Talk:Paper - The staged sequential development of the anus and rectum in human embryos and fetuses

By Pieter A. de Vries and Gerald W. Friedland

1 of anorectal development. For example, does an external cloaca and a proctodeum exist in human embryos? When and how do the cloacal membrane and the urorectal septum form and fuse? What role do the folds around the developing anus play in its separation from the urogenital tract? All these ques- tions are of great practical importance, because the current International Classification of anorectal malformations is based upon some of these con- tentious concepts. For this reason, we undertook a detailed study of the embryology of the anorectal region.

MATERiA'L3 AND METHODS As part of an ongoing study of hindgut and urinary tract development, we examined our own collection of human embryos and fetuses as well as the remarkable Carnegie Collection.

Streeter divided the human embryos of the Carnegie Collection into “Age Groups”l"5 based

on the stage of development of several organ systems.‘ We found Streeter’s “Age Groups" in- valuable because they offer a means of studying serial development which is not necessarily

related to the overall length of the embryo. RE"U'LTS

We have divided our findings in the embryonic Age Groups into three periods: (1) cloacal development; (2) the development of a urogenital sinus and rectum; and (3) the establishment of an anus.

(1) The Development ofthe Cloaca (Fig. 1)

This period extends from Age Group VI (12-15 days) when the early embryo has developed a distinct yolk sac, to the beginning of Age Group Xlll (about 28 days) when the mesonephric ducts join the cloaca. We will describe the develop- ment of the embryonic germ layers in human embryos during this period in some detail for two reasons: first, prior to organ development, cells move from

one area to another and from one plane to another; second, interference with either cell proliferation, morphogenetic movements, or tissues which produce

-1 &— -1 1 .1 :nj njj-jjur-_jj & 1 1 j— x— 3 _- I— -

"'Estimation of ovulation age has required revision.6 Recently O’Rahilly7 substituted the

term “Stage” for'“Age Group.” However, we wish to perpetuate Streeter’s idea of “horizons" which indicates that as development progresses from one stage to the next there are only arbitrary

quota-

demarcations in what is a continuous development or series. From the Departments of Surgery and Radiology. Stanford University Medical Center, Stanford,

..a!if. and the Department a,fSa.rgery, Santa Clara Valley Medical Center. San Jose. Calif.

Presented before the Fifth Annual Meeting of the American Pediatric Surgical Association, New Orleans, La., April 4-6, 1974.

Address for renrint requests.‘ Pieter A. de Vries, M.D.., D pertinent of Sargerv, Santa Clara

1' Valley Medical Center, San Jose, Calif. 95I25. © I974 by Grune & Stratton, Inc.

Journal of Pediatric Surgery, Vol. 9, No. 5 (October), I974

Fig. l. (A) The open bar denotes the range of lengths in embryos in each of the Age Groups. The bar is at the approximate age (days after ovulation) of embryos in the designated Age Group. The period extends from early appearance of the gut primordium to the formation of the cloaca. (3) Similar to Chart I covering a period from the formation of cloaca to the end of embryonic development (the beginning of fetal stages).

organizer factors may give rise to congenital anomalies much earlier than is often appreciated.

By Age Group VI the embryo has developed into a central blastodisc where the spherical amniotic cavity and yolk sac come into contact. The palisade of tall columnar cells which forms the epiblastic layer of the embryonic disc curves upward at its periphery: this layer, which becomes one cell thick, lines the roof of the amniotic vesicle. An irregular layer of large cuboidal foamy cells, beneath and opposite the epiblast, constitutes the hypoblast or endoderm of the embry- onic disc—the primordium of the gut epithelium. This layer is continuous at its periphery with the squamous cell lining of the spherical yolk sac. Beyond the embryonic disc the second cell layer of the yolk sac and amniotic cavity is an external coat of primary mesoderm. Since the primary mesodermal cells arose at an earlier stage primarily from the chorionic (outer, extraembryonic) layer 3-9 of the ovum, they also surround the peripheral edges of the disc where they form the body stalk.

