Talk:Paper - development of mammalian ova and placenta formation in mammals 3

From Embryology


Lecture III. The Placenta of the Hedgehog.

IN the hedgehog, the small ovum enters a crypt in the mucous membrane on the antimesometrial side of the uterus, there it becomes a didermic blastocyst in the manner already described, and the maternal epithelium of the walls of the crypt disappears. The trophoblast forms a thick syncytial layer which fuses with the decidua and, after the fusion has occurred, spaces appear in the trophoblast, practically in the same manner that they appear in the cone of external trophoblast in the mouse and rat. Whilst the thickening and vacuolation of the trophoblast is proceeding the capillaries of the decidua, in its immediate neighbourhood, enlarge, and eventually they open into the spaces in the trophoblast. In this manner maternal blood begins to circulate in fœtal tissues, and it continues 80 to circulate to the end of the gestation. At first the placenta is chiefiy an omphalic placenta ; but, as development proceeds, the mesoderm extends between the entoderm and trophoblast and cleaves into two layers. Thus the yolk-sac is gradually separated from, whilst, simultaneously, the allantois attains connection with the inner surface of the chorion, and the placenta becomes an allantoic placenta.

The completed allantoic placenta is of discoid form, and it lies on the antimesometrial side of the uterus. The opposite or omphalic portion of the ovum, covered by decidua, projects into the uterine cavity and converts it into a narrow cleft.

In the course of the further development processes of fœtal mesoderm pass into the trophoblast and penetrate between the spaces containing maternal blood. From this period onwards the only structures which intervene between the fœtal and maternal blood-streams are the fœtal trophoblast and the mesodermal walls of the fœtal vessels.

The form of placentation met with in the hedgehog is not entirely characteristic of the placentæ of the other insectivora which have been investigated, at all events in the early stages ; for in Tupaja javanica, the common shrew (Sorex vulgaris), and in the mole (Talpa europea), the blastodermic vesicle does not enter a crypt in the mucous membrane but becomes attached to one or more ridges of decidual tissue. These ridges in the first instance are covered by 486 DR ARTHUR ROBINSON.

uterine epithelium. The uterine epithelium persists and is converted into a syncytium which covers the outer surfaces of the fœtal villi according to Strähl ; but in Sorex and Tupaja, according to Hubrecht, the maternal epithelium is replaced by the fœtal trophoblast, and Vernhout asserts, in opposition to Strähl, that in the mole also the trophoblast replaces the maternal epithelium, in which case the final result in all the insectivora which have been examined is practically the same ; for in Erinaceus, Sorex, and Tupaja spaces appear in the proliferating trophoblast which gain communications with fœtal vessels, and thereafter serve as channels for the maternal blood. Villi of the fœtal mesoderm project between these blood-filled spaces of the trophoblast, and the tissues which intervene between the maternal and the fœtal blood are entirely fœtal, trophoblast, and mesoderm. In the mole fœtal trophoblast covered villi penetrate the decidua and surround the maternal capillaries. The latter lose their endothelium, and thenceforth the maternal blood circulates in the fœtal trophoblast. The only important difference which exists between the placenta of the mole and the placenta of the other insectivora mentioned, is the presence, in the placenta of the former, of numerous dilated glands whose widely open mouths lie on the surface of the placenta immediately beneath the chorionic epithelium. The glands contain a secretion which coagulates by the action of the reagents used for the preservation of the specimen, and then presents a granular appearance, and it is intermingled with fragments of broken- down cells, and, in some cases, also with extravasations of maternal blood. The portions of the degenerated cells and fragments of the blood extravasations are absorbed into the interior of the chorionic cells, and are no doubt used as pabulum. In the mole also an omphalic placenta of simple form persists throughout the whole gestation, whilst in the other insectivora the omphalic placenta is entirely replaced by the allantoic in the later stages.

A very peculiar and interesting form of placentation has been discovered in Centetes by Strähl. In this animal a huge blood extravasation occurs in the middle of the allantoic placenta, and entirely destroys its central portion, leaving à peripheral ring, The margin of the latter is surrounded by a deep groove, and this is bounded externally by a ridge of epithelial cells. Outlying branches of the allantoic vessels leave the placenta, cross the groove, and plunge into the ridge of cells. Strähl is inclined to look upon this form of placentation as comparable, so far as the chief placenta is concerned, with the typical mammalian placentæ, and he compares the epithelial ridge with the placenta of the pig.

The Placentæ of Bats.

The placentæ of bats have been examined chiefly by Frommel, Van Beneden, Duval, and Gühre, and their observations show, in spite of some differences of opinion in regard to special points, that a general uniformity is to be traced in the formation of the placentæ of those cheiroptera which have been investigated, and this general uniformity is in no way seriously marred by the peculiarities which appear in the pteropidæ.

In Hippocrepis vulgaris, Vespertilio murinus, Esperugo noctula, and Rhinolophus ferrum-equinum, the blastoderm attaches itself to the mesometrial side of the uterine chamber, the thickened tropho- blast over and around the inner cell mass destroying and replacing the maternal epithelium of the uterine surface. In the region of attachment the inner portions of the glands are obliterated, but their outer parts remain and form a glandular layer in the outer part of the decidua. The interglandular tissue proliferates, its capillary vessels become dilated and their epithelium hypertrophies. Even in the early stages many of the maternal capillaries lie in close relation with the outer surface of the trophoblast, which in the meantime has differentiated into an inner cellular layer, for which Van Beneden has proposed the term cytoblast, and outer syncytial layer, which the same observer has called the plasmodiblast (PL LIV. fig. 1). In the course of its extension the plasmodiblast encloses the most internal maternal capillaries, and at the same time processes of the cytoblast, containing cores of vascular fœtal mesoderm, project outwards into the plasmodiblast, repeating, almost step by step, certain stages in the formation of the allantoic placenta of the rat. The villous processes of the cytoblast penetrate between the portions of the plasmodiblast which have surrounded the maternal capillaries, and, as the endothelial walls of the latter have now disappeared, the only tissues intervening between the fœtal and the maternal blood- streams are fœtal tissues, plasmodiblast, cytoblast, and the mesodermal walls of the fœtal vessels.

