On August 31st, 1902, an elephant belonging to Messrs. Sanger and Sons gave birth to a calf in the gardens of the Zoological Society of London under circumstances related by the Society’s Prosector, Mr. F. E. Beddard, F.R.S., in the ‘Proceedings’ of the year 1902, vol. ii.

The birth of an elephant in this country is so rare an event that only two former occasions have been recorded ; and at no previous time has an elephant been born in the gardens of the Zoological Society. The opportunities of investigating the minute structure of the placenta have therefore occurred but seldom ; in fact, the only accounts we have of the microscopic features of the elephant’s placenta are those by Sir William Turner in 1876, and by Dr. H. C. Chapman in 1880.

The latter author, writing in 1899 (Chapman, 4), referring to the specimen described by him in 1880, remarks that it is the sole specimen in existence of an elephant’s placenta at full term; that of Owen’s being “d’un fœtus de six mois environ d’après mon examen …. Je puis dire que l’occasion d’obtenir un plácente d’éléphant ne s’est pas encore présentée, même en Orient, ni dans l’Inde ni dans le Siam ; j’ai reçu des lettres des hommes qui gardent les éléphants royaux qui affirment que la naissance d’un jeune éléphant est tout à fait inconnue.” Though this may be an exaggeration of its rarity, still it is clear that the difficulty of obtaining a placenta is so great that we ought to take full advantage of the present opportunity, especially as the records of previous observations are meagre, and do not afford us any information upon matters which are now of special interest, on account of the advance which has been made in the knowledge of the minute anatomy, and of the development of the mammalian placenta during the last twenty years.

Through the kindness of Mr. Beddard some pieces of the fœtal placenta of the elephant born in the Zoological Gardens came into our hands, and we herewith record the results of our investigations and the interpretation we place upon the structures observed. There can be little doubt that the calf, although it never lived after birth (Beddard, 2), was born at full term. The actual date of impregnation was not known, but it was believed to have been in April, 1900. Accordingly the birth of the calf was expected in February, 1902. The event, however, did not occur until more than six months later. The histological characters of the placenta, as explained below, also point to the placenta as being fully formed.

Previous Observations

The fœtal membranes and placenta of the elephant were described by Owen (13) in 1857, who obtained from Dr. Morton in India the membranes of a fœtal elephant. These membranes, which had been preserved in “a keg of arrack,” were supposed to have been of the period “at about the middle of the period of utero-gestation.”

Owen describes the general features of the zonary nature of the placenta, and alludes to the presence of two other villous patches covering the ends of the oblong sac or chorion, which he regarded as being also areas of attachment to the uterus. “The chief points of attachment of the chorion to the uterus are, at the equator, by the annular placenta, and at each pole of the elongated sac by the subcircular villous patches.” Owen seems to have made no examination into the more minute anatomy of these patches of attachment.

Twenty years later, however, Sir William Turner received from Professor Flower “a slice of the zone for microscopical examination” of this same specimen, which had been preserved in the Museum of the Royal College of Surgeons. Turner described the result of his examination in his “Lectures on the Comparative Anatomy of the Placenta” on pages 101-102. He describes how he “succeeded in passing some injection into the vessels of the chorion and the large trunks in the stem of the villi,” and concluded that “there could be no doubt, however, that in this separated placenta of the elephant a large amount of uterine mucosa was inextricably locked in between the fœtal villi.”

A few years later Dr. H. C. Chapman (3) described the birth of an elephant and the nature of the placenta at full term, remarking that the chief points of difference between his description and that of Owen are to be regarded as due to the fact that his specimen was one at full term and Owen’s was one of about half the period of gestation. Chapman made some observations on the more minute structures of the placenta, and states that “the examination of the injected blood-vessels in my specimen leaves little doubt that at least one fourth of the girdle-like placenta of the elephant consists of hypertrophical mucous membrane of the uterus.”

Our specimen would appear to agree with that of Chapman with regard to the absence of any allantoic cavity.

Owen found in his specimen (about half time) a distinct allantoic cavity or series of cavities, the allantois being “so interposed between the chorion and the amnios as to prevent any part of the amnios attaining the inner surface of the placenta.”

Chapman, however, found no allantoic cavity, but from the position of the fœtal vessels with regard to the inner surface of the placenta, and to certain bodies which he identified with structures named subcircular bodies by Owen, and stated to be upon the inner surface of the allantois, he concludes that the cavity of the allantois has disappeared through fusion of its walls together and to the amnion.

Our specimen similarly exhibited a fusion of amnion with allantois and chorion in the region of the zonary belt, but elsewhere the amnion was fully separated from the endochorion. Nowhere could an allantoic cavity be found.

Chapman also noticed and figured subcircular areas of villous tissue at the poles of the chorionic sac, already described by Owen.

Both authors allude to the morphological interest of those villous patches of connection with the walls of the uterus. Owen writes : “The most important modification in the vascular structures connecting the chorion with the uterus in the elephant is their combination of two forms of the placenta, viz., the ‘annular’ and the ‘diffused,’ which forms are restricted in other Mammalia to distinct kinds of quadrupeds.”

Neither author gives any account of the more minute structure of this diffuse and probably non-deciduous part of the placenta. It is to be regretted that this portion of the fœtal membranes was not preserved on the occasion of the birth of the young elephant in the Zoological Gardens last year and that no observations were recorded upon this point (Beddard, 2). It is, however, certain that these villous patches if present must have been very inconspicuous, for although not actually looked for, they were not so evident as to attract attention.

There would seem to be some variation in the breadth of the zonary placental area. Owen speaks of its being “partially divided by opposite constrictions into two moieties, one measuring 12 inches, the other 10 inches in length, and the extreme breadth being 5 inches in each ; the connecting isthmus is 3 inches in breadth.” Chapman finds that “the placenta preserves the same average width all round; there is no constriction dividing it into ‘two moieties ‘described by Professor Owen.”

He also notes another difference, namely, that “the villous processes are as well developed at the edges of the placenta as in the middle.” In our specimen the belt was divided by two double constrictions into three chief areas, which were approximately 10 inches in width.

Like Chapman’s specimen the whole area was villous and “somewhat broken up into cotyledons as one finds in the human placenta” (Stevens, v. Beddard, 2).

Through the kindness of Professor Stewart we have been able to examine the historic placenta described by Owen and Turner which is now in the Museum of the Royal College of Surgeons.

We find the specimen exactly as described by Owen, with its three points of attachment, the zonary belt and the two villous patches, near the poles. These latter are quite obvious though the villi are short, none exceeding 2 mm. in length.

There is one point which we should like to add to Owen’s description.

On page 348 he writes : “A thin brown deciduous layer is continued from the borders of the placenta, from a distance varying from 1—3 inches, upon the outer surface of the chorion.”

Chapman says of this in his specimen : “On each side of the placenta there is an indistinct brownish granular layer four inches in width and about a line thick, which runs parallel with the whole circumference of the placenta, and in some places even overlaps it slightly. This same granular matter was found even scattered over the surface of the placenta and was easily rubbed off with the fingers.” He suspects it is of maternal origin.

This is present in Owen’s; but what we want to draw attention to is the fact that there are numerous arborescent villi along this region, the smallest of which are exactly similar to the small villi of the subcircular villous patches of the poles.

These villi have the appearance of having come out of their maternal crypts without bringing away any maternal tissue. Nor is there any trace of maternal tissue to be detected among them unless the brown granular matter is maternal ; so that this region is as diffuse and non-deciduate as the subcircular polar patches.

Histology of the After-birth of the Elephant

With reference to the macroscopic characters of the fœtal membraues we have nothing beyond the remarks made above to add to the descriptions of Owen and Chapman.