Toward the end of Age Group VI, a short radial line, known as the primitive streak, appears on the surface of the epiblast at the periphery of the disc. A similar event occurs in other amniotes, and is most easily seen in bird eggs. This streak, which arises at the caudal end of the embryo, defines the vertical axis. When first seen, it is a caudal groove in the upper surface of the epiblastic disc which narrows into a slit more centrally. The base of the slit projects down- ward (inward) like a short narrow keel to contact the endoderm. The streak is probably homologous to the dorsal lip of the blastopore of anamniotes; it is the C! -12 n'1' ,... C) 1: 3:

g. .. . . hey invaginate

I“ to the interior of the evolving embryo to form the mesoderm. Primary ecto- dermal cells remain in the epiblast.

In embryos between Age Group VI and VIII we could see no area between the primitive streak and the body stalk which we could define as the cloacal membrane. By definition,” the cloacal in mbrane is th t area betwee th primitive streak and the body stalk where endoderm and ectoderm fuse without intervening mesoderm. Interestingly, during Age Groups VI and VII, some of the large vesicular endoderm cells in front of the primitive streak contain vacu- oles and appear to have lost some of their substance. Others have noted this phenornenon"‘” but have been unable to explain it. In Age Group VIII, the notochord, which is proliferating and elongating in front of the primitive streak, fuses with the underlying endoderm. Its lumen communicates with the amniotic cavity through the primitive pit. During Age Group VIII this common wall breaks down, thereby forming the neuroenteric canal between the amniotic cavity and the yolk sac. Since cellular vacuolization and lost of substance precede the breakdown of the endodermal cells, we believe that these phe- nomena indicate that the endodermal cells are degenerating. This form of cellular degeneration is peculiar in that it is associated with cellular prolifer- ation. A similar process appears to occur in later Age Groups when endodermal cellular proliferation results in thickening of the cloacal membrane which then breaks down.

A ductal structure emerges from the most caudal end of the gut endoderm in Age Group VII (Fig. 2A), and runs into the body stalk. Its cells, like those of the most caudal endoderm, are cuboidal. They differ significantly from the squamous cells of the yolk sac and from the cells of the primary mesoderm. This duct, known as the allantois, is important because it becomes part of the bladder and also serves a landmark indicating the future site of the posterior intestinal portal until Age Group X; thereafter it marks the ventrocephalic limit of the cloaca.

In the middle of the third week, Age Group VIII, the primitive streak occu- pies the caudal 25% of the embryo. This ratio persists during the following two same as that of the soma above it. Just beyond the yolk sac caudally in No. 9251, the allantoic duct contacts the terminal epiblast and amnion directly.* At the end of the third week, the para-axial mesoblast starts its metameric segmentation to form somites. In the Carnegie Collection there are only three histologic quality to be of value. These are No. 1878 (Ingalls)‘5 which has two somites, and No. 7650 which has two to three somites. In No. 7650, we first observed unequivocal fusion of ectoderm and endoderm without any interven- ing mesodermal cells (Fig. 2B, No. 7650). This early primordium of the cloacal

According to Florian“ V. Mollendorf (1921) described a similar fusion of allantois with amnion in an embryo resembling one of this Age Group.

Fig. 2. (A) Top row, left: Embryo 9222 borderline between Age Group VI and VII. Section through the caudal portion of the embryo at the primitive streak. Upper right: Section through embryo 7802 borderline between Age Group VII and VIII through the primitive streak. Middle left: Same embryo, at the most caudal portion of the primitive streak. Middle right: Some embryo showing the allantois iunctioning with the endoderm at the caudal end of the embryo. lower left: Section through the primitive streak in Age Group VIII embryo. lower right: Some emhryo sectioned through the gent: loyets ot their ioootion with the 'ooo'y stoiii. hiesohiast inter- venes between epiblast and hypoblast. No cloacal membrane. (3) Top three photographs are of sections through the caudal end of the 2-3 somite embryo, 7650. In the center photograph epihiast and hy,-L-ehlest unite to e cieecei ntetnhmne nheve the yeiit sec end heie-er the slit like end of the amniotic cavity. Top left, mesoblast can be seen intervening between the two primary layers. Top right, the allantois can be seen in the body stalk. lower left section through the posterior intestinal portal of Age Group X embryo. lower right, some embryo, through the evolving tail bucl showing the amniotic space above and behind it. The allantois can be seen in the body stalk below the amniotic cavity. (C) Transverse sections through embryo No. 2053, at the end of Age Group XI. Upper left, allantois coming off ventrally from the upper cloaca into the body staiii. ‘Upper right, a few sections further caudai, just beneath the body staik. ifiidciie photograph, section through the very top of the cloacal membrane. Note the lateral folds and thin membrane. Bottom photograph, section through the middle portion of the cloacal membrane she‘.-ring the thick endedernte! cern-,'.-enent and the thin ectederrne! centnenent of the cie-ace! inerti-

