and x" remain back to form the loops marked the same in Fig. 11. In addition to this the loop y in Fig. 9 has become bent over to the left side to form y in Fig. 11.

All these twists and curves in the small intestines of the four embryos just descril)ed can be followed fairly well in the figures, and the reader may think that there is considerable imagination required to do this. Any one, however, who may study the models in which the corresponding loops have been marked as in the figures will not doubt regarding the accuracy of this description. It is a most remarkable fact that four specimens should correspond as well as they do here. Were the whole affair more or less haphazard no comparisou whatever could have been made.


A comparison of embryos II, IX, X, VI, His's Pr and KO, shows that the change in position of the intestine and its future twisting is due to the descent of the abdominal viscera, accompanied by the relatively rapid growth of the small intestine. The following table gives the measurements of the intestines in these embryos as well as the level of the stomach

ind the caecum. The measurements of the intestine of embryos Pr* and KOf have been taken from the illustrations given by His and are only approximations. The position of the stomach here given is its lowest measurements including the omental bag.

Table giving tue Position and Length of the Small Intestine.

Length of Intestine.

Number of Embryo- Position of Stomach.

II 1 Dorsal

3 Lumbar 5 Lumbar 1 Sacral

Position of Csecum. Small. Large.

(• Dorsal 1.7 mm. 1.5 mm.

10 " 3. 1.5

13 '< 4. ;i

5 Lumbar 9. 3.7

3 Sacral I'J. 7.

1 Sacral ' 34. 8.

52. 8.

His, Atlas, Theil III, Taf. 1, Fig. 4. fHis, Abhandl. d. K. S. Ges. d. Wiss., U

Bel. XIV, No. 7.



[Nos. 90-91

A study of this table by comparing it with the illustrations shows that the intestine is gradually elongating and at the same time being pushed towards the pelvis by the large liver and other organs descending upon it. In No. II the intestinal canal is still a comparatively straight tube, but in Pr and KO it is already well bent, is much larger than in No. TI and is pushed into the cord. In No. IX it is located in the cord. From No. II to No. IX the small intestine has increased its length five times and the large intestine over two times, and the space which they should occupy within the body has remained the same. Under these conditions the intestine must escape if it has a chance, and the coelomic space within this cord naturally receives it. This movement of the intestine is due to mechanical causes and, were the ccelom of the cord not there to receive it, the intestine no doubt would make room for itself within the body. I do not think that the umbilical ducts had anything to do with it any more than to keep the opening in the cord open, for, before the intestine begins to enter the cord, its connection with the duct is severed.

After the intestine has entered the cord. No. IX, Fig. B, the small intestine grows rapidly, as the table and Figs. C and D show. Embryos IX, X and VI are all about the same size, but no doubt VI is considerably older than IX ; the organs are all firmer and more developed, and the small intestine has increased its length considerably more than the large intestine. The organs have all been pressed down to the pelvis as far as they will go, as Fig. D. shows. In so doing the large intestine makes a sharp bend in the neighborhood of the fourth lumbar segment, in all of the embryos given in the above table. This bend, therefore, may be looked upon as a fixed point toward which the viscera descend, but beyond which they do not go. Of course after the intestine is in the cord the loops may descend lower, but within the body this is a very fixed point.

After the intestine is within the cord its further elongation and its mesenteric attachment causes it to be thrown into coils, as shown in the plates. The large intestine lies, however, in the sagittal plane of the body, partly within the body and partly within the cord. It does not grow as rapidly as the small intestine ; and, as the small intestine is folded into coils, the whole begins to rotate around an axis which is identical with that of the large intestine. By this process the small intestine is gradually turned from the right to the left side of the body, and in so doing is rolled under the superior mesenteric artery. This takes place while the large intestine has an antero-posterior direction and before there is any transverse colon. This latter is the result of a kinking which is to follow, and is in no way formed by a shoving of the large intestine over the small, as given in the Hertwig diagrams.


Although it is comparatively easy to understand how the intestine leaves the peritoneal cavity to enter the cord, it is extremely difficult to see how and why it returns. When the intestine enters the cord the communication of coelom with the body cavity is very free and the intestine is small, but when

the intestine returns to the body cavity the intestine is large, while the opening is small. Fig. D.

In embryo No. II, and in younger embryos, the belly stalk is very large and contains within it no muscles nor permanent blood-vessels of the abdominal walls of the future individual. It is not until the muscles wander, carrying with them their nerves and to a certain extent their blood-vessels, that the belly wall is finally completed.* In embryo No. VI, for instance, the rectus abdominis is about half-way around from the dorsal to the ventral median line, thus leaving a large area between the two recti, which is little more than a membrane. It seems that, until the abdominal walls are fairly completed, the intestine remains within the cord, and, at the last moment before the two recti come together in the middle line, the intestine returns to the peritoneal cavity.

In very young pigs' embryos, when the mammary ridge is still over the muscle plates, I have found that the segmental arteries form an anastomosis with one another throughout the extent of this ridge. This artery goes through a series of muscles which have just been split off from the muscle plates. As the embryos grow, the mammary ridge wanders towards the ventral median linef and carries with it this anastomosis of segmental arteries and the portion of muscle plates which are destined to form respectively the internal mammary, deep epigastric arteries and the rectus abdominis muscles. The nerve connections of the various segments of the rectus are formed as the muscle is splitting off from the muscle plates, and in this way the origin of the different parts of the rectus is indicated, as already shown by Nussbaum. What I have here described for the pig can also be verified for the human embryo, and this will make it plain how the lateral body walls are formed from the belly stalk.

But the closing off also takes place from above downwards. In an early stage, while the septum transversum is still in the neck, the umbilical vesicle also extends upwards. The heart is first closed off by the beginning of the membrana reuniens, and the ventral wall is completed by the amnion moving over the embryo from left to right.J Then the umbilical vesicle is pinched off from above downwards, corresponding with the descent of the liver and other viscera. In embryo II the stalk extends from opposite the third dorsal vertebra to opposite the second sacral, while in embryo IX it extends from opposite the second lumbar segment to opposite the fourth sacral. In other words, the oral end of the stalk has receded eleven segments, and its aboral end two segments ; or the whole stalk is moving away from the head, and its attachment to the body is rapidly becoming smaller and smaller. Later the growth of the abdominal walls is greater between the cord and the pelvis, as shown by the sections made by Merke].§

At one time I thought of the possibility of the expansion of the coelom of the cord and its incorporation with the abdominal cavity, and this was also carefully investigated. Were this

See also Nussbaum, Verhandl. d. anat. Ges., 1894, '95 and '96. to. Schultze, Anat. Anz., Bd. 7 ; and Verhandl. d. phys. med. Ges. Wurzburg, Bd. 26. t Mall, Journal of Morphology, vol. XII. i Merkel, Abhandl. d. k. Ges. d. Wiss. in Gottiugen, Bd. 40.

Septembee-Octobbe, 1898.]



the case it would be necessary to find stages in which the rectus abdominis had wandered up into the stalk to incorporate it and to enclose the intestine within it. No such stage has ever been found, while on the contrary the recti nearly close the communication between the stalk and peritoneal cavities, before the intestine slides back into the body.

The return of the intestine into the body must take place very rapidly, for I have never seen a specimen in which it is in the process of returning. Embryos 40 mm. long either have the intestine in the cord or in the peritoneal cavity, and, if it is in the latter, the communication between the cord and peritoneal cavity is open and surrounded by a thin membrane, showing that it also is being closed. This membrane now closes the whole opening, and later the recti muscles wander into it to complete the abdominal walls.

Since I was unable to find the desired stage in the human embryo, I examined a number of pigs' embryos, hoping in this way to find stages in which the intestine is returning to the peritoneal cavity.

In a pig's embryo 12 mm. long a single loop of intestine extends into the extra-embryonic cojlom, beyond the cord and on the right side of the body. It is still in communication with the umbilical vesicle, which, in turn, is attached to the ventral body wall over the heart. As the loops of the intestine increase in number in older embryos, they make room for themselves below the liver and in front of the Wolffian body, for, unlike the human embryo, there is considerable space in this region. In general the greater number of loops remain within the body cavity of the embryo, and the loops within the cord are not numerous. The increase of loops within the body cavity and their rotation seem to draw upon the loops within the cord, so that when the embryo has reached 35 mm. in length, the loops have all returned to the body cavity.

No doubt, in the human embryo some similar mechanism is present in the return of the loops to the peritoneal cavity, but, as the critical stage has not yet presented itself, this question must be left open for future observation.


Although it is extremely difficult to understand how the intestine returns to the peritoneal cavity, it is not difficult to recognize the various loops after their return. Unfortunately I have not been able to study carefully a good stage between embryos XLV and XXXIV, as the various specimens of this stage at my disposal were not perfect, or if so, the series of sections were broken. It was not until a number of good specimens had been spoiled for this present purpose that I found that, by removing the ventral abdominal walls, good series could be obtained. If this is not done, the intestine is very liable to be imbedded poorly, and as a result the sections are not perfect. Dissected specimens, on the other hand, cannot be relied upon, for, when once handled, it is impossible to replace the intestines to their original position with certainty, unless they have been modeled as soon as the abdominal walls were removed. Of course dissection is an extremely good method of control, but for a comparison of the loops I think that no method will improve the model. Then it was found

that all the loops represented in No. XLV are again recognizable in XXXIV, and on this account it is believed that an intermediate stage is unnecessary, unless that stage is one in which the intestine is in process of returning from the cord to the peritoneal cavity.

A comparison of Figs. 10 and 13 shows that the loops of XLV are again represented in XXXIV. The marked change is that the mesentery of the large intestine has increased greatly. The loops of the upper part of the intestine have rolled completely to the left of the superior mesenteric artery, and the loops which were formerly within the cord have now been transferred in Mo to the right side of the body. While this has been taking place the stomach has been enlarging also, and by the tilting of the intestine the pyloric portion of the stomach, d, has come nearer the cscum, about to the point marked d' in Fig. 10.