The portions which we have had for microscopical investigation are pieces of the thick zonary placental area and certain curious disc-like bodies which are found scattered over the extra-placental parts of the fœtal membranes and noticed by Owen, who spoke of them as “the subcircular bodies. “

This material had been preserved in a saturated solution of corrosive sublimate and kept subsequently in strong alcohol.

In a section taken through the whole thickness of the zonary region of the after-birth we distinguish three well-marked layers or regions which, commencing from the chorionic surface next the fœtus and representing the inner surface of the amnion, are, firstly, a layer of some 2-3 mm. in thickness, composed of fibrous tissue containing fœtal blood-vessels, fig. 1 (A). Secondly, a middle layer, by far the thickest of the three, being 30-35 mm. thick, made up of a close tangle of fœtal and maternal blood-channels, fig. 1 (B), and thirdly, the more maternal surface along which the rupture between after-birth and uterus has occurred, fig. 1 (C), with a thickness of about 8 mm.

We will take these three layers separately. Region A fig. 1 is made partly of gelatinous and partly of fibrous connective tissue, and contains the blood-vessels, arteries, veins, and capillaries having a distinct endothelium. We can find no constant line of demarcation that would indicate the line of fusion between the amnion and allantois or between the walls of the allantois, but there is very generally a split, which may represent one or other of these presumably originally separate organs. At other points there is however undoubted fusion throughout the whole thickness of this region.

Except in the walls of the blood-vessels only a very few nuclei can be found; and of these the greater number lie close to the foetal surface. The walls of the blood-vessels are greatly thickened and are built up of interlacing bundles of large non-striated muscle fibres embedded in the same sort of gelatinous and fine fibrous tissue just alluded to.

These muscle fibres seem to be somewhat peculiar ; they are in some cases very large. They show a wavy outline with elongated nucleus which in transverse section is clearly seen to be subcentral in position. Each fibre has a thicker middle part which tapers suddenly at each end to a long fine fibre.

The middle layer (region B, fig. 1) is made up of flattened plates of fibrous tissue which run from the fibrous layers (A) and branch repeatedly in all directions in an arborescent fashion. Many branches end in small flattened foliations within the region (B), others pass through this layer and end in thicker less flattened terminations within the region (C). It seems possible from the fact that we find very thick trunks in region C that some of these lobate branches may have been torn off and left in the walls of the uterus (fig. 1 v”, fig. 3 fb,, fig. 15 v” and st. v.). Wherever these fibrous ramifications are found they contain blood-vessels which in the foliation, i.e., in region B, break up into a fine network of. capillaries, and in the lobate terminations, i.e., region C, into a network of vessels which are mostly of rather larger calibre.

In region B there is a complicated system of wide blood channels (m. ch.), containing the maternal blood, and forming a series of interlocking loops around and between the foliations which contain the capillaries of foetal connection (f. c.) (fig. 2).

The maternal blood flows in sinuous channels, having very definite walls, whose character we describe in a subsequent paragraph.

Between the fibrous blood-vessel-carrying foetal plates and their foliations on the one hand, and the sinuous maternal-blood-carrying loops on the other, there is a well marked layer of tissue of a syncytial character which is present throughout, and which cannot from purely anatomical considerations be said to belong to the one or to the other.

For reasons which will appear in the sequel we believe that it is wholly of fœtal origin.

This layer is not indicated in the diagrammatic fig. 1. Fig. 2, which is a careful camera drawing of a part of the region B, fig. 1, represents the histological nature of the parts just named. At f. c. one of the terminal foliations of a fœtal villus is thus seen in transverse section, the line ending in the fibrous groundwork, and at (c) a fœtal capillary cut across transversely. Let it be noted that the size of the lumen is only just sufficient to accommodate a single red corpuscle. These capillaries, like all the fœtal blood-vessels, are lined by a well marked endothelium. Occasionally a section will pass parallel to the flattened surface of the foliation and show the network of the capillaries (fig. 4).

A maternal blood-channel packed with blood corpuscles may be seen at m. ch. These channels are loops of sinuous vessels which open into larger ones as indicated in fig. 2, m. ch. These channels have very well defined walls, but the walls are of a peculiar nature, and it is not possible to speak with certainty about them without a knowledge of their development.

The walls are homogeneous, with a thickness of about diameter of a red corpuscle. They stain with the same dyes that stain the fibrous tissue of the fœtal villi, but less strongly. We believe the walls to be quite devoid of nuclei; although in one or two places nuclei have the appearance of extending into these walls, yet such cases are equally open to other interpretation. (Fig-11-End.)

Cells with deeply staining nuclei can be found on the inside of the homogeneous wall of the maternal blood-channel, which cells lie flattened up against the wall. It is not possible to follow any connecting strands between the cells, but the general appearance is that of a much attenuated endothelium. In parts especially nearer to the foetal side of the organ there can be little doubt of its endothelial appearance, but in the deeper parts and in the larger channels the matter is more doubtful. (Compare figs. 8 and 13.)

In places one is almost inclined to suggest that these cells are leucocytes flattened up against the homogeneous wall (v, figs. 13 and 14).

The thickness of these walls is very uniform throughout the system of loops, but the walls of the wider channels with which the loops communicate are thicker; the outer parts of these show a less firm consistency, and stain rather differently ; they are really intervillous spaces, as explained in Section II.

As to the origin of these walls and endothelium containing the maternal blood we can say nothing definite, but we have discussed below some of the possibilities, in our reference to the microscopic details of Owen’s specimen. The whole of this region (B) is composed of foetal vessels and the fibrous tissue which carries them, and these maternal blood-channels with their endothelium—if such it is—together with a third tissue, a syncytial tissue, which everywhere separates the fœtal and maternal vessels.

This syncytial layer, fig. 2 (cy), fig. 11 (cy), is a loose granular cytoplasm with faintly staining nuclei (n), scattered at rather wide intervals.

This syncytial layer at one place appears more closely applied to the foetal tissues, at another more closely applied to the maternal blood-channels. At almost all points it is crowded with brown and yellow pigment granules, so that in an unstained section the meandering course of this intervening layer stands out clearly.

It may be noticed that so complicated and complete is the branching of the foetal plates that a section taken horizontally through the region B is but slightly different to one taken vertically through the same region.

In region C of fig. 1 we find a very different state of affairs. This region is that bordering upon the line of rupture whereby the after-birth has become separated from the walls of the uterus. The general character of this layer may be gathered from fig. 3. Naturally enough the tissues here show signs of damage and are more difficult to interpret. There are, however, certain details of considerable interest which can be clearly defined. There is a complete absence of the “meandering” appearance. The maternal loop-like channels are not visible as such.

The lobate terminations take the place of the foliation described above. These are of much interest. Fig. 12 is a drawing of a section through one of these. Compared with a foliation it may be noticed that the whole is thicker—there is a greater amount of fibrous ground tissue—the blood capillaries are more internally placed (of. fig. 2, c.), and are really more of the nature of small arteries and venules than true capillaries. And, lastly, over the surface of the whole there is a well marked layer of almost columnar cells with well marked cell boundaries and nucleus. The layer of columnar cells cannot be found everywhere as clearly as in the upper part of the fig. 12, but traces of it can generally be seen as in the lower surface of the villus in fig. 12.

These villi in some cases seem to have passed through the whole of the after-birth and possibly may have become torn during parturition, and small portions may have been left behind in the uterus. Generally, however, they end as in figs. 3 and 12, lying in a loose detritus of cellular tissue. In this detritus we detect large numbers of cells which have the same general characters as those forming the columnar epithelium just described.

It is in this region C, if anywhere, that one would expect to find a trace of uterine glands and other undoubted maternal tissue. We cannot, however, offer any evidence of the presence of glandular epithelial cells, or indeed of other maternal tissues, except blood cells.

