Studies on Sterility and Prenatal Mortality in Wild Rabbits: II. The Occurrence of Fibrin in the Yolk-Sac Contents of Embryos During and Immediately After Implantation

It has been shown (Brambell, 1942,1944) that all the embryos in many of the litters of wild rabbits from Caernarvonshire die between the 11th and 15th days of gestation. The heaviest mortality is on the 11th and 12th days. A precisely similar mortality has been found to occur in wild rabbits from Anglesey, Norfolk and Dumfriesshire (unpublished), so that it must be regarded as widespread throughout Britain. The incidence of the mortality varies both from one locality to another and also from season to season in the same locality. The embryos that die at this stage of development are not aborted but undergo a gradual process of removal by autolysis in situ, and even after the embryos have disappeared the placental sites, where such ‘reabsorption ‘has occurred, can be distinguished readily from post-partum placental sites or abortion sites by the remnants of the maternal decidual tissue projecting into the lumen of the uterus from the mesometrial side.

This particular kind of prenatal mortality is characterized both by the remarkable constancy of the stage of development at which it occurs and by the fact that it affects all the embryos in the litter almost simultaneously. The statistical significance of these characteristics has been discussed elsewhere (Brambell, 1942,1944 ; Brambell & Mills, 1944, 1946). Attempts to investigate the immediate cause of the mortality embryologically have met with difficulty. The material was obtained largely by trapping and partly by shooting and netting (Allen, Brambell & Milk, 1946), and the pregnant animals had been dead for periods varying from an hour or two to a day or more before dissection. Further, the mortality could only be identified with certainty when the embryos had not only died but had begun to autolyse before the death of the mother. It has proved impossible in these circumstances to distinguish the autolytic changes consequent on the embryonic mortality and the putrefactive changes ensuing from the death of the mother from the irregularities in development preceding and leading to the death of the embryos. Since the incidence of the mortality is so high in some series it is reasonable to suppose that a substantial proportion of the embryos obtained on the 8th, 9th and 10th days of gestation would have died on the 11th or 12th days if the mother had survived, and that therefore irregularities in development that would ultimately result in death might be detected in them without the complication of autolytic changes. An embryological investigation of such embryos and of their embryonic membranes was undertaken on this assumption. The investigations resulting, embryological and experimental, are the subject of this paper.

A brief outline of the salient points in the development of the rabbit from the 7th to the 15th day post-coitum, in so far as they are material, is necessary for clarity. The blastocysts reach the uterus from the Fallopian tube at the end of the 3rd day post-coitum when they are still very small and at about the time when the primitive yolk-sac cavity appears. They remain free in the uterine lumen for 3 days, during which time they are expanding with increasing rapidity. The blastocysts are very thin-walled vesicles, 3–4 mm. in diameter, still surrounded by the stretched and attenuated zona pellucida by the end of the 6th day. The fluid contents of the primitive yolk-sac cavity, the only cavity so far developed, accounts for the greater part of their bulk, and it is the expansion of this cavity which is mainly responsible for their rapid increase in size. At this time they fill and expand the uterine lumen, so that each is lodged in a small spherical chamber, the implantation cavity. Early on the 7th day the zona pellucida is shed and the naked trophoblast of the lower hemisphere of the yolk-sac adheres to the uterine mucosa on the antimesometrial side. The embryonic disk is orientated mesometrially. Thus implantation is of the central type and occurs early on the 7th day when the blastocyst has attained a diameter of 3–4 mm. It expands very rapidly during the two succeeding days and is about 12 mm. in diameter on the 9th day. This expansion is brought about partly by growth of the embryonic cells but mainly by increase of the yolk-sac fluid. It results in the antimesometrial wall of the implantation cavity being stretched thin and blown out like a bubble so that this part of the uterus becomes semi-transparent. Meanwhile the embryonic entoderm has been extending around the primitive yolk-sac cavity, forming with the overlying trophoblast the bilaminar omphalopleur, and converting the cavity into the definitive yolk-sac cavity. The establishment of the bilaminar omphalopleur is completed when the entoderm has extended to the ventral pole of the yolk-sac on the 8th day. Then the yolk-sac cavity is lined throughout with a very thin endothelium of entoderm cells, whereas previously a part of its wall consisted of trophoblast only. The bilaminar omphalopleur of the ventral or antimesometrial hemisphere of the yolk-sac wall is a transitory structure which disappears by the 14th day, thus placing the yolk-sac cavity in continuity with the uterine lumen and exposing the entoderm of the vascular yolk-sac splanchnopleur forming the dorsal hemisphere. This hemisphere of the yolk-sac wall meanwhile has been invaginated within the yolk-sac cavity, owing to the growth of the embryo and the extension of the exocoele, and it now forms an absorptive yolk-sac placenta.