This shifting of the loops, half to the right and half to the left side, as well as the sliding down of the stomach towards the Cfficum, has finally locked the duodenum (loop 1) around the root of the mesentery, as shown in Figs. 13 and 13. As the loops come out on the left side. Fig. 13, we have the beginning of the second group of loops (3) of the intestine. The deeper layer of these loops, not shown in the figures, is a single curve lying immediately in front of the mesocolon. The loops 3 together can easily be imagined as arising from the same as in Fig. 10 by a simple bending of the portion on the dorsal side of the large intestine towards the ventral median line. The loop 3 which lay formerly on the right side of the body is now altogether on the left side. In the illustration, with the exception of its ending in loop 4, it cannot be recognized as the loop 3 of Figs. 10 and 11. In the models, however, the loop 3 forms a distinct cluster situated between the loops 2 and 4, and therefore, by exclusion, it must represent the loop 3 of embryo XLV.

The similarity of the loop 4 in embryos XLV and XXXIV is very striking. Its beginning, the arrangement of convolutions and their position are much the same in both embryos, if the change in position of the intestine in the older embryo is taken into consideration. The next cluster I have marked 5, and the remaining portion 6. The loops 6 are again smaller, and their diameter is less than those of the upper part of the intestine.

In this embryo we can see pretty well the adult form of the intestine, only that the mesentery is transverse to the body rather than diagonally downward toward the right side. In this embryo we also see the relation of the intestine to the mesentery better than in the adult. At this time the mesentery is relatively simple, as but few secondary adhesions have taken place. Unraveling this stage into the one represented by embryo XLV, we can fully understand the relation of the mesentery to the abdominal viscera. These two stages represent beautifully the arrangement of the intestine in the dog and monkey respectively.

The following table gives the length of the various loops of intestine in the embryos described. The measurement was taken along the distal border of the intestine and therefore is considerably longer than along the mesenteric border. But this is a border easily measured, and the length of the hard



[Nos. 90-91.

ened iutestiue is considerably less than where it is taken out fresh and stretched. The table shows that the lower end of the intestine grows more rapidly than the upper end until the intestine is returned to the peritoneal cavity, when the upper end of the intestine grows more rapidly. ^


IN Millimeters of









Small Intestine.


Total Length.

No. of Enil>ryo.







Intee- Small tine. Iniestine. I

wliole itestine.




J. 5































































Infant '






•After Weinberg

'After Sernoff

It is not difficult to follow the development of the intestine from embryo XXXIV to the adult, by simple dissection ; but, in order to be more certain of the relation of the deeper layers of the intestine of an older embryo, I had the intestine of a four months' foetus modeled. In this, however, the mesentery was not included, as the loops only were desired. The intestine was removed from the body in toto, and carefully imbedded in paraffin, after which it was cut into sections 100 /a thick. These were drawn upon wax plates, at a scale of ten, thus making each plate one millimeter thick. The intestine outlines were then cut out, and the remaining frame-work of wax plates was carefully piled, and the cavities cast with plaster of Paris. After the plaster had set, the wax was melted in hot water, leaving the plaster cast of the intestines enlarged ten times.

Figs. 14 and 15 are drawings of this model, and they show about what would be expected in a stage more advanced than embryo XXXIV. The large intestine has become more extensive, the transverse colon having become bent forward, and the descending colon having a very marked S in it before it passes over to the rectum. The intestines, as a whole, are shifted more to the left side of the body, so that the colon encircles the intestine rather than simply marks its border. The lower part of the small intestine is filled with a great quantity of meconium at this stage, showing that vermicular action must take place at this early time, as all this substance has been propelled downward to the cfficum. This same condition I have noticed in other embryos of the same stage.

The loops have shifted somewhat over one another, but one could not unwittingly separate the model so that it would not fall into its respective groups. This separation is given in Fig. 15.

It is evident that the loops are now shifting and adjusting themselves to the space they have to occupy. The loops 2 and 3 are still recognizable, while the loop 4 has been pushed back of them and extends over about as great an area as loop 6. The loop 5 lies about in the middle line, is more to the left than in the younger specimens, and is destined, ultimately, to

lie in the left iliac fossa. The loop 6 will descend into the pelvis when the pelvis becomes large enough to hold it, making room for the green, which is shifted to the right side and to the umbilical region. All this will be accomplished with the descent of the caecum to form the ascending colon, thus bringing about the re-arrangement of the position of the loops by a rotation of the lower end of the intestine toward the pelvis.


It is relatively easy to follow the intestines in an older fcetus or in a new-born child after they have been hardened in formalin or other substances which keep the intestines sufficiently in place while they are being handled. I have examined the intestines by this method in a number of new-born children, and have found them much the same as in fcetus XXXIV and XLVIII. The intestine passes over and back along its mesenteric attachment from left to right, while in foetus XXXIV and XLVIII, the direction of the mesentery of the intestine is at right angles to the axis of the body, in the new-born this attachment is from the left hypochondriac region diagonally downward towards the right iliac fossa, with a curve somewhat towards the right fossa. This makes its course a curved line, which is also curved spirally around the body. While above it is left and deep, below it is right and superficial. The intestine now is attached along this line, crossing and recrossing it, over and back again from duodenum to csecum. In so doing, the convolutions above lie to the left of the mesentery, and are piled upon one another, making the planes of their circles at right angles to the body, while below and to the right they lie in front of the mesentery, and the planes of the convolutions are jjerpendicular to the body.

For the adult I have examined the intestines of about 50 cadavers, in which 41 were not diseased nor adherent in any way. Of them, one-half were negroes. The intestines were all coagulated in position, the cadaver being on its back, with about 1.5 to 2 kilos of pure carbolic acid. It was injected in 33 per cent, solution to preserve the subjects for dissection. After the abdominal cavity was opened the position of the intestines was either sketched or photographed and then the intestines removed, loop by loop, making a tracing of their course at the same time. In this way the general course of the intestine was followed. In removing the loops it was found that in nearly all specimens they came out as distinct groups, as for instance the group on the right side of the body was usually one loop crossing the middle line but twice; once to communicate with the loops above, and once with those below. To follow the intestine in this way the method was amply sufficient, but free-hand modeling or corrosion gives a much more satisfactory result. The models of three specimens which I have made in this way have proved to be of great value in gaining a clear idea of the position of the intestines. A large number of diagrams, sketches, photographs, and models were compared with one another till I was finally able to convert them into a common scheme, by which I have been in the habit of demonstrating the course of the intestines to students, and then immediately verifying

September-October, 1898.]



it on the cadaver. In doing this, it is necessary to be prepaied for the variations, and these again can be classified.


It was generally believed that the intestines within the abdominal cavity had no definite position until a few years ago, when Ilenke* demonstrated that this was not the case. A glance at the various standard test-books on anatomy shows that there is a tendency among them to locate the main groups of the small intestine in fairly definite portions of the abdominal cavity. Gegenbaurt gives an illustration copied from Luschka in which the jejunum and ileum are located respectively in the upper and lower portions of the abdomen. In the text he expressly states that the jejunum is located in the upper portion of the abdominal cavity and extends down to the left iliac fossa. The ileum, however, is located in the lower portion of the abdominal cavity and in the pelvis, and extends over to the right iliac fossa. HoffmannJ gives an excellent illustration of the coils of the small intestine, locating the jejunum mainly in the umbilical and left iliac regions, with the ileum within the pelvis and lower abdominal regions. Similar descriptions are given by Testnt§ and by Q,uain,|| with the exception that they are more cautious about locating definite loops. Quain states, "the jejunum lies above and to the left side of the ileum, but the coils are so irregular that the position of any individual loop offers but little clue to the part of the intestine it belongs," while Testut states that tlie position of the intestine changes, due to the muscular contraction, and so on.

The first decided step in advance to locate the position of the intestine was made by Henke when he studied carefully the spaces in which the intestine may lie. He found that the abdominal cavity may be divided into four compartments, the greater of which lies within the concavity of the diaphragm and is filled with the organs which are more or less firmly fixed with ligaments. The other three compartments are separated from one another by the ridge formed by the two psoas muscles and by the vertebral column. This makes a right and left compartment and a lower compartment which extends into the pelvis. Into these three compartments the intestine must accommodate itself, and Ilenke thinks it has a fairly definite position. He is cautious enough, however, to state that under certain conditions the loops may shift from one space to another, but what the regularity of the position is, or what the rule of the shifting, is difficult to determine from Henke's paper. His illustrations, however, are very good, but, according to my experience, do not represent the normal type of the intestine. Of course, we could not expect them to do so, for the number of cadavers he studied carefully appears to have been but three.

Henke's method of study was to make sketches of loops of intestine and then to remove them, sketching again the loop

•Henke, His's Archiv, 1891.

tGegenbaur, Anatomie, 1890, Bd. 2, S. .59.

^ Hoffmann-Rauber, Anatomie, 188G, Bd. 1, S. 557.

<> Testut, Anatomie, 1894, t. 3, p. 505.

II Quain, Anatomy, 189«, vol. 3, pt. 4, p. 103.

below. By combining the drawings he finally outlined the course of the intestine from the duodenum to the ca3cum. Of course, he examined a great number of intestines in fresh cadavers, but it is difficult to trace the course of the intestine in them, as the slightest disturbance will make one's result uncertain. While, therefore, he gives very little certainty regarding the course of the intestine, he states definitely that the course of the loops on the left side is horizontal to the body, while on the right side it is perpendicular.

A few years later Sernoff* studied a few cadavers more carefully and with more accurate methods than Henke, but did not verify Henke's result. Sernoff injected the cadaver with a large quantity of chromic acid, and in this way the intestine and mesentery were hardened in position. Then, alter opening the abdominal cavity, a cast was made of the intestine, and finally the surface loops were stained with fuchsiu. In this way the surface loops were marked after disturbing the intestine for purpose of exploration and measurement. Next the intestine was removed, showing the form and position of the mesentery which was left within the body. I'he method throughout is accurate, but the number of specimens is not numerous enough for any generalization of the position of the loops. Only two records of intestines in normal position are given, and although Sernoff believes that these are diametrically opposed to each other in position, I think it is not difficult to see that there is a great similarity in them. The fact that a higher loop may be on the right of a lower one in one case and the reverse in another does not necessarily overthrow a general scheme. Also, it is not of much significance regarding the general course of the intestine that in specimens 2 and 3 (he does not picture No. 2) the direction of the convolution is not horizontal above and to the left and perpendicular below and to the right.