The transition from region B to region 0 is sudden, and forms a very striking feature. The connection between the fœtal trunks in B and those in C is quite clear. Here and there it is possible to trace main fœtal trunks giving off foliaceous branches in region B, and passing into region C and giving off the more lobate terminations in the latter layer. These terminations, as well as the main trunks, have a tendency to lie horizontally, that is to say parallel with the lines separating region B from region C, and in between these we find layers of coagulum mixed with blood corpuscles and cell detritus.

We regard this space as an irregular cavity or series of narrow chinks filled with blood which bathes the terminations of the villi directly in some places, but at other points it is separated from the trophoblast layer of the villi by the remains of a curious homogeneous material, whose nature we discuss in our Section II. upon a half-term placenta.

In fig. 3 some of these points are illustrated. The larger fœtal trunks are shown at f. b., the smaller branches at f. c. The investing columnar epithelium (tr.) is seen to be drawn away from the villi in most places, though their relation to the villi is clearly proved by the general distribution of the epithelium and by the position of the nuclei in the cells, which lie close to the fœtal border. The opposite border of the cells has a marked tendency to adhere to the coagulum, which fills up most of the space between the several villi.

This coagulum is partly of a reticulate nature, and is probably true blood clot, but other parts are more solid and contain remains of oval nuclei, and are to be regarded as of either trophoblastic or maternal origin or of double origin, as will be shown under Section II. Fig. 3 illustrates the difference between coagulum and detritus.

We regard this space containing a coagulum and blood corpuscles and cell detritus as one of probably many large blood sinuses into which the lobate terminations hung freely. The characters of the epithelium covering these parts of the villi are those of an epithelium perfectly free from pressure. Moreover, we can find no cells amongst the detritus which we cannot ascribe in origin to these epithelial coverings. There is no trace of any other kind of cell, as, for instance, cells which might line an expanded uterine gland.

We, should expect to find communications between this space and the maternal channels of region B, and also evidence of connecting pieces which joined the afterbirth to the uterine walls if the condition described above resembles that of the Carnivora as described by Duval (5), etc.

In fig. 7 (x) we believe there is such a communication. It is a drawing of a vessel just inside region B, which apparently crosses the boundary layer and opens out into the space between the fœtal villi in region C. There is no doubt that this is one of the larger maternal blood-vessels, as it has the same characteristic appearance as all the others already described. It lies within region B because smaller maternal vessels can be seen between it and region C, as shown at m. ch’. in the diagram. The fœtal plate which appears to divide B from C is shown at f. c., and it will be noted that the wall of the vessel depicted seems to end abruptly at this line. Up to this point inside the vessel we find the usual red corpuscles and very numerous leucocytes, but projecting from the region C into the vessel a coagulum will be found apparently identical with the coagula seen in various parts of region C.

We think that there can be very little doubt that the epithelium covering the lobate terminations in region C (fig. 12) is continuous with and has a similar origin to the syncytial layer of region B, in spite of the rather different character of nucleus; because of the connection which we can in certain places trace between the two, and of the identity of relation of the layers to the fœtal villi, and of the constantly lighter staining of the nuclei. The syncytial layer is less evident in the parts of region B nearer the fœtal surface, and close to this surface may perhaps be absent. In fig, 11, which is nearer the maternal surface, the syncytial layer is evident and shows a decided marking out into cell areas. Compare figs. 11 and 12, which are of the same magnification. We find blood-corpuscles abundant between the maternal vessels and the syncytium in the parts close to the maternal surface of region B, and in many places also towards the fœtal surface. The amount, however, diminishes very much the more distant the spot is from the region C, which is the line of rupture. We regard most of this extravasated blood as being probably due to the violence of parturition.

The maternal blood-channels are filled with blood-corpuscles, except the sinus in which the lobate terminations hang, which being open to the torn surface is not distended with blood. It contains coagulum and some corpuscles, together with the detritus as described above.

We can detect no difference between the red bloodcorpuscles, which are spherical within the maternal channels, and those in the fœtal capillaries. There are no nucleated red corpuscles. The maternal stream contains many more white corpuscles than can be found within the fœtal vessels.

In the fibrous tissue of the larger fœtal villi there are irregular lymph spaces, which contain some coagulum, and numerous leucocytes (fig. 6), but no red blood corpuscles.

Pigment

The general brown colour of an unstained piece of the placenta is due to the presence of dark brown or yellow pigment granules. These are of various sizes, and the larger ones are distinctly spherical with clear centres ; that is to say, the pigment is deposited somewhat unevenly upon the periphery of minute spheres.

They can be found only sparingly distributed in region A, not at all in region C, but abundantly in region B. Here they may be found in the fibrous tissues of the fœtal villi, and occasionally in the fœtal blood in leucocytes. They occur chiefly, however, in the syncytial layer, which is crowded with them, especially around the nuclei (fig. 11), and in the leucocytes in the maternal blood (figs. 13 and 14) and in the leucocytes in the lymphatic spaces in the large trunks of the fœtal villi. The leucocytes of the maternal blood are nearly always loaded with this pigment.

We find that the pigment is free of iron. When sections are treated with ferro-cyanide of potassium and dilute hydrochloric acid, there is a general blue coloration after a while, the blood-corpuscles and certain parts of the fœtal fibrous tissue being particularly strongly stained. The pigment granules are absolutely untouched by the blue. They are insoluble in ether and alcohol and soluble in caustic potash 1 per cent., and in ammonium sulphide.

There are several points of interest in connection with this pigment, which may be considered to be an excretory product. Fig. 13 is a drawing of a part of a maternal channel lying close against a fœtal villus. The syncytial layer is attached chiefly to the latter, but parts are adherent to the maternal channel wall. The maternal channel contains red corpuscles of different sizes, and large leucocytes, which are loaded with pigment, some sticking to the walls, others free within the blood-stream and some (l1) apparently undergoing disintegration, the pigment being set free in the stream. Frequently these cells may be found flattened up against the walls, as in fig. 14 (end) and, indeed, sometimes it is difficult to say whether they do not penetrate the walls.

Again, it is possible to find all stages between flattened pigment-laden cells as (l) in fig. 14 and (end) of the same drawing, which are not to be distinguished from the cells we find in other parts of the channels, and have spoken of as possibly endothelial.

It is likely that the leucocytes play some part in the trans-ference of this pigment. The appearances seem to favour the idea, namely, that leucocytes convey the pigment, which is formed or deposited by the syncytial layer, into the maternal layer, where they disintegrate, and so discharge their load of pigment.

It seems possible that these leucocytes are derived from the endothelial layer, as suggested by the forms shown in fig. 14, and by the probable method of development as indicated under Section II. of this paper.

Another possibility suggests itself. The large number of leucocytes in the fœtal fibrous tissue alluded to above is very remarkable. These are spherical cells with small, deeply-staining nuclei, and are mostly without pigment granules.

The size and character of the nucleus correspond with the size and character of the nucleus in the pigment-loaded cells of the maternal channels. Do these leucocytes migrate from the fœtal fibrous tissue through the syncytial layer, and there pick up the pigment and pass on to the maternal stream ?

It is possible to find in the syncytial layer small nuclei which correspond in size and staining affinity with these leucocytes, which nuclei are certainly not the true nuclei belonging to the syncytium.

It is rather significant that in the immediate neighbourhood of one of these fœtal leucocyte-containing spaces the adjoining maternal channels are more crowded with pigmentladen leucocytes (u, fig. 6) than elsewhere.

The Subcircular Bodies

The position of these peculiar structures was described by Owen and Chapman, and a more minute examination was made by Turner, who correctly describes them “as composed for the most part of fine fibres.”

In transverse section figure (9) they are seen to be composed almost entirely of fine fibres, and probably some gelatinous matrix. The fibres are especially abundant, and arranged as though forming a capsule round the middle region (cap.).

There are blood vessels, mostly capillaries, which are more numerous towards the chorionic surface. Only very few nuclei lie embedded among the fibres, and these are chiefly near the surface. Most of the nuclei one sees are nuclei of the endothelium of the capillaries.