The placental formation in Pteropus edulis, which has been described by Dr Gühre, differs in some interesting details from that met with in other Cheiroptera. The ovum becomes attached to a large mushroom-shaped outgrowth from the uterine wall, which consists both of glandular and interglandular tissue, and eventually this extends completely, or almost completely, round the outer surface of the chorion of the ovum. The trophoblast of the latter proliferates and forms à syncytium which replaces the uterine epithelium, and the inner parts of the glands are obliterated. In the later stages the outer surface of the bell-shaped decidual mass fuses with the adjacent part of the uterine wall, and à disc of placental tissue results, which differs little, if at all, from the placentæ of other bats.

The Placentæ of Monkeys.

Itis well known from the investigations of Sir William Turner and Prof. Waldeyer that the structure of the placenta in monkeys and the higher apes is generally similar to that of the human subject ; and the observations of Strähl on Selenka’s specimens have shown that the well-formed placenta of the orang differs less from the human placenta than it does from the placenta of the gibbon. The differences which are discernible between the placentæ of the anthro- poids and the human placentæ are mainly differences of the decidual 483 DR ARTHUR ROBINSON.

tissues and do not affect to any extent the fœtal components, if we except mere deviations in the form and size of the villi. In both the old- and new-world monkeys no decidua capsularis surrounds the ovum as it lies in the uterus. Therefore a portion of the surface of the chorion in these animals is in contact with the fluids which lie in the uterine cavity, whilst in anthropoids and the human subject a decidua capsularis is present, and, in both the latter, the changes which affect that portion of the uterine mucosa which is not at first in direct relation with the ovum-—that is, the portion which is known as the decidua vera—are more marked than they are in the monkeys ; and in the higher forms they are least marked in the gibbon, and most marked in the human subject. In the new-world monkeys only a single discoid placenta is developed as à rule, but in the old world monkeys generally possess two discoid placentæ whose vessels are continuous with each other, a primary placenta on the dorsal wall of the uterus and a smaller secondary placenta on the opposite wall. Therefore in old-world monkeys two main groups of villi are formed on the opposite poles of the chorion. In the anthropoids and in the human subject only one discoid placenta is formed, but the ovum is enveloped in a decidua capsularis, into which villi project from all parts of the chorion—at least, in the earlier stages of the development. In the gibbon, however, there are two main groups of villi which are situated at opposite poles of the ovum, recalling, and possibly representing, the villi of the primary and secondary placentæ of the old-world monkeys; and the intermediate or tertiary villi between the two main groups are of smaller size than the others. In the human subject this division of the villi into three groups is not very evident, even if it can be distinguished at all. In no case, however, can the structure of the adult placenta give any very clear indication of the formation of the organ, and indeed most of the errors of description of placental structure which have appeared in the past are due to the fact that the earliest stages of development had not been studied. It is unfortunate that even to-day much information is still wanting concerning these early stages in monkeys, apes, and man, for the material is difhicult and expensive to obtain. We are indebted for what we know of the earlier stages of placental formation in monkeys and the anthropoid apes almost entirely to the late Prof. Selenka, whose magnificent work has excited the admiration of all those who are acquainted with it, not only on account of the rare material with which he has dealt, but more on account of the care and skill with which he has carried out his investigations, and the splendid figures with which he has illustrated his descriptions.

Unfortunately, even Professor Selenka’s observations do not provide us with a knowledge of those earliest stages in the placental formation of monkeys which are so desirable before positive conclusions can be formed with regard to the actual origin of some of the most interesting parts ; for even in the youngest specimens of his collection the embryonic area of the ovum is differentiated and covered in by the amnion, and a small closed yolk-sac is attached to the ventral MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 489

surface of the area, whilst the mesoderm has extended not only over the outer surface of the amnion, but also over the whole inner surface of the chorion. The embryonic area, with its attached amnion and yolk-sac, is connected with the inner surface of the chorion by a short and comparatively thick stalk of mesoderm, along which the allantoic vessels afterwards pass to and from the embryo ; consequently it represents the allantoic mesoderm of other forms, but at this period there is no diverticulum of entoderm in its interior.

In Selenka’s youngest specimen of Semnopithecus nasicus the ovum is attached to one surface of the uterus by a number of large villous processes which contain cores of mesoderm at their bases (P1. LIV. fig. 2). Outside the mesodermal cores, and projecting well beyond them, is a layer of ectodermal cells, which is considerably thickened over the apices of the cores, and the cellular ectoderm is covered externally by a syncytium with regard to the origin of which there are no available data. At the apices of the villi the syncytium is blended with the maternal decidua, and in the outer part of the syncytium, and between the villi, there are spaces which are in direct communication with the enlarged capillaries in the decidua. The non-villous portion of the chorion is covered by the layer of cellular ectoderm, but is entirely devoid of syncytium.

In the decidua in the immediate neighbourhood of the most peripheral villi of the chorion there are nests of epithelial cells which appear to be derived from the uterine epithelium. The uterine glands in the placental area are closed and dilated, whilst those at its margins are also dilated ; but they maintain their connection with the uterine cavity, some of them opening close to the most peripheral villi.

In another specimen the rudiment of the fœtal part of the secondary placenta is present as a thickening of trophoblast on the ventral part of the ovum.