Recently Weinberg! has studied a number of specimens in new-born infants very carefully. He gives good illustrations and descriptions of ten specimens which were studied with the method employed by Sernoff. In general, Weinberg's specimens are all after the same plan, showing that the intestine goes over and back, antero-posteriorly, beginning at the upper left side and ending in the lower right side. In seven specimens out of the ten the large upj)er segment of the jejunum lay in the left upper part of the abdominal cavity. In three specimens, only a short portion of the jejunum lay in the left hypochondriac fossa and the rest came up in contact with the ventral abdominal wall. The direction of the loop ill this region was mainly transverse to the body, and, in general, the extent of them was about two-fifths of the length of the small intestine. Following these loops there is a group of irregular convolutions which lie in the left iliac fossa, and include the middle fifth of the intestine. Then the intestine crosses the left psoas muscle, and the remaining two-tifihs of the small intestine lie between this and the right psoas, as well as ou the right side of the abdominal cavity. The direction of the convolutions of this portion is mainly perpendicular. The extent of the intestine which comes in contact

Sernoff, Internat. Monatsch. f. Anat. u. Phys., 1894. t Weinberg, Internat. Monatsch. f. Anat. u. Phys., 1896.



[Nos. 90-91.

with the anterior abdominal wall is about one-third of its whole length, which corresponds with the measurements given by Sernoff for the adult.

In general, Weinberg's results confirm Henke's and give for my purpose the important intermediate stage between the foetus and the adult. I have also examined the intestine of a number of new-born babies after they had been hardened in situ in formalin or in carbolic acid and can confirm the work of Weinberg.

In describing the direction of the loops of the intestine in the various portions of the abdominal cavity it is not well to state the direction of the external loop to the body cavity, for this loop may be only the connecting link between two more important loops above and below. The main loops must be isolated whether they are superficial or deep, and the plane of the circle which they describe makes the general direction of the loop. If the intestine makes a continuous spiral, then the direction of any one of the loops is about parallel with the circles the loops describe ; but if the spiral reverses itself, then the connecting loop is at right angles with the plane of the circles. If this is not considered, it may be that the suj)erficial loops are perpendicular, while the main loops are transverse to the body axis.


The cadavers had all been carried in the sujiine position for at least two kilometers over the rough cobblestones of Baltimore before they were delivered at the laboratory, and this shaking may account for the regularity of the arrangement of the intestine. They were then injected with about 1.5 to 2 kilograms of carbolic acid crystals in the form of a 33 per cent, solution into the femoral artery. This coagulates completely the abdominal and thoracic viscera. After that the bodies were frozen and some of them were not studied until two years later, while most of them were opened at about the end of a year. The older bodies are preferable, as the surplus water has evaporated and the tissues are fairly dry and somewhat hard.

In all of these specimens the general direction of the intestine was diagonally across the abdomen from the left hypochondriac space towards the right iliac fossa, usually diverging once or sometimes twice towards the right side of the abdomen and always towards its end, into the cavity of the pelvis.

The general form and position of the mesentery is well shown in Fig. 17, as well as in Figs. 5 and 8 by Sernoff. These figures show the large curves made by the mesentery to attach itself to the loops, first on one side of the root of the mesentery and then on the other. I have tried to follow rather the greater groups of couvolutions, for it is hopeless to attempt to number every individual loop. .

In 31 of the specimens the arrangement of the loops was after the same plan ; therefore I shall consider this the normal, and the arrangements of the intestine in the other specimens as variations of this plan. In these specimens the jeji;num first arranged itself into two distinct groups of loops situated well up in the left hypochondriac region. Each group made more than a complete circle, and both of them

came in contact with the anterior abdominal walls. They are marked 2 and 3 respectively in Fig. 16, the loop 2 being the one which communicates with the duodenum. After this the intestine passes through the umbilical region to the right side of the body. This loop is marked 4 in the figure. Then the intestine recrosses the median line to make a few convolutions in the left iliac fossa (5), after which it fills the pelvis and lower abdominal cavities between the psoas muscles (6). The course of the intestine which has been pictured in Fig. 16 is given in Fig. 18.

When now this arrangement of the intestine is compared with that of foetuses XXXIV and XLVIII, as well as with Weinberg's specimens, it is fairly easy to see the gradual transformation of XXXIV into the adult type. Fig. 12 still shows the intestine about equally distributed on both sides of the body, with the caecum still very high. In Fig. 14 there is already a marked descent of the cfficum towards the future pelvis. In comparing these two figures it is to be observed that Fig. 12 is a ventral view and Fig. 14 a view from above. The outlines of the stomachs in the two figures will show from what point the models have been drawn.

When we pass from these two specimens to the figures of Weinberg, we see a similarity between them and most of his figures, but in a number of them the intestines have begun to shift more and more. In general there is a tendency for the irregular lines of mesentery to bend towards the left iliac fossa, for, with the descent of the csecum, the whole mesentery is rotating towards the left side. Weinberg's Fig. 18 shows this well. Hand in hand with this movement one or more loops move towards the right side of the body, as his Figures 11 and 19 show. As yet there is no marked pelvic cavity to take the lower end of the ileum, and as soon as the pelvis is large enough to hold it, we can easily imagine the intestine pictured by Weinberg in Figs. 5, 9, 11, 16, 17, 18 and 19 to be converted into my Fig. E by a simple descent of the ileum into the pelvic cavity. The other few specimens may be considered as variations.

In a shifting of this sort it is probable that the middle loops of the intestine would be transferred to the right side, while the upper half would remain on the left side, and the lower half in the pelvis and lower abdominal regions. According to Sernoff's three measurements, on an average 41 per cent, of the length of the intestine is on the left side, 41 per cent, in the pelvic cavity and about 18 per cent, on the right side. In embryos IX, X, VI and XLV the loops 2, 3 and 4 together are shorter than the loops 5 and 6, while in XXXIV and XLVIII, these first three groups are somewhat greater in length than the lower two. In the younger embryos it was the ileum which grows more rapidly, while in the older embryos the jejunum is beginning to overtake the ileum. So from these measurements, as well as from the indication in Figs. 12 and 14, it is the loop 4 which is destined to cross the middle line and to take its position on the right side of the body.

Heuke showed that it is not difficult to separate the intestine into two great groups, the dividing line of which is the left psoas muscle. This, usually, is also the limit between the loops 4 aud 5, as shown in Fig. 16. The diagrammatic



September-October, 1898.]



Fig. E shows this still better between loops 4 and 5. When the loops 5 are within the pelvis this separation made by the psoas is still more marked. Also in those specimens iu which the position of the intestine is as iu Fig. E, the loop 4 can be lifted towards the left side, making a beantiful demonstration of the attachment of this loop. A glance at Fig. 17, as well as at Sernoff's Figs. 5 and 8, will easily explain why this should be so.


Fig. E. — Scheme of the intestine. The arrows indicate the Tariations. The variations a and bb were most frequent ; the variation e least frequent.


One of the most common variations I have found is one in which there were no intestines on the right side of the peritoneal cavity. The loop 4 had been transferred to the left side of the body, as indicated in Fig. E by the arrow a. Otherwise the intestine had its usual position. It is this loop which is so easily isolated and, probably on this account, it is readily displaced by an enlarged caecum or ascending colon. It may possibly be that this loop can be shaken to the left side, and this could easily be tested by experiment. In one of these specimens the sigmoid flexure of the colon filled the right side of the abdominal cavity.

The second variation, as w^ell as the first, occurred six times. It is marked by the arrows bb in Fig. E. The loop 4 was again displaced to the left side, and the loops 2 and 3 were displaced to the right side. In other words, the very upper part of the jejunum formed the loop on the right side of the body.

The third variation occurred five times. In these specimens the intestine had its normal position, with the exception that an additional loop arose out of the pelvis and filled a portion of the right side of the body, as indicated by the arrow c. Sernoff's specimen, pictured in Figs. 3, 4 and 5, is to be included with this group.

The next variation occurred two times. It is indicated by

the arrows d in Fig. E. The large loop 4 was again drawn to the left side, and the space it formerly occupied was filled by two loops, one from the upper part of the jejunum, and the other from the lower part of the ileum.

The last variation occurred once. It is marked by the arrows e. The loop 4 was displaced and its place taken by a large loop which arose from the ileum within the pelvis. Henke's specimen B seems to belong to this group, if I can judge by his illustrations.


The variations given above all fit within the common scheme and can easily be explained. In all of my specimens I found but one extreme variation, and in this the intestine crossed the middle line at the beginning of the jejunum and then filled the right fossa. From here it descended immediately into the pelvis and filled it and the lower abdominal cavity completely. Then it left the pelvic cavity and filled the left fossa, extending up to the beginning of the duodenum. When it hud reached this point it took a fairly direct course along the descending colon over the floor of the pelvis, and passed directly to the cascum. Henke* has also described a variation practically identical with this. WeiubergI has also described one similar to this, only that the jejunum descends immediately to the pelvis and then gradually rises to the left side, and finally over to the right side. What kind of a mechanism can bring about this extreme variation is not possible to state.

It could be asiBerted that these few instances of marked variations indicate the normal, but in my specimens it is one in forty-two; in Weinberg's, one in ten; and in Henke's specimens it is not stated how many cadavers were examined carefully.


