Gonad Grafts in Embryonic Chicks and Their Relation to Sexual Differentiation

The present series of experiments were designed with a view to substantiate, if possible, the claim made by Minoura in 1921, that the effect of grafting ovary or testis on to the chorio-allantoic membrane of the developing chick embryo is the production, in the host, of modifications in the character of the developing gonads and the associated ducts towards an intersexual type. Such modified chicks were referred to as intersexes and their causation explained by postulating, that in effect, the graft elaborates a specific secretion or hormone. This substance obtaining access to the tissues of the host through the common circulation acts, in the case of a testis graft on female chick, on the developing urogenital organs and produces a modification in the development of these organs approximating them to the male type A similar explanation served in the case of male embryos, whose gonads and accessory sexual ducts were modified in the female direction through the activity of a specific secretion from the ovarian grafts.

As is well known in cattle the female in twins of the opposite sex is frequently sterile. Willier, who has made a complete histological study of the gonads from these sterile females, has shown that the gonad is primarily an ovary which has been modified to a more or less greater extent in the direction of a testis. The sexual ducts are male-like and the external genitalia, though usually female in type, may be modified also in the direction of the male.

The theory put forward by Lillie to explain these cases is briefly, that the sex hormones elaborated by the gonad of the male co-twin obtain access to the tissues of the female embryo through the anastomosing blood vessels of the fœtal membranes and produce a modification in the development of the sexual apparatus of the female in the male direction. Such a theory of course does not readily lend itself to experimental verification.

For the study of growth and differentiation of grafted tissues many workers have shown that the embryonic membranes of the developing chick offer exceptional facilities owing to the rapid vascularisation and incorporation of the graft.

This was the experimental method adopted by Minoura in order to reproduce artificially in the chick those abnormalities which occur naturally in the case of the female co-twin in cattle. A piece of testis or ovary was grafted on to the chorio-allantoic membrane of the developing chick. In favourable cases rapid vascular connection between the graft and the host was established. Any specific secretion of the implanted gonad was, therefore, through the common blood supply, accessible to the tissues of the host.

In the course of the research, however, one difficulty presented itself. To quote from Minoura, “The interpretation of the intersexual individuals is rendered very difficult, since, we do not know with certainty the sex of the embryos on which the graft was implanted.” To obviate the difficulty in the present series of experiments, birds with a single sex-linked character difference were crossed and the eggs obtained were used as experimental material. This renders the determination of the sex of the embryo by an external examination an easy matter.

The facts that, primarily, the work of Minoura provides the only experimental evidence substantiating the theory of Lillie on the free-martin in cattle and therefore a confirmation of his results is required ; and secondly, that in the present series improvements in technique have been devised, are held to offer sufficient justification for a repetition of this work.

As previously noted, the eggs used to provide the experimental material for this research were obtained from a sex-linked cross. The particular cross used was Light Sussex females × Rhode Island Red males, and the sex-linked character involved was the colour of the plumage. The down colour of the male chick is white and that of the female reddish-yellow. Sex can be determined by the down colour of the embryo as early as the tenth day of incubation. The chicks used to provide the tissues for grafting were, for the most part, closely related to the experimental ones and had been hatched in the laboratory. Some of the older birds used as a source of material for grafting were, however, unrelated to the embryos on which such tissues were grafted.

After incubating for seven days the egg for operation was removed from the incubator and candled. At this stage in the development of the embryo the position of the main blood vessels was clearly discernible. A mark was made on the shell over the intersection of two large vessels and the surface of the shell in this region well sterilised by rubbing with a piece of cotton-wool soaked in 70 per cent, alcohol. The removal of a piece of the shell was found to be done most expeditiously and without injury to the underlying shell membrane by the use of a small circular saw, 1 cm. in diameter, attached to a dental drill worked by a small electric motor. The fragments of the shell were brushed off and the egg placed on a pad of cotton-wool over the candling box to minimise the cooling during the progress of the operation. A flap of the shell membrane was carefully lifted up and the piece of tissue to be grafted was ‘placed on the chorio-allantoic membrane in the intersection between two of the larger blood vessels. It should be noted that no intentional injury to the embryonic membranes was attempted when grafting, but the tissue was placed on to the membrane as carefully as possible. The shell membrane was replaced and the opening in the shell sealed with a piece of sterilised paper previously immersed in fresh egg albumen. The egg was then returned to the incubator and not rotated for a period of eighteen hours. At the end of this time the egg was turned as usual twice daily.