In Cercocebus, at a somewhat older stage of development, practically the same conditions are present, both in the primary and the secondary placenta ; therefore it seems clear that although in the earlier stage of Semnopithecus no syncytium was developed on the pole of the ovum where the secondary placenta has formed in Cercocebus, nevertheless such a syncytium does appear when that pole comes into contact with the uterine mucosa ; this fact, however, gives no indication as to the origin of the syncytium, which is a point of importance and dispute.

In the gibbon, Hylobates concolor, Selenka’s youngest specimen is one in which the ovum is completely enclosed in the decidua, but there is no evidence as to how this enclosure has come about. The outer surface of the ovum is separated from the inner surface of the decidua by the number of intercommunicating spaces —the intervillous spaces ; and these, in Selenka’s specimen, are said to be filled with lympb (PI. LIV. fig. 3).

Projecting from the surface of the chorion is à comparatively small number of branched villi, and the apices of the longest of these are


attached to the outer wall of the intervillous spaces, whilst the others hang free in the lymph. The outer surfaces of the villi are covered with a syncytium, and by means of this the apices of the longer villi are connected with the decidua (PI LIV. fig. 4); but it is not clear, from the specimens, what the nature of the tissue is that forms the outer walls of the intervillous spaces, except that it is not maternal uterine epithelium. In older specimens the syncytium disappears from the apices of the villi which are attached to the outer walls of the intervillous spaces, and both the outer portions of the villi and the outer walls of the spaces are covered by a layer of fibrin.

From Strähl's tigures of the orang’s placenta, the exact nature of the immediate walls of the intervillous spaces is not evident; but apparently they do not differ in any essential respect from those met with in the human placenta.

The Human Placenta.

The youngest specimen of human placental development is that described by Peters. The whole ovum, which is calculated as being about à week old, is embedded in the maternal decidua, which, at the point of embedment, projects slightly into the uterine cavity from its posterior wall. On the apex of this projection of the decidua there is a small break in the continuity of the uterine epithelium, and over this area à small flattened mass of fibrin-like material is deposited (PI. LIV. fig. 5). These appearances have led to the conclusion that immediately after entering the uterus the fertilised ovum penetrates the uterine mucosa, destroying the surface cells and embedding itself in the subepithelial tissues in à manner very similar to that which has been traced by Spee in the guinea pig. As soon as it is embedded, the ovum commences to expand, forcing the superficial decidua towards the uterine cavity and thus forming for itself the decidua capsularis or reflexa, whilst that portion of. the decidua which lies between the ovum and the uterine muscle becomes the decidua serotina or basalis ; the remainder of the uterine mucous membrane forming the decidua vera. In the portion of the decidua which lies internal to the ovum there are no glands, therefore glands are absent also in the decidua reflexa as it is projected into the cavity of the uterus. In the deeper parts of the decidua basalis the remains of dilated glands are present, but they have entirely lost their communications with the cavity of the uterus. ‘l'he glands of the decidua vera are elongated and tortuous ; they are dilated some distance from their outer extremities, forming the spongy layer of the decidua, and they open by funnel-shaped mouths on the surface of the mucous membrane.

The most interésting features, however, are those of the ovum itself. It consists of an outer wall of trophoblast lined internally by mesoderm, and to this, on the side opposed. to the patch of fibrin which closes the orifice in the maternal epithelium, the rudiment of the embryo is attached. This rudiment is a plate of ectoderm covered by a small and completely closed amnion. It is connected, ventrally, MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 491

with a small closed yolk-sac, and all these structures are embedded in mesoderm, which, continuous with the mesodermal lining of the chorion by a strand of mesodermal cells, represents the allantoic mesoderm.

From the whole of the outer surface of the ovum villous processes of the trophoblast project into the surrounding decidua, and in the trophoblast there are spaces containing maternal blood. The apices of the trophoblastic processes embrace maternal capillaries which are in direct continuity with the spaces in the trophoblastic syncytium, and through these maternal blood enters the tropho- blastic spaces. Some of the villous processes of the chorion contain cores of fœtal mesoderm, but others are merely masses of trophoblastic ectodermal cells. The conditions, therefore, are very like to those met with at a similar stage in the development of the hedgehog.

At the end of gestation the conditions present are very different from those which exist in the early stages. The decidua reflexa is reduced to à thin membrane which has fused with the atrophied decidua vera, and the trophoblast covered villi which were present in it, in the earlier stages, have disappeared.

In the region of the decidua basalis à disc-shaped placenta has formed, which is connected on the one hand with the allantoic vessels of the fœtus, and on the other with the maternal arteries and veins of the uterus. The inner or fœtal surface of this placental mass is covered with amnion and is relatively smooth, but the outer surface, which is continuous with the uterine wall, is rough and lobulated. The sulci between the lobules are situated at the bases of septa which run inwards towards the the fœtal surface of the placenta, some only for short distances, before they terminate in pointed and irregular processes, and others as far as the surface of the chorion, to which they become attached. In the early months of development many of these septa contain maternal arteries which open on the surfaces of the septa. Between the septa are large intervillous spaces, bounded on the fœtal side by the chorion and on the maternal by a tissue the nature of which is not certain. These spaces inter-communicate, but are partially separated from each other by the septa previously described. The fœtal villi spring from the surface of the chorion and branch repeatedly in the intervillous spaces which are filled with maternal blood ; therefore the walls of the spaces, both where they are formed by the chorion and its villi and where they are formed by the septa and the maternal decidua, are bathed by circulating maternal blood, which enters and leaves the inter-villous spaces through the outer or maternal walls of the spaces or through the septa.