Turner evidently regarded these as corresponding to the structures in the mare and pig, which gives rise to the “hippomanes “in the former. On the chorionic surface there are curious small protuberances, with constricted stalks, which are especially vascular (i>, fig. 10), and are probably minute villi, although in the specimen which we had there is no trace of a trophoblastic epithelium.

It seems to us extremely doubtful whether these structures are comparable to the hippomanes producing bodies. The capsule-like internal structure suggests an origin comparable to the curious recesses described by Hubrecht in the diffuse non-deciduate placenta of a lemur. [Hubrecht in his Spolia Nemoris on Nycticebus Tardigradus.]

Histology of the “Half-term” Placenta

Owen’s specimen was supposed by him to be of about halfterm, though Chapman considered it to be of not more than six months.

In spite of the fact that it has been over fifty years in spirit, it is possible to make out a good deal from certain parts. The more central parts are less well preserved, and have shrunk, but the edges are remarkably good and of very great interest.

The central part resembles in all essential points the condition already described for the full-term placenta, but between the centre and the edges we find conditions which have every appearance of being those of the earlier stages in development.

Except for the important absence of the walls of the uterus, we believe we have in this specimen of Owen’s all the stages of formation of the fœtal villi and their connection with the maternal blood supply.

Fig. 15 is a figure of the cut surface across the zonary belt of Owen’s specimen. On the fœtal surface the subcircular bodies are to be seen on the zonary belt as well as on the other parts of the chorion (sub-circ.). On the maternal side there are several points of interest. On each side of the thicker belt there are many branched villi, and beyond these many very small villi which are not branched. These latter were not discovered until an examination had been made by means of sections (fig. 15). They are, however, just visible by help of a pocket lens, and by a careful search over the whole chorion we have been able to recognise them for a distance of some five or six inches on each side of the zonary belt, and also for some three inches around the two villous patches near the poles.

Over the remainder we believe there are no villi, though we have not been able to examine all parts by sections. This, then, reduces the quite smooth part of the chorion in the half-term placenta to a couple of bands some five or six inches in width, between the poles and the zonary placental belt.

Turning to the maternal surface of the zonary placental belt, it will be seen that on either side there is an area completely devoid of villi and quite smooth.

These smooth bands are covered with the same “brownish granular matter “which is found extending outwards over the branched villi at the sides of the zonary belt alluded to above, as described by Chapman, who suspected it to be of maternal origin. Whether it is so we will discuss later.

Lastly, the more median portion of the maternal surface is on the whole rough, though here and there smooth patches can be made out by a surface examination of the whole specimen which resemble the lateral zones. The ragged state is due to the projection of villous tufts (V”) of long stalks ending in tufts of villi (V”), of similar stalks apparently broken (st.v.), and of very occasional large blood-vessels, which can hardly, from their nature, be anything else than maternal arteries.

On the face of the section the branching fœtal trunks, and the larger blood channels which contain the maternal blood can be easily seen ; and a more careful inspection will show that the smooth zones forming the sides of the placental belt are formed by a thin homogeneous material which seems to prevent as yet the fœtal villi gaining access to the maternal tissues along these areas, This material extends across the central zone, bat is here broken up by the penetration of the villi (which no doubt fitted into corresponding crypts in the walls of the uterus), and constitutes our region 0 in the description of the full-term placenta.

The microscopic examination of this smooth area throws a flood of light upon what we have found in the full-term placenta if, as we think, the edges of the zonary belt represent younger phases of development than the more central part, which resembles so closely our full-term specimen. It will be remembered that these smooth bands are not present in the full-term placenta (Chapman, 3, Beddard, 2).

The nature and origin of the homogeneous material forming the smooth surfaces of the lateral bands present us with a problem of much difficulty. Probably it is impossible to solve it completely in the absence of the uterine part of the placenta. There would seem to be two layers—one, the thinner, lying next the uterus, in which, or more often on the outside of which, we can find a few large rounded cells with small, dark nucleus (figs. 19 and 20) -, and the other a thicker layer which generally stains very slightly differently from the thin layer, and contains more or less degenerated nuclei (fig. 23).

Abutting upon, and in many cases embedded in, this are the ends of fœtal villi. These fœtal terminations form a well-marked layer (fig. 16), in which they lie closely apposed to one another, and only sometimes separated by an intrusion of the homogeneous material just spoken of.

The villi consist of a core of mesoblast with capillaries (which in this region stain deeply, almost black with haematoxylin, and give the Prussian blue reaction with ferrocyanide of potassium), and are covered by a typical cubical epithelium, whose cell-boundaries are distinct and whose nuclei lie close to the fœtal border of the layer. The cytoplasm generally stains very slightly. This, however, is only the case when the villi, crowded together, lie against one another. Where they abut upon the homogeneous material the epithelium has a very different appearance. The cell boundaries are less distinct ; the nuclei are much larger, and towards the apices of the villi the cells and nuclei are much elongated and the epithelium shows signs of proliferation (fig. 17).

As the cells become separated from their point of origin all trace of cell-Wall disappears, and the nuclei get fainter as they are absorbed into the homogeneous material, and ultimately lose all special affinity for nuclear stains, and are stained only by such diffuse stains that colour the homogeneous material, if they do not even totally vanish (figs. 17 and 23).

We may say here that although the nuclei throughout all parts of Owen’s specimen stain readily as a whole with such stains as haematoxylin, thionin, saffranin, carmalum, yet the preservation is not sufficiently good to allow of any detail of chromatic grains. In no case have we seen an undoubted mitotic figure.

There is no trace of any maternal blood channel outside the inner margin of this area consisting of terminations of fœtal villi.

To what conclusion must we come as regards this homogeneous material ? Is it all fœtal : a kind of plasmoditro-phoblast ? or is it partly maternal : either cell detritus—e. g. uterine epithelium—glandular secretion, or both ?

It stains moderately with hæmatoxylin carmalum, eosin, nigrosin, and deeply with saffranin ; it is not affected by osmic acid nor by ferrocyanide of potassium, whereas the mesoblastic portions of the terminations of the fœtal villi are stained an intense blue.

With fuchsin and orange there is a slight differentiation between the outer thinner and inner thicker layers, which are further distinguishable by the presence of the rounded cells in the outer layer and the elongated nuclei in the inner layer (fig. 23).

It seems probable, therefore, that it is of double origin, partly from fœtal trophoblast (the inner thicker portion) and partly from maternal tissues (the outer thinner), in which case the large rounded cells with small nuclei would be of maternal origin.

It is difficult to understand what the function of this layer may be, unless it is to prevent for the moment—for some not discernible reason—the entrance of the fœtal villi into the maternal tissues. The fact that cells are being budded off from the trophoblast seems to indicate a non-absorbent epithelium ; that the nuclei of these budded-off cells should die and disappear shows it is not a phagocytic edge that is being formed, and this is also rendered improbable by the smooth regularity of its superficies. The ready response to the iron test noticed above suggests the absorption of iron from this layer. We have searched in vain, however, for blood remains, which, perhaps, one might expect to find if this layer was the seat of destruction of hæmoglobin-bearing corpuscles.

A glance at the cut end of Owen’s specimen (fig. 15) will show that the surface of the section is roughly marked out into areas by the longer and stouter branches of the fœtal villi. Within the boundaries of each such area we find on the whole a uniform stage of development, but one area will often show a different state to its neighbours, and roughly speaking the nearer the area is to the maternal surface and to the edge of the placental zone the earlier is the stage represented.

On the supposition that these stages which undoubtedly occur really represent the earlier phases in the normal development through which the more central region has already passed—for the central region resembles in all essential features the full-term placenta—we have here all the stages showing how the maternal blood gains access to the fœtal part of the placenta.