brane. The configuration of the lateral cloacal folds and their relation to the cloacal membrane also varies along the cloacal membrane.

‘CI Xlll,no. 8066 TRANSVERSE

Fig. 3. (A) Age Group XII embryo 8066. Upper left, section through the junction of the mesonephric duct with the cloaca. Upper right, first section below the body stalk showing cloaca and beginning of cloaca membrane. Bottom right, the cloaca and membrane near the most caudal portion. The relatively thin cloacal membrane is composed largely of endodermal cells. (3) Serial sections through the caudal portion of the embryo 8314, Age Group XIV. Upper left, at the junc- tion of the hindgut with the cloaca. Upper right, just below the junction. Bottom row, sections left to right, progressing from cephalad to caudad. At the left, the cephalic end of the cloacal membrane. On the far right just caudal to the future anal region. Lower right middle, shows the hindgut coming off the cloaca and the anal tubercles are seen on either side of the terminal cloaca. A v’./'

- - .- ' »- .- -' - . . . I .

-.0 V -_ . .-. I. . . I _ ~

-_' _.', r_ ‘H.-._. _.~.. ‘ll. . -._,-.-'._-,...l -'

65 -' " ~ 920 - . .I" ‘ -

«i . I _. -- _ ., - I-

Fig. 3. (C) Similar sections through embryos in Age Groups XV, XVI, and XVIII at the same magnification except for the photograph in the lower right which has an increased magnification. The sections are from cephalad to caudad as one goes from left to right. Center left section is iust above the iunction of hincigut and cioaca. Center right is at the iunction. Far right are near or at the anal end of the cloaca. (D) Sagittal sections through embryos of Age Group XIV through XVIII. They show the evolution of urorectal septum as seen in the midsagittal plane. The thiclx

cienc-.:! tnetnhtn e, nssccinteci with the tngenitn! sinus, can he seen .;enttn!!-,= with the thinner membrane associated with the anus seen dorsally. In Age Group XVI through XVIII the develop- ment of the ventral abdominal wall beneath the body stalk can be seen. The indentation on the

dorsal wall of the cloaca is the anal bulb, the future site of the columnar portion of the ano- rectum. 762 DE VRIES AND FRIEDLAND

embryo folds so that the allantoic duct and evolving cloacal membrane shift ventrally. This occurs as a result of greater dorsal than ventral growth. We believe that this difference in growth is due to at least two factors: first, fewer cells have, as yet, migrated ventrally from their middorsal convergence in the epiblast; and even if the rate of cell division in the two regions were the same, exponential cell growth would account for a greater cell mass dorsally (we see no regional difference in the number of mitotic figures); second, there appears to be little growth in the yolk sac at its junction with endoderm. As a conse- quence of this folding, the primitive streak becomes the most caudal portion and lies at the apex of the fold.

In Age Group X (4-12 somites), the hindgut starts to develop in the hollow of the U-shaped tail fold. The primitive streak, now only 18% of the embryo length, grows away from both the dorsal soma and the elongating cloacal mem- brane. By the end of this age group, the differentiating somites extend from the anterior to the posterior intestinal portal (growth and development of the cephalic end has progressed ahead of the caudal end). The hindgut above the body stalk elongates; the posterior intestinal portal migrates toward the head.

The period which extends from the appearance of the first somite to the de- velopment of the last (Age Group X to XIII), coincides with the inception and maximal caudal extension of the hindgut. During this same time the orifice formed by the junction of the yolk sac and gut endoderm constrict while the fore- and hindgut elongate.