From the points where the vessels enter the spaces their endothelium can be traced for a short distance on the walls of the space, and then it entirely disappears. The fœtal boundaries of the spaces, where they are formed either by the surface of the chorion or the villi, are covered by a thin layer of syncytium, in which cell- outlines are occasionally visible ; but at the apices of the villi, which 492 DR ARTHUR ROBINSON.

are attached to the uterine surface of the spaces, the syncytial layer disappears. The uterine walls of the spaces, except where the endothelium of the vessels is prolonged over them, are formed by large cells called decidual cells, or they are covered by a layer of fibrin which may also be deposited between the cells, and a similar layer of fibrin is described as covering the surfaces of the villi in some cases. Our knowledge of the stages which intervene between the youngest and the oldest specimens of placental formation in the human subject is still incomplete, but we know that in the earlier months of their development the mesoderm of the fœtal villi is covered by two layers of tissue. Immediately next the mesoderm is a layer of cells, Langhans’ layer, and this is covered externally by a syncytial layer of nucleated protoplasm. With regard to the nature of these two layers there has been much difference of opinion. Indeed, some observers deny the presence of two layers, and others assert the presence of three layers. Amongst those who investigated and only recognised one layer, are Sir William Turner and Prof. Ercolani ; the former was inclined to the belief that the layer was maternal uterine epithelium, and the latter that it was formed by maternal decidual cells.

Of those observers who recognise two layers, some believe that both are fœtal and of ectodermal origin, or that one is ectodermal and one mesodermal. Others assert that both are maternal, the outer being formed by decidual cells and the inner by the endothelium of the maternal vessels. Still others believe one layer to be fœtal and the other maternal, and these are all agreed that the inner is fœtal and ectodermal, but they differ amongst themselves as to the nature of the outer layer, some asserting that it is the endothelium of the maternal vessels, and others that it is maternal glandular epithelium.

When three layers are recognised, the two outer are said to be fœtal and the inner maternal endothelium ; and one observer alone who recognises three layers ascribes to them all an origin from maternal decidual cells.

The differences of opinion, therefore, are not only numerous but important ; and the only conclusion that can be drawn from them is that no satisfactory solution of the question is likely to be derived from an examination of the placenta in any of the later stages of its development, and that no absolutely conclusive opinion can be formed until more young stages in good preservation have been examined. Still, perhaps, something like an approximation to the truth may be obtained by a comparison of the early with the later stages, and a comparison of both with the conditions which are known to exist in other mammals. It is to be regretted that our knowledge of those animals most closely allied to man is too incomplete to give any great help in this direction, if they alone are considered ; but it is not necessary that we should look to them alone, or even to them at all, for one of the striking features in association with placental formation is the difference which often exists between closely allied forms, and the similarities which can be traced between the processes MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 493

which occur in mammals of widely divergent groups. Looked at from this point of view, and comparing the phenomena, as we know them in the rat and mouse, the hedgehog, bats, and some of the apes, with those which have been described in association with the formation of the human placenta, there can be but little doubt that the syncytium and the cell layer of the human villi in the early months, and the single layer of cells or syncytium in the later months, are both of fœtal ectodermal origin—that is, they are both trophoblastic ectoderm. The fibrin lining the spaces is obviously adventitious and of secondary nature ; but how are we to account for the covering of the uterine walls of the intervillous spaces? The weight of the evidence seems to prove that the endothelium of the maternal vessels terminates à short distance from their openings into the spaces, and that the remainder of the uterine walls of the spaces are formed by decidual tissues. It is obvious, by comparison with the rat, that such a condition is quite possible, for in that animal the outer surface of the coue of external trophoblast is only connected at irregular intervals with the surface of the decidua from which the maternal epithelium has disappeared, and between the points of connection there are spaces between the fœtal and maternal tissues which are either empty or which contain maternal blood. If we imagine that the portions of the trophoblast between these spaces are enlongated into villi, a condition essentially similar to that figured by Selenka in Hylobates would be produced. The phenomena presented by Peters youngest human ovum tend, however, towards a different explanation, for those features indicate conditions much more closely allied to the conditions met with in the early stages of the formation of the placenta of the hedgehog, where spaces appear in the rapidly proliferating trophoblast which are in direct continuity with the cavities of the maternal vessels. An exactly similar condition is met with, but in a more limited area, in the case of the bat’s ovum, and in a still more limited area in the trophoblastic cone of the rat. In the rat, after a time, villi, consisting of the cellular internal trophoblast with cores of mesoderm, are projected into the syncytial cone of the trophoblast, and the cells of the internal trophoblast soon blend with the syncytium of the trophoblastic cone ; obviously, the cell layer of these villi is comparable with the cytoblastic layer of the trophoblast of bat’s ovum, and if we were to imagine that the spaces in the syncytial trophoplast were to become greatly enlarged, and that the cell-covered villi, each Carrying a thin layer of the syncytial wall, were to project into the spaces, we should have a condition similar to that found in the gibbon, except that the outer walls of the intervillous spaces would consist of fœtal trophoblastic ectoderm, and it is exactly with regard to the structure of these outer walls of the intervillous spaces both in the human placeuta and the placentæ of apes, that we are in doubt. The matter, however, cannot yet be settled, and only careful observation of other stages of development will serve to show which of the two explanations I have suggested is the more correct, or if another explanation of entirely different character is to be accepted. 494 DR ARTHUR ROBINSON.

The Placenta in Marsupials.

The fœtuses of the Marsupials form the main parts of all their important organs and systems during the period of their intrauterine life ; and although the ova of those forms which have been examined contain a large amount of yolk granules, yet the food so stored, though it may suffice for the very earliest developmental purposes, is not sufficient for the nutrition of the developing embryo during the whole period of gestation, and placental structures appear. In the opossum, Didelphys virginiana, these placental structures are in their simplest form, for the outer surface of the omphalic portion of the ovular wall is merely folded, and the folds are embedded in corresponding sulci of the mucosa. Apparently, so far as the observations of Selenka show, there is no intimate adhesion between the trophoblastic ectodermal cells on the outer surface of the chorion and the uterine epithelium which remains intact. Still, the characters, both of the maternal and fœtal chorionic cells, which are in contact with each other, appear to indicate that they are playing an active rather than a passive part during the time of their apposition.