The details of these processes are exemplified as far as we can determine them with our limited material by the sections drawn for figs. 16-24. Fig. 18 shows a more internally-placed area where all trace of cell boundary has disappeared ; the mesoblastic portions of the fœtal villi (f. v.) stand out clearly, and between these the broken-down trophoblastic syncytium. Nuclei are perceptible, but are less easily stained.

In this, we see several clear spaces (m. ch.) apparently within the common syncytium or plasmodium formed by this breaking down of the trophoblast of adjoining villi. These spaces have exactly the appearance of having been produced by the bursting in of a stream of fluid containing corpuscles and so crushing up the fine reticulum of the plasmodium, thereby forming the well defined non-cellular walls so characteristic of the full-term maternal channels.

The original nuclei of the trophoblast can be seen as fairly large oval highly staining nuclei lying on either side of, but usually well away from the newly formed channels. Possibly they are less numerous than in the supposed earlier stage, but of this we are not sure. The nuclei of the corpuscles which float in the fluid in these channels are much smaller and are more deeply stained.

Whence are these nuclei ? The two alternatives are that these corpuscles are derived from certain of the nuclei of the plasmodium, dividing and causing a local liquefaction comparable to the process described by Hubrecht in the placenta of Tarsius spectrum (Hubrecht, 9) or they have had some more foreign origin and have been brought into their present position in a stream driven along where resistance is least by the pressure of the maternal blood behind it.

We incline to the latter hypothesis. Firstly, because there are no cells or nuclei in the plasmodium which seem to be the parent cells of these corpuscles. Secondly, at the heads of these new channels the contained corpuscles are not, as far as we can judge, undergoing multiplication, while we can find places, in the larger channels, where they are certainly being formed. Thirdly, we can find no intermediate stages between the unbroken plasmodium and the clear space containing separate corpuscles.

As favouring the other view there is the fact that there are no non-nucleated red corpuscles at all in these newest channels. We find them only in the larger or older channels. Although we cannot speak positively about the origin of these corpuscles, we have some evidence of their immediate formation.

In the larger channels there are large numbers of red non-nucleated blood corpuscles, and large numbers also of nucleated corpuscles, some with more than one nucleus and quantities of bodies of all sizes which stain deeply which seem to be fragments of nuclei (fig. 22).

Also near the margin are numbers of large cells, sometimes elongated and almost endothelial in position, which in some cases appear to be breaking up, and in others to have more than one nucleus. We regard these as the parent cells of the nucleated corpuscles within the channels, but the question, which must remain unanswered for the present, is what is the actual origin of these mother cells ? We do not find them in the small channels, but we can trace them back into the largest channels and even into the large vessels (fig. 25), which occasionally occur and are evidently the connecting vessels between the placental blood system and the vessels of the uterus, which we are inclined tp consider maternal (fig. 15,and fig. 1 b. U. V. A.). An hypothesis not without some evidence to support it would be to suppose these cells to be of maternal endothelial origin, and to become gradually washed along the newly formed channels, adhering for a while (after the manner of white blood corpuscles) to the sides of the larger vessels (figs. 22 and 24), and growing and multiplying.

In the smallest newly formed channels there is clearly no endothelium. We see the non-cellular wall as described in the first section, and think it probable that the endothelium therein described is formed by these migrating large cells just mentioned. On our hypothesis this apparent endothelium would be of maternal origin, though different in nature to the endothelium, both of the maternal channels of the carnivore and of the rodent type of placenta.

There are still traces of this proliferating endothelium in some few of the larger channels in the full-term placenta.

In some places there are characters which seem to suggest the former hypothesis, but they afford evidence even less convincing than what we have just advanced on the other side.

In fig. 21 an interesting section is illustrated wherein the new channels and their blood contents are seen to be in chinks between the trophoblast layers of adjoining villi. This is very clearly recognisable in the larger circular channel (M. C. H.) in the wall next to the mesoblast of a villus, and in the long, narrow channel at the bottom of the drawing.

In these the trophoblast has hardly changed from its cubical character which is typical of the villi near the smooth edge (v. fig. 16), and the nuclei and cell divisions are distinctly recognisable.

We found on examining the rough surface of the placenta of Owen’s specimen that we could only make out a few of the vessels which connect the placental circulation with the vascular system of the uterus. Of these, which were less than half a dozen in number, we were able to take a piece of one. It is in a bad state of preservation, but examination of it by section enabled us to be sure that it is a blood-vessel, and probably arterial. Its endothelium undoubtedly shows signs of proliferation (fig. 25).

These vessels pass through region C before giving off branches in region B.

In fig. 15 certain long villous tufts may be seen. These are villi which have passed through the homogeneous layer and presumably have penetrated into the uterine mucosa (or into glands). Some, we believe, have been torn off and left in the uterus (fig. 15, st. v.) (this, we believe, has also occurred in our full-term specimen), others have come out from their crypts (fig. 15, V”).

It is extremely difficult to make out any epithelium covering these villi. They are much branched and have small pieces of something identical or similar in appearance to the homogeneous material, which perhaps contains cells. Anyhow there is no definite epithelium. The mesoblastic core, which is thick and lobate, contains a more or less central vessel (artery ?) and a number of small capillaries arranged round the periphery—separated by a large amount of connective tissue, The pigment, which is so obvious a feature of the full-term placenta, is here hardly recognisable. A few fine granules may, however, be found over most parts.

The branched villi (fig. 15, V V) close to the edge of the placental zone have a covering of trophoblast similar to that on the crowded villi of the zonary belt. They also are covered by the curious homogeneous layer which at the apices of the villi seems to be receiving nuclei from the foetal epithelium.

The only difference between such a villous patch and the villi of the placental zone is that in the former the villi are less numerous and less crowded ; in the latter they are much crowded. It is not difficult to conceive how an increase in size and number of villi such as those on the border (fig. 15> V V) might lead to a, close tangle of foetal villi originally diffuse and non-deciduate, especially if (as seems to be the case) they were prevented from penetrating the uterine tissues. Into this tangle a bleeding at the surface of the uterus might percolate, and so give rise to the system we have described above, by flowing at first between the villi (figs. 16, 21, 24) and ultimately perhaps within the substance of a syncytium (fig. 18).

We must leave for the present any further consideration of the fate of the cells L. C. (figs. 21, 22, 24) and be. (figs. 21, 20), etc., which we have spoken of as “endothelial cells/’ “maternal cells,” “blood cells,” as we are doubtful whether the condition of preservation is sufficiently satisfactory in Owen’s specimen to enable us to come to a reliable conclusion upon so difficult and important a point. We think we have strong evidence that they are derived from the endothelium of maternal blood-vessels. There are also appearances which resemble stages in haematopoietic areas, as described for other mammals (e. g. Hubrecht, 9), but we certainly have been unable so far to follow the process.

There are no non-nucleated blood corpuscles in the small developing channels; which fact points strongly to the conclusion that there is no circulation as yet in these channels.

The cells floating in the fluid in these channels are of various shapes; squamous, fusiform, or subspherical. The cytoplasm of the smaller ones is clear, and the nuclei, which are nearly spherical, stain deeply.

We have assumed in our conclusions that the processes observed in the lateral regions of the half-term placenta are essentially similar to the processes by which the middle region, which resembles the full-term condition, is formed.

We are probably justified in this supposition since we know that at full term the lateral smooth regions have disappeared, and the whole breadth is similar to the central part.

Conclusions

Those who have devoted any time to the study of the development of the placenta know well enough how difficult a task it is to determine the origin of certain of the tissues of the fully formed organ, even when a rich series of developmental stages is procurable.

Since we have only the foetal part of the placenta at its final and one other stage before us, it follows that there must be many points upon which we cannot speak with certainty. It is perhaps a hazardous undertaking to attempt an interpretation of the facts as given above, but seeing how improbable it is that more suitable material will be obtainable in the immediate future we must risk the attempt.