Henke has stated that the loops of the intestine may shift from one of the abdominal fossas to the other, and, no doubt, this is true. We are familiar with the fact that a distension of any of the pelvic viscera pushes all of the loops of intestine out of the pelvis,J and emptying it again allows the loop to descend to the floor of the pelvis. So likewise a distension of the colon or a certain number of loops of small intestines will displace a certain member of loops from their natural position. Since, however, the intestine was located after one plan in 41 cadavers, I do not think it probable that ordinary shaking will displace any number of loops. Pure mechanical disturbances, as by returning the intestines after operation, will also be overcome by the intestines shifting about to their normal position, guided by the attachment of the mesentery, its length and the space within the abdomen. To give this last question a thorough test I made a number of experiments upon dogs. In these animals the intestine is closely rolled up in a very regular fashion below the stomach, and the whole is carefully tucked in by the very large omentum. Upon opening an animal, one is struck with the neatness and accuracy of the

»Henke, His's Archiv, 1891, p. 101, Taf. IV, Fig. 12. \ Weinberg, Internat. Monatsch. f . Anat. u. Phys., 1896. I Among others, Garson, His's Archiv, 1878.



[Nos. 90-91.

adjustment of the omentum, and it is easily disturbed by handling. When, however, the intestine and omentum are withdrawn through an abdominal wound, they are disturbed to such an extent that it is impossible to return them to the abdominal cavity as they were found, with the omentum covering them. After the intestines have been pushed into the abdominal cavity in a haphazard way and the animal sewed up, using all antiseptic precautions, the loops as well as the omentum readjust themselves as they were before, provided no marked inflammation takes place. So in the dog, the intestine and omentum seek their normal position after they have been disturbed.


Fia. 1. — Profile view of Embryo No. XII. Enlarged 38 times. The body wall over the heart has been cut out. Am, amnion ; UV, umbilical vesicle; OV, optic vesicle ; AV., auditory vesicle; Oa, third occipital muscle plate ; Cb, eigthth cervical muscle plate ; H, heart; P, pericardial cavity ; VOM, omphalo-mesenteric vein ; MR, membrana reuniens ; D, D', openings which connect the peritoneal cavities of the two sides with each other.

Fig. 2. — Same as Fig. 1, but half of the model has been removed to show the extent of the ectoderm and entoderm. Br', Br", first and second branchial pouches; M, mouth ; S, Seessel's pocket ; T, thyroid ; L, liver ; NC, neurenteric canal.

Fig. 3. — Intestine of Embryo II, viewed from the right side. Enlarged 34 times. C, caecum; M, mesentery ; y, remnant of yolk sac.

Figs. 4 and 5. — Intestine and liver of Embryo IX. Enlarged 25 times. C, ciecum ; OMA, omphalo-mesenteric artery ; HV, hepatic vein ; UV, umbilical vein ; PV, portal vein ; FW, foramen of Winslow ; GB, gall bladder.


Figs. 6 and 7. — Intestine and liver of Embryo X. Enlarged 12i times. U, position of umbilical vessels ; C, caecum ; FW, foramen of Winslow.

Figs. 8 and 9.— Intestine and liver of Embryo VI. Enlarged 12i times.


Figs. 10 and 11. — Intestine and stomach of Embryo No. XLV. Enlarged 16 times.

Figs. 12 and 13.— Intestine of Embryo No. XXXIV. Enlarged 4 times. Viewed from the ventral side. In Fig. 13 certain loops have been lifted off to show the deeper loops.


Figs. 14 and 15.— Intestine of Embryo No. XLVIII. Enlarged 2* times. The view is from the ventral and cephalic side of the model. The mesentery was not included in the model. Fig. 15 is a dissected model to show the deeper loops. The lower part of the intestine is enormously distended with cell debris, etc., showing that vermicular action is present at this early stage.

Fig. 16. — Usual position of the intestine in the abdominal cavity. Although this is an actual specimen, it represents the condition in twenty-one out of forty-one cadavers. The numbers in the figure mark the parts which are homologous with the loops correspondingly numbered in the other figures.


Fig. 17. — Usual position of the mesentery.

Fig. 18.— Course of the intestine. This figure is taken from a model made from the same cadaver from which Figs. 16 and 17 were drawn.



By John Bruce MacCallum.

[From the Anatomical Laboratory of the Johns Hopkins University, Baltimore.)

In a previous paper* I described the structure and histogenesis of the heart muscle cell of mammals. The process of development, as demonstrated in embryo pigs, was found to be quite definite, and in order to complete this work I have studied the heart muscle of a series of lower vertebrates, taking representatives from the various large classes. The definite stages in the histogenesis of the muscle cell, confirmed by its comparative histology, made it seem possible to determine the method of growth of a muscle as a whole ; that is, to learn by what means a small embryonic muscle gradually becomes large, as in the adult. Such a study of the heart seemed to present many difficulties which that of a simpler system of muscles would not. I therefore turned to the striated muscle of other parts of the body, the so-called voluntary muscles, and with this problem in view I studied the histogenesis of the striated

J. B. MacCallum, On the Histology and Histogenesis of the Heart Muscle Cell. Anatomischer Anzeiger, Jena, Bd. xiii. No. 23, 1897, S. 609-620.

muscle fibre of the thigh in embryo pigs. In these a process was demonstrated which proved to be essentially the same as that found in heart muscle, with differences made necessary by the difference between the respective adult tissues. Knowing, then, the course of growth of the individual muscle cells, I have endeavored to determine the relation of this to the growth of the muscle as a whole. In order to begin with the simplest problem, I have chosen the sartorius muscle, as it seemed to show less complexity of structure than many other muscles. Two sides of this problem presented themselves. On the one hand it was necessary to determine how far the growth in size of tiie muscle as a whole was caused by increase in the number of fibres, and how far it was due to a mere growth in size of the individual cells. On the other hand it was necessary to make out the relation between the various stages in the histogenesis of the muscle cell and these two methods of growth. These problems I have endeavored to solve by the study of a series of human embryos, using the sartorius as the

September-October, 1898.]



type of a simple muscle. For the sake of clearness I will give the results of this work under three headings, as follows:

I. Synopsis of the work already done on the histogenesis of the heart muscle cell, supplemented by a study of its comparative histology.

II. The histogenesis of the voluntary striated muscle cell.

III. The growth of the sartorius muscle and its relation to the histogenesis.


As described iu the article referred to above, the adult human heart muscle is made up of rhomboidal branching cells whose processes come together end to end. The protoplasm of each cell consists of a number of darkly staining columns, running longitudinally, which are separated by unstained substance. The columns are commonly called fibril bundles, and the unstained substance is the sarcoplasm. With special methods of staining, particularly by Kolossow's osmic acid method, a definite relation can be made out between these two parts of the cell. The fibril bundles present regular striatious in the form of darkly staining lines. Narrow striations, the so-called Krause's membranes, alternate with broader bands which are known in voluntary muscle as Briicke's lines. The Krause's membranes, however, do not belong to the fibril bundles alone. They can be seen also extending across the sarcoplasm, as shown in Fig. 1. The sarcoplasm is thus divided into compartments which are limited horizontally by membranes continuous with the narrow striations on the fibril bundles. In cross section (Fig. 2) the muscle fibre is seen to be made up of dark masses, the cross-sections of fibril bundles, separated by sarcoplasm which is divided into definite circular or polygonal areas. The compartments of the sarcoplasm, then, are disc shaped, and I have proposed for them the name sarcoplasmic discs. As described, they are bounded by membranes which are continuous with the fibril bundles at definite points, namely, at the narrow striations or Krause's membranes. It will then be seen that there is a definite network in the cell made up of the fibril bundles and the membranes bounding the sarcoplasmic discs.

Fiii. 1. — Longitudinal section of Adult Human Heart Muscle. -ff, Krause's membrane; .S', sarcoplasmic disc ; F, fibril bundle.

Fir.. 2. — Cross-section of Adult Human Heart Muscle. C, central sarcoplasm; F, fibril bundle; S, sarcoplasmic disc.

The stages in the embryo leading up to this adult structure e.\'pluin the relation of the fibril bundles to the sarcoplasm.

The earliest stage iu the development shows an irregular network in the cell protoplasm with no fibril bundles. This network tends to become more and more regular until the meshes are of the form of large discs. Some of these break up into smaller ones, and in the nodal points of the network there is an accumulation or differentiation of its substance, giving rise to longitudinally disposed masses. These become what in the adult are known as fibril bundles, and the discs left are the sarcoplasmic discs. This formation of fibril bundles takes place first at the periphery of the cell, so that those which are the latest to appear are nearest the centre of the cell. It is apparent, then, that the continuous network spoken of in the adult fibre, made up of the fibril bundles and the membranes bounding the sarcoplasmic discs, is developed directly by a process of differentiation from the primitive protoplasmic network of the embryonic cell. The gradual acquirement of a special function has so altered this network that a complicated structure is formed out of an extremely simple one.

The study of the heart muscle of lower animals gives an interesting repetition of some of the stages of this process. Although the structures met with iu the comparative histology of the organ do not make up an uninterrupted sequence as in its histogenesis, yet there is a sufficient resemblance to confirm the results obtained in embryonic tissues. Hearts from animals representing the various large groups of vertebrates were studied. Of the fishes several species were obtained, among which were the larval lamprey (Ammoccetes), the adult lamprey, and several kinds of Teleosts. The Amphibians were represented by the frog and the toad, while of the reptiles the snake and turtle were used. Several species of birds were studied, such as the English sparrow, the crow, aud the common fowl.