The coelom appears by the beginning of the 8th day and is followed immediately by the formation of the amniotic folds. These close, completing the formation of the amnion, at days post-coitum. The chorionic trophoblast of the outer walls of the amniotic folds adheres to the uterine epithelium, invading it and growing into the mouths of the uterine glands, even before the amnion closes. This meso-metrial attachment is the beginning of the allanto-chorionic placenta, which is rapidly established thereafter. Blood-islands differentiate in the area vasculosa at 8 days post-coitum, and blood first appears in the heart early on the 9th day. Presumably the heart has begun to beat at this time.

The embryological material sectioned is shown in Table 1. Complete series of sections were not prepared with the majority of the embiyos, but only a few sections through the middle of each uterine swelling were mounted. A few complete series were employed, mainly of the tame rabbit embryos. Most of the sections were stained with Ehrlich’s haematoxylin and eosin.

During 1944, attention was directed particularly to the examination of embryos at stages of development between implantation and the time of maximum mortality. This was done with a view to distinguishing abnormalities in 7–10-day embryos that might result in their subsequent death. Consequently, all the swellings were opened under saline and the blastocysts examined in the fresh state with a binocular dissecting microscope. Previously it had been the practice to fix the swellings before opening them.

It was found that in a substantial proportion of pregnant uteri examined at these stages the contents of the yolk-sac cavities of the embryos were gelatinous rather than fluid. When the uteri were opened under saline and the blastocysts exposed, these often appeared translucent, rather than transparent, and when they were punctured they did not collapse, as do normal fresh blastocysts. It was the practice to pipette a few drops of absolute alcohol on to opened blastocysts of these ages, as this rendered the embryos more clearly visible and so assisted in ageing them accurately. However, this treatment, when applied to a gelatinous blastocyst, rendered the yolk-sac contents opaque and completely obscured the embryo which lay below it on the mesometrial side. Closer examination then revealed an irregular network of white strands and trabeculae traversing the yolk-sac cavity. The earliest stage of development at which such gelatinous blastocysts were observed macroscopically was 7 days, and the latest was 12 days, but they were most frequent at 9 and 10 days. Sometimes all the blastocysts in a pair of uteri were similar in this respect, but in others a few only were affected.

Another abnormality was observed in many uteri consisting of semi-opaque opalescent strands, attached to the embryonic membranes and extending up and down the uterine lumen, but not adhering to the uterine wall. Sometimes a strand ended freely, but frequently it stretched from the membranes of one embryo to those of the next, connecting-them together. Often the membranes of all the embryos in a uterus were so connected. These strands were sufficiently solid to remain intact while the uterus was pinned out and opened under saline, and they could then be stretched slightly and tested gently with a forceps. They varied in thickness up to a maximum of approximately 3 mm. in diameter. In a few instances they were tubular and filled with blood, though there was little or none free in the uterine lumen. The presence of the strands was associated commonly with deflation of the embryonic membranes. This deflation of the embryonic vesicles resulted in change in shape of the uterine swellings containing them, which, from being almost spherical and sharply delimited from the adjoining regions of the uterus, became elongated and tapered down to the diameter of the uterus between the implantation sites. Such swellings were often irregular in size and tended to appear dull and opaque, lacking the limpid translucence antimesometrially of swellings containing healthy embryos. It was often possible, with experience, to predict correctly from these indications that strands would, or would not, be found when the uterus was opened. The earliest stage at which strands were observed was 8 days and the latest was 12 days. They were most frequent at 8 and 9 days. They occurred at the same stages as the gelatinous blastocysts, and both abnormalities were frequently encountered together in the same uterus.

