A COMPARATIVE STUDY O F THE VAS SUBINTESTINALE I N THE VERTEBRATES P A U L I N E KIMBALL Department of Anatomy. New Pork University NINE FIGURES CONTENTS Introduction ........................................................ A . General considerations ........................................ ............ B. Scope of the present investigation . . . . . . . . . . . . . C Material ..................................................... Present investigation ................................................ A Reptilia ..................................................... 1 Chelonia (turtle) ...................................... 2 Crocodilia (alligator) .................................. 3 . Squamata ............................................. Ophidia (snake) .................................... Lacertilia (lizards) ........... .................... B. Aves ................................. .................... 1. Struthiones (ostrich) .................................. 2 . Gaviae (tern) ........................................ 3 . Anseres (duck) ........................................ 4 . Gallinae (grouse) ...................................... C Mammalia ................................................... 1. Marsupialia (opossum) ................................ 2 . Carnivora (cat) ....................................... 3 Insectivora (mole) ..................................... 4 Edentata (armadillo) .................................. 5 Primates (man) ....................................... Conclusions and summary ............................................ Bibliography ........................................................ . . . . . . . . 371 373 374 375 376 376 377 379 379 379 380 380 381 382 383 383 383 384 386 387 388 389 393 397 INTRODUCTION This investigation was concerned originally with the question as to whether o r not a subintestinal artery occurs during the development of all of the mammals having an allantoic 371 THE A M E R I C A N J O U R N A L OF ANATOMY. VOL . 42. N O . 2 372 PAULINE KIMBALL placenta. A subintestinal connection between the umbilical and the omphalomesenteric arterial circulations has been known for some years to occur in embryonic mice, rats, and pigs, but apparently no effort has been made to find out whether the occurrence of a subintestinal artery is exceptional in these forms or characteristic of placental mammals in general. The question of the presence of a subintestinal artery in man was raised by Senior ( '25) in his discussion of one type of anomaly of the human umbilical artery. Anomalies reported by Kuliga ( '08) and Hafferl ( '19), Senior considered, were examples of this type. In a newborn infant described by Kuliga ('08), blood reached the placenta through the coeliomesenteric artery. This vessel, after giving off the usual intestinal branches, left the mesentery to pass around the left side of the ileum 10 em. above the caecum to become continuous with the single umbilical artery of the umbilical cord. Examination of the hypogastric arteries revealed the absence of the intra-abdominal part of both arteries. Senior thought this anomalous vessel consisted of three parts : a proximal portion (omphalomesenterit-:),a middle portion (a. subintestinalis), and a distal portion (umbilical). This interpretation was strengthened by Hafferl's description ('19) of an artery in a woman of middle age which arose from one of the ileal branches of the superior mesenteric artery and joined the right lateral umbilical ligament. It would appear in this case that the subintestinal artery had retained both its anterior and its posterior connections and had loosened itself from the gut wall when the gut began to lengthen. To test Senior's hypothesis of the presence of a subintestinal artery in man, as well as its probable presence in the embryos of all the Eutheria, the subintestinal artery was sought in the embryos of all placental mammals of which material of suitable age was available. Upon consulting the literature, it was found that a subintestinal vein, although it has been described as occurring in VAS SUBINTESTINALE IN T H E VERTEBRATES 373 all of the Anamia, has been found in only two of the Amniota, the chick and the opossum. It has never been described in Reptilia. It was decided to extend this investigation to include a study of the subintestinal vein in as many forms as the material available would permit. Since the a. subintestinalis and the V. subintestinalis are identical as far as position is concerned, they are, no doubt, homologous vessels. They differ only in their connections. The name, vas subintestinale, is used to indicate their morphology ; the names a. subintestinalis and v. subintestinalis are used to indicate their function. A . General considerations The primary vascular system of vertebrate animals consists of a median dorsal artery running dorsal to the alimentary canal and a ventral or subintestinal vein. I n animals in which the anus is not terminal, that portion of this simple longitudinal circulatory loop lying behind the anus is called caudal (artery and vein), while the portion anterior to the anus constitutes the v. subintestinalis. This type of vascular system, which is almost completely retained by Amphioxus, has been recognized for many years as making its appearance in the embryos of Pisces and Amphibia. The vas subintestinale is venous in the lower vertebrates, but arterial in the vertebrates with an allantoic placenta, with the possible exception of some Marsupialia (Perameles. Hill, '97) and some of the Lacertilia (Scincid lizards),l in which conditions have not been reported. The a. subintestinalis of the higher vertebrates is, however, a true homologue of the v. subintestinalis of the lower vertebrates. It is identical as f a r as position is concerned, but its connections have ceased to be venous and have become arterial. Position, it would seem, rather than size or connections (and therefore function), An allantoic placenta has been reported as present in the following Scincid lizards : Chalcides tridactylis and C. ocellatus (Giocomini, '91), Tiliqua seincoides (Flynn, '23)' Lygosoma entrecasteaux (Harrison and Weeks, ' 2 5 ) , L. (Hinulia) quoyi, Egernia whitei, and E. striolata (Weeks, '27). 