In the heart of Ammoccetes the muscle cells are small and spindle shaped. In some of them no fibril bundles can be seen, but most of the cells show a single row of very narrow fibril bundles at the periphery. The rest of the cell is made up of sarcoplasmic discs. In the lamjjrey the structure is similar, while in the higher fishes, such as the pickerel, the cells are considerably larger and the fibril bundles are more conspicuous. In longitudinal sections the individual fibril bundles are quite similar to those in adult mammalian muscle. The same striations are present, and the same relation of these to the sarcoplasm exists. In the frog and toad the structure of the heart muscle differs only slightly from that described in fishes. The cells are larger, aud the fibril bundles are closer together, tending to form large irregular columns. The nucleus is in the centre of the cell, and the fibril bundles are only at the periphery. The heart muscle of the Keptilia, as shown in the turtle and snake, is made up of cells which are not markedly different from those of the Amphibian heart. The fibril bundles at the periphery of the cell are large and conspicuous, and in many cells there are smaller ones more centrally placed. In the birds, however, there is a very decided advance on the structures described so far. The cells are very large, almost as great as the adult mammalian fibre. In cross-section they resemble the structure described iu mammals. The nucleus is in the centre aud the rest of the



[Nos. 90-91.

cell is almost filled with fibril bundles. These are in the form of flat bands around the periphery, as in mammalian muscle, and are separated by sarcoplasmic discs. There is a very great difference between these cells and those found in the hearts of Fishes, Amphibians and Eeptiles. As regards the structure of the heart muscle, then, the classes of Vertebrates are divided into two groups, one comprising the Fishes, Amphibians and Eeptiles, and the other the Birds and Mammals. It will be noticed that these groups are the cold-blooded and warm-blooded animals respectively ; and it is possible that in addition to the control of the temperature by the central nervous system, there is a relation between the heat regulation and the degree of development of the circulatory system. The mammalian embryo in utero resembles the cold-blooded animals in the fact that its temperature is the same as that of its surroundings. They are under similar conditions in this respect, and the structure of the heart muscle is almost identical. Although the transition from the structure of the heart muscle of low^er to that of higher animals is not perfectly gradual, yet the comparative histology corresponds roughly with the histogenesis of the cell. The Fishes, Amphibians and Reptiles, which in other respects are clearly lower than the Mammals and Birds, possess also au embryonic type of heart muscle. The heart of an adult fish, for example, is made up of cells which are almost identical with an early stage in the development of the mammalian fibre. While the comparative anatomy and the embryology of many organs run parallel to one another, it is interesting to note that this same relation holds good even in the internal structure of a single cell, and that the most minute details of the cell structure in the adult heart muscle of the lower animals are identical with those in embi'yonic mammalian tissue.


In order to determine the course of development of the voluntary striated muscle cell I have used a series of human embryos and one of embryo pigs. Of the former I used the sartorius muscle in each case, and in the latter the muscles in the front of the thigh. The human series consisted of embryos of the following lengths in millimeters from vertex to breech: 10, 30, 75, 102, 130, 170 and 200. The series of pig embryos was made up of specimens of the following lengths in millimeters: 25, 34, 45, 57, 64, 70, 75, 100, 125 and 150. The sartorius of an embryo rabbit 35 mm. in length was also studied.

The embryos, which were obtained iu a fresh condition, were treated according to Kolossow's osmic acid method,* and the sections afterwards stained iu safranin. Those which were already hardened in formalin, alcohol, or Miiller's fluid, were cut iu paraffin, and the sections treated by a method somewhat similar to Kolossow's. They were immersed in 2 per cent, osmic acid for three or four minutes, and then transferred to Kolossow's reducing fluid and left until the

precipitation wa« complete. They were then stained in safranin to differentiate the nucleus. Sections stained in this way show the protoplasmic structure with great clearness.

In an embryo pig 25 mm. long the voluntary muscle has quite an undifferentiated character. The sartorius in a crosssection of the leg cannot be recognized with certainty. There are merely small groups of spindle shaped cells with a loose connective tissue between. In the muscle cells there are no fibril bundles, and the protoplasm is very scanty as compared with that in the adult cell. If in a cross-section the cell has its nucleus cut through, the protoplasm is hardly visible. It is seen only as a narrow rim around the centrally placed nucleus. The cross-section of a cell above or below the nucleus shows a definite network which divides the cell into small circles. These correspond with the structures spoken of as sarcoplasmic discs. This is shown in Fig. 3.





'^IG. 3. — Cross-section of Volui. 7 Muscle from the Thigh of an Embryo Pig, 25 mm. loug. .1, cell showing the nucleus; B, cell showing sarcoplasmic discs.

Fig. 4. — Cross-section of Voluntary Muscle from the Thigh of an Embryo Pig, 45 mm. long, showing fibril bundles at the periphery of the cells.

•A. K0I088OW. Ueber eine neue Methode der Bearbeitung der Gewebe mit Osmiumsiiure. Ztschr. f. wissensch. Mikr., Brnschwg., Bd. ix, 1892-3, S. 38-43.

The muscle of an embryo pig 34 mm. in length is somewhat farther advanced in development. The bundles of cells making up the individual muscles are quite distinct and separate. The cells themselves show a centrally placed nucleus, and protoplasm which is very slightly differentiated. In some places fibril bundles can be made out at the periphery of the cell, but these are not very distinct.

In an embryo pig 45 mm. in length the muscle cells contain definite fibril bundles. These are present in a single row around the periphery of the cell, as shown in Fig. 4. The nucleus is still situated in the centre of the cell. The protoplasm near the nucleus contains no fibril bundles, and is divided into sarcoplasmic discs as described before. In longitudinal sections the fibril bundles are seen to be definitely related to the sarcoplasm in the way described above. The horizontal boundaries of the sarcoplasmic discs are membranes continuous with the fibril bundles at the narrow striations.

Very similar cells are found in the muscle of an embryo 57 mm. long. The fibril bundles, however, are somewhat more conspicuous. In embryos 64 mm. and 70 mm. in length the muscle cells contain, iu addition to the peripheral row of fibril bundles, scattered fibril bundles nearer the nucleus. In many cells nuclei are seen, both at the periphery and in the centre.

September-October, 1898.]



In the muscle cells of an embryo pig 75 mm. long, the peripheral and central nuclei are seen very clearly. These diiier somewhat in appearance. The peripherally placed nucleus stains deeply and uniformly, and has a solid appearance. The central nucleus, however, is large and vesicular. It possesses a definite nuclear membrane, and the chromatin network is delicate and distinct, as shown in Pig. 5. In some places the outline of this nucleus grows irregular. What finally becomes of it it is difficult to say. The nuclei in adult muscle cells are all peripherally placed, and have an af)pearance resembling the peripheral nucleus described here.


Fig. 5. — Cross-section of Voluntary Muscle from the Tliigh of an Embryo Pig, 75 mm. in leugtli. A, central vesicular nucleus; B, peripheral solid nucleus.

In a pig 100 mm. long the appearance of the muscle cells is not essentially different. The fibril bundles are more numerous and somewhat closer together. The muscle cells of an embryo 150 mm. in length, however, are entirely filled with fibril bundles. The sarcoplasm is not abundant. In some of the cells the central nucleus cannot be made out, while in others it is still present.

In adult muscle the nuclei are all at the periphery of the cell. In longitudinal sections the fibre consists of fibril bundles separated by a very small amount of sarcoplasm. The fibril bundles possess alternating narrow and broad striations. In many cases the narrow striation (Krause's membrane) can be seen extending across the sarcoplasm, dividing it into compartments. This is approximately the same structure as that which has been described in heart muscle. In the latter, however, the sarcoplasm between the fibril bundles is so abundant that the narrow striations form disc-shaped compartments, or sarcoplasmic discs. In voluntary muscle the sarcoplasm is very scanty, and the discs are so encroached upon by the growth of the fibril bundles that they are hardly recognizable. As in heart muscle, the lines dividing the sarcoplasm are continuous with the fibril bundles at the narrow striations.

The course of the development, as shown in the series of human embryos, is made up of stages which are very similar to those described in the embryo pig. In human embryos 10 mm. and 30 mm. in length the muscle cells contain no fibril bundles. In embryos 75 mm. and 102 mm. long there is a single row of fibril bundles around the periphery of the

muscle cell. The nucleus is of a vesicular character and is situated in the centre of the cell. In an embryo 130 mm. long the muscle cells contain a vesicular central nucleus and one or more solid deeply staining peripheral ones. There are fibril bundles in the central part of the cell as well as at the outside. The muscle of an embryo 170 mm. in length is made up of cells which are very similar to those of an embryo pig 150 mm. long. The cells are filled entirely with fibril bundles, and the centrally placed nucleus is not often present. In older embryos, in the new-born, and in the adult subjects, the nuclei of the muscle cells are all at the periphery.

The course of development, then, seems to be quite definite. Beginning with a cell which contains only a protoplasmic network, the extremely complex voluntary muscle fibre is gradually formed. The first step is an accumulation of the network at definite places around the periphery of the cell, giving rise to fibril bundles. This accumulation takes place in the angles of the network, so that the meshes remain as discs between the fibril bundles. The membranes, bounding these discs horizontally, are continuous with the narrow striations on the fibril bundles. The formation of fibril bundles goes on until they occupy a large part of the cell. The sarcoplasm in this way becomes gradually replaced, and in the adult cell it is very inconspicuous. At the same time the nuclei undergo changes. The cell which at first contains only a centrally placed nucleus acquires other nuclei which are situated at the periphery. It is possible that the latter are derived from the centrally placed nucleus which is finally lost.

It seems that the same hypothesis is applicable here as was suggested for the development of heart muscle. It simplifies the conception of the structure of striated muscle very greatly to consider the fibril bundles and the membranes bounding the compartments in the sarcoplasm, as derived from the primitive network found in the muscle cells of very young embryos.

These results show that the same process of growth takes place in all kinds of striated muscle cells. The adult structure of voluntary muscle is, however, somewhat different from that of heart muscle, and as a consequence the later stages in the development are different. One of the most noticeable differences is the position of the nucleus. Why it should remain in the centre of the cell in heart muscle, and in the voluntary muscle cell be situated at the periphery, is difficult to determine. Until more is known concerning the relation between the function of the cell protoplasm and that of the nucleus, this can only be the subject of hypothesis.