The chick used to supply the grafting material was killed by decapitation, skinned, and placed in a sterilised petrie dish when the abdominal cavity was opened up ; the dish was then covered and set aside until the egg was ready to receive the tissue to be grafted. The usual precaution of sterilising all instruments and dishes used during the operation was taken.

Out of each batch of eggs many were incubated normally and served as controls. In others the square piece of shell was removed and the shell membrane lifted up as if in preparation for the implantation of a graft. The membrane was then replaced in position and the opening sealed as described previously.

The seventh day from the beginning of incubation of the egg was chosen as the most suitable age to implant tissue on to the chorio-allantoic membrane of the developing embryo. At this stage the ectoderm of this membrane is supplied with a well-developed vascular network, and large blood vessels permeate the mesoderm ensuring rapid vascularisation and incorporation of the graft. The developing gonads at this stage have reached the end of the period of sexual indifference and sexual differentiation has begun. From this time onwards the epithelial cords of the first proliferation begin to degenerate in the female—in the male they are destined to give rise to the spermatic tubules. The conditions for the production of successful grafts at earlier stages in the development of the chick are unfavourable, owing to the fact that the blood vessels of the fœtal membranes are not sufficiently well developed to ensure rapid vascularisation of the piece of grafted tissue. In this series of experiments no embryos were used as hosts for grafted tissue after the ninth day of incubation.

The age of the grafted gonad tissue varied from that of 14th-day embryos up to tissues from cockerels or pullets ten weeks old. The size of the piece of grafted tissue was usually one quarter to a whole embryonic gonad, and in the case of the older tissue about 2 mm. square.

Altogether 540 embryos were used as hosts for the reception of grafted tissue. In the majority of cases the graft consisted of ovary or testis. Pieces of thyroid, liver, spleen, heart, lens, Wolffian body, adrenal, kidney, pancreas, and gall bladder were also used as grafting material ; these latter were only used to ascertain whether the above-mentioned tissues had any effect on the development of the gonads or the urogenital ducts of the chick embryo on which they were grafted. The effect on the chick embryo as a whole, or on any part not connected with the sexual apparatus, has not been taken into consideration in the scope of this paper.

Most of the eggs were removed from the incubator for examination at the seventeenth day of incubation ; a few were opened on the fourteenth, fifteenth, and sixteenth day. The longest period that the graft remained on the chorio-allantoic membrane was ten days, and the shortest period seven days.

At the end of the necessary period of incubation the eggs were removed from the incubator and the paper sealing the opening in the shell scraped off. The shell was then carefully chipped away from the opening until half the egg was uncovered. It was found that many of the eggs had become infected and had failed to develop. In the living embryos removal of the shell membrane exposed to view the position of the graft. In all cases a large white or pinkish growth, which varied greatly in size in different embryos—in some it was nearly 1 cm. in diameter—was seen, situated on the membrane at the point of implantation of the grafted tissue. The presence of this growth was demonstrated not only in the case where grafts were made but also in those embryos in which the operative technique was carried out up to the point of lifting the shell membrane, but where the egg Was re-sealed without an actual graft being made. The grafted tissue may sometimes be seen situated in the centre of the mass, but it was found on histological examination of sections of the growths that in many cases grafted tissue could not be located. It is therefore evident that the presence of the growth does not indicate the presence of grafted tissue, and also that even when the grafted tissue can be located in the growth, the relative size of the growth is not an indication of the rate of growth in the various grafts, but really measures the amount of damage done to the chorio-allantoic membrane at the time of grafting.