374 PAULINE KIMBALL should be the criterion of individuality of the blood vessels and of nomenclature. A comparative study of the formation of blood vessels shows that it is not unusual for a welldetermined blood route connected in one way in one or more forms to acquire new connections and to assume an entirely different function in other forms. The present investigation was undertaken at the suggestion of Prof. H. D. Senior, Department of Anatomy, New York University. I wish to take this opportunity to thank Professor Senior for his advice and criticism of this paper and for the facilities he has placed at my disposal. B. Scope of the presefit ifivestigatioN As f a r as the original question as to whether o r not a subintestinal artery occurs in all of the mammals with an allantoic placenta is concerned, its presence is described in Carnivora (cat), page 386 ; Insectivora (mole), page 387 ; Edentata (armadillo), page 388, and Primates (man), page 359. Material for its study in Sirenia, Cetacea, and Cheiroptera has not been available. Ravn ('94) was the first to describe a subintestinal connection between the umbilical and omphalomesenteric arteries iii mammals. He observed this connection in rat and mouse embryos and called it the 'omphalomesenteric artery. ' His nomenclature is based on function-the fact that this vessel transmits blood from the single umbilical artery to the yolk prior to the time when the proximal part of the definitive omphalomesenteric arteries are established. The subintestinal position of this early connection between the umbilical and the omphalomesenteric arteries entitles it to be regarded in the morphological rather than in the physiological sense, for it represents the embryonic subintestinal vein of the lower vertebrate animals which in placental mammals has acquired arterial connections. Evans ('12) reported a similar arrangement in pig embryos measuring 7.5 mm. to 9 mm., and Sabin ('17) was the first t o VAS SUBINTESTINALE IN THE VERTEBRATES 375 apply the proper nomenclature-subintestinal artery-to this arterial connection between the umbilical and the omphalomesenteric circulations. Evans merely makes a statement of the existence of this vessel, while Sabin gives a detailed account of it. Bartfeld ('25) observed an arterial connection which he called the ' rete periintestinale caudale ' between the omphalomesenteric and the umbilical arterial systems in Talpa europea (mole). He did not associate this 'rete' with the vas subintestinale. It is doubtful that he is familiar with Sabin's work ( '17), for he makes no reference to it in his paper. The v. subintestinalis, as f a r as is known, has never been described in Reptilia. A description of this vessel in Chelonia (turtle), Lacertilia (lizard), Ophidia (snake), and Crocodilia (alligator) is given, pages 376 to 380. A subintestinal venous connection between the omphalomesenteric and umbilical circulations has been described in Aves (chick) by Hochstetter ( '88). Other writers, notably Lillie ( '08), Evans ( '09), and Sabin ( W ) , have also described it. Sabin's account is the most complete. As f a r as is known, it has not been described in any other bird. A description of it as it occurs in Struthio (ostrich), Anas (duck), Bonasa (grouse), and Sterna hirundo (tern) is given, pages 380 to 383. The subintestinal vein has not been described as occurring in the monotremes. It has been reported t o be present in one of the marsupials (opossum) by Senior ( '25). C. Material The material used in making this investigation was obtained from the following embryological collections : California Embryological Collection (C. E. C.). Carnegie Institution Embryological Collection (C. I. E. C.). Harvard University Embryological Collection (H. E. C. ). Huntington Embryological Collection, Columbia University (Hun. E. C.). Texas Embryological Collection (T. E. C.). Wistar Institute Embryological Collection (W. I. E. C.). ~ 376 PAULINE KIMBALL I wish to express my appreciation and thanks to Profs. H. $1. Evans, G. L. Streeter, J. L. Bremer, A. Elwyn, and J. T. Patterson, and Dr. M. J. Greenman for their generosity in placing their material at my disposal. The embryos studied were for the most part cut either into cross or longitudinal sections. Wax reconstructions were made from drawings of the sections projected t o scale by the Edinger apparatus. In the case of the Evans human embryo, no. 168, graphic and celluloid reconstructions were made from photographs of cross-sections of the embryo. PRESENT INVESTIGATION A . Reptilia A subintestinal vein has never been described in Reptilia. Ruckert (’06) bases much of his account of the early yolk vessels in Reptilia upon Strahl’s study of Lacerta. At the stage of development illustrated in Hertwig’s Handbuch, figure 870, the area vasculosa resembles that of a similar stage in selachians. It will be noticed, however, that there is no curving upward of the venous net in the caudal portion of the embryo. Ruckert calls attention to this fact and points out that the external ring vein is not connected with the vessels (subintestinal veins) within the posterior part of the body of the embryo. Figure 871 (Hertwig’s Handbuch) is that of a little later stage, showing the appearance of the aa. vitellinae. These are very numerous. They are lateral branches of the aorta and arise ab initio. They are “in toto homolog den dorso-ventralen Darmgefassen der Amphibien, die von der Aorta hinab zum Dotter ziehen.” It is of