Unlike the more centrally placed somites, the lateral plate mesoderm never segments. It, however, divides into two laminae, with a space, the coelum in between. In the central portion of the embryo, surrounding the yolk sac where there is as yet no ventral body wall, the space within the embryo (embryonic coelum) communicates freely with the space surrounding the embryo (extra embryonic coelum). In the lower portion of the embryo, the inner lamina sur- round the endoderm and elongates the hindgut above the cloaca; the outer lamina builds the body wall. However, in the region developing after the in- ception of the tail fold, the lateral mesoderm does not split and the coelum extends only to the upper part of the cloaca; below this level, the unsplit meso- blast surrounds the cloaca and tailgut. Because the cloacal membrane contains no mesoblast, the mesoblastic proliferations on either side appear as folds sur- rounding it. These cloacal folds” (genital fold'3'”) first appear at the upper margin of the cloaca and extend progressively caudally.

By Age Group XI, the endodermal walls of the cloaca have thickened a little. This thickening is greater on its ventral than its dorsal wall and greater on its cephalic than its caudal portion. Where the cloaca joins the allantois, the mem- brane is narrow and composed of thin layers of endoderm and ectoderm (Fig. 2C). Below this, the membrane widens and thickens, but thins again near its most inferior portion. In the central thickest portion the endoderm clearly accounts for by far about three quarters of the thickness. [Keibel’° estimated endoderm to comprise four-fifths, ectoderm one-fifth in 3-mm embryo (Age Group XII).] Embryo No. 6344, 13 somites, has a cloacal membrane of 102 a, whereas, in No. 5072, 17 somites, it has increased in length to 180 p.

During Age Groups XII (21-29 somites) and XIII (Fig. 3A) (30 or more somites) the gut extends beyond the cloaca into the tail; therefore, we could not determine the exact length of the cloacal membrane. In Age Group XIII, for instance, our measurements range between 195 and 344 1;. Nevertheless, its length has increased progressively through the four age groups during which the cloaca has developed completely as a definable structure. It continues to thicken.

The mesonephric ducts join the superolateral wall of the cloaca just inside the cloacal membrane at the end of Age Group XII or beginning of Age Group XIII, and the folds surrounding the cloaca become most prominent at‘ its superior and inferior margins.

(2) Development of the Urogenital Sinus and Rectum (Age Groups X I V-X VIII )

At the beginning of Age Group XIV (Fig. 3B), the cloaca is a relatively large chamber into which a hindgut of smaller caliber empties superiorly and from which a tailgut, also of smaller caliber, exits inferiorly. Just in front of the hindgut the allantoic outlet projects ventrosuperiorly.

During Age Groups XIV and XV (about 32-33 days) the ventral abdominal wall beneath the body stalk evolves; in Age Group XIII, in the ventral midline, the allantois extended directly into the body stalk. New growth of mesoblast beneath the body stalk displaces the upper end of the cloacal membrane in— feriorly. The descent of the membrane is, in effect, like that of a door hinged on the anal tubercles, which swings down through an arc of about 105°. The anal tubercles,“ which become prominent in Age Group XIV, cause surface eleva- tions on either side of the lower end of the cloacal membrane. They form as a result of enlarging mesoblastic nodes on either side of the junction of the tailgut with the cloaca and are most prominent laterally as surface elevations. More significantly, however, they progressively impinge upon the lumen of this junc- tion which was once wide in Age Group XIII. The mesoblastic tubercles fuse centrally therby displacing the cloacal orifice of the involuting tailgut dorsally, away from the membrane. In some sagittally sectioned embryos of Age Group XV the residual epithelial tract of the tailgut is visible within the mesenchyme, arching over these fused anal tubercles. It extends from the dorsal cloacal wall, well above the cloacal membrane, to the ectodermal surface well posteriorly.