In Perameles and Dasyurus, Hill has described much more advanced placental formations ; but the two differ from each other in essential points, and therefore require separate consideration.


In Perameles the placenta is formed partly by the allantoic and partly by the omphalic portion of the chorion, but the strncture of both parts is essentially the same. In the placental region the maternal epithelium proliferates and forms a thick syncytium into which the maternal capillaries penetrate and form a series of dilated blood-spaces directly beneath the superficial surface of the syncytium. The trophoblastic ectoderm of the chorion unites with the inner surface of the syncytium, and for a time portions of the maternal syncytium and the fœtal ectodermal cells separate the maternal blood- spaces from the fœtal blood-vessels in the mesoderm on the under surface of the fœtal ectoderm. At a later period the fœtal ectoderm cells disappear, and the fœtal vessels are merely separated from the maternal spaces by the maternal syncytium ; indeed, in some places this disappears also, and the mesodermal walls of the fœtal vessels lie in contact with the endothelium of the maternal spaces. The uterine glands in the placental area are enlarged ; many of them are closed by the maternal syncytium, but some remain open, and Hill figures à fœtal vessel covered with mesoderm alone lying in the open mouth of a uterine gland.


In Dasyurus the placenta is entirely an omphalic placenta, and in its most important details it differs entirely from the placenta of MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 493

Perameles. In the latter the placenta is formed in the region of the chorion over which the mesoderm has extended, but in Dasyurus it is in the non-mesodermal part of the chorion that the most intimate union between the chorion and the decidua occurs; though a less intimate union is also found over the vascularised portion of the yolk-sac. The most striking difference, however, between Dasyurus and Perameles is that in the latter the epithelial syncytium which encloses the maternal vessels of the placental area is formed by the proliferation and fusion of the maternal uterine epithelial cells, whilst in Dasyurus the maternal epithelium is destroyed by the fœtal ectoderm, the cells of which fuse into a syncytium which invades the decidua and surrounds the dilated maternal capillaries. Ât à later period of gestation the entoderm also forms a syncytial tissue, which blends largely with the ectodermal syncytium, but here and there between the two syncytial layers spaces appear, and into these maternal blood is poured. We have, therefore, in Dasyurus à direct relationship of maternal blood, not only to fœtal ectoderm but also to the entoderm of the yolk-sac, and every opportunity is thus afforded for the passage of nutritive material from the maternal blood into the yolk-sac, whence it can easily pass into the alimentary canal of the fœtus.


At first sight it may seem a serious omission that no reference has hitherto been made to the placentation of the lemurs, which are placed, in systematic classification, in near relationship with the monkeys. There are, however, two reasons for paying no special attention to their placentation, for, in the first place, excepting Tarsius, about whose position in the animal scale there is some doubt, the placentæ of the lemurs, so far as they have been described by Milne Edwards, Sir William Turner, and Hubrecht, are of the simple villous type met with in many of the ungulata—that is, they are formed by fœtal trophoblast covered villi which fit into crypts lined with maternal epithelium ; and the placenta of Tarsius, whatever be the proper morphological position of that animal, corresponds more closely with the placentæ of some of the insectivora than it does with the placentæ of the lemurs or of the apes.

General Considerations.

In the description of the placentæ which have been chosen as specimens of placental formation I have avoided introducing, as far as possible, any terms which would indicate an attempt to classify the placentæ into groups : but now the details of formation have been considered, it may be advisable to attempt to find some means of grouping the various forms into divisions or classes, and to apply to each division or class an appropriate name—that is, a name which shall indicate the main peculiarities of the placentæ which are grouped in that special division. Such attempts have previously been made, but the classifications hitherto adopted have become 496 DR ARTHUR ROBINSON.

more and more inadequate as our knowledge of placentæ has become more and more precise, and I cannot help thinking that even Strähl’s most recently suggested classification is not quite all that could be desired.

It is well known that at the period of birth in some animals the fœtal and maternal portions of the placenta separate from each other, the former being cast out of the uterus either simultaneously with or shortly after the fœtus, whilst the maternal decidua returns to the ordinary conditions of uterine mucous membrane, and in such cases there is no tearing or destruction of the maternal tissues and no necessary loss of maternal blood. This is the case in the pig, in the mare, and in the equidæ generally. In the other mammals a certain amount of blood is lost at the birth of the fœtus, and it was urged that in these latter cases the union of the fœtal and maternal tissues in the placenta was so intimate that some of the latter were torn off and removed together with the fœtal portions. It was upon this account that Prof. Weber, in 1835, suggested the term caducous with regard to those placentæ which were supposed at the period of birth to carry with them some parts of the maternal decidua. Eschricht and von Baer both accepted the suggestion, and classed placental mammals into the caducous and non-caducous, the former being those in which a portion of the maternal decidua was believed to be cast off at birth, and the latter those in which the fœtal and maternal tissues merely separated from each other.