interest to call attention to a certain description by Harrison and Weeks (’25) of the fetal blood circulation in Lygosoma entrecasteauxi ( Scincid lizard). These fetuses measured 30 to 40 mm. They state, “as the yolk-stalk is fol lowed outwards, it is seen that a second and smaller vitelline vein enters the umbilical vein.” VAS SUBINTESTINALE I N THE VERTEBRATES 377 The Scincid lizards are peculiar in that they have both a yolk-sac placenta and an allantoic placenta. The former precedes the latter, both phylogenetically and ontogenetically. Later, they fuse to form a common stalk containing the yolk and allantoic vessels. The writer offers this interpretation of Harrison and Weeks ’ description : A subintestinal connection does occur between the two circulations when the two placentae are separate. Later, when the yolk and allantoic stalks fuse to form a common yolk-allantoic stalk, the subintestinal vein is carried out with the vessels of each stalk, so that “as the yolk-stalk is followed (outward, it is seen that a second and smaller vitelline vein enters the umbilical vein” (the umbilical and omphalomesenteric veins are single). I n this case, the subintestinal vein is not involved in the formation of the later gut vessels, as is the case in the other vertebrates, but is carried out in toto into the stalk where it remains throughout fetal life in its primitive condition. It is to be regretted that material f o r the study of the complete history of the v. subintestinalis in each of the orders, Chelonia, Crocodilia, and Squamata, is not available. 2 . Chelouzia. Four embryos of Thalassoclrelys (loggerhead turtle) examined have a v. subihtestinalis. They are no. 826 (8.5 mm.), no. 827 (9 mm.), no. 841 (10 mm.), and no. 842 (10 mm.) of the Hun. E. C. All have undergone incubation for eleven days. The subintestinal vein is collapsed in spots in no. 826 (8.5 mm.) and no. 827 (9 mm.). These embryos resemble no. 841 (10 mm.), which was reconstructed, and are of interest because they serve as a guide to the time and length at which the subintestinal vein appears. The subintestinal vein in turtles is double at first, but later becomes single. No. 842 (10 mm.) has a double subintestinal vein which extends posteriorly from the yolk sac more than half the length of the hind gut. Anteriorly, each of these veins opens separately onto the yolk; posteriorly, they coalesce to form a single channel, which soon splits at the cloaca to join the subcardinal veins. 378 PAULINE KIMBALL The condition of the subintestinal vein in no. 841 (10 mm.) is a little further advanced than in the embryo just described. The wax reconstruction of the posterior end of this embryo (fig. 1)shows that the v. subintestinalis is a single continuous channel. It arises from the peri-anal ring just anterior to the cloaca and ventral to the hind gut. The subcardinal veins swing around the hind gut to join the peri-anal ring laterally. Fig. 1 Reconstruction of loggerhead-turtle embryo, no. 841, Hun. E. C., 10 mm., 11 days. X 16.6. Two allantoic veins from the allantois join the peri-anal ring just anterior to the latter. As the subintestinal vein is traced forward along the ventral surface of the gut, it tends to lie t o the left. There is a short right subintestinal vein near the yolk sac which the reconstruction of this region revealed. This vessel is very short and is closely related to the left subintestinal vein by numerous cross branches. Apparently, this vessel is being rapidly displaced by the left subintestinal vein. VAS SUBINTESTINALE I N THE VERTEBRATES 379 The postcardinal veins are present ; but the subintestinal vein is not related to these veins as in Aves, but are related to the subcardinal veins. 2. Crocodilia. The v. subintestinalis is present in an embryo of Alligator mississippiensis, no. 1504, H. E. C. (4 mm.). This embryo is of interest in that it is the youngest reptilian embryo examined. A subintestinal vein (fig. 2 ) arises from each of the two allantoic veins at the junction of the allantois with the hind gut. They run forward ventral to the gut and soon coalesce to form a single vein which, upon reaching the yolk, joins a tributary of the left omphalomesenteric vein. Fig. 2 Reconstruction of alligator embryo, no. 1504, H. E. C., 4 mm. X 16.6. The postcardinal veins a r e present, but are difficult t o see in some of the sections. They extend posteriorly from the sinus venosus as very thin thread-like vessels and are not related in any way to the subintestinal venous system. No trace of the subcardinal veins could be found. 3. Squamnta. Ophidia. Two embryos of the snake, Eutaenia radix (H. E. C.) were examined, and both have a subintestinal vein. They are no. 1348 (5.8 mm.) and no. 1361 (6 mm.). The reconstruction of no, 1348 (5.8 mm.) shows a subintestinal vein ventral to the hind gut. A large caudal vein is present, but it is collapsed at the point where it should join 380 PAULINE KIMBALL the subintestinal vein. The subintestinal vein is easily followed from the cloaca posteriorly to the yolk anteriorly, where it opens out upon the yolk (fig. 9, A ) . The postcardinal veins are present anteriorly, but have not yet reached the peri-anal ring posteriorly. Lacertilia. Four specimens of Lacertilia (H. E. C.) belonging to two different genera, Lacerta muralis and Iguana tuberculata, were found to have a subintestinal vein. The two embryos of Lacerta muralis are no. 827 (6.4 mm.) and no. 815 (6.4 mm.) ; the two embryos of Iguana tuberculata were no. 1540 (6.5 mm.) and no. 1532 (6.1 mm.). In no. 815 (6.4 mm.) the peri-anal venous ring is formed posteriorly by the caudal vein, laterally by the postcardinals and subcardinals, and anteriorly by the subintestinal vein which runs forward ventral to the hind gut toward the yolk. In no. 827 (6.4 mm.) a similar arrangement of the vessels occurs. The actual termination of the