Some of the embryos of 8 or 9 days which displayed one or other or both of these abnormalities appeared to be otherwise normal and healthy, but in other instances no embryo could be distinguished macroscopically. The majority of uteri at 10, 11 or 12 days which contained strands or gelatinous material had retrogressing embryos. There are therefore grounds for suggesting that the gelatinous blastocysts and strands, though tending to occur at slightly earlier stages, are related to the mortality that attains a maximum on the 11th and 12th days.

Although many examples of both these abnormalities were observed in series 3–9 inclusive, complete data are available only for series 7 and 8. These are summarized in Table 2.

The shift of maximum numbers from the group of normal litters on the 7th day to the group showing one or both abnormalities on the 8th and 9th days, to those showing both regression and abnormalities on the 10th day, and finally to those showing regression only on the 11th and 12th days is significant, and clearly suggests that the abnormalities are a phase in the process of regression. It is equally possible, however, to interpret these data as showing that the gelatinous condition of the yolk-sac contents and the formation of strands are either a change that precedes the death of the embryos at 10, 11 or 12 days, or that they are characteristic of embryos which have already died and are regressing.

Microscopic examination of sections of blastocysts revealed that the gelatinous condition of the yolk-sac contents was due to the presence of a branching network of fine fibrils (Pl. 11) which stained faintly with eosin after Bouin’s fluid fixation. Sometimes they formed a very open-meshed reticulum throughout the yolk-sac. Sometimes a number of spherical foci could be discerned scattered unevenly in the yolk-sac cavity, each consisting of a denser central mass of fibres with irregularly branching fibres radiating from it (Pl. 11, fig. 2). Sometimes what appeared to be an autolysing cell or the pycnotic remains of a nucleus could be discerned in the centre of such foci, in others it was represented by a small cavity, and in many no central structure could be discerned. Sometimes the reticulum was compacted into a dense felting of fibres in which the individual ones could scarcely be distinguished (Pl. 11, fig. 1). Such felted masses were usually confined to the periphery of the yolk-sac cavity, frequently in the antimesometrial hemisphere, and the rest of the cavity was occupied by a much more open reticulum. Sometimes strands or trabeculae of tangled or felted fibres stretched across the cavity, and it was plainly these which were visible macroscopically.

The individual fibres varied in diameter, the thickest being of the order of 0 · 5 μ. They branched irregularly and the free extremities tapered gradually. No visible structure could be discerned in them. They were visible after treatment with several different fixatives, but Bouin’s fluid gave the best results. They were difficult to stain, the most satisfactory results being obtained with eosin, and with water blue in Passini’s stain. They were found in smears of yolk-sac content, as well as in sections, both air-dried and fixed by several methods, and sometimes reacted positively, at other times negatively, to vigorous Gram staining.

The distribution of the reticulum was remarkable. It was present in the majority of embryos of 8 and 9 days examined microscopically, and in some of those of 7 and 10 days. It was not found in any blastocysts of 7 days or earlier which were still surrounded by the zona pellucida. The few 8- and 9-day embryos in which it was not found were badly preserved, and so it was not possible to be sure that it was absent. It varied very greatly in the degree of development and in many was slight and localized, consisting only of a few fibres near the periphery of the yolk-sac cavity. Clearly these would not have been classed as gelatinous blastocysts on macroscopic examination. Within the embryo it was confined to the yolk-sac cavity and the gut cavity, in which it could be seen in a few instances. Since the mid-gut opens freely into the yolk-sac cavity at these stages that is not surprising. It was never observed in the exocoele or amniotic cavity, although these are only separated from the yolk-sac cavity by thin membranes. It was sometimes present also in the uterine lumen. It was found also, at slightly later stages, in io- and n-day embryos in the cavities or crypts of the placental region formed from the enlarged uterine glands after their necks had been invaded and blocked by the trophoblast.