In the lectures on classification which Prof. Huxley delivered, as Hunterian Professor, in 1863, he introduced the terms deciduate and non-deciduate as preferable to caducous and non-caducous ; the term non-deciduate indicating that no part of the maternal tissues was thrown off at birth, and the term deciduate indicating that parts of the maternal tissues were shed at that period. At the same time, Prof. Huxley drew attention to the fact that he did not intend to indicate that no homologue of the decidua was to be found in non-deciduate animals, for he recognised that in them also the uterine mucosa hypertrophied during the period of gestation. Year by year the supposed basis of this classification is gradually falling away, and it is becoming more and more evident how equally incorrect were some of the observations which Owen put forward in opposition to a placental classification, and those which Huxley advanced in contra- diction of Owen’s opinions. In the fifth lecture on classification Huxley quotes à paragraph from one of Owen’s papers which runs as follows : —‘The degree of resemblance in outward form between the placenta of the rat or hare on the one hand, and the mycetes and macacus on the other, seems to me to be more than counterbalanced by the difference of structure. The pedunculate and cotyloid placenta of the rat consists of fœtal parts exclusively ; the maternal areolar portion is as distinct from it as it is in the cotyledon of the ruminant, and is a persistent structure of the uterus. The discoid placenta of the monkey includes a large proportion of maternal cellular structure, which comes away with the fœtal portion. The difference in the organic interblending of the circulatory organs of mother and offspring MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 497

between the rodentia and quadrumana is of much more real importance than the degree of superficial similarity.” In opposition to this statement Huxley drew attention to what he describes as a magnified view of a section of the placenta and uterus of a pregnant rat, and he describes it in the following words :— Fig. 42 represents a section of the uterus, chorion, and partially injected placenta of a fœtal rat, one inch and a quarter long, taken in à direction per- pendicular to the long axis of the uterine cornu ; a, is the mesometrium traversed by a large uterine vein ; d, is the wall of the uterus, becoming looser in texture and traversed by large venous channels in its inner substance; «, is the decidual layer of the uterus of a cavernous structure, whence vascular processes are continued towards the chorionic surface of the placenta. A large vein (£) passes directly from the decidual layer (4), and the uterine sinuses beneath it, to near the chorionic surface of the placenta, bencath which it branches out horizontally. The chorion (/f), rendered vascular over its non-placental part by the omphalomeseraic vessels (4), only begins to exhibit villous processes and folds at the point (y). These outer-most villi appear to me to be free ; but more internally they become closely connected with the upper surface of the placenta ; and over the central face of a third of the fœtal face of the placenta, the umbilical vessels ({) ramify in à radiating fashion, and send prolongations down between the decidual lamellæ. The slightest traction exerted upon the cord causes the placenta to separate along the line, e, m, e, bringing with it, of course, the cup-shaped decidua 4.”

‘It is obvious, from the above description, that the ‘ pedunculate and cotyloid” placenta of the rat does not ‘consist of fœtal parts exclusively, but that, on the contrary, as Eschricht has well pointed out, ‘the organic interblending of the circulatory organs of mother and offspring ” is as complete in the rat as in the man.”

These quotations make it evident that both Huxley and Owen believed they saw in the human placenta an organic interblending of the circulatory organs of mother and offspring, and that Huxley contended for à similar interblending in the rat, whilst Owen asserted that the rat’s placenta consisted of fœtal parts only.

The figure on which Huxley bases his attack upon Owen’s state- ment regarding the rat’s placenta is a correct representation of the actual conditions ; but his interpretation of it is not correct, for what he describes as the chorion is the invaginated wall of the yolk-sac, and the villi upon it, which he describes as being free externally but connected with the placenta internally, have no connection with that organ. The decidual lamellæ are in reality formed by fœtal trophoblast. Indeed, there is every reason to believe that Owen’s statement that the placenta of the rat consists of fœtal parts exclusively is absolutely correct, and it is extremely probable that no maternal parts, except the blood which is circulating in the fœtal spaces, is shed at birth.

Extending our investigations to other placenta, it must be noted that Vernhout has shown, in the case of the mole, that not only are no maternal tissues shed at birth, but that some of the fœtal 498 DR ARTHUR ROBINSON.

tissues remain in the uterus and are gradually absorbed; and Hills observations on Perameles and Dasyurus prove that a similar condition exists in those animals. Indeed, the latter investigator has proposed to call such placentæ contra-deciduate, to indicate that fœtal tissues remain attached to the decidua after the termination of the gestation.

In the carnivora the only maternal tissues which are cast off with the placenta are endothelial walled capillaries between the fœtal villi and the blood which they contain. In a sense, therefore, the carnivora are more deciduate than the rodentia and the insectivora, for there is reason to believe that the conditions which occur in the rat and the mole are common to all the placentæ of both those orders. In the cheiroptera there can be but little doubt that the main part of the placenta, if not the whole, is formed by the trophoblastic syncytium, and that it is therefore entirely a fœtal structure, with the exception of the blood in the sinuses. Indeed, the structural differences between the adult placentæ of the insectivora and those of the cheiroptera are so slight that it is scarcely to be expected that at the end of gestation the method of separation will differ to any important extent ; but observations upon this point are still wanting.

When we turn to the placentæ of the monkeys, anthropoid apes, and man, we are struck with the resemblances which occur between them and the placentæ of the rodentia, insectivora, and cheiroptera ; and the resemblance is not that organic interblending which Eschricht, Owen, and Huxley believe they saw, but, on the contrary, the preponderant part played in them all by the fœtal ectoderm. Comparison of the youngest specimens of Semnopithecus and Cercocebus placentæ in Selenka’s collection with the early placentæ of rodents, insectivora, and man at once suggests the probability that the syncytial layer on the villi of the monkeys’ placentæ is fœtal trophoblastic ectoderm which is gradually surrounding either spaces which existed between the chorion and the decidua, similar in nature to the spaces in the rat and mouse between the cone of external trophoblast and the decidua, or it is surrounding dilated maternal blood-vessels, as in the cheiroptera, insectivora, and dasyurus.