Histologically the growth (fig. 6) consisted of a mass of inflammatory tissue, myxomatous mesenchyme with intricate network of capillaries, the walls of which showed layers of proliferating endothelial cells. The darker rounded areas (fig. 6) were infoldings of the surface epithelium which have been eût across tangentially. The dark central zone from which the capillaries are seen to radiate outwards is probably the zone of origin of the new-formed blood vessels. The spaces between the meshes formed by the anastomosing protoplasmic processes of the mesenchyme cells were normally filled with inflammatory exudate and a few polymorph cells.

The histology of the growth confirms the view that the pinkish mass of tissue results from an injury to the fœtal membrane at the time of grafting. The growth was dissected off the membrane and removed to a tube of fixative. (The growths and gonads obtained during the course of the experiments were fixed in Allen’s modification of Bouin, and after sectioning the tissues were stained by Heidenhain’s iron-alum haematoxylin method). The egg content was poured out into a dish and a careful search of the chorio-allantoic membrane made. It was noted in several instances that where the presence of grafted material could not be clearly demonstrated in the centre of the growth, careful search of the chorio-allantoic membrane revealed the graft attached to the membrane at a spot some distance from the pinkish mass of the growth. Many of these grafts were quite satisfactory. Vascular connections to the graft were noticed and microscopical examination proved the presence of healthy grafted tissue.

The embryo was then dissected, the urogenital system examined macroscopically, and the gonads fixed for histological study.

Before proceeding to the description of the experimental embryos obtained and the results of the histological study of the growths and gonads, the actual experimental data will be given.

In the fowl the sex-ratio normally reveals a slight excess of females. The number of males per thousand females as ascertained by various authors is as follows : 806 (Field), 947 (Darwin), 944 (Pearl). In these experiments the sex-ratio in the chicks surviving after operation, as shown by the tables, may be taken as normal. Thus the power of a chick to survive the operation is independent of the sex of the individual. Where gonad tissue was grafted, either testis or ovary, no significant difference in rate of mortality was shown in the two sexes.

The gonads are formed early in embryonic life from a thickening of the cœlomic epithelium in the angle between the mesentery and the mesonephros. They soon assume the form of ridges lying longitudinally along the inner face of the Wolffian bodies. The formation of the genital ridge results from the proliferation of the sub-epithelial mesenchyme. The germinal epithelium is composed of tall cylindrical cells between which are found oval or rounded cells—the primordial germ cells. After this point the germinal epithelium gives rise to local thickenings. These grow rapidly, push into the substroma, and gradually lose connection with the germinal epithelium giving rise to a network of sub-epithelial cellular cords. This stage marks the end of the period of sexual indifference, for up to the seventh day (according to Firket) it is not possible to determine whether the embryonic gonad will develop into ovary or testis. The epithelial cords abort in the female and are replaced by another proliferation at the eleventh day of epithelial cords from which the oocytes of the functional ovary are derived. In the male, however, there is only one proliferation of sexual cords from the germinal epithelium. These enlarge* and anastomose to form a sub-epithelial network of tubules from which ultimately the mature spermatic tubules are derived. In the right gonad of the female the medullary cords abort and are not replaced by a second proliferation of epithelial cords as in the case of the left gonad. The microscopical changes taking place in the early period in the development of the gonad become apparent macroscopically at approximately the eighth day of incubation.

Plate I. gives critical stages in the development of the urogenital organs of the embryonic chick. In fig. 1 a stage of sexual indifference is shown. The gonads are thin elongated organs situated anteriorly on the median ventral surface of the Wolffian body. Wolffian and Müllerian ducts are present on both sides. Further development of the uro-genital system marks the end of the period of sexual indifference. From now on the gonads in the two sexes show considerable variation in size and shape. In the female the left ovary increases in size and becomes somewhat flattened. The right ovary regresses as development proceeds, until at the time of hatching a mere trace of the gonad is usually to be seen, lying obliquely across the anterio-ventral surface of the disappearing Wolffian body. The cases illustrated are from typical embryos ; there is, however, comparatively wide variation in the rate of regression of the right gonad. As previously described, no secondary proliferation of sexual cords takes place in the right ovary of the female, and the abortion of the primary sexual cords accounts for the regression of this organ. Some growth in size of the right gonad takes place during the second week of incubation.