subintestinal vein in the omphalomesenteric veins is a little more clearly shown. I n no. 1532 (6.1 mm.) the posterior relation of the subintestinal vein could not be determined accurately, owing to the collapse of the vessel; it terminates in the conjoined omphalomesenteric veins. The subintestinal vein and its connections in no. 1540 (6.5 mm.) resembles that of no. 1532 (6.1 mm.). B. Aves Popoff ( '94), through injection, determined the nature of the blood vessels on the area vasculosa of the chick. The area vasculosa resembles that of the reptiles. Both the sinus terminalis (rand Sinus) and the internal ring veins (innere Randvene) are venous. Two yolk-gut veins (vv. vitellinae laterales) derived from the internal ring veins collect the blood from tbe vascular network at the sides and above the head. Two yolk arteries (aa. vitellinae laterales) connect the aorta with the yolk. The left yolk vein appears to be the larger of the two yolk veins, for, in addition t o draining the left half of the area va'sculosa, it received two large VAS SUBINTESTINALE I N T E E VERTEBRATES 381 branches, the anterior vitelline vein from the cephalic portion of the vascular net and posterior vitelline vein from the caudal portion of the vascular net. Aves, in addition to having a well-developed yolk sac, have a large allantois with a well-developed circulation. Hochstetter ('88) was the first to notice a subintestinal connection between the allantoic and the omphalomesenteric venous circulations. Lillie ( 'OS), Evans ( '09), and Sabin ('17) have described this vessel. According to Sabin, the subintestinal vein appears on the third day of incubation. It develops out of an extensive plexus which surrounds the gut. At first there are two veins, a right and a left, but these soon coalesce to form a single vein. The subintestinal vein a t this stage receives the most caudal branches of the posterior limb bud (Evans) and drains the entire posterior end of the gut, the allantois, the rectum, and the intestine up to the yolk sac (Sabin). As to the fate of the subintestinal vein in the chick, Sabin says that, after the umbilical circulation appears on the fourth day, the subintestinal vein ceases to be a single channel and develops into the plexus in the wall of the gut. The presence of the v. subintestinalis in four additional species of Aves-Struthio (ostrich), Sterna hirundo (tern), Anas (duck), and Bonasa (grouse)-is of interest and of value from a comparative standpoint, f o r there can now be little doubt but that the v. subintestinalis is present in the embryos of all birds. I n the stages examined of all of these forms the subintestinal vein is at the height of its development. The enormous size of the subintestinal vein in Struthio and in Anas is most striking. A comparison of the subintestinal vein in these forms with the same vessel in the chick shows it to be at least five times as great. 1. Struthiones. Reconstruction of Struthio (ostrich) no. 2236 (H. E. C., 7.2 mm.) (fig.3 ) shows the subintestinal vein to be the direct continuation of the single allantoic vein which runs along the dorsal wall of the allantoic stalk to the junction of the allantois with the hind gut. At this point the - 382 PAULINE KIMBALL vessel turns to continue ventral to the gut as the subintestinal vein. Upon reaching the yolk sac, it joins a branch of the left omphalomesenteric vein. The postcardinal veins have taken over the caudal vein and the veins of the posterior limb in this embryo. The subcardinal veins have not yet appeared. The dorsal roots of the umbilical arteries are present. 2, Gauiae. I n an embryo of Sterna hirundo (tern), no. 2167 (H. E. C., 5 mm.), there are two allantoic veins present. The reconstruction shows the subintestinal vein arising from the anastomosis of the two allantoic veins, which occurs ventral to the juncture of the cloaca with the allantoic stalk. The subintestinal vein runs forward ventral t o the hind gut from this point to the yolk sac, where it enters a branch of the left omphalomesenteric vein. The two caudal veins do not join the subintestinal vein in this embryo, f o r the postcardinal venous system has reached the peri-anal ring and has taken over the drainage of the tail and posterior limbs. VAS SUBINTESTINALE I N THE VERTEBRATES 383 The subcardinal veins appear in the region of the fore part of the hind gut, but have not yet reached as far as the perianal ring. 3. Anseres. The subintestinal vein was observed in two duck embryos (Anas) (H. E. C.), no. 2194 (95 hours) and no. 2195 (104 hours). I n the older embryo the subintestinal vein arises in the usual way-from the anastomosis of the two allantoic veins and runs forward ventral to the gut. The two caudal veins join the two allantoic veins on either side of the cloaca, thus forming the peri-anal ring from which the subintestinal veins arise. The postcardinal veins have taken over the drainage of the posterior limb. There are no subcardinal veins present in the region of the peri-anal ring. Figure 9, B, is a photomicrograph of a cross-section of this embryo. The enormous size of the subintestinal vein is of interest. It is much larger in Anas and Struthio than in the chick. The way in which it bulges from the ventral wall of the hind gut resembles the state of the a. subintestinalis in Ravn’s case, thus suggesting the possibility of separation from the gut wall. 4. Gallinae. I n an embryo of Bonasa (grouse), no. 2215 (H. E. C., 7.3 mm., 97 hours), the subintestinal vein arises from the peri-anal ring formed by the anastomosis of the two caudal veins with the two allantoic veins. It runs forward ventral to the hind gut and terminates in a branch of the left omphalomesenteric vein. The postcardinal venous system has taken over the drainage of the posterior limbs. C. Marnrnalia The area vasculosa of the mammals differs from that of the other