This being granted, the probability that the walls of the later and more fully developed intervillous spaces of the monkeys, anthropoid apes, and man are entirely, or almost entirely, in the fœtal part of the placenta is great, and it is questionable whether or not the trabeculæ of the placenta, which have generally been believed to be of maternal tissue, are not in reality fœtal, little, if at all, dissimilar in character to the septa between the villous processes in the rodent’s placenta.

Until further stages of placental formation in the monkeys, anthropoid apes, and man are at our disposal, it is impossible to make any positive assertion as to the termination of the fœtal and the commencement of the maternal tissues in them, or as to the exact position of the line of separation at birth in relation to the MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA. 499

constituent parts of the organ. All that can be said at the present moment is that. bearing in mind the facts which have been brought to light in recent years, the statements previously made as to the deciduate nature of many placentæ require careful reconsideration ; the probability being that as some placentæ which were believed to be deciduate have been shown to be non-deciduate, and as other so- called deciduate placentæ appear to be formed on the same lines, it is very possible that the latter also are non-deciduate. I would urge, therefore, that the terms deciduate and non-deciduate be avoided, and that terms be adopted which do not suggest a process taking place at the end of gestation of which we have no definite certainty.

To avoid the difficulty which has arisen with regard to the terms deciduate and non-deciduate, Prof. Strähl has proposed the terms True or Complete Placenta and Semiplacenta or Half Placenta; the former to be applied to those placentæ in which, at the end of gestation, maternal blood-vessels or spaces containing maternal blood are opened, and the latter of those in which this opening up of maternal spaces filled with maternal blood does not occur.

These terms, however, to my mind, are not at all satisfactory, for à given portion of tissue is or is not a placenta. If it is a placenta, it must be a true placenta and at the same time a complete placenta, otherwise it could not perform its proper functions. As the terms caducous and non-caducous, deciduate and non-deciduate, are misleading, if not entirely incorrect, so far as the former term is concerned, I would propose to adopt in their place, at least tentatively, until our knowledge, which is at present in a transition stage, becomes more exact and complete, the terms Apposita and Conjuncta ; the former term to indicate that the layers are merely in apposition and can be separated without any necessary rupture or tearing, and the latter to indicate that the fœtal and maternal sections of the organ are united together or conjoined. These terms would indicate nothing more than actually exists—the close apposition of the placentæ of the ungulates, and the union of the fœtal and maternal tissues as in the majority of the mammalia.

À classification on placental basis must take into account all mammals, and at once two great groups appear—-the oviparous, and the viviparous, the former class including the egg-laying monotremes and the latter all the remaining mammals. Prof. Strähl proposes to divide the vivipara to achorial and therefore aplacental, and chorial or placental groups. The achoriata is to include those forms in which there is no firm union between the fœtal and the maternal tissues. But where is the line to be drawn between the firm and not firm union? Certainly it is suggested that in many of the marsupials there is no firm union ; but in all those which have been examined by satisfactory methods, the chorion is folded between corresponding folds of the uterine mucosa, as in the opossums described by Selenka. And although that author states that there is no firm union between the fœtal and the maternal tissues, nevertheless his figures show an interlocking almost as close as that formerly described in the pig; and it is quite certain that the chorion functions during à very 500 DR ARTHUR ROBINSON.

important period of development as à nutritive organ, therefore there does not appear to be any valid reason for retaining the term ‘achoriata.”

If this term is rejected, then all viviparous mammalia are choriate or placental, and they are separable into those with apposed placentæ and those with conjoined placentæ.

Apposed Placentæ | Conjoined Placentæ are are Avillous. Some marsupials, | Zonary (Carnivora) — and the pig. (a) Simple : Cat. Villous— (b) Compound (zone with scattered (a) Diffuse : Equidæ. villi): Elephant. (b) Cotyledonary: Sheep and Cow. | Zono-discoidal : Ferret. (c) Zonary : Halicore Dugong. Discoidal—

(a) Simple: Man, Anthropoids, Rodents, Insectivora, Cheirop- tera.

(b) Duplex : Old-world Monkeys.

(c) Perforate : Centetes.


Looking back over the points which have arisen in connection with ‘the phenomena which occur during the early stages in the development of mammalian ova and the formation of the placenta, the most important features may be summarised as follows :—

(1) At some period during the development of the ovum and before its maturation commences, one half of the chromosomes originally present in its nucleus disappear.

(2) During the development of the ovum in the ovary there is formed around it a membrane, the zona pellucida, or oolemma ; and whilst it is not quite clear whether this membrane is formed by the ovum or by the membrana granulosa, the probability appears to be that the latter is the case.

(3) The function of the zona pellucida is also uncertain ; but as it is thickest in cases of ova which develop in the cavity of the uterus, and more especially thick and long persistent in those in which the embryonic ectoderm soon assumes à superfcial position on the surface of the ovum, it is possible that its function is to prevent the contact of the trophoblast with the uterine mucosa until the ovum has attained a considerable size or a favourable situation, and to prevent the contact of the embryonic ectoderm with the decidua, either until the cells of that portion of the ectoderm have become specialised, or until circumstances have arisen which preclude any possibility of their contact with the decidua.

(4) As à general rule, during the process of maturation, the ovum undergoes two unequal divisions, but in certain cases one of these divisions may not occur, and apparently its omission has no evil effects.

(5) After fertilisation has occurred the number of chromosomes in the ovum is restored to that which originally existed. Journ. of Anat. and Physiology, July 1904.) [PLATE LIV,

Fic. 4. Fic. 8.


(6) During the segmentation which follows fertilisation qualitative changes occur in the cells; but though it is not yet quite clear whether these are merely physiological differentiations, due to the inter-relations of the cells, or whether they are due to the distribution of different portions of the idioplasm of the original ovum to different cells or groups of cells, the latter seems to be most probable.