There is a dramatic increase in the thickness of the cloacal membrane, in Age Group XIV and XV. Coincidentally, the urorectal septum develops and clearly demarcates the boundaries of the future rectum and anus dorsally and orogenital sinus ventrally. That portion of the cloacal membrane associated with the urogenital sinus is significantly thicker. In a frontal view in Age Group XV, the evolving urorectal septum has the configuration of a horseshoe. Meso- blasts proliferates on either side of the cloaca along a line extending from the point above, where the ventral wall of the hindgut meets the allantois, down- wards dorsal to the junction of the nephric duct with the cloaca; this meso- blastic proliferation extends down into both lateral cloacal folds at the margins of the cloacal membrane. Continued mesoblastic and endothelial proliferation constricts the cloacal lumen within the plane defined by the above~mentioned line, toward a focal point on the inner surface of the cloacal membrane. This

Fig. 4. (A) Series of sagittal sections through two Age Group XV embryos. The upper series a younger embryo, shows the lesser development of the anorectum and urogenital sinus. The series progresses from the left to right from midsagittal plane to the embryos right lateral region. The lower series of an older embryo, similarly sectioned—trom rnidsagittal to the right lateral region. As one progresses from me_dial to lateral one can see the apparent fusion of rnesoblast which is, in fact, an illusion for the tissue is continuous in a horseshoe configuration about the cloaca. (B) All but the right lower photograph are of midsagittal sections and show the anorectum after dissolution of the anal membrane. The primordium of the rectal musculature can be well seen in Age Group XIX. The internal longitudinal musculature can be well visualized. The puborectalis can be seen dorsally in Age Group XX. In Age Group XXIII all the muscles both of the anorectol wall and the pelvis can be seen. lower right, pubococcygeus is seen streaming

upward in front of the sacral nerves decreasing from out of the spinal cord. (C) Frontal sections of Age Group XXIII embryos, on the Iett, the inner circumferential muscle of the rectum is partially cut through. The longitudinal muscular coat of the rectum terminates within the external

sphincter. The levator musculature can be seen iunctioning with the external sphincter. In the right hand photograph, the future crypts associated with the columns of Mergagni are clearly seen together with the external sphincter below the termination of the levator ani muscles.

constriction progresses towards the midline. The mesoblastic proliferation within the folds imparts a figure—eight configuration to these folds around the cloaca.

The pattern of growth, which is well established in Age Group XV (Fig. 3C), progresses during Age Groups XVI, XVII, and XVIII (Figs. 3D, 4A), so that by the middle of the seventh week (the beginning of Age Group XIX) the anus and rectum are completely divided from the urogenital sinus (Fig. 4B, No. 6432).

Complete separation of the anorectum from the urogenital sinus, in Age Group XIX by epithelial (endoderm) fusion, occurs at a small central area just above the membrane. At the time that the anal membrane breaks down, (some- 766 DE VRIES AND FRIEDLAND

what after the urogenital membrane breaks down.in later Age Group XIX), the mesodermal perineal body extends to about the same level as the anal folds (hillocks). In sagittal sections of Age Group XIX, this fact is well seen (Fig. 4B). Ectoderm covers the inferior edge of the narrow perineal body before the anal membrane breaks down. At this time, the cloaca is completely divided and no “external cloaca” exists.

(3) Development of the Arms

The anal tubercles are first seen in Age Group XIV. After they unite behind the cloaca in Age Group XV, they continue to grow, thereby forming a U-shaped fold dorsally and laterally between the tail and the anus. Internally, the dorsal cloacal wall in Age Group XVI evaginates just above the margin where the tubercle impinges on the lumen. This evagination subsequently be- comes the bulbus analis portion of the anorectal wall. The future site of the crypts and columns of Morgagni is, therefore, clearly defined long before rup- ture of the anal membrane in the middle of Age Group XIX (about 48 days, 16- to 18-mm length) and arises from endoderm. The anorectal columns, as such, are first seen in Age Group XX and XXI (51-52 days). From Age Group XIX through XXIII (Fig. 4C), the end of the embryonic period, the anorectal musculature becomes defined. In Age Group XIX the termination of the cir- cumferential muscle of the rectum is well seen. The anal portion of the rectum is relatively quite long extending for a relatively great distance above the site of the anal membrane. This endodermally lined anorectum has clearly a hindgut origin and at the close of the embryonic period the epithelium like the rest of the gut is stratified cuboidal. Before the start of the fetal period, the beginning of the ninth week the external sphincter, levator ani, and particularly the puborectalis, and even the ganglionic plexi of the rectum are well defined and remarkably developed.