vertebrates in that it is more highly arterial. The innere Randvene remains venous, just as it does in all of the other vertebrates, as far as is known, except elasmobranchs ; the sinus terminalis, which is venous in all of the other vertebrates, becomes arterial in the mammals. During the earlier stages of development there are numerous aa. vitellinae as in 384 PAULINE HIMBALL reptiles, but soon the number is reduced to two aa. vitellinae laterales or t o one, the single a. vitellina media. The vas subintestinale has been reported t o be a vein in Marsupialia (Senior, '25) and an artery in Rodentia (Ravn, '94) and Ungulata (Evans, '12, and Sabin, '17). An arterial rete periintestinale caudale has been described in Insectivora (Bartfeld, '25). Nothing more need be said about the work of Ravn, Evans, and Bartfeld than has already been mentioned on page 375. Further mention of Sabin's work is necessary. Sabin states that the subintestinal artery of the pig arises from ventral branches of the aorta whieh form a sheet of capillaries on either side of the alimentary canal. This vessel is apparently well developed in a pig embryo of 20 somites and can be seen in these embryos as late as 9 or 10 mm. After the caudal flexure has formed, the subintestinal artery breaks up into the capillary plexus within the wall of the gut. I n 15- to 17-mm. embryos a new longitudinal artery appears in the dorsal wall of the gut and takes over the work of the subintestinal artery. The mammals are peculiar in that the vas subintestinale may be a vein o r an artery, depending upon the connections which it established. The presence of an allantoic placenta is apparently instrumental in causing the vas subintestinale to become an artery. It is to be regretted that material is not available t o study the complete history of the vas subintestinale in all the mammalian orders. Material, however, is available for the study of its complete history in the cat. 1. Marsupialia. Senior ('25) reported the presence of the subintestinal vein in opossum embryo, no. 16174, W. I. E. C. (6.92 mm.), but did not describe the vessel or its connections. A subintestinal vein is present in that embryo and in no. 16173, W. I. E. C. (7.6 mm.). These embryos are of about the same age; the 7.6-mm. embryo is perhaps a little younger than the 6.92-mm. embryo. I n both of them the subintestinal vein is in the process of being transformed into a plexus in the wall of the gut. Posteriorly, the subintestinal vein has lost 385 VAS SUBINTESTINALE I N THE VERTEBRATES its connection with the umbilical veins and has become plexiform; anteriorly, it remains as a single vessel' which terminates in a branch of the omphalomesenteric vein. No. 16173 (7.6 mm.) is especially favorable f o r tracing the subintestinal vein to its termination in the yolk vein and for following the yolk vein to its termination in the right omphalo- lsl"larnl.rir. i".li. x Fig. 4 16.6. Reconstruction of opossum embryo, no. 16174, W.I. E.C., 6.92 mm. mesenteric vein near its junction with the left omphalomesenteric vein. No. 16174 (6.92 mm.) was reconstructed not because it was the most favorable embryo for study, but because it came into my possession before no. 16173 (7.6 mm.). Figure 4 illustrates the condition of the other arteries and veins as well as that of the subintestinal vein in the region of the hind gut. 386 PAULINE KIMBALL Examination of figure 4 shows that the postcardinal veins have joined the peri-anal ring and have taken over the drainage of the hind limbs. The subcardinal veins have appeared anteriorly in the region of the anterior part of the hind gut. The connection between the postcardinal veins and the subcardinal veins is seen in figure 4. 2. Carnivora. Abundant material was available for study of the complete history of the vas subintestinale in the eat (Hun.E. C.). Embryos ranging from 3.5 mm. to 8.5 mm. were examined. Fig. 5 Reconstruction of mole embryo, no. 990, Hun. E. C., 4 mm. x 16.6. I n no. 687 (3.5 mm.) and no. 135 (4.5 mm.), the subintestinal artery is double. I n these specimens the two subintestinal arteries arise on either side from a series of ventral roots of each umbilical artery. These arteries run forward ventral and lateral to the hind gut to the yolk sac, where they join the omphalornesenteric arteries. The youngest embryo reconstructed which contains a single subintestinal artery is no. 889 (6 mm.). The subintestinal artery (fig. 5) arises posteriorly from the peri-anal ring. The umbilical arteries arise by several ventral roots and assist in the formation of the peri-anal ring. The dorsal root of each umbilical artery is present, but very small. Several ventral-gut vessels arising from the aorta join the subintestinal artery posteriorly. VAS SUBINTESTINALE I N THE VERTEBRATES 387 The second reconstruction made was that of no. 261 (6.5 mm.). I n this embryo the subintestinal artery is no longer connected with the umbilical arteries posteriorly. The subintestinal artery a t this stage receives all of its blood from the ventral-gut vessels, several of which can be seen posteriorly. The subintestinal artery, as a single channel, could be traced to its termination in the omphalomesenteric arteries f o r at least three-fourths the length of the hind gut. Figure 9, C, is a photomicrograph of this vessel as it appears in a cross-section taken through the fore part of the hind gut. No. 752 (7 mm.) contains the largest subintestinal artery, as far as actual diameter of the vessel is concerned. The posterior connection with the umbilical circulation by means of the ventral root has entirely disappeared. This specimen appears to be in practically the same stage of development as no. 261 (6.5 mm.). The last trace of the subintestinal artery is found in no. 285 (8.5 mm.). The anterior connection of the subintestinal artery with the omphalomesenteric is still present and can be traced to the yolk sac postero-anteriorly for nineteen sections. This embryo is in the same stage of development as the opossum embryo, no. 16174 (6.92 mm.). 