(7) When the segmentation is completed and the blastocyst is formed, the outer layer of the ovum, in the majority and probably in all cases, consists of trophoblastic ectoderm, or trophoblastic and amniotic ectoderm, whilst the inner mass contains the rudiments of the embryonic ectoderm and the entoderm, but may also include a certain amount of trophoblast.

(8) The trophoblastic ectoderm plays a most important part in the formation of the placenta, attaining in placentæ appositæ a very close association with the cells of the uterine epithelium, and possibly with the walls of the maternal vessels, and in the placentæ conjunctæ it invades the decidua, destroying portions of it and utilising them as pabulum, and it attains a close association with the endothelium of the maternal capillaries, or with the maternal blood itself.

(9) That in the cases of perameles, dasyurus, and the mole, it is certain not only that no portions of the maternal decidua are cast off at birth, but also that portions of the. fœtal part of the placenta remain in the uterus and are gradually absorbed. In the rat and mouse it is most probable that no maternal tissues are cast off at the end of gestation. In cheiroptera, monkeys, anthropoid apes, and man it is doubtful if any maternal tissue except blood is separated from the uterus at birth, and in the carnivora the only maternal structures cast off are those portions of the endothelium of the maternal vessels which lie between the fœtal villi and their contained blood.


In figs. 1, 2, and 4 the deep black layer represents the plasmodiblast or external portion of the trophoblast. It consists of nucleated protoplasm in which cell outlines are not distinguishable. In the same figures the black or shaded layer with white dots is the cytoblast or inner layer of the trophoblast. As a rule, the cytoblast is distinctly demarcated into cell territories, not represented in the figures, but in some cases, even in this layer, the cell outlines are not visible, at ail events for a time. .

In fig. 3 no attempt has been made to differentiate the cytoblast from the plasmodiblast, but in fig. 4, which represents a portion of the chorion of fig. 3 more highly magnified, it will be seen that the plasmodiblast is practically confined to the apices of the villi which are united to the decidua, and that the plasmodiblast, therefore, constitutes the connecting medium between the fœtal and maternal tissues.

In fig. 5 the deep black layer represents the whole of the trophoblast which has not yet differentiated into plasmodiblastic and cytoblastic portions, 502 MAMMALIAN OVA AND DIFFERENTIATION OF PLACENTA.

Letters common to all figures :—

A. Amniotic cavity. IVS. Intervillous space.

Cy. Internaltrophoblast or cytoblast. 3/BS. Space containing maternal £E. Ectoderm of embryonic area. blood. EN. Entoderm. AD. Maternal decidua. ET, External trophoblast or plas- T. Trophoblast.

modiblast. UC. Uterine cavity.

FM. Fœtal mesoderm. UE. Uterine epithelium.

FP. Plug of fibrin. UG. Uterine gland.

Fig. 1. Section through the embryonic area and the adjacent part of the chorion of a bat’s ovum (after Van Beneden). The diagram shows that the plasmodiblastic portion of the trophoblast or extra-embryonic ectoderm is in contact with and has commenced to surround the dilated superficial blood-vessels of the decidua. At the stage depicted the decidual blood-vessels still retain their endothelium, but this is destroyed as the plasmodiblast surrounds the spaces, until ultimately the maternal blood circulates in spaces which are directly surrounded by the plasmodiblast alone.

Fig. 2. Diagram representing a section through a young ovum of Semnopithecus nasicus (after Selenka).

At the stage represented, the ovum is attached to one side of the uterine cavity by a series of large villous processes which are covered with plasmodiblast, and the spaces between the villi, which are lined with plasmodiblast, are in direct continuity with the cavities of superficial dilated blood-vessels of the decidua. The opposite side of the ovum, at the stage represented, is free, and is covered with cytoblast only, but at a later period it becomes attached to the wall of the uterus and enters into the formation of a secondary placenta.

Fig. 8. Diagram of a young ovum of the gibbon (after Selenka). The ovum is entirely embedded in the decidua, and the large intervillous space which inter- venes between the ovum and the decidua, except at the points of attachment of the villi, is filled with fluid, but from the specimen and the description it does not appear certain that the fluid was blood. There are numerous dilated blood- vessels in the decidua in the immediate neighbourhood of the ovum, but it is not evident that they open directly into the intervillous space, though it is extremely probable that they do so open. The ovum is covered with tropho- blastic ectoderm which is mainly cytoblastic in character, but on the apices of the villi which are attached to the decidua it is distinctly plasmodiblastic in character. There are no traces of uterine epithelium on the decidual surface of the intervillous space. The chorionic villi form two groups, a large or primary group round the neïghbourhood of the embryo and amnion, and a smaller secondary group at the opposite pole of the ovum. The former represent the primary, and the latter the secondary placenta of Semnopithecus.

Fig. 4 Diagram of a more highly magnified portion of a gibbon’s placenta, showing the dilated blood-vessels, the remains of the uterine glands in the decidua, and the cytoblastie and plasmodiblastic portions of the chorionice villi. Some of the maternal blood-vessels are shown opening into the intervillous spaces.

Pie. 5. Diagram of a section of a young human ovum (after Peters). The ovum, which is not more than a week old, is entirely embedded in the decidua. It probably passed into the decidua through an opening in the uterine epithelium, which is plugged by a mass of fibrin. The decidual tissues around the ovum contain many blood-filled spaces, some of which are apparently devoid of any endothelial lining. The surface of the ovum is covered by a thick syneytial layer of ectoderm, which is not yet separated into plasmodiblastic and cytoblastic portions ; in other words, the cytoblastic portion is not yet differentiated. Numerous and irregularly formed villous processes are projected from the surface of the ovum into the decidua. Some of the processes consist of trophoblast alone, and others contain cores of vascular fœtal mesoderm. Both groups are gradually surrounding the immediately adjacent blood-filled spaces of the decidua.