Sections of strands joining blastocysts showed that these consisted of felted cords of similar fibres, crumpled remains of embryonic membranes, cell debris, leucocytes and red blood corpuscles of maternal origin, compacted together by the pressure of the uterine walls. Their structure was consistent with the view that they represent a later stage in the evolution of gelatinous blastocysts.

The fibrous network never appeared to have penetrated into the tissues, either embryonic or maternal, being always confined to the cavities named.

Examination of blastocysts of tame rabbits at corresponding stages of development did not reveal any that could be identified macroscopically as gelatinous blastocysts, nor were any strands connecting the blastocysts found. Microscopic examination of sections showed that the yolk-sac cavities of many of them were completely free from the fibrous network. It was present in others but was much less developed than in the majority of the wild embryos. Only in a few instances did the network attain the density and extent met with in unmistakable gelatinous blastocysts of wild animals. The fibres, when present, were so similar in structure and staining affinities to those in wild material that the identity of the two need not be questioned.

The suggestion that the fibrous reticulum was an artefact produced by the histological technique was inconsistent with facts, since the fibres were present both in sections and smears, and after a wide variety of fixatives had been employed. It was obvious that they were not tissue fibres formed by fibroblastic cells, since no such cells are present in the yolk-sac cavity. Two possibilities remained : first, that the reticulum was the mycelium of a micro-organism living in the yolk-sac cavity and the uterine lumen; secondly, that it was a network of organic fibres formed by some biochemical process without the intervention of fibroblastic cells.

The morphological resemblance to the mycelium of an actinomycete when grown in culture media was so close that the possibility of the reticulum being an invasive organism belonging to this group could not be eliminated microscopically. Since the ultimate objective was to elucidate the course of the prenatal mortality, the possibility of the presence of a causative organism could not be dismissed lightly and no other structure that could be interpreted as such had been observed in the uteri. Consequently it was decided to undertake culture experiments to determine if a micro-organism was involved.

Cultures were made from the yolk-sac contents of embryos from eighteen wild rabbits 6–13 days post-coitum. Several of the embryos used were typical gelatinous blastocysts, and the presence of the reticulum in the others was confirmed either by smear preparations from the same embryo or by sections of other embryos from the same uteri. Both plate and stab cultures were prepared, using a variety of media, and were incubated aerobically and anaerobically. Sterile conditions could not be obtained, since post-mortem material was employed, the animals having been gutted by the trappers and having been dead several hours before they were received. Therefore a variety of the usual contaminants appeared in the cultures, but no one type of micro-organism was present consistently. An actinomycete appeared in two cultures from one rabbit soon after the experiments were begun. It belonged to the group of Micromonospora and was microaerophilic when first obtained, but it rapidly became aerobic on subculturing. This positive result of the culture experiments, taken in conjunction with the close morphological resemblance of the reticulum in sections and smears, encouraged further attempts to isolate a causative organism, but it proved impossible to repeat the result.