3. Insectivora. The three mole embryos (Hun. E. C.) examined have a pair of subintestinal arteries. They are no. 402 (3.5 mm.), no. 407 (3.5 mm.), and no. 990 (4mm.). The reconstruction of no. 990 (4 mm.) shows two subintestinal arteries arising from the peri-anal arterial ring (fig. 6). These arteries run forward parallel to each other and ventral to the hind gut, to terminate in the omphalomesenteric arteries, where they anastomose before forming a loop about the gut on their way to the aorta. The ventral roots of the umbilical arteries are very small; no dorsal roots have yet appeared. Upon examining the literature, it was found that Bartfeld ('as), in Bis account of the development of the gut arteries of Talpa europea, described a ' rete periintestinale caudale ' arterial connection between the umbilical and omphalomesenTHE AXERICAN JOURNAL O F ANATOMY, VOL. 42, NO. 2 388 PAULINE KIMBALL teric circulations. Examination of figures 6 and 7 of his paper shows a reconstruction of this ‘rete.’ Judging from these figures, the most ventral vessel of this gut plexus must be the subintestinal artery. The umbilical arteries arise from the aorta by several ventral roots (fig. 6). The subintestinal artery takes origin from these ventral roots and runs forward to the yolk (fig. 7). This arrangement of the vessels, as shown in Bartfeld’s figures, corresponds very much to the arrangement which the writer has observed in the cat embryo, no. 889, 6.5 mm. (fig. 5). Fig. 6 Reconstruction of cat embryo, no. 889, Hun. E. C., 6 mm. X 16.6. Bartfeld compares this ‘rete periintestinale caudale’ of the mole with a similar ‘rete’ which he found in pig (Bartfeld, fig. 15). Bartfeld’s bibliography contains a reference t o Evans’ work ( ’12, Keibel and Mall), but since Evans’ description of the ventral ‘arterial route’ in pig occurs in one of the footnotes, it is possible that Bartfeld did not see it. There is no reference to Sabin’s work in his bibliography. 4. Edentata. The armadillo embryos are of special interest, because the subintestinal artery has separated itself from the hind gut and lies a short distance ventral to, and connected with, the gut by a short thread-like mesentery (fig. 9, D). The embryos are no. 275, no. 356 A, and no. 356 B (T. E. C.). The two latter embryos are litter mates. VAS SUBINTESTINALE I N THE VERTEBRATES 389 The subintestinal artery in all three of these embryos is about to be transformed into a plexus in the wall of the gut. It is plexiform posteriorly, while anteriorly it persists as a single vessel which terminates in the conjoined omphalomesenteric arteries. No. 275 is a little younger than no. 356A and no. 356B. I n this embryo it is possible to trace the anterior persisting portion of the subintestinal artery f o r more than half the length of the hind gut. Figure 9, D, is a photomicrograph of a cross-section of this embryo. The thin thread-like mesentery which connects the subintestinal artery with the hind gut is readily observed. I n some sections this mesentery appears t o have been lost, but this could not be positively determined. The subintestinal artery in no. 356A and no. 356B is almost completely transformed into a plexus in the wall of the hind gut. Only a short, degenerated anterior portion terminating in the conjoined omphalomesenteric arteries remains. These embryos are older than any of the other mammalian embryos studied. The yolk sac is retracted into the abdomen ; the omphalomesenteric circulation is greatly re duced and is on the verge of disappearing (probably nonfunctional) ; the allantoic circulation is now the functioning circulation (the dorsal roots of the umbilical arteries are very large ; the ventral roots have disappeared). 5. Primates. The study of the subintestinal artery in man met with the usual difficulty-absence of suitable material. Consequently, numerous embryos had t o be examined before the proper material for making this investigation could be found. All the embryos which were likely to have a subintestinal artery from the Carnegie Institution of Washington, Cornell University Medical College, Huntington Collection of Columbia University, Harvard Medical School, and photographs of one embryo, the Evans embryo, no. 168, of the University of California Embryological Collection were examined. No. 836 (C. I. E. C., 4 mm., 30 somites) was the youngest embryo studied (fig. 7 ) . I n this embryo the a. subintestinalis 390 PAULINE KIMBALL is double f o r the posterior two-thirds of the length of the hind gut. It arises on either side from the ventral roots of each umbilical artery. The dorsal root is present, but is very small. The hind gut is very short in this embryo. The two subintestinal arteries join just before the hind gut begins to widen out to become continuous with the yolk sac. On either side and ventral to this neck there is an arterial plexus. The /--- A Fig. 7 Reconstruction of human embryo, no. 836, C. I.E. C., 4 mm. X 16.6. right plexus appears to be in the process of being formed; the left plexus is by far the most prominent, and it is through this plexus that the subintestinal artery can be traced to its termination in the left omphalomesenteric artery. No. 1075 (C. I.E. C., 6 mm.) is only a little older than no. 836 (4mm.). According to Streeter, the arm buds are more distinct, the tail longer, and the lungs are a shade further along. The vas subintestinale (fig. 8) is double for almost VAS SUBINTESTINALE I N T H E VERTEBRATES 391 the entire length of the hind gut. These vessels arise posteriorly on either side from the ventral root of the umbilical artery, run forward parallel to each other and ventral to the hind gut to the yolk sac, where they