Since the beginning of the differentiation is the same in both heart muscle and voluntary muscle, the development of the power of contraction must run a somewhat similar course. If this be so, it is conceivable that contractions in definite directions begin when the irregular network of the primitive cell becomes made up of regular membranes at right angles to one another. When there is need of more perfect and stronger contractions, it is probable that the power for this is brought about by a strengthening of the network in the direction of the contractions. This strengthening takes place by an accumulation of the substance of the network, which gives rise to



[Nos. 90-91.

fibril bundles. Why this should always take place first around the periphery of the cell is not clear. It is true, however, not only in the development of the heart mnscle and the voluntary mi;scle cells, but also in the evolution of the heart muscle in lower animals. It is possible that there is a relation between this peripheral disposition of the fibril bundles and the spindle-like form of the cells. The fibril bundles, on account of their position, are curved in correspondence with the outline of the cell. On contraction, therefore, the pull is not only in the long axis of the cell, but also at an angle to this. Thus the cells acting together exert an influence on one another, which would not be the same if the fibril bundles, which are probably the chief agents in contraction, were situated in straight lines in the long axis of the cell. If this be so, the force which acts at an angle to the long axis of the cell must control the whole contraction, just as two muscles opposing one another make a more delicate mechanism than a single muscle. It is possible that this is of advantage in the early stages of the development. An hypothesis concerning such a subject must necessarily be vague and unsatisfactory. All that can be said with certainty is that there seems to be both in the histogenesis and comparative histology of the muscle cell a tendency for the fibril bundles to be formed first at the periphery of the cell. If the fibril bundles are the special contractile elements, it seems that it is of advantage to have the mechanisms for contraction as near the outer part of the cell as possible. The same thing is seen in the adult heart muscle fibre, for although nearly the whole cell is filled with fibril bundles, those at the periphery are many times as large as the ones which are nearer the centre.


In applying the process of development of the voluntary muscle fibre as described above, to explain a special muscle, I have endeavored to determine the I'elation between the growth of the muscle as a whole and the growth of the individual cells. The sartorius muscle was chosen for study because of its apparent simplicity of structure, and because its general characters in the adult vary within narrow limits. The fact that this definite and simple structure is developed gradually from a small group of embryonic cells, indicates that the growth takes place by a definite and simple process. In order to detei-mine the exact nature of this process I have made estimations of the number of fibres to be found in crosssections of the sartorius muscles taken from adults and from a series of human embryos.

Methods of Study. — Sections were taken from adult human sartorius muscles at the upper, middle, and lower thirds. Similar sections from the sartorius of a new-born babe were cut. Muscles were also used from a series of human embryos varying in length from 74 mm. (vertex to breech) to the stage shortly before birth. In these only the middle third was studied. Sections were cut in both celloidin and paraflBn, celloidiu being used mainly for the larger muscles. Various methods of staining were employed, but chiefly hematoxylin and eosin or safranin.

In obtaining the number of fibres found in a cross-section no effort was made to actually count them. All that was

aimed at was to obtain as accurate an estimate as possible with methods which would entail the least number of errors. The most simple procedure seemed to be to compare the area occupied by the muscle-section as a whole and the average area occupied by a single fibre. This ratio must represent approximately the number of fibres. Thus the exact area of the whole cross-section was obtained and from this was subtracted the area of the connective tissue, giving the area of true muscle substance. The average area occupied by a single muscle fibre was then determined. The number representing the area of the muscle substance divided by that representing the area of one muscle fibre must be the number of fibres contained in the cross-section.

The sections of the larger muscles, adult and new-born, were projected on a screen with a projection lantern, and the outlines of the muscle and connective tissue carefully traced on a sheet of paper fastened to the screen. A slide ruled in millimeters was projected at the same time in order to obtain the exact magnification. ' This gave only the amount by which the length of a line was magnified, and in order to obtain the magnification in area this number was squared. A planimeter was used to determine in square millimeters the area of the muscle bundles in the tracing. This number divided by the magnification in area gave the actual area of the cross-section of the muscle.

For the smaller muscles a somewhat different method was employed. The sections were projected by means of a camera lucida and Leitz 3 objective, on a paper ruled in square millimeters. An exact tracing was taken as before of the muscle and the connective tissue in the section. At the same time a scale of hundredths of a millimeter was projected with the same power in order to obtain the magnification. The area of the tracing was got by counting the millimeter squares contained in it. At the edges only those squares were counted outside whose centre the line of tracing passed. The area of the connective tissue was similarly determined and subtracted from the area of the entire section. This gave the area of the muscle substance contained in the section. As a check on this the area of the muscle substance was directly counted. The actual area of the muscle was got as before by dividing the area of the muscle tracing by its magnification in area.

The actual area of one muscle fibre was determined by projecting the section upon a known ruled area by means of a camera lucida and a Leitz 7 objective. With a scale of hundredths of a millimeter a square was projected whose sides were .1 mm. in length. Its area therefore was .01 sq. mm. With the same power the muscle fibres were projected into this square and traced off. The fibres falling in the square were then counted and the number was divided into .01 sq. mm. This gave the average area occupied by one muscle fibre.

These two determinations were made for each section, so that all that was I'equired to obtain an estimation of the number of fibres contained in the section was to divide the area of the muscle by that of the fibre.

Results of the Estimations. — The sections of the adult sartorius muscles were ])rojected with a magnification of 14.5

September-October, 1898.1



diameters, that is, the area was magnified by 210.25. lu the two muscles taken from the same subject the area in each case was greater in the middle third than in either the upper or lower third. In the right sartorius the area of the muscle tracing of the upper third was 19,880 sq. mm.; that of the middle third 24,490 sq. mm. and that of the lower third 18,390 sq. mm. In the left muscle the tracing of the muscle substance in the upper third contained 16,000 sq. mm., the middle third 22,760 sq. mm., and the lower third 18,250 sq. mm. In the new-born the muscle tracing of the upper third contained 2260 sq. mm., of the middle third 2430 sq. mm., and of the lower third 2040 sq. mm. These were all magnified to the same degree. This seems to show that the muscle substance in the sartorius is in the form of a spindle. There is a greater amount of muscle in the middle than at the two ends.

For comparison with the muscles of the embryos, only the middle of the sartorius was used in each case. The tracing of the middle third of the first adult sartorius contained, as mentioned above, 24,490 sq. mm. The magnification in area was 210.25, so that the actual area was 24,490-4-210.25, or 116.48 sq. mm. The number of fibres in .25 sq. mm. of the muscle section was found to be 305, so that the area represented by one fibre .25 -^ 305 = .0008196 sq. mm. The number of fibres, then; in the cross-section would be 116.48 divided by .0008196, which equals 142,118.

By a similar calculation the actual area of the section of the second adult sartorius muscle was found to be 108.252 sq. mm.

The number of fibres in .25 sq. mm. was 315, and the area occupied by one fibre was .0007936 sq. mm. It follows from this that the number of fibres in the section is 108.252 -f.0007936, which equals 136,406. The number of fibres in the sartorius of the new-born babe was estimated by both the methods described. Theresults were slightly different. With a magnification of 14.5 diameters, and measured by the plauimeter, the actual area was found to be 11.557 sq. mm. Magnified 14 diameters and projected on millimeter paper, the result was 12.316 sq. mm. The number of fibres in .01 sq. mm. was 98, and the area occupied by one fibre .000102 sq. mm. Then the number of fibres, according to one calculation, would be 113,304, and according to the other 120,745. This shows that the estimations are subject to an error of about 1 to 17, or somewhat less than 6 per cent.

Considering the individual variations, the number of fibres in the muscle of the new-born babe is approximately the same as that in the adult muscle. After birth, then, the growth in the number of fibres cannot be very great.

The largest human embryo studied was one which measured 200 mm. from vertex to breech. The actual area of muscle in the cross-section of the sartorius was 8.417 sq. mm. The area represented by one fibre was found to be .0000555 sq. mm. The number of fibres is 151,657. The greatness of this number can be accounted for only by supposing that the muscle would have been an unusually large one if it had become adult. It suggests, however, that there is little or no growth in the number of fibres after birth.



Number of Subject.

Lenath of Body.

Dimensions of Muscle in Millimeters.

Area of Musclc-lracing in CrossSection.

Miignification in Areas.

Actual Area.

Number of Fibres in a known Area.

Area represented by one Fibre.

Number of Fibres.

Adult No. 74 (right) . .

180 cm.


24490 sq. mm.




in .25 Bq. mm.

.0008196 sq. mm.


Adult No. 74 (left)....

" "


22760 " f 2430 "



108.252 " 11.557


" " "


136406 i 113304

50 "

140x7 52x6

1 2414 " 1212 "

196 144

1 12.316 8.417

98 180

" .01

.000102 .0000555 "


( 120745

Vertex Breech.

Nape Breecli.

Embryo No. A

200 mm.

130 mm.


No. 98

170 "

106 "


1474 "


5.7578 "


" "

.00U04484 "


" No.65

130 "

75 "


5901 "


1.02164 "


" " "



No. B

102 "

65 "


6565 "


1.02578 "


" "

.00003937 "


No. 64

74 "

46 "


4142 "


.382858 "


.00005882 "


The younger embryos were studied in the same way. One measured 170 mm. from vertex to breech. The actual ai'ea of muscle in the cross-section of the sartorius was 5.7578 sq. mm. The area occupied by one fibre was .00004484 sq. mm. The number of fibres, then, was 128,408. This number is approximately the same as that determined for the new-born and adult subjects. The muscles described thus far contain practically the same number of fibres in a cross-section. The differences can be accounted for by individual variation and by the nature of the estimation.

In the sartorius of an embryo 130 mm. from vertex to breech, the actual area of muscle in a cross-section was found to be 1.02164 sq. mm. The area occupied by one fibre was .0000226 sq. mm. The number of fibres, then, was 45,205. There is a very decided difference between this and the muscle last described. The number is less than one-third of that present in the adult muscles and in those of older embryos. It seems fair to conclude that in embryos between 130 mm. and 170 mm. in length the fibres found in the cross-section of the sartorius muscle increase in number as well as in size.



[Nos. 90-91.

In the sartorius of an embryo 103 mm. from vertex to breech there was found a still smaller number of fibres. The actual area of muscle in a cross-section was 1.03578 sq. mm. and the area occupied by one fibre .00003937 sq .mm. The number of fibres, then, was 36,055. The difference between 45,205, which was the number found in the embryo 130 mm. long, and 30,055 is too great to be accounted for in any way but by an actual growth in number.