Since the reticulum was known to be present in the yolk-sac content of some tame rabbit embryos, it was decided to attempt to obtain cultures from them under aseptic conditions. Cultures were made from the yolk-sac content of several embryos from each of eight tame rabbits 8 or 9 days pregnant. Aseptic precautions were taken and upwards of a hundred stab cultures made on blood agar and semi-solid glucose nutrient agar which normally prove suitable for actinomycetes. These were incubated both aerobically and anaerobically. The aseptic precautions proved adequate, and accidental contaminants only developed in less than 1 % of the cultures, a satisfactorily small proportion. All the other cultures were sterile. No actinomycete was obtained in any of the cultures, although the presence of the reticulum was demonstrated in some of the smears of each animal used. The possibility of an actinomycete being present and unculturable with the technique employed was negligible. It was concluded therefore that the actinomycete originally obtained from the wild material was an accidental contaminant, and the reticulum observed in the yolk-sac was not a micro-organism culturable by the usual techniques. These results not only disposed of the theory that an actinomycete was present in the yolk-sac fluid but they provided very strong evidence that this fluid is bacteriologically sterile.

It appeared probable that the network in the yolk-sac cavity was composed of protein fibres, possibly fibrin, since it was neither an artefact nor a micro-organism. Consequently clots of fibrin were prepared from citrated rabbit plasma by the addition of calcium. These were fixed in Bouin’s fluid, embedded in paraffin, sectioned and stained with eosin, using the same technique as was employed for the embryos. The fibrin network prepared in this way closely resembled the reticulum in the yolk-sac both in microscopic appearance and in histochemical properties.

Experiments were planned to test whether fibrinogen was present normally in the yolk-sac fluid of embryos of tame rabbits. Since thrombin is specific in its action in converting fibrinogen into fibrin it was decided to employ it. Rabbit embryos of approximately 9 days post-coitum were used, as the content of the yolk-sac is then at a maximum. The embryos were obtained either from rabbits killed for the purpose or by unilateral hysterectomy under ether anaesthesia in the course of experiments designed for another purpose. The yolk-sac fluid was withdrawn separately from each embryo by means of a hypodermic syringe, the needle being inserted through the uterine wall antimesometrially. The contents was withdrawn into 0· 1 ml. of 3 % sodium citrate solution in 4 graduated 1 ml. syringe, and the citrated yolk-sac fluid measured and transferred to a Wasserman tube. Quantities of up to 0· 75 ml. of yolk-sac fluid were obtained from each embryo. Care was taken to avoid contamination of the fluid with blood, and any samples in which this occurred were rejected. It was impracticable to ensure by this means that the fluid contents of the yolk-sac so withdrawn was not contaminated with traces of other embryonic fluids from either the exocoele or the amnion, if this happened to be already formed, or with tissue exudation from the uterine lumen. It is certain that the amount of such contamination, if any, was small because the amount of fluid obtained was commensurate with the size of the yolk-sac cavity and because the deflation of the uterine swelling was observed as the fluid was withdrawn. A solution of human thrombin in saline, containing 2 units per 1 ml., was added, either to the undiluted citrated yolk-sac fluid, or to the fluid after it had been diluted with an equal quantity of saline. The quantity of thrombin solution added was such that the final concentration was 1 unit per 1 ml. This resulted in the formation of a clot in from 30 to 60 sec. at room temperature in all cases. The experiments are summarized in Table 3. The samples of yolk-sac fluid from two or more embryos from the same rabbit were combined in some instances.