In the male, the gonads develop more or less equally, though the right testis is frequently somewhat smaller than the left. The testes do not become flattened as in the case of the ovary. It is worthy of note that, as development proceeds, the anterior end of the right testis tends to bend to the right and assume an oblique position with regard to the Wolffian body in the same way as the right ovary. The change in position of the degenerating right ovary and of the anterior half of the right testis may be due to a more rapid regression of the right Wolffian body in both sexes, caused possibly by differential blood supply to the urogenital organs on both sides of the body.

The Müllerian ducts in the male embryo begin to degenerate usually about the twelfth day of incubation (fig. 3), and by the fourteenth day completely disappear. In the female, the left Müllerian duct increases in size and at the twelfth day signs of a posterior swelling are visible. A slight increase in diameter of the right Müllerian duct at its posterior end is also noticeable. Soon after, degeneration of the right Müllerian duct from its anterior end takes place. The duct, however, does not completely disappear, the posterior portion still being apparent at the time of hatching. Finlay gives in his paper a figure showing that in exceptional cases this fragment retains the power of growth. The left Müllerian duct continues to develop and becomes a functional oviduct.

From the development of the urogenital system it is seen that the sex of an embryo can be determined macroscopically by examination of the gonads from at least the eighth day of incubation by: (1) the difference in shape between the gonads in the two sexes ; (2) variation in size between the right and left ovary in the female.

As development proceeds, further sexual characteristics become apparent, owing to the complete disappearance of the Müllerian ducts in the male and the partial disappearance of the right Müllerian duct in the female.

The presence of Wolffian ducts occurs in both sexes throughout embryonic life. In the female, they are quite prominent at the time of hatching, though possibly not quite as much as in the male. The difference in diameter in the ducts in the two sexes is too small, however, to be used as a distinguishing sexual characteristic in these experiments.

The great majority of the embryos were examined at the seventeenth day of incubation, when the sexual characteristics of the two sexes were very well marked. As previously stated, the sex of any of the experimentally obtained embryos was identified without any possible doubt by an external examination of the plumage of the embryo. In the female the plumage colour of the chick was reddish-yellow, and in the male white. It was found on internal examination of the urogenital organs and accessory ducts of the experimental embryos, no matter whether gonads or any other tissue had been implanted, that they showed no deviation from the normal structure peculiar to the sex as identified by an examination of the plumage, with two exceptions to be described.

Of the two abnormal forms obtained, one was a control embryo and the other had received a gall-bladder graft. Externally the chicks were structurally abnormal. They were cross-beaked and possessed no osseous formation at the top of the skull. Judging by the plumage colour they are females and the internal examination confirmed this. The gonads in both the chicks were normal, but a peculiarity of the Müllerian ducts was exhibited in that the right Müllerian duct had attained as high a stage of development as the left. Those who use incubators extensively know that such abnormalities are somewhat frequently met with, and result from a rise of temperature a few degrees above the optimum due to some error in the regulating mechanism.

As a general rule the appearance of the organs of the experimental chicks is unhealthy. The liver is green instead of reddish-brown, and the green-bile pigment is found to colour even the Wolffian body. Apart from the two abnormal forms described above no abnormalities of the urogenital system were produced in the presence of grafted tissue, whether the material used for grafting had been testis, ovary, or any other of the tissues mentioned.

It was found that on opening the egg at the seventeenth day of incubation for examination a pinkish mass of tissue at the opening of the shell was disclosed. In many cases this contained no evidence of grafted tissue and its histological features have been previously described. In some cases the presence of a graft was detected,, attached to the embryonic membrane and well vascularised, at some distance from the mass of granulating tissue. The occurrence of such grafts was to be expected from the fact that, when grafting, the material was just placed on the embryonic membranes, and no attempt was made to localise the graft by breaking the external layer of ectoderm and embedding it in the mesenchyme.

The grafts are found embedded in the mesenchyme underlying the ectoderm of the embryonic membrane. They show a continuous range from the presence of well-developed and healthy grafted tissue, with little or no areas of degeneration, passing through a stage in which the healthy grafted tissue is restricted to its peripheral areas, the centre being occupied by a degenerating mass of tissue, to that shown in fig. 7 in which no healthy grafted tissue can be discerned.