The sartorius from an embryo 74 mm. from vertex to breech gave in the cross-section an actual muscle area of .383858 sq. mm. The area represented by one fibre was .00005883 sq. mm. The number of fibres, then, was found to be only 6509. This number is less than one-quarter of that found in the embryo 103 mm. long, and about one-twentieth of the number in an adult.

These figures are put together in the preceding table. The adult muscles are placed at the top, and following these the embryonic muscles in the order of their measurements.






Fig. 6. — Diagram illustrating the Growth of Muscle. It shows the possible growth in the number of cells in a cross-section of a muscle, without a correspouding increase in the number of cells in the entire muscle. At the right tlie cells are represented as cross-sections taken in each case on the plane A.

It seems clear from this table that the growth in number of fibres, or at least the increase in the number found in a crosssection, takes place at a definite period. This increase in the number of fibres cut in any given cross-section might be brought about in two ways. There might be an actual increase in the number of fibres in the muscle as a whole, or there might be merely a growth in length of the fibres. This would cause them to grow past one another, so that a greater number would be cut in any one cross-section. This is illustrated in the diagram above, Fig. 6. In I and II there is the same number of fibres, but in II they are longer and have grown past one another. As a consequence the section A contains many more fibres in II than it does in I.

It is very possible that both of these processes go on at the same time. All that can be determined, with such methods as have been used here, is the growth in the number of fibres found in a cross-section of the muscle. This growth begins at a definite place in the development of the embryo. If the

fibres ran the whole length of the muscle it would be a simple matter to determine the point at which the multiplication of the cells ceased; for then the number of fibres found in a cross-section would be the number contained in the whole muscle. In an embryo 140 mm. in length, however, longitudinal sections of the sartorius show that the fibres certainly do not run the whole length of the muscle. This is seen also in teased prepai-ations. All that can be said, then, is that at a certain stage in the development, namely, in embryos between 130 and 170 mm. in length, the increase in the number of fibres found in a cross-section ceases. Since this increase is brought about by the two processes spoken of above, it is probable that the actual multiplication of cells ceases some time previous to this stage. If this be true, there are as many fibres present in the sartorius muscle of an embryo 170 mm. long as there are in the adult muscle. After this stage the growth in size of the muscle must be a growth in the size of the fibre.

It is to be emphasized, however, that the estimations given here are not of the number of fibres in the muscle as a whole. They are merely estimations of the number found in crosssections, so that any suggestions as to the growth of the muscle as a whole are based on the supposition that the increase of the cells is uniform throughout the muscle. I have observed no indication whatever of special growing points in the sartorius muscle. In cross-sections the number of fibres is approximately the same in all parts of the muscle, although the tendons at each end make the number there somewhat less, as shown above. In the various embryos all the cells in any one section seem to be at the same stage of development. In longitudinal sections no growing points can be made out, and when karyokinetic figures are present they are not confined to any one part of the muscle. If the growth is proved to be uniform throughout the muscle, an increase in the number of the cells in the whole muscle must cause an increase in the number found in a cross-section. Indeed, the number found in a cross-section might increase after the actual number of fibres in the muscle had ceased to grow; for, as shown above, the cells in growing in length might pass one another and more of them would be cut in any section. If this be true, a small embryonic muscle, after a certain stage, becomes a large adult muscle, not by an increase in the number of cells, but by a growth in their size. This bears indirectly on other questions, and may prove of interest in connection with muscular hypertrophy. If the very enormous increase in size which the sartorius muscle undergoes during development can take place simply by a growth in the size of the individual cells, it is probable that such a small increase as is present in a muscle hypertrophied through exercise must be due to the same cause. It also seems probable that hypertrophy of the heart is due to growth in the size of the fibres rather than to an increase in their number. These are problems, however, which require special methods for their solution.

It is interesting to note the relation between the histogenesis of the cell and the growth in the number of cells in the sartorius. It will be seen on comparison of the figures given above, that the increase in the number of fibres ceases approximately when the fibre has become filled with fibril bundles, and the nuclei have become situated at the periphery of the

September-October, 1898.]



cell. Ill a hiuiiau embryo 170 mm. long the muscle cells resemble the adult cells in every way excejjt in size. This stage also marks the point at which the increase in the number of fibres found in a cross-section ceases. Embryos between 130 mm. and 170 mm. in length show the transition between the embryonic and adult types of muscle cells, and this is also the period at which the number of fibres stops growing. It will be remembered that in embryos 130 mm. in length the nucleus is centrally placed and vesicular in character, while in an embryo 170 mm. long the muscle cells contain only peripherally disposed nuclei. There is, possibly, a relation between the position of the nucleus and the power of the cells to produce new fibres.

The course of growth of the sartorius muscle may be epitomized somewhat as follows : At an early stage in the development, the cells are small and spindle-shaped, and scattered in loose bundles. At first there are no fibril bundles, and the nucleus is centrally placed. Subsequently the

fibril bundles appear around the periphery of the cell. The cells multiply and increase in bulk until the embryo is between 130 mm. and 170 mm. in length from vertex to breech. At this stage the bundles of cells become more compact, and the cells themselves are filled with fibril bundles as in the adult. The nucleus is situated at the margin of the cell. The fibres now grow in length and thickness, but probably no longer increase in number. In embryos smaller than 170 mm. in length there is a regular increase in the number of fibres found in a cross-section. After this, however, the number remains approximately constant.

The study of the sartorius as the type of a simple muscle can only be a step towards the explanation of more complex muscles. When accurate methods of estimation have been employed in the study of muscles of more intricate character, and the results considered in connection with the process of histogenesis of the muscle fibre, a more definite idea of the growth of muscle in general may be arrived at.



By John J. Abel, M. D.

[From the Pharmacological Laboratory of the Johns Hopkins University .'\

In my first paper on the chemistry of the suprarenal capsules, in which I reported in detail on researches carried out with the help of Di: A. C. Crawford, I was able to show that the blood-pressure raising constituent can be separated from aqueous extracts of the capsules in the form of a benzoate, and that this remarkable substance is not, as has been maintained, either pyrocatechin or an immediate derivative of it, and V. Fiirth,* in an interesting paper published after, announces the same conclusion. We also gave as our opinion that this substance is to be classed with the alkaloids, founding this opinion on facts stated at length in our paper.

It is my purpose to give in the following brief paper an outline only of certain new observations made in the past year on this chromogenic substance or blood-pressure raising constituent.f

The extract used had been prepared with warm water slightly acidulated with sulphuric acid, and it was then concentrated in vacuo until the extract from 50 kg. of fresh suprarenals was reduced in volume to about 10 liters. This condensed extract was then heated to 80°, the coagulated proteids were filtered off and the clear filtrate benzoated in fractional poi'tions. It was found to be unnecessary to remove the proteids entirely.

Hoppe-Seyler's Zeitschr. f. physiol. Chem., vol. 24, p. 1-42. f It is a pleasant duty to acknowledge that this rpsearch would have been impossible but for the liberality of Messrs. P. D. Armour & Co., of Chicago, who have supplied me with large quantities of a concentrated aqueous extract of the suprarenals of the beef prepared according to my direction. My thanks are also due to Prof. A. G. Manns, chief chemist of the firm, forthe care he has taken in preparing these solutions and for the interest he has taken in the scientific aspects of tlie subiect.

The crude, sticky mixture thus obtained, which consisted of the benzoates of our chromogen, of inosit, possibly also of carbohydrates, creatine and other substances, was then washed thoroughly with water and then dissolved as far as possible in warm glacial acetic acid. A considerable residue remained undissolved. The acetic acid solution was poured into much ether, and again a great deal of material was precipitated. The acid ether solution was first repeatedly shaken out with water, causing a further deposition of resinous matter, and then with a solution of sodium hydrate until all the acetic acid was removed and only a clear but slightly colored ether solution of a benzoate remained. These repeated washings caused copious deposits to fall out.

The ether solution was again washed with water and then once or twice with a 10 per cent, solution of sulphuric acid, followed with water. This washing with acid was now discontinued as it caused the benzoate of the chromogen to fall out in the form of a sticky resin.

It will be seen that by the above processes a number of foreign benzoates are removed; thus, the benzoate of inosit being insoluble in glacial acetic acid and that of grape sugar in ether.

When the benzoate of the chromogen had been treated as stated the ether was removed by distillation, and a yellowish, sticky benzoate remained which became brittle when allowed to dry in the air in thin layers. By boiling its alcoholic solution with animal charcoal, further purification was effected, so that when small quantities of this alcoholic solution were allowed to evaporate bunches of prismatic crystals were deposited. Many different solvents have been tried, but from none does it crystallize with enough difficulty to leave a mother-liquor.

Nevertheless, I have been able to learn something as to the



[Nos. 90-91.

composition and natnre of the cliromogen, the assumed bloodpressure raising constituent. lu order to isolate this substance, the benzoate as obtained from the washed ether solution was decomposed with water in an autoclave under a pressure of 8-13 atmospheres. A clear, slightly straw-colored solution is thus obtained, which, when freed from benzoic acid and from a certain amount of a black resin which is deposited here as well as in other methods of decomposing the benzoate, gives all the well-known color and reduction tests of a fresh aqueous solution of the glands with one diiference, which is that the addition of a little ammonia and iodine water no longer gives the characteristic rose-pink color, but instead, a vivid green. In all other respects the chromogen appears to be unaltered. A little ammonia, however, is set free during the hydrolytic decomposition just described, but whether this is derived from our substance or from some benzoate still contaminating the benzoate of the chromogen cannot as yet be stated.

When all the benzoic acid has been removed from the solution of the benzoyl product as taken from the autoclave, the cautious addition of very dilute ammonia, drop by drop, causes a copious precipitation of a substance which falls out in a flocculent precipitate much as does casein when precipitated from milk with acetic acid. The precipitate rapidly darkens and must be removed with the help of a suction filter as rapidly as possible.