In fetuses of 55-mm the anal portion of the rectum has become relatively much shorter by a gradual relative shortening and broadening. This process continues in later fetal life.

DISCUSSION

In this study, we, as othcrs,2°'23*2‘ have failed to find either an external cloaca or a significant proctodeum. Some claim that such entities exist.”"9'25'2° How can we reconcile these divergent viewpoints‘? An understanding of how some authors derived their facts may explain these differences of opinion.

Embryologists prior to 1880, commonly referred to as the “classical embry- ologists,” had few mammalian embryos for study, so they used lower forms, particularly birds, instead. They claimed” that, during anorectal development, the ectoderm invaginates to form a cloacal bud which grows into the mesoderm, develops a lumen, and communicates with the “internal cloaca”; they named the structure “the external cloaca.” We believe that these earlier observers probably examined chick embryos which were not closely staged. As a result, they mistook the development of the bursa of Fabricius for the subsequent development of the external cloaca in the chick. Placental mammals including humans, unlike the chick,” monotremes,” and even marsupials” lack an external cloaca. Attempts to find in man a structure homologous with an external cloaca or proctodeum have, we believe, led to misinterpretation of normal development.

We have enlarged on an earlier study” of how the cloacal membrane and urorectal septum form and fuse. Neither Rathke,-3‘ who believed in central approximation of lateral folds, nor Torneaux,” who believed in a descent of a urorectal septum, accurately described cloacal division. We might well have accepted Keibel’s contention that it does not really matter exactly how the division takes place,” were it not for the fact that it becomes significant regard- ing the nature of the development of congenital abnormalities. For example, Stephens” alluded to Wood-Jones’3“ statement that the cloaca is normally un- related to the anus and rectum and that a “proctodeum” contributes sig- nificantly to the development of the anus. Wood-Jones’ descriptive embryology is unacceptable. However, Bill on the other hand cited Felix” and Retterer” in his description of the development of “cloacal” or high anorectal defects with rectourethral or rectovaginal fistulae. His description has limited value, for it does not accurately portray the division of the cloaca, nor does it fully elucidate the genesis of anorectal malformations. Bill’s view of the genesis of perineal fistulae associated with imperforate anus and the International Classification of covered anus appear to be largely derived from Tench’s view of perineal embryology.* Tench” revived the concept of an external cloaca. He derived this hypothetical structure we believe for two reasons: First, he did not recognize the normal variation in thickness of the anal and urogenital portions of the cloacal membrane, probably because his study was largely on fetuses; second, despite his correct observation that the urorectal septum appeared to extend down as far as the cloacal tubercle in a sagittally sectioned embryo (Age Group XXIII), he curiously insisted that the separation of the urogenital sinus from the anus was the result of forward growth and medial fusion of anal tubercles in later fetal life. We believe this view is the result of the manner in which he con- ducted his study.

Tench’s study was based on 17 fixed unsectioned human embryos which he “photographed to show the region in which the anus develops.” He then serially sectioned them in a sagittal plane. Although his findings on sections were in conflict with those derived from his photographs he chose to accept the latter. What accounts for the discrepancy in findings?

In studying whole fixed embryos we have found, as did Streeter,' “specimens are quite transparent after formalin fixation." They are not, however, uni- formly transparent. An apparent absence of tissue in the perineal area, or ventral anal wall, is striking when viewed through a binocular microscope. However, on probing the area we found it to be transparent rather than absent.

Bill’s references are interesting because of the conflicting views of the authors. Retterer36 agreed with Rathl-re“ as regards the urorectal septum fusing from side to side whereas Felix” accepted Torneaux’s33 view of descent. Bill’s description more closely resembles that described first by Torneaux.13 In regard to the anus, Tench stated” (p. 339) . . it is difiicult to understand how these tubercles [anal] could bend posteriorly, reversing this process as claimed by Retterer (l890).'” Yet, Bill” who appeared to agree in 1953 with Tench‘s description recently cited only Reichel39 who also agreed largely with Torneaux.