The graft was normally found situated in the mesenchyme below the ectoderm of the chorion. Surrounding the graft there was a sheath of connective tissue corresponding to the normal tunica albuginea of the testis. Just outside this was a vascular layer composed of numerous blood vessels and capillaries each of which was the centre of a large area of polymorpho-nuclear leucocytes. Occasionally, as in fig. 13, a large blood clot surrounded by a broad zone of large, foreign-body giant cells could be seen in close proximity to the graft. This evidently results from a rupture of one of the blood vessels of the embryonic membrane during the implantation of the graft. In good grafts the seminiferous tubules were quite normal and mitotic figures were numerous, denoting active growth. The interstitial tissue was apparently normal but polymorpho-nuclear cells were numerous. In one or two of the grafts the deposition of large amounts of deeply stained granular pigment had taken place between the spermatic tubules. The writer is not prepared to attach any great significance to the presence of this pigment, as in the course of examining chicks used to provide grafting material the presence of a testis, either completely pigmented or pigmented in one area only, was not unfrequently met with. Fig. 7 shows an almost completely degenerated testis graft. The outlines of the tubules were still faintly visible, but in nearly all the basement membrane had broken down and the tubules were filled with a mass of debris in which only a few degenerate nuclei were found.

Very few successful ovarian grafts were obtained. The majority were highly necrosed and showed no evidence of growing gonad tissue. The successful graft was usually attached to the fœtal membrane of the host embryo by means of the medullary cord region of the graft. The typical ovarian structure was retained, consisting of an outer layer, the cortex, and inner layer of degenerate medullary cords. The oöcytes were normal (fig. 22) and little evidence of degeneration was seen.

It was found that the percentage of successful testis grafts obtained far exceeded that of successful ovary grafts, independently of the sex of the embryo in which the graft was made. It is difficult to account for this, and it is suggested that the structure of the ovary may not be favourable for rapid vascularisation of the graft.

In Table III. details are given of the successful grafts of gonad tissue obtained on chicks of the opposite sex to those from which the grafting material was taken. All the cases listed had successful grafts, in that these contained, at the most, minute centres of necrotic material scattered through the bulk of healthy active gonad tissue.

The following is a brief description of the normal histological structure of the testis at the seventeenth day of incubation. The gonad is enclosed in a tunica albuginea several cells thick, consisting of loose connective tissue elements with flattened elongated nuclei. Blood vessels occur at frequent intervals. The immature spermatic tubules are surrounded by a well-defined basement membrane lined with epithelial cells. Numerous primordial germ cells with large nuclei are present in the tubules and a great many of them are undergoing mitosis. No synaptic stages are to be seen. At this stage of development the tubules show signs of branching. The intertubular tissue is abundant and small blood vessels and haematopoietic foci occur occasionally throughout the mesenchyme.

No deviation from this normal type of structure was found in the testes of embryos obtained after ovarian grafts, but for an increase in the haematopoietic foci of the number of cells belonging to the myeloid group.

The normal ovary at the seventeenth day of incubation consists of an external layer of germinal epithelium ; under this layer are situated the greatly enlarged cortical cords containing the definitive oocytes, some of which are in various stages of synapsis, while others are in the resting stage. Owing to the enlargement of the cords the band of connective tissue between the germinal epithelium and the degenerating medullary cords has been almost completely obliterated. The medullary cords are all canaliculised, the process being more advanced at the proximal end, where they are represented by an open network in the meshes of which are found islets of luteal cells ; these are also numerous at the distal end. Haematopoietic foci occur as in the testis.

There was no deviation from this type in the structure of ovaries, embryos obtained after testis grafts, except that, as in the testis, the cells belonging to the myeloid group were increased in number.

The results of the experiments given in this paper do not lend any support to the view that in the embryonic chick a grafted gonad of the opposite sex secretes definite hormones which, obtaining entrance to the body of the host, cause the differentiation of the urogenital system to be modified so that intersexual forms result. The results do not agree with those recorded by Minoura. Critical examination of Minoura’s method and his interpretation of results shows the latter to be open to serious criticism.