anastornose before Fig. 8 16.6. joining the omphalomesenteric arteries. The dorsal roots of the umbilical arteries are present, but are very small in comparison to the ventral roots. The Evans embryo, no. 168 (C. E. C., Japanese), is considerably older than either no. 836 or no. 1075. Several reconstructions, based on the photographs of cross-sections of 392 PAULINE KIMBALL this embryo, were made-celluloid, graphic, and a wax model of a few of the sections taken through the region of junction between the subintestinal artery and the left omphalomesenteric artery. This embryo is of interest from a comparative standpoint, for the vas subintestinale is in the same stage of its developmental history as the opossum embryo (6.92 mm.), the cat embryo (8.5 mm.), and the armadillo embryo-the phase of its disappearance. I n this embryo the subintestinal artery is plexiform in the posterior three-fourths of the hind gut and exists as a single channel anteriorly, where it joins the left omphalomesenteric artery. The ventral roots of the umbilical arteries are no longer present; the dorsal roots are quite large and are now the functional roots. CONCLUSIONS AND SUMMARY The history of the vas subintestinale is now practically complete. There can be no doubt that the v. subintestinalis and the a. subintestinalis are morphologically the same vessel. Their position is identical, only their connections differ. The presence of an allantoic placenta is apparently the factor which determines whether the vas subintestinale shall be a vein or an artery. With the reduction of the yolk-sac circulation, the allantoic circulation is necessarily increased. I n making this change the arterial system appears to outstrip the venous system in development. This results in an extensive arterial peri-anal ring, which gives rise to the a. subintestinalis. Sabin ('17) calls attention to the fact that in the gut of the pig the arteries precede the veins. Based upon the study of the a. subintestinalis in the cat, of which abundant material was available for the study of its complete history in this form, it can be said that three stages are recognizable in the developmental history of the vas subinntestinnabis. Whether with respect to vein or artery, all of the embryos studied fall in one or another of these three stages. VAS SUBINTESTINALE I N THE VERTEBRATES 393 I n the first stage the vas subintestinale consists of a pair of vessels, one on either side and ventral to the hind gut. Examples of this stage have been mentioned: v. subintestinalis-turtle (no. 842, Hun. E. C., 10 mm.) ; a. subintestinalis -cat (no. 687, 3.5 mm., and no. 135, 4.5 mm., Hun. E. C.), mole (no. 402, 3.5 mm., no. 407, 3.5 mm., and no. 990, 4 mm., Hun. E. C.), and man (no. 836, 4 mm., and no. 1975, 6 mm., C. I. E. C.). The second stage may be regarded as the stage at which the vas subintestinale is at the height of its development. In this stage the two paired vessels coalesce to form a single channel. The best examples of this stage are: v. subintestinalis-turtle (no. 841, 10 mm., Hun. E. C.), alligator (no. 1504, 4 mm., H. E. C.), snake (no. 1348, 5.8 mm., H.E. C.), ostrich (no. 2236, 7.2 mm., H.E. C.), tern (no. 2167, 5 mm., H.E. C.), duck (no. 2194 and no. 2195, H. E. C.), and grouse (no. 2215, 7.5 mm., H. E. C.); a. subintestinalis-cat (no. 889, 6 mm., Hun. E. C.). The third stage consists of the stage of transformation of the vas subintestinale. I n the adult the v. subintestinalis is retained in some forms as the vein of the spiral valve. I n Amphibia, Reptilia, Aves, and the lower Mammalia it develops into the plexus in the wall of the gut. The a. subintestinalis has a similar fate. When the vas subintestinale begins to transform, the posterior portion breaks up first ; the anterior portion is retained (v. subintestinalis as a branch of v. omphalomesenteria; a. subintestinalis as a branch of a. omphalomesenteria). Examples of this stage are : v. subintestinalis-opossum (no. 16174,6.92 mm., W. I. E. C.) ; a. subintestinalis-cat (no. 285, 8.5 mm., Hun. E. C.), armadillo (no. 275, no. 356 A, no. 356 B, T. E. C.), and man (no. 168, C. E. C.). Previous writers, notably Hochstetter ( '88, subintestinal vein, chick) and Sabin ('17, subintestinal artery, pig), have described the first two stages in the development of the vas subintestinale, i.e., stage I, double vas subintestinale ; stage 11, single subintestinale. Evidently, no previous writer has 394 PAULINE KIMBALL observed that the anterior portion of the vas subintestinale is retained f o r quite a period of time after the posterior portion has been lost. The tendency for the vas subintestinale to produce a bulge in the wall of the gut has been pointed out and can be observed Y #"#I 6 Fig. 9 Photomicrographs showing the subintestinal vein in cross-section. x 25. A, snake embryo, no. 1348, H. E. C., 5.8 mni. B, duck embryo, no. 2195, H. E. C., 104 hours. C , cat embryo, no. 261, Hun.E.C., 6.5 mm. D, armadillo embryo, no. 275, T.E.C. in the photomicrographs (fig. 9). Actual separation of this vessel from the wall of the gut has been observed in the armadillo. These observations lend weight to the suggestion made by Senior ( '25) that the anomaly reported by Kuliga ('08) was due to the disappearance of the proximal part of both umbilical arteries and the retention of the subintestinal artery VAS SUBINTESTINALE I N THE VERTEBRATES 395 by the proximal part of the omphalomesenteric artery and the anomaly reported by Hafferl ( ’19), which Senior thought was due to the retention of the a. subintestinale in toto without loss of any part of the umbilical arteries. It is suggested that in the Scincid lizards a subintestinal venous connection exists between the omphalomesenteric and umbilical venous circulations when the yolk stalk and the allantoic stalk are separate and spring separately from the intestine. Later, when the yolk stalk and the allantoic stalk fuse t o form a common yolk-allantoic stalk, the subintestinal vein is carried out in toto into this common yolkallantoic stalk, so that “as the yolk-stalk is followed outward, it is seen that a second and smaller vitelline vein enters the umbilical vein” (Harrison and Weeks, ’25). The peculiarity of the Reptilia in the precocious development of the subcardinal veins has been noted. These veins seem to develop almost synchronously with the postcardinals. The result of such an arrangement is that the subintestinal vein takes origin from these veins instead of the postcardinals, as in Marsupialia and Aves. On the following page is a tabular summary of the results of the present investigation and those of previous investigators. 