Serial dilutions with saline were made containing of yolk-sac fluid and were clotted by the addition of equal quantities of the thrombin solution. These were compared visually with clots formed in similar serial dilutions of rabbit-blood plasma. It was possible to estimate by this means that the concentration of fibrinogen in the yolk-sac fluid was of the order of 50% of that in the blood plasma. The quantities of yolk-sac fluid that could be obtained from the embryos of one fitter would be ample, on this basis, for the estimation of nitrogen in the clot by microanalyses. Accordingly clots were prepared from the combined yolk-sac fluid of several embryos of the one litter. The citrated fluid was diluted with equal quantities of saline, so as to reduce the density of the clot, and thrombin solution added. After the clot had formed it was frozen in a refrigerator and subsequently thawed at room temperature. This caused the clot to contract to a comparatively small size. The residue was filtered off on a fine sinter-glass filter and the clot washed several times with saline and then with distilled water. The wet clot was dried in a vacuum. The filtrate and washings were combined and the nitrogen estimated both in the residue and in the dried clot. The results are given in Table 4, the estimate of nitrogen in the residue being corrected for the small amount added in the thrombin solution. Estimates of nitrogen in a single sample of rabbit-blood plasma treated in the same way are given for comparison. The sample of plasma was taken from a rabbit in late pregnancy (29 days). Another sample, not included in the table, was taken from a rabbit in early pregnancy (⁠days) and yielded 5· 78 mg. of dried clot per ml. of plasma, an amount which agrees closely with that given in the table. Weight for weight of dried clot the mean concentration of fibrin in the yolk-sac fluid is 37% of that in the sample of plasma, and the corresponding figure based on the nitrogen content of the clot is 33 %. The difference in the concentration of fibrin in the yolk-sac fluids of the embryos of E102 and E103 respectively is well beyond the experimental error of the determinations and is clearly significant. The differences in the percentages of nitrogen in the dried clots obtained from yolk-sac fluid may be accounted for by a reasonable margin of error in the determinations, but the difference between these and the percentage of nitrogen in the plasma clot appears significant, but may be due to inefficiency in the technique of separation of the clot from the residue. It can scarcely be due to inequality in the drying of the clots, since all three of the determinations on yolk-sac fluid agree reasonably well, while that on plasma is in excess of them. It would be necessary to assume that none of the clots from yolk-sac fluid had been completely dried, and that approximately the same amount of moisture had been left in each to account for the difference on this basis. The mean quantity of nitrogen in the residue of the yolk-sac fluid, after the removal of the clot, is 57% of that in the serum. Further inquiry is necessary before any suggestion can be offered as to the nature of the nitrogen-containing compounds in the residue of the yolk-sac fluid. It may be noted, however, that the proportion of fibrin-nitrogen to non-fibrin-nitrogen in the yolk-sac fluid and in the plasma is different.

The results reported in this paper demonstrate that fibrinogen is a normal component of the yolk-sac fluid of rabbit embryos and that at 9 days post-coitum it is there in quantities of 30–40% of that in the blood plasma. Since the appearance of fibrinogen in the yolk-sac cavity precedes and accompanies the establishment of the embryonic vascular system it is highly probable that it is of maternal origin. It would be very difficult to believe that it could be produced in such quantities by the embryo at so early a stage of development. If this assumption is justified then the fibrinogen must be absorbed through the yolk-sac wall. The trophoblast of the yolk-sac wall at least must be readily permeable to it. At this time the trophoblast of the antimesometrial hemisphere of the yolk-sac adheres to and is actively invading and destroying the uterine epithelium, presumably passing the products of the destruction of the maternal tissues into the yolk-sac cavity. The results indicate that among the proteins maternal fibrinogen at least can traverse the trophoblast as such. It is not clear whether the fibrinogen can also pass through the entoderm of the yolk-sac wall for it has been found in the cavity soon after the establishment of the bilaminar omphalopleur has been completed. It is possible, therefore, that all the fibrinogen enters the yolk-sac cavity before this has been completely enclosed by entoderm.

The nutriment derived by the mammalian embryo in the uterus from the mother is divided conventionally into histiotrophe and haemotrophe. The histiotrophe consists of the secretions of the tubal and uterine glands, transudation, extravasated blood, cell debris and other products of destruction of the maternal endometrium by the trophoblast. The haemotrophe consists of nutritive materials carried in the maternal blood and transferred from the maternal to the foetal circulations in the placenta. Therefore until the foetal circulation is established and the placenta is functional the nutrition of the embryo is entirely histiotrophic. After the placenta has begun to function it is predominantly haemotrophic. Histiotrophic nutrition may continue after the establishment of haemotrophic nutrition but is of progressively declining importance.