In order to demonstrate successfully the production of intersexual forms due to the action of specific sex-hormones secreted by a gonad graft, several essential conditions must be fulfilled:—

• (1) If the theory of the modification of embryos by the action of definite sex-hormones elaborated by the grafted gonad is to be accepted, then the presence of healthy grafted tissue on the chorio-allantoic membrane must be adequately demonstrated. The fact that on opening the egg at some period subsequent to grafting a white or pinkish mass may be observed having vascular connection with the embryonic membrane, has been shown to bear no direct relation to the graft ; it has also been shown that even if the graft is present the size of the growth does not indicate activity in the growth of the graft, but is merely the measure of the damage done to the delicate chorio-allantoic membrane during the operation. The fact that such growths are prçsent when the operative technique is carried out up to the point of lifting up the shell membrane and sealing the opening in the shell without implanting grafting material, supports the view that the presence of such tissue masses is a response of the injured membrane.

  Minoura does not state whether all growths were histologically examined, and in his description of the intersexual forms obtained no histological evidence of the presence of grafted tissue, healthy or necrotic, was given except in two cases, one a testis and the other an ovary graft.

• (2) The original sex of the embryo on which the graft was implanted must be identified, as otherwise the interpretation of abnormal forms is necessarily rendered very difficult. The use as experimental material of eggs from a mating, the male and female offspring of which show a single sex-linked character difference, makes it possible to determine whether the experimental embryo is a genetic male or female. Such material was obtained for the experiments described in this paper, i.e. eggs from Rhode Island Red males × Light Sussex females. The identification of the genetic sex of the embryo by an external examination was thus rendered infallible, even as early as the tenth day of incubation.

  In the female there is in all embryos considerable variation in the degree of regression of the right gonad. Slightly atypical females with lesser regression than usual may not necessarily represent an intersexual condition. Similarly, the presence of Müllerian ducts associated with male type gonads may not indicate intersexuality : such individuals may be exceptional males in which growth and differentiation of Müllerian ducts has occurred ; these ducts are present in male embryos during the early stages of development, but usually degenerate about the twelfth day of incubation. In the absence of evidence as to the genetic sex of the abnormal forms obtained by Minoura, it is possible that his cases may be classified under these two heads.

• (3) It is essential to submit the gonads of abnormal embryos to careful histological examination. Minoura’s brief histological description of the gonads of intersexual embryos does not indicate any striking deviation from the normal type : male type gonads, considered as ovary modified by testis graft, and female type gonads, regarded as testis modified by ovary graft, do not differ greatly from normal testis and ovary respectively. The presence of inactive or degenerating cells in the seminiferous tubules, and the presence of necrosis or partial disappearance of the cortical cords of the ovary could possibly be due to pathological conditions induced in the embryo by the presence of necrosed grafted tissue on the embryonic membrane. As previously described by the present writer, histological examination of embryos in which successful gonad grafts of the opposite sex were present did not reveal any deviation from the typical structure.

• (4) The effect of the production of a number of intersexual forms on the sex ratio in the surviving chicks must be considered. It has been shown in the experiments described in this paper, that the normal sex ratio in the chicks living after the operation is maintained, practically equal numbers of males and females being obtained. If intersexual forms are produced as a result of the action of testis grafts on female chicks, then a corresponding reduction in the number of unmodified females below the normal sex. ratio may be expected. Similarly, in the case of ovary grafts, the production of intersexual forms resulting from the interaction between the grafts and the male embryos will reduce the number of unmodified males below the normal ratio.

Examination of Minoura’s figures shows that out of 104 individuals living after testis grafts, 16 were classified as intersexual and therefore, presumably, were originally females whose sexual differentiation had been modified in the direction of the male by the action of the graft. Of the remaining chicks 52 were females and 36 were males, i.e. in spite of the production of 16 intersexual forms there was no decrease in the number of unmodified females below the normal. If it is considered, however, that most of the so-called intersexes were not females modified by the effect of the graft, but males in which an unusual development of the Müllerian ducts had occurred, then the original sex-ratio of the chicks would not show any appreciable divergence from the normal. In the chicks living after ovary grafts the relative numbers of the sexes are not so striking as in the case of chicks living after testis grafts. Out of the 83 chicks obtained 14 were classified as intersexual forms, the remainder consisted of 37 females and 32 males. The number of unmodified males obtained is therefore somewhat lower than would be expected in a normal population.