396 PAULINE KIMBALL REPTILIA: Thalassochelgs (caretta) No. 826, 8.5 mm., 11days No. 827, 9 mm., 11days No. 841, 10 mm., 11days No. 842, 1 0 mm., 11days Lacerta muralis No. 827, 6.4 mm. No. 815, 6.1 mm. Iguana tubwculata No. 1540, 6.5 mm. No. 1532, 6.1 mm. Eutaenia radix No. 1348, 5.6 mm. No. 1361, 6 mm. Alligator mississippiensis No. 1504, 4 mm. AVES: Gallus domtsticus Struthio No. 2236, 7.2 mm. Anas No. 2194, 3 days, 2 1 hours No. 2193, 4 days, 8 hours Bonasa No. 221.5, 4 days, 1 hour, 7.3 mm. Sterna hirundo No. 2167, 5 mm. MAMMALIA: Didel phis virginiuna No. 16174, 6.92 mm. Nus (rat and mouse) Sus scrofa 7.5 t o 9 mm. 20 somit es to 9 or 1 0 mm. Talpa mropaea No.990, 4mm. Felis domesticus No. 687, 3.5 mm. No. 135, 4.5 mm. No. 889, 6 mm. No. 261, 6.5 mm. No. 752, 7 mm. No. 285, 8.5 mm. Tatusia No. 356 A No. 356 B No. 275 Homo sapiens No. 1075, 6 mm. No. 836, 4mm. No. 168 Collection Vessel Author H.E. C. Vein Kimball H.E. C. Vein Kimball H.E. C. Vein Kimball H.E. C. Vein Kimball €1. E. C . Vein Kimball H.E. C . Vein Vein Hochstetter ( '88) Eimball H . E . C. Vein Kimball H.E.C. Vein Kimball H.E. C. Vein Kimball W. I. E.C . Vein' Senior ( '25) Artery Ravn ( '94) Artery Artery Rete caudalis Hun. E. C . Artery Hun. E. C. Artery Evans ( '12) Sabin ( '17) Bartfeld ( '25) Kimball Rimball T. E. C. Artery Kimball C. I. E. C. Artery Kimball C. E. C. VAS SUBINTESTINALE IN THE VERTEBRATES 397 BIBLIOGRAPHY S. 1925 Zui. Entwicklungsgeschichte der Darmarterien der WirbelBARTFELD, tiere. Die Entwicklungsgeschichte der Darmarterien bei Talpa europea. Zeitschr. f. Anat. u. Entw., Bd. 75, S. 226-257. EVANS, H. M. 1909 On the development of the aorta, cardinal and umbilical veins and other blood vessels of the vertebrate embryos from capillaries. Anat. Rec., vol. 3, pp. 498-518. 1912 The development of the vascular system. Manual of Human Embryology, edited by Keibel and Mall, vol. 2, p. 657. FELrx, W. 1910 Zur Entwicklungsgeschichte der Rumpfarterien des menschlichen Embryo. Morph. Jahrb., Bd. 41, Heft 4, S. 577-614. FLYNN, T. T. 1923 On the occurrence of a true allanto-placenta of the conjoint type in an Australian lizard. Rec., Aust. Mus., vol. 14, pp. 72-77. GIACOMINI,E. 1891 Materiaux pour 1'6tude du dbveloppement dn Septs chal. cides. Arch. Ital. de Biol., T. 16, pp. 332-359. 1891 Uber die Entwickelung von Septs chalcides. Anat. Anz., Bd. 6, S. 548-551. - 1906 Sulla maniera di gestazione e sugli annessi embrionali del Gongylus ocellatus. Forsk. Mem. Accad. Bologna, ser. 6, 111. pp. 131-175. HAFFERL,A. 1919 Uber ein Fall von Persistenz der Arteria omphalomesenterica. Zeits. f. angew. Anat. u. Eonstitutionsl., Bd. 4, S. 292-297. AND WEEKS 1925 On the occurrence of placentation in the Scincid HARRISON lizard, Lygosoma entrecasteauxi. Proc. Linn. Soe. N. S. W., vol. 1, pp. 472-486. HILL, J. P. 1900 On the foetal membranes, placentation and parturition of the native cat (Dasyurus viverrines). Anat. Ana., Bd. 18, S. 364-373. 1897 The placentation of Perameles. Quart. Journ. Micr. Sci., vol. 40, no. 159, pp. 385-447. HOCHSTETTER, F. 1888 Beitrage zur Entwicklungsgeschichte des Venensystems der Amnioten. Morph. Jahrb., Bd. 13, S. 575-586. 1893 Entwicklungsgeschichte des Gefasssystems. Ergebnisse d. Anat. und Entw., Bd. 3, S. 460-489. 1906 Die Entwiekelung des Blutgefasssystems. Handbuch d. vergl. u. exp. Entw.-lehre d. Wirbeltiere, Bd. 3, pl. 2, S. 111-149. EULIGA, P. 1905 Uber Sirenenmissbildung und ihre Genese. Monatssch. f. Geburtsh. u. Gynak., Bd. 27, 8. 297-321. LILLIE,F. R. 1908 The development of the chick, p. 367. PATTEN, B. M. 1925 Embryology of the chick, p. 367. POPOFF, D. 1894 Die Dottersack-Gefasse des Huhnes. Wiesbaden. RAVN,E. 1894 Uber die Arteria omphalo-mesenterica der Ratten und Mause. Anat.'Anz., Bd. 9, S. 420-424. RAFFAELE, F. 1888 Le uova galleggianti e le larve die Teleostei nel golfo di Napoli. Mitth. a. d. Zoolog. Stat. zu Neapel, Bd. 8, 5. 1-85. RUCKERT, J., UND MOLLIER,6. 1906 Die erste Entstehung der Gefasse und des Blutes bei Wirbeltieren. Handb. d. vergl. u. exp. Entw.-lehre d. Wirbeltiere, herausg. von 0. Hertwig, Bd. 1, T. 2, S. 1019-1273. 398 PAULINE KIMBALL SABIN, F. R. 1917 Origin and development of the primitive vessels of the chick and of the pig. Contrib. t o Embryol., Carnegie Inst. Washington, vol. 6, no. 18, pp. 61-125. SEMON,R. 1894 Die Embryonalhiillen der Monotremen und Marsupialier. Zool. Forschungsreisen in Australien u.s.w., Bd. 2, S. 19-58. SENIOR,H. D. 1909 The development of the heart in shad (Alosa sapadissima, Wilson). Am. Jour. Anat., vol. 9, no. 2, pp. 211-262. 1925 An interpretation of the recorded arterial anomalies of the human pelvis and thigh. Am. Jour. Anat., vol. 36, no. 1, pp. 1-47. STRAHL, H. 1382 BeitrLge zur Entwickelung von Lacerta agilis. Arch. f. Anat. u. physiol. Anat., S. 242-279. 1883 Beitrage zur Entwickelung der Reptilien. Ebenda, S. 1-40. TANDLER,J. 1903 Zur Entwickelungsgeschichte der menschlichen Darmarterien. Anat. Hefte, Bd. 23, S. 187-209. WEEKS, H. C. 1927 A note on reproductive phenomena in some lizards. Proc. Linn. Soc. N. S. W., vol. 52, pt. 2, pp. 25-33, no. 211.