As a result of our investigation, certain questions at once present themselves. Do our investigations confirm Turner’s conclusion that the elephant’s placenta is deciduous ? What is the origin of the syncytial layer; does it belong to the fœtal or maternal tissue ? What is the origin of the walls of the maternal channels; are they fœtal or maternal ? How does the placenta compare as regards its minute structures with other known placentas ?

Turner defined a deciduate placenta as one in which there is a “shedding of the vascular part of the maternal placenta during parturition.” He came to the conclusion that the elephant’s placenta is without doubt deciduate.

The placenta, as will be gathered from our description above, is at first sight more like the type of placenta found in those orders called deciduous than that of the true ungulates, lemur, etc., and known as non-deciduous. There can be no doubt about the shedding of at any rate part of the maternal vascular system. But for all that it is impossible to speak with certainty, except as regards the blood, from so scanty a supply of material ; and it is not absolutely certain that any maternal tissue except the blood, which obviously comes away in large quantities, is lost during parturition. In the after-birth the great mass of tissue is without doubt fœtal.

Of the three regions, A, B, and C, of our description of the full-term specimen, region A is wholly and region 0 mostly fœtal. That this is true of region A needs no explanation.

As regards region 0 the matter is less obvious. We find, however, some tissue which we must regard as a remnant of the homogeneous material described in Section II. on the Owen placenta. This contains nuclei, which are probably of the thick inner layer of that material and therefore trophoblastic ; and very occasionally we think we can detect one of the rounded cells with small nucleus which we have ascribed to maternal tissue.

There also (though not shown in our section fig. 3) lie the arterial vessels which connect the maternal system with the vascular system of the after-birth as suggested in our diagram fig. 1 (ó).

We can, however, find no trace of glandular epithelium or of other tissue of any kind except blood-corpnscles and coagulum, which do not show strong evidence of foetal origin.

The detritus which lies between the villi of this region is composed of coagulum, blood-corpuscles, and cells, nearly all of which are in every respect similar to those covering the foetal villi.

In region B we find some evidence of maternal tissue besides blood, but it is by no means conclusive. We have no doubt that—on the hypothesis of the epithelial layer covering the villi being trophoblastic—the maternal blood circulates at first between the trophoblast layers of adjoining villi ; (that is to say, really between the fœtus and the mother), and ultimately probably within a plasmodium caused by the breaking-down of the trophoblast cells. The evidence is clearly against the endothelium of these channels being an ordinary maternal capillary endothelium, though we think it is for the reason given above very likely the product of cells derived from maternal blood-vessel endothelial cells. If this is not so, then the only maternal tissue in region B is the blood.

The placenta as regards its zonary belt is deciduate. Owen wrote : “The most important modification in the vascular structures connecting the chorion with the uterus, in the elephant, is their combination of two forms of the placenta, viz. the annular and the diffused, which forms are restricted in other mammals to distinct kinds of quadrupeds.” Aud even in the zonary belt there is a combination of deciduous and non-deciduous placentation. The lateral bands of this belt are formed of long branched tufts which seem to form no interlocking system of blood-vessels destined to be thrown off at birth, Their appearance in the half-term placenta (fig. 15) tempts the suggestion that they form an edging which tacks down, so to speak, the lateral bands of the belt whose smooth maternal surface indicates but slight means of attachment.

A similar function amongst others must be assigned to the long villi which pierce region C and enter the uterine walls.

Although there is an obvious resemblance between the elephant’s placenta and the zonary placenta of the Carnivora, and even between the meandering character which we have described, and the labyrinthine lamellæ of the dog or cat’s placenta, yet we believe that this resemblance is superficial— for in reality the two types are very different.

Bearing in mind the possibility of affinity between Ungulates, Carnivora, and Proboscidea through some such forms as Phenacodus, it is especially interesting to compare the placenta in these three types.

Firstly, the elephant’s placenta differs from that of the dog in having a distinctly non-deciduous part, which resembles closely the diffuse type of the mare.

The carnivorous placenta is a broad central band leaving the two poles bare, while in the elephant the central band is comparatively narrow, and the poles are villous, leaving two broad bare zones between the poles and the central deciduous belt.

There is in the carnivora nothing comparable to long villi which pass through the “coagulum “region and extend deep into the maternal tissues, and are, so far as the projecting parts of the villi are concerned, non-deciduate.

There is no sign in the elephant’s placenta of anything like the angio-plasmode. The mode of vascularisation of the after-birth is quite different in the two types.

In the Carnivora the trophoblast advances into the maternal tissues according to Duval and surrounds the maternal capillaries. Other authors, Strahl, Fleischmann, Heinricius, etc., give different accounts, but in no case is there anything comparable to the conditions we find in the elephant, where the trophoblast is comparatively inactive and extravasated blood from the maternal system seems to force its way between the mother and fœtus aud ultimately into the trophoblastic plasmodium. In this it resembles far more closely the process as it is known to occur in Rodents, Insectivora, Tarsius, Primates, etc.

There is, however, this striking difference. In the Rodents, Insectivora, Tarsius, etc., the vascularisation of the trophoblast by the maternal blood occurs before the advent of foetal capillaries. In the elephant the fœtal villi are fully developed with their capillaries before any trace of maternal vascularisation occurs.

In a recent interesting paper in the ‘Philosophical Transactions of the Royal Society/ by C. W. Andrews, on “The Evolution of the Elephant,” the writer discusses the probability of affinity between the Proboscidea and the Sirenia, and gives as one of the points of resemblance the supposed fact that the placenta in each group is non-deciduate and zonary. The author would seem to have derived this statement from Flower and Lyddeker’s “Mammals,” where the statement occurs that the elephant’s placenta is non-deciduate and zonary, which is misleading, for the zonary part is incomparably the more important, and is, as Turner and Owen asserted, deciduate.

Turner very clearly emphasises these differences in his paper on the placenta of the Dugong. In discussing the possibility of affinity between the two groups, he says, “In this connection, therefore, it is interesting to observe that, as regards its form, the placenta, both in the elephant and the dugong, is zonary; though they differ in this very important particular, that in the elephant the zonary placenta is deciduate, in the dugong it is almost entirely, if not entirely, non-deciduate.”

There is an interesting point in Turner’s description of the dugong’s placenta which suggested the explanation of the very large thick villi with broken stems, many of which pass obliquely through our region C (fig. 3, fb, fig. 15, V”).

Turner writes of the dugong (p. 655) : “But my description has shown that, in addition to the multitude of short villi and shallow crypts, the dugong also possessed a small proportion of larger villi, which were implanted in larger and wider and more deeply seated crypts, passing for some distance in an oblique direction subjacent to the layer of short crypts.”

In the specimen of the dugong described by Turner the fœtal placenta had been torn away from the maternal tissues in such a way that the layer of short villi were.drawn out intact from their uterine crypts, but the longer, less numerous villi had been broken off and left within their crypts.

Turner regarded this as abnormal, and believed that in normal parturition the longer villi would also come out, perhaps bringing some of the uterine tissues with them. He does not seem to have thought it likely that they should be naturally torn off and left, and absorbed by the maternal tissues, which is an alternative by no means improbable. Such an absorption of fœtal tissues by the uterine tissues is common enough—as, for instance, in the case of the placenta of the mole, of Perameles (Hill), mouse (Jenkinson), etc. A similar absorption may frequently be noticed in rabbits, where embryos, if too numerous, are crowded out, die, and become absorbed together with the placenta. Such also is the probable fate of the long villi of the elephant and dugong.