It is washed with a little water, then with cold absolute alcohol and ether, and immediately ground up in agate mortars while it is still moist with ether. On drying it becomes a light grayish powder. This is the free chromogen with such slight modification as has occurred during the hydi'olysis of its benzoate. When dry it is almost insoluble in water as also in a whole series of organic solvents ; it is very soluble in warm dilute acids, in cold glacial acetic acid and in acetic anhydride. Dilute solutions in slightly acidulated water give an intense green color with ferric chloride or with ammonia, and they reduce ammoniacal silver solutions. Such solutions, exposed to the air, gradually deposit a brown precipitate, and this goes on until but little of the chromogen is left.

The behavior of the substance toward the halogens, which all precipitate it from its solutions, and toward the numerous alkaloidal reagents, I hope to report on at some future date. I shall only say here that a little of the dried chromogen obtained by breaking up the benzoyl product with acids as described in my first paper and which still gives the rose-pink color with ammonia and iodine water, strikes a rich plum color when treated with a drop of sulphuric acid or with Mandelin's reagent, reminding one of the effect of similar tests on strychnine. The chromogen as derived from its benzoate by hydrolysis in the autoclave does not give this color, but an olive-green followed by pink, which gives place to dirty hues.

Strong alkalies decompose the substance, boiling it with alcoholic solutions of potassium hydrate and chloroform brings out the nauseating odor of a carbylamine. On attempting to isolate this volatile substance by distillation, it was found to be decomposed, and on again treating the distillate with alcoholic potash and chloroform, the carbylamine was regenerated, thus showing that a primary amine had been split off when the chromogen was treated in this way.


On fusing the substance with powdered potassium hydrate and then diluting with water, the penetrating odor of skatol rises from the solution. When this solution is shaken out with ether and the ether allowed to evaporate, little globules remain having an intensely foecal odor, and when these are dissolved in concentrated hydrochloric acid the solution at once takes on the fine characteristic pink color always seen when even small quantities of skatol are thus treated.

An alcoholic solution of these globules gives to a pine sliver, moistened with hydrochloric acid, a rich dark red color; a solution in benzol to which picric acid in benzol is added immediately deposits a picrate, not in crystals but in the form of reddish droplets, and an aqueous solution treated with sulphuric acid and potassium nitrite gives the whitish turbidity seen when skatol is similarly treated. Salkowski's reaction was also obtained, though imperfectly, as the production of intense colors in this test demands more substance than was left at my disposal.

The characteristic odor of this decomposition product, together with its chemical reactions, would make it appear that we have either skatol itself or one of the isomeric indols.

Some importance must be attached to this discovery, since, taken with the various reactions of the chromogen, the results of the elementary analyses and such facts as that dry distillation yields benzoic acid, amines, etc., and heating with zinc dust yields pyrol, it clearly enables us to classify the chromogen, in a preliminary way at least, among complex aromatic bases not very dissimilar from the alkaloids. The results of combustion analyses show that its empirical formula is C17H15NO4, thus approaching in elementary composition some of the alkaloids.

The composition of pseudomorphine, for example, is represented by CnHi9N04, that of cocaine by CnHsiNO., that of sanguinarine by CioHisNOi, and that of benzylideue collidine dicarboxyacid by CnHisNO., and among these alkaloids sanguinarine is noteworthy for its power to raise the blood pressure.*

In this connection, too, it is of interest to note that Stohrf has shown that skatol is liberated when strychnine is heated with calcium oxide, and that Hoffmann and KonigsJ have obtained indol from tetrahydroquinoline by passing its vapor through a tube heated to redness.


The results of the elementary analyses are as follows:

0.145 gm. of substance, dried in vacuo at 100° C, gave 0.3675 gm. CO. and 0.0681 of H2O or 69.1 3,'?; C and 5.24,-?; H.

A second anlilysis made with 0.1863 gm. of substance, prepared at a different time and with slight modifications, gave 0.473 gm. of CO.2 and 0.0103 gm. of H=0 or 69.28^ C and 6.09^ H.

A nitrogen estimation, using substance prepared at the same time as that used in the second carbon and hydrogen analysis, gave the following results :

H. Meyer, Arch. f. exp. Pathol, u. Pharmakol., XXIX, 426. t Berichte d. Deutsch. chem. Gesellsch., vol. 20, p. 1108. t Ibid., vol. 16, p. 738.

Septembek-October, 1898.]



0.1784 gm. of substance gave 7.8 cc. of N at 21° C. and under a barometric pressure of 761 mm. of mercury. In this estimation, tlierefore, the N amounts to bfc.

Putting these results in tabular form, we have



69.28 6.09 5.00

C — 69.12 H= 5.21

N =

Calculating for an empirical formula, we find that the expression CuHisN'Oi meets the requirements, since theory demands for C 68.68

H 5.05 N 4.71 The agreement between the percentages demanded by this formula and the results obtained by analysis is as close as could be expected, since we are dealing with an amorphous substance and one in which the percentage of both H and N is very low.

In all of the above analyses a correction of 9.2 per cent, in the weight of substance given is made for ash. In spite of the fact that the beuzoate itself is entirely free of ash, the amorphous chromogen obtained from it has carried down much mineral matter derived from the utensils used iu the cleavage experiments and in subsequent manipulations.

The acetate of the new substance was also prepared and analyzed. The results thus far obtained are not fully in accord with the above formula, but this is diie to the fact that the acetate decomposes during drying at 100° C. in vacuo. Acetic acid appears to be given off under these circumstances, and thus the C, H and N" content is much changed from that required by the formula thus given. I do not doubt that when analyses are made with the avoidance of this loss the results will be concordant.

Not being able to repeat this part of the work at present, owing to lack of material, I here append the results of analyses made with an acetate which was constantly losing weight. The acetate was prepared by dissolving the free base in glacial acetic acid, and allowing this acid solution to flow in a thin stream into ether. The acetate is at once completely precipitated and may readily be collected, washed and dried. The percentages obtained on analyses were

C = 58.16

H= 5.82

]Sr= 5.04 whereas the diacetate CnHuNO^CiH.OO: requires

C = 60.43

H= 5.51

N= 3.36

The analytic data for the above percentages of C, H and N are as follows :

0.153 gm. of acetate gave 0.3263 gm. of 00= and 0.0802 gm. of HjO ; 0.2046gm. of the same material gave 9.1 cc. of Nat 20.5°C. and under a barometric pressure of 754.6 mm. of mercury.

The method of preparation of the acetate does not tend to diminish the ash, and fusing the substance on platinum foil showed its presence in at least as large amounts as in the free base. In the absence of direct estimates for ash, it was thought fair to assume its presence to the extent of at least 10 per cent., and the weights here given have been corrected in accordance with this assumption.

I have already remarked that an analysis of the benzoate of the chromogen as thus far prepared showed it to contain C = 72.54^ H = 5.54$i^, N=i3.46^.

The analytic data are as follows :

0.2966 gm. of substance dried iu vacuo at 80" C. gave 0.78895 gm. of CO^ and 0.14785 gm. of HjO. 0.29656 gm of the same material gave 8.7 cc. of N at 18.25° 0. and under a barometric pressure of 760 mm. of mercury.

The mouobenzoate of CmHhNO. is CHhNO-.CO.OoH^ and requii'es that C = 71.82j^

H= 4.74^ N= 3.49^ whereas our analysis gives

C = 72.54 H= 5.54 N= 3.46

This discrepancy in the carbon and hydrogen percentages is readily accounted for as the amorphous resinous benzoate analyzed is exceedingly difficult to dry to constancy of weight, and is, furthermore, perhaps not quite free from foreign benzoates. The results of analyses, nevertheless, point to the conclusion that we have the monobenzoate of the new base before us.

The above-named methods of isolating the active jirinciple are far from being as satisfactory as could be desired. The resinous substance found in the autoclave on decomposing the benzoyl product always retains a considerable amount of the base. This may be extracted with dilute sulphuric acid and may then be precipitated with ammonia. This precipitation is, however, incomplete — a considerable amount of the base always remaining in solution. A considerable loss also occurs during the washing of the free base with water and alcohol, the latter agent especially dissolves considerable of the moist precipitate. The high ash content of the free base and of the acetate is also a most undesirable feature of the methods above described. Had there not been a tolerably fair agreement in the analytical results for the free base, its benzoate and iicetate, with good reasons for the divergence in the case of the acetate, I should have hesitated to publish my results at this time.

I have lately found in sodium picrate a good agent for the complete precipitation of the base from its solutions in dilute mineral acids. The picrate is fairly soluble in a number of organic solvents, as, for example, alcohol, acetic ether and methylal, and may be precipitated from its solutions in these agents by the addition of much ether. On redissolving and reprecipitating a yellow picrate is obtained which leaves no ash when burned on platinum foil, and which, I believe, can be made to crystallize. It is my intention to give, in the near future, a more detailed description of the properties of this picrate with analyses and molecular weight determinations.


To summarize the results of this investigation in a few words : The active principle of the suprarenal capsule has been isolated iu the form of a powder of a light gray to brownish color, whose percentage composition is expressed by the formula CnHuNO*. A primary amine and a methylindol are easily split off from its molecule by treatment with powdered alkalies.

Should molecular weight determinations prove that the above formula correctly expresses the molecular weight of the new base, it will be seen that its molecule can contain



[Nos. 90-91.

only one substituted benzene ring in addition to the nitrogenous complex of atoms from which the skatol is derived. Oxidation and substitution experiments are, however, still necessary before more definite statements can be made as to the constitution of this compound.

In its native state, as found in the suprarenal capsule, this substance differs by one chejjiical reaction only from its state as described in this paper. Chemically considered, the difference in composition between its native state and as here

described must be very slight. And yet this difference which is just marked enough to give a greater stability to the substance is also great enough, apparently, to deprive it of its power to raise the blood-pressure, for, jn the physiological experiments, thus far made, small quantities of the new base were found to be inactive in this respect.

I wish to express my thanks to my assistant. Dr. Walter Jones, for the valuable assistance rendered in making the analyses recorded in this paper.