The stage of development at which haemotrophic nutrition begins varies in different groups of mammals according to the type of implantation and placentation. It tends to be late in species with central implantation, in which the blastocyst remains free in the uterine lumen for a comparatively long time and reaches a relatively large size before it becomes attached to the uterine wall. The rabbit is one of these, and in it haemotrophic nutrition cannot begin until late on the 9th or early on the 10th day after copulation. The efficiency of the placenta as an organ of haemotrophic nutrition will depend on the barrier presented by the intervening tissues to the transfer of nutritive materials from the maternal to the foetal circulations, which are never in direct contact. Primitively six tissues are involved in this barrier; the maternal capillary endothelium, connective tissue and uterine epithelium and the foetal trophoblast, mesenchyme and capillary endothelium. The classification of placentae suggested by Grosser (1927) and commonly employed is based on the extent of the reduction in these tissue layers. The greatest possible efficiency is attained in the haemo-endothelial type of placenta in which only the foetal endothelium intervenes between the maternal and foetal blood. This type of placentation was recognized first by Mossman (1926) in the rabbit. It results from the destruction of all the intervening maternal tissues by the placental trophoblast, so that the maternal blood circulates through lacunae in the spongy trophoblast, bathing its surface directly, and in the subsequent thinning out and disappearance of the trophoblast and mesenchyme covering the foetal capillaries at the points where they adjoin these lacunae. According to Mossman (1926) this condition is achieved in the rabbit on the 13th or 14th day. Thus the maximum efficiency of the haemotrophic method of nutrition is attained at the time when the bilaminar omphalopleur disappears and the entoderm of the vascular splanchnic wall of the yolk-sac is exposed to the uterine lumen.

The problem of placental permeability has been the subject of many researches, which are admirably summarized and discussed by Needham (1931) and need not be recapitulated. Extraordinarily little is known, however, regarding the passage of substances into the embryo at earlier stages, when the nutrition is exclusively histiotrophic. Evidently fibrinogen can enter at this stage. Much larger quantities of nitrogen also are present in the yolk-sac fluid in some other form than fibrinogen, but the nature and derivation of these nitrogen-containing compounds awaits elucidation. Yet if fibrinogen can enter as such it is quite probable that other maternal proteins can do so as well. Presumably the fibrinogen is utilized by the embryo, possibly being digested in the yolk-sac cavity and absorbed in soluble form by the entoderm.

The increase in size of the blastocyst from the 7th to the 9th day is very rapid and is due mainly to increase in volume of the yolk-sac fluid. The blastocyst is not more than 4 mm. in diameter at the time of attachment on the 7th day and attains a size of approximately 12 mm. in diameter on the 9th day, when as much as 0· 75 ml. of fluid can be aspirated from the yolk-sac. The increase in volume during the two intervening days therefore is of the order of twenty times. Throughout this period of rapid expansion the blastocyst remains almost spherical, being only slightly elongated in the longitudinal axis of the uterus and somewhat flattened on the mesometrial side, presumably these being respectively the directions of least and greatest resistance to expansion by the uterine tissues. How a sufficient hydrostatic pressure is generated to expand the uterine chamber and stretch thin its muscular wall is difficult to imagine. Yet this occurs within the vesicular blastocyst bounded only by a very thin cellular membrane that is highly permeable to protein.

It is not practicable to aspirate fluid from the cavities of the amnion and exocoele, and to test it with thrombin for the presence of fibrinogen, at such early stages of development of the embryo when these cavities are newly formed and still very small. Microscopic examination of sections of embryos of wild rabbits in which the yolk-sac contents were clotted has shown that fibrin is always absent from the amniotic cavity and the exocoele. It must be concluded that their walls are impermeable to fibrinogen. This difference in permeability to protein of the omphalopleur and the other embryonic membranes is interesting. It may be suggested that the mesoderm which encloses both the exocoele and the amniotic cavity, but does not enter into the bilaminar omphalopleur, provides the barrier.