  1. The full-term after-birth of the elephant consists of a chorion from which spring many much-branched villi, which spread out in all directions into plate-like branches. These end in (a) proximal foliaceous terminations, in which the fœtal blood vessels ramify, which interlace with a complicated system of much larger blood channels filled with maternal blood, having well-defined but non-nucleated walls ; (b) more distal lobate terminations, which are covered by a well-marked columnar or cubical epithelium — presumably the trophoblast — which are partly embedded in a kind of coagulum or detritus, and partly appear to hang loosely in irregular blood spaces without walls ; (c) the stems of still more prolonged villi, which have been torn off and probably left embedded in the walls of the uterus; (d) a few torn ends of blood-vessels.

  2. The main trunks of the villi and their foliaceous terminations are everywhere separated from the maternal bloodchannels by a syncytial layer, which is continuous with the epithelium covering the lobate terminations, and is presumably trophoblastic.

  3. The half-term placenta originally examined by Owen in 1850 shows, in its more central region, characters which are essentially similar to those of the full-term specimen, and goes far to prove the existence of longer villi which penetrate deeply into the uterine mucosa. The lateral areas of the zonary belt exhibit many most interesting previous conditions. We are able to see in these the simple terminations of the fœtal villi covered with a single layer of trophoblast separated from the uterine tissues by a layer of material partly maternal and partly of fœtal origin.

    There is no process of growth round existing maternal capillaries to form an angio-plasmode, nor apparently any phagocytic action on the part of the trophoblast. The vascularisation of the after-birth is effected by the invasion of the trophoblast by extravasated maternal blood, which flows at first in intercellular and intervillous passages which form the larger channels of the after-birth maternal vascular system, and then makes its way along intra-cellular or intrasyncytial canals through a plasmodium produced by the breaking down of the trophoblast of two adjoining villi.

    We think the evidence is in favour of considering the corpuscles floating in this invading stream, which contains no red non-nucleated corpuscles in its more advanced portions, to be of maternal rather than trophoblastic origin.

  4. The tissues of the full-term placenta contain pigment granules, which are deposited chiefly in the syncytial layer. This we regard as an excretory product ; it is almost quite absent from the tissues of the half-term specimen. Leucocytes, either of maternal or fœtal origin, seem to be concerned in the transference of this pigment into the maternal blood stream.

  5. The subcircular bodies of Owen we find as described by him and Turner, though we note the presence of minute villi on their outer surface.

  6. We confirm the opinion of previous writers that the zonary band in part is a “deciduous “form of placenta, although there is not much maternal tissue except the blood. It is not correct to speak of the after-birth being composed of a “much hypertrophied mucosa layer of the uterus.”

  7. The placenta of the elephant shows by its long villi, which tend to remain embedded in the uterus wall, a resemblance to the condition found in the Sirenia; by the villous patches at the poles and other villi which come out from the uterus, either with or without their trophoblastic covering, but with no maternal cells attached, a resemblance to the ungulata vera of the Perissodactyl type ; by the invasion of the trophoblast—if such it is—by the maternal blood stream, a resemblance to the Discoplacental type, although the actual manner by which this invasion occurs would seem to be—so far as our very limited material affords us opportunity of observation—unlike anything hitherto described.1

  8. The resemblance, at first sight obvious enough to the zonary placenta of the carnivora, is superficial. The elephant’s placenta differs from that of the carnivora in (a) consisting of three areas of attachment instead of one, two of which are wholly in the non-deciduous type, the other partly deciduous, partly non-deciduous. (b) There is nothing formed comparable to an angio-plasmode. (c) The maternal capillaries do not directly become the maternal vessels of the after-birth.

1.
Andrews
,
C. W.
—“
On the Evolution of the Proboscides
,”
‘Phil. Trans. Roy. Soc. London,’
1903
, vol.
clxlvi
.
2.
Beddard
,
F. E.
—“
Report on the Birth of an Indian Elephant in the Society’s Menagerie
,”
‘Proc. Zoo). Society, London,’
1902
.
3.
Chapman
,
H. C.
—“
The Placenta and Generative Apparatus of the Elephant
,”
‘Journ. Acad. Nat. Sci. Philad.,’ 2nd series
, vol.
viii
,
1880
.
4.
Chapman
,
H. C.
—“
La Gestation et le Placenta de l’éléphant (Elephas Asiaticum)
,”
‘Comptes Rendus et Mémoires de la Société de Biologie,’ llth série
, tome
i
,
1899
.
5.
Duval
,
M.
—“
Le Placenta des Carnassiers
,”
‘Journ. de l’Anat. et de la Physiol.,’
1893—1895
.
6.
Hill
,
J. B.
—“
The Placentation of Perameles
,”
‘Quart. Journ. Micr. Sci.,’
vol.
xl
.
7.
Hubrecht
,
A. A W.
—“
Studies in Mammalian Embryology
,”
‘Quart. Journ. Micr. Sci.,’
vols,
xxx, xxxi
,
1890
.
8.
Hubrecht
,
A. A. W.
—“
Spolia nemoris
,”
‘Quart. Journ. Micr. Sci.,’
vol.
xxxvi
,
1894
.
9.
Hubrecht
,
A. A. W.
—“
Ueber die Entwickelung der Placenta von Tarsius und Tapaja, nebst Bemerkungen über deren Bedeutung als Hæmatopoietiscbe Organe
,”
‘Proceedings of the Fourth International Congress of Zoology,’
1899
.
10.
Hubrecht
,
A. A. W.
—“
Furchung und Keimblattbildung bei Tarsius Spectrum
,”
‘Verhandelingen der Koninklijke Akademie van Wetenschappen te Amsterdam, (Tweide Sectie),” Deel
,
viii
, No.
6
,
1902
.
11.
Jenkinson
,
J. W.
—“
Observations on the Histology and Physiology of the Mouse
,”
‘Tijdschrift’d. Nederlandische Dierkundige Vereeniging,’ 2nd ser., D
.
vii
,
1902
.
12.
Lüsebbink
. —
“Die erste Entwickelung der Zotten in der Hundepla-centa Anat. Hefte,” 1st
Ab.
i
,
1891
.
13.
Owen
,
R.
—“
Description of the Fœtal Membranes and Placenta of the Elephant (Elephas Indians Cuv.), with Remarks on the Value of Placentary Characters in the Classification of the Mammalia
,”
‘Phil. Trans. Roy. Soc. London,’
1857
.
14.
Owen
,
R.
“Anatomy of Vertebrates,”
London
,
1866
.
15.
Turner
,
W.
“Lectures on the Comparative Anatomy of the Placenta,”
Edinburgh
,
1876
.
16.
Turner
,
W.
—“
On the Placentation of the Lemurs
,”
‘Phil. Trans. Roy. Soc. London,’
1876
.
17.
Turner
,
W.
—“
On the Placentation of Halicore Dugong
,”
‘Trans. Roy. Soc. Edin.,’
1890
, vol.
xxxv
.

Illustrating Mr. Richard Assheton’s and Dr. Thomas G. Stevens’s paper, “Notes on the Structure and the Development of the Elephant’s Placenta.”

Complete List of Letterings

A. Arterial vessel or placental afferent, be. Blood cells (derived from maternal endothelium?), bl. Maternal blood corpuscle, cap. Capsule-like structure, co. Coagulum. eg. Syncytial layer—trophoblast, cycy. Syncytium formed by fusion of two adjoining trophoblasts, end. Endothelial-like cells, f.b. Fœtal blood-vessels, f.c. Fœtal capillaries, f.v. Fœtal villi. i. Deeply staining iron-containing tissues. L. C. Large cells presumed to be of maternal origin. I. Leucocyte. I’. Leucocyte undergoing disintegration. Me. Maternal cells. M. G. H. Large maternal channel inter-cellular. m. ch. Small maternal channel intra-cellular. Mes. f. Fœtal mesoblast, BI./. Nucleus of non-striated muscle fibres, n. nuclei, pl. plasmodial remnant. shr. Space caused by shrinkage, sh. v. short villus, si. v. Stem of villus broken off and presumably left in the walls of the uterus. Sub. circ. Subcircular bodies of Owen. tr. Trophoblast (probably). U. Walls of the uterus. U.V.A. Uterine blood-vessel (arterial). U.V.V. Uterine bloodvessel (venous). V. Branched villus of non-deciduous border. V’. Villus of deciduate portion. V”. Villus of zonary belt which penetrates into the walls of the uterus. V”. Minute villi on the chorion beyond the zonary belt. w. Non-cellular wall of the maternal blood-channels, x. Point of communication between the maternal blood-channels of region B inside the space of region C. z. Layer of homogeneous material containing degenerating nuclei of trophoblastic origin, zz. Layer of homogeneous material containing no trophoblastic nuclei, but occasional rounded cells (of maternal origin?).