Embryologists of the last century believed that the cloacal membrane is formed ventrally from the primitive streak. Florian described the cloacal mem- brane arising in early embryos (Age Group VI) separate from the primitive streak, by primary mesoderm. Fusion, he believed, occurred at a later time. Other embryologists have also subsequently proclaimed the existence of a cloacal membrane in Age Groups VII and VIII. We, on the other hand, have studied some of the same embryos, and do not confirm their findings. We, as Keibel, could not definitely identify a cloacal membrane until Age Group IX. Florian believed that the primitive streak mesoderm contributes to the body stalk. We have studied the development of the infraumbilical abdominal wall, and find his view compatible with our observations.

A detailed analysis of hindgut and anorectal anomalies together with our views on the time and nature of their genesis will be subsequently presented.

SUMMARY

Human embryos do not normally have an “external cloaca”; they have at the most, a slight anal depression, but no proctodeum. The division of the cloaca into the anus and urogenital sinus is a process which involves spatially continuous mesoblastic and endodermal proliferations of the cloacal wall. This results in the establishment of a horseshoe-shaped septum, which constricts onto the cloacal membrane because of cellular growth. The cloacal folds repre- sent the heaped-up mesoblast around the cloacal orifice. Within the folds at the locus of the urorectal septum in front of the anus, the mesoblast is an insepar- able part of the septal structure. In contradistinction, the separated anal tubercles fuse medially to separate the cloaca from the tailgut. These folds do not, however, contribute to the formation of an “external cloaca.” The ap- parent cephalocaudal and dorsoventral growth is explicable within the recog- nized parameters of axial cell migrations and proliferations.

ACKNOWLEDGMENT

We wish to thank Dr. James D. Ebert, Director, Department of Embryology, Carnegie Institute, Washington, and Dr. Ernest Gardner, Associate Director, Carnegie Collection of Embryos, De- partment of Anatomy, University of California, Davis, for making the unique collection of human embryos totally available to us for study.

REFERENCES

l. Streeter, GL: Developmental horizons in human embryos. Description of age group XI, 13 to 20 somites, and age group XII, 21 to 29 somites. Contrib Embryol Carnegie Inst 301211, 1942

2. Streeter GL: Developmental horizons in human embryos. Description of age group XIII, embryos about 4 or 5 millimeters long, and age group XIV, period of indentation of the lens vesicle. Contrib Embryol Carnegie Inst 31:27, 1945

3. Streeter GL: Developmental horizons in human embryos. Description of age groups XV, XVI, XVII, and XVIII, being the third issue of a survey of the Carnegie Collection. Contrib Embryol. Carnegie Inst 32:133, 1948

4. Streeter GL (prepared for publication by C. H. Heuser and G. W. Corner): Developmental horizons in human embryos. Descrip- tion of age groups XIX, XX, XXI, XXII, and XXIII, being the fifth issue of a survey of the Carnegie Collection. Contrib Embryol Carnegie Inst 34:l75, 1954

5. Heuser CH, Corner GW: Developmental horizons in human embryos. Description of age

group X, 4 to 12 somites. Contrib Embryol Carnegie Inst 36:29, 1957

6. Olivier G, Pineau H: Horizons de Streeter et age embryonnaire. Bull Assoc Anat 47:.573, 1962

7. 0’Rahilly R: The timing and sequence of events in human cardiogenesis. Acta Anat (Basel) 79:70, 1971

8. Hertig AT: Angiogenesis in the early human chorion and the primary placenta of the macague monkey. Carnegie Inst Wash Publ 459, Contrib Embryol pp 37-81, 1935

9. Hertig AT, Rock J: Two human ova of the pre-villous stage, having a developmental age of about seven and nine days respectively. Carnegie Inst. Wash Publ 557, Contrib Embryol 31:65, 1945

10. Kolliker A: Entwicklungsgeschichte des Menschen. 1869 p 271

11. Stieve H: Bin 13 1/2 Tage altes, in der Gebéirmtitter erhaltenes menschliches Ei. Z Mikr Anat Forsch 7:295, 1926

12. Fetzer M, Florian J: Der Embryo “Fetzer" mit beginnendes axialmesoderm- bildung and bereits angelegter Kloakenmembran. Z Mikr Anat Forsch 21:35l, 1930