The conversion of the fibrinogen into a clot of fibrin in the yolk-sac cavity must be regarded as an abnormality, since it had not occurred in many obviously healthy tame rabbit embryos examined. The problem of when the clot forms in the yolk-sac cavities of embryos of wild rabbits is difficult to solve. Some of the wild rabbits in which gelatinous blastocysts were found were examined very soon after death, while the bodies were still quite warm. Clots only form in the yolk-sac cavities of healthy tame rabbit embryos when these are left in the gutted carcass of the mother for many hours, and even then they tend to be much less apparent. These facts indicate that clotting of the yolk-sac fluid often occurs in wild rabbits before the death of the mother. Moreover, the variability often experienced in the degree of development of the clot from embryo to embryo in the same uterus of a wild rabbit bears out this conclusion. Whether or not clotting occurs before the death of the embryos, when this precedes the death of the mother, is still more obscure. Yet this problem is important in relation to the prenatal mortality, for if, as might seem the simplest assumption, it only occurs after the death of the embryos, then all the gelatinous blastocysts observed in the wild material were already dead and would represent a heavy prenatal mortality, occurring at 8, 9 and 10 days, not previously recognized, in addition to that which undoubtedly occurs about 11 and 12 days. If, on the other hand, the clotting precedes the death of the embryos, then the gelatinous blastocysts observed can be regarded as abnormalities in development preliminary to the subsequent death and regression of the embryos on the 11th and 12th days, and it is not necessary to assume a mortality additional to that already recognized. It would appear that this problem can be solved only by experimental means, and investigations are in progress along such lines. Preliminary results indicate that the embryo can survive clotting of the yolk-sac contents at least for a short time. It follows that should clotting prove to be preliminary to the death of the embryos around the 11th and 12th days, then knowledge of the factors which bring it about might throw much light on the cause of this important fraction of the prenatal mortality in wild rabbits.

Finally, it will be apparent that knowledge of the nutrition of the mammalian embryo during the histiotrophic phase is extraordinarily scanty. Little is known regarding the composition of the histiotrophe or the physiological factors governing its production, the manner of its passage through the limiting membrane of the embryo, and the way in which it is subsequently utilized. It is hoped that this paper will serve to draw attention to these problems and to show that the yolk-sac fluid of the rabbit embryo provides convenient material for physiological and biochemical investigations of them. There is here a wide field for fundamental research of great potential significance in relation to prenatal health.

We are grateful to Prof. J. P. Hill, F.R.S., and to Prof. D. M. S. Watson, F.R.S., for their advice and for the interest they have taken in the work. We are greatly indebted to Miss Dagne Erikson for undertaking the bacteriological work summarized in §4 and for the interest she has taken in the problem throughout. We are also indebted to W. T. Rowlands, Esq., for providing us with bacteriological facilities in his laboratory and to H. W. Smith, Esq., both of this College, for advice and assistance in making the initial cultures. We are indebted to Dr R. A. Kekwick for his advice on the haematological work and to the Medical Research Council’s Serum Unit at the Lister Institute for the provision of the human thrombin employed. We wish to thank also Prof. E. D. Hughes, of the Department of Chemistry and Dr W. McLean of the Department of Agricultural Chemistry of this College for their assistance with the micro-analyses of the yolk-sac content. We are indebted to our colleagues Miss Megan Henderson and Miss Patricia Allen, who are engaged on other aspects of the problem, for their assistance, and also to Mr R. A. Lansdowne for skilled technical assistance and for micro-photographs. The work has been financed by a grant from the Agricultural Research Council for investigations on the reproduction of wild rabbits, for which we wish to express our thanks.

Fig. 1. Fibrin reticulum in the yolk-sac cavity of an embryo (6/271, 1R) of a wild rabbit, circa early 8 days post-coitum. The reticulum is felted into a dense mass at the bottom of the figure, × 680.

Fig. 2. Fibrin reticulum in the yolk-sac cavity of an embryo (6/129, isolated blastocyst) of a wild rabbit, circa 8 days post-coitum. Two foci are seen with fibrillae radiating irregularly from them, × 680.