Plate 1

Fig. 1 a.A diagram of a section through the full-term zonary placenta of an elephant to illustrate the relative position of the several regions A—D. Region A is the chorion from which villi pass towards the walls of the uterus. Some pass through the whole thickness of the after-birth, others only extend a short distance. Region B, which forms the great bulk of the after-birth, is made up of fœtal villi and channels containing maternal blood. Region C is a much looser layer in which the fœtal villi are lying freely in irregular blood-spaces. There are no walls to these blood-spaces. This region is traversed by a few blood-vessels which carry blood probably towards the fœtal placenta. Some of the ends of the villi in this region are covered by a plasmodial remnant (pl.). The region lettered D indicates the probable relation of the uterine surface to the after-birth. The longest villi penetrate into this maternal tissue (V”).

Fig. 1 b.A diagram of a similar region to show the probable relation of the blood channels to the maternal vascular system. A is an arterial vessel carrying blood from the uterine vessels U V A to region B, where it circulates in the channels round the villi and collects into large efferent vessels, which open into irregular spaces in region C, whence it is supposed the blood passes into maternal veins, U V V.

Fig. 2.—A camera drawing of a portion of a section taken vertically through the placenta showing the features characteristic of region B. The wide maternal channels (m. ch.) are filled with corpuscles. The mesoblast tissue of the fœtal villi and their branches are coloured darkly. Many fœtal capillaries are seen cut across transversely (f. c.). Between the fœtal mesoblast and the blood-channels containing maternal blood an irregular layer of syncytium (ay.) is seen which can almost without doubt be called trophoblast, × 340.

Fig. 3.—A camera drawing of a piece of region C, from a section taken vertically through the placenta. The large fœtal blood-vessels (/. b.) in the large trunks of the villi and the small terminations are clothed by a columnar epithelium (tr.), probably the trophoblast. These epithelia have shrunk away in many places and adhere to a coagulum, which is partly blood-clot and partly remnants of a trophoblastic plasmodium and cell detritus, × 75.

Fig. 4.—A section of a foliaceous termination of a fœtal villus in region B cut parallel to its broad surface, showing the network of fine capillaries. × circ. 200.

Plate 2

Fig. 5.—A group of leucocytes in a space among the tissues of a fœtal villus. × circ. 200.

Fig. 6.—A section taken across a fœtal villus cutting a fœtal vein and artery transversely. This shows the large spaces in the mesoblastic tissues filled with leucocytes. Adjoining this is a maternal blood-channel of moderate size also crowded with leucocytes. × 95.

Fig. 7:—A camera drawing of a section through the boundary between region B and region C. At the point x a communication between the vascular channels of region B and the irregular blood-space of region C is shown, × 75.

Fig. 8.—A piece of a maternal blood-channel from region B to show the sinuous baggy nature of these channels and the endothelial character of cells found lining the walls (end). × circ. 200.

Fig. 9.—A transverse section through one of the subcircular bodies, ch. chorionic surface ; all. allantoic surface ; f. c. capillaries, × 95.

Fig. 10.—A piece of the chorionic surface of the above showing minute vascular villi ; f. c. capillaries. There is no trophoblast, × 200.

Plate 3

Fig. 11.—A highly magnified drawing of a small piece of region B to show the characters of the syncytial layer (cy.) It is taken near to region C, and shows the syncytial layer more distinctly than would a piece nearer to the fœtal surface. The syncytium is distinctly marked out into cell areas. The distribution of the pigment granules is shown, and also the very definite walls to the maternal blood-vessels. × 920.

Fig. 12.—A highly magnified drawing of the terminal portions of fœtal villi in region C, showing the larger size of the capillaries and the more columnar character of the investing epithelium, not here a syncytium. Compare this with fig. 11. × 920.

Fig. 13.—A small piece of a maternal blood-channel (from region B), and an adjoining fœtal villus ®’. This is to show the leucocytes (l) in the channel crowded with pigment granules. Some near the middle seem to be disintegrating, or at any rate getting rid of their load of pigment V. × 920.

Fig. 14.—Another piece similar to that of fig. 13, but showing the flattening-up against the wall of these pigment-collecting (?) leucocytes, × 920.

[All the above are from the full-term placenta. Those that follow are from Owen’s specimen, namely the half-term placenta.)

Fig. 15.—This is a drawing of a section through the whole zonal belt of Owen’s specimen. The comparatively small area over which the villi penetrate is clearly seen, leaving nearly two-thirds of the maternal surface smooth. Note also the bands of non-deciduous villi on either side V, the minute simple villi V’”, and the large trunks V’ with their long tufts V”. At pl. the homogeneous material, and at 16 the region of the crowded together terminations of the villi abutting upony)Z. Nat. size.

Plate 4

Fig. 16.—A piece just within the homogeneous material layer. It shows the ends of fœtal villi covered with well-defined cubical epithelium. Between these in some places small pieces of detritus and of the homogeneous layer have been enclosed, × 300.

Fig. 17.—An apex of one of the villi abutting against the homogeneous layer. The cells of the trophoblast seem to be passing out into the plasmodium when they lose their affinity for staining and appear to die. × 400.

Fig. 18.—A rather more internally placed area. Here the trophoblast has degenerated into a syncytium, and the adjoining syncytia fuse. In the lefthand corner an advanced stage is illustrated. Within the syncytium certain spaces are recognisable (m. ch.) in which numerous cells are floating (A c.). × 325.

Fig. 19.—Some of these large rounded cells mentioned under fig. 23, seen on the maternal surface outside a sharply defined membranous layer. × 400.

Fig. 20.—Similar cells just within the membranous layer. × 400.

Fig. 21.—A piece taken from about the spot where the line 16 ends in fig. 15. Here the new blood-ehannel is clearly seen to be between the trophoblast cells, not within them, × 100.

Fig. 22.—A large maternal channel derived from inter-cellular (intervillous) space. The walls are well marked, and on their outer side are supported by the columnar cells of the trophoblast. Compare fig. 24. Inside are many blood-cells, and along the border large cells which give rise to smaller ones. Are they maternal in origin ? × 490.

Plate 5

Fig. 23.—A piece of the limiting homogeneous layer, showing one apex of a villus abutting against it. The nuclei passing off into it are seen getting less and less distinct as they pass further away. Beneath this layer is an outer layer destitute of the above-mentioned oval nuclei, but some tissues containing loose rounded cells with small nuclei probably of maternal origin. × 400.

Fig. 24.—Part of a large intervillous space forming one of the chief blood-channels for maternal blood. At/, c. the fœtal villus with its capillaries; at ir. the trophoblast layer shrunk away from its mesoblast and forming the wall to the large blood-channel C. U. in which many large cells (end) and masses of fragments can be observed, × 400.

Fig. 25.—A portion of a transverse section of one of the few blood-vessels found attached to the rough median area of Owen’s specimen. The endothelium has the appearance of being in a state of proliferation. It is assumed to be a maternal artery, × 400.

1

There is a very marked resemblance to the conditions found in the sheep’s placenta during the final stages of development. This I hope to describe in detail in a paper now in preparation.—R. A.