13. Hertig AT, Rock J: Two human ova of the pre-villous stage, having an ovulation age of about eleven and twelve days respectively. Carnegie Inst Wash Publ 528, Contrib Embryo] 291127, 1949

14. Heuser CH, Rock J, Hertig AT: Two human embryos showing early stages of the definitive yolk sac. Contrib Embryol Carnegie Inst 31:85, 1945

15. Ingalls NW: A human embryo at the beginning of segmentation with special refer- ence to the vascular system. Contrib Embryol Carnegie Inst 11:61, 1920

16. Florian J: The early development of man, with special reference to the development of the mesoderm and cloacal membrane. J Anat (Lond) 67:263. 1933

17. Felix W: The development of the uro- genital organs, in Keibel F, Mall FP (eds): Manual of Human Embryology, vol 2. Phila- delphia, Lippincott, 1912

18. Torneaux F: Sur les primiers developpe- ment du Cloague du tubercle genital et de l‘anus chez l"embryon de mouton. J Anat Physiol 241503, 1888

19. Tench EM: Development of the anus in the human embryo. Am J Anat 591333, 1936

20. Keibel F: Zur entwickungsgeschichte des menschlicher Urogenital apparates. Arch Anat Physiol 1896, p. 55

21. Reichel P: Die Entwicklung des Darmes und ihre Bedeutung Ptir die Entstehung gewisser Missbildungen. Z Geburtsch Gynak Bd 142582, 1888

22. Johnson FP: Development of the rectum. Am J Anat 16:1, 1914

23. Pohlman AG: The development of the cloaca in human embryos. Am J Anat 12:1, 1911

24. Strahl H: Zur Bildung der Kloake des Kaninchens. Arch Anat Phys Anat Abt. 1886

25. Nagel W: Ueber die Entwicklung der inneren und ausseren Genitalien bein menschlichen Weibe. Arch Gynéiko145:453,

26. Minot CS: Human Embryology. New York, Macmillan, 1897, p 259

27. Gasser E: Cloaca. Arch Anat Physiol Anat Abt, 1880. p 297

28. Goppert E: Beitrage zur vergleichenden Anatomic des Rehlkopfs und seiner Umgebung mit besonderer Bertickaischtegung der Monotremen. Semon’s Zool Forsch in Australian, 3, part 2, 1901

29. Burns RK: The effect of male hormone on the differential of the urogenital sinus in young oppossums. Carnegie Inst Wash Publ 557, Contrib Embryol 2062163, 1945

30. de Vries P, Friedland GW: “H-Type“ ano-urethral fistula. Radiology. In press.

31. Rathke H: Abhandlungen zur Bildungs und Entwickelungsgeschichte des Menschen und der Thiere. Leipzig, 1832

32. Tourneux MF: Mecanisme survant lequel s‘operent la disjonction du rectum d‘avec le bouchon cloacal, et la formation de l‘anus chez l‘embryonen du mouton. C R Soc Biol (Paris) 2:207, 1890

33. Stephens FP, Smith EP: Anorectal mal- formations in children. Year Book, Chicago. 1971, p 143

34. Jones FW: The nature of malformations of the rectum and the urogenital passages. Br MedJ 2:1630, 1904

35. Bill AH, Johnson RJ: Failure of migra- tion of the rectal opening as the cause for most cases of imperforate anus. Surg Gynecol Obstet 1061643, 1958

36. Retterer E: Sur l’origine et Pévolution de la région anogenitale des mammiferes. J Anat (Paris) 262126, 1890

37. Bill AH, Johnson RJ: Congenital median bond of the anus: Report of 6 cases with results of surgical events leading to abnormality. Surg Gynecol Obstet 971307, 1953

38. Johnson RJ, Polken M, Derrick W, et al: The embryology of high anorectal and associ- ated genitourinary anomalies in the female. Surg Gynecol Obstet l35:759

39. Reichel P: Die Entwickelung der Harn- blase and I-Iarnrohre. Verh Phys Med Ges Wiirzburg 27:147, 1893