If we accept Minoura’s interpretation of the original sex of his “intersexes,” then in the case of the chicks with testis grafts the ratio of genetic males to genetic females was 36:68, while in the case of the chicks with ovary grafts the ratio of genetic males to genetic females was 46:37. These ratios are strikingly divergent, and statistical analysis shows that the odds against this divergence being due to random sampling are approximately 200:1. Indeed it is so unlikely that such divergence could arise by random sampling as to throw doubt on the validity of Minoura’s identification of the original sex of his modified chicks.

Since the evidence produced by Minoura cannot be said to demonstrate conclusively that the production of intersexual individuals was due to the action of the sex hormones of the graft, it seems desirable to consider the possible causes responsible for the occurrence of abnormalities in the reproductive system described in some of the embryos. Granting the occurrence of sexually abnormal forms, these can be divided into two classes.

Fluctuations in temperature, moisture content, failure to turn the eggs daily during the process of incubation of the eggs, must all be considered as possible agents affecting adversely the normal development and differentiation of the embryo.

Cases of naturally occurring intersexuality in fowls have frequently been described ; in such individuals an ovary or ovo-testis is present on the left side and a testis on the right. This, according to Benoit, is due to insufficient functioning of the ovary at a very early stage in development, permitting the sexual cords of the first proliferation to develop into testicular elements before their regression has proceeded too far. This only takes place in the female, no evidence of intersexuality in the male having been recorded. This is not a likely cause of the occurrence of intersexual forms in such experiments.

It is impossible to determine whether the intersexual condition as described by Minoura was caused by the specific activity of the small active portion of the graft, or was due to the presence of the large necrotic growth on the embryonic membrane affecting normal differentiation of the urogenital system. In the experiments set forth in this paper it is shown that many healthy active grafts were obtained which did not in any way interfere with the normal processes of sexual differentiation. Operated eggs were examined in the majority of cases on the seventeenth day of incubation when the graft had been present on the fœtal membranes for ten days. Since at the seventeenth day of incubation the differentiation of the urogenital system is almost as complete as at the time of hatching, any abnormalities of the gonads or their associated ducts are easily discernible. In all the cases listed in Table III. well-developed active grafts were present. If, as Minoura affirms, such grafts influence the sexual differentiation of the operated embryo, we should expect to find intersexual forms among the cases quoted. As stated above, however, such forms are entirely absent.

It may be argued that antagonism between the hormones produced by the graft and those produced by the host embryo (since at the time of grafting sexual differentiation from the indifferent stage has already begun) may have rendered ineffective the action of the graft hormone. This may possibly be the case, but it is not likely in view of the established fact that ovarian and testicular tissues can co-exist and exert their typical specific physiological influences in one and the same individual. Such artificial hermaphrodites have been induced only in a fully sexually differentiated individual and in such the conditions are not the same as in the embryo.

It has been claimed that the method of grafting gonads on to the chorio-allantoic membrane of the developing chick reproduces the conditions obtaining in the case of the free-martin in cattle (Lillie), where, owing to anastomosis of the blood vessels, a common circulation is established between bisexual embryos. The two cases, however, are not analogous. In the latter, it is the multiple secretions of all the endocrine organs of the male embryo which are poured into the common blood stream and circulated in the body of its female co-twin, while in the case of the egg grafts, the isolated organ exerts only its own specific physiological influence. Moreover, in the interpretation of the bovine free-martin it is assumed with reason that phenotypic maleness and femaleness depend upon the presence of functional testis and ovary respectively. Such interpretation does not explain why the male comes to possess testes, the female, ovaries. However, in the case of the fowl the sexes are genetically different, as is clearly shown by the facts of sex-linked inheritance. The cytological evidence is as yet insufficient to enable us to demonstrate the exact difference in the chromosome complex of the two sexes, but it is established that the female is the heterogametic sex. In the beginning of the differentiation of the embryonic gonad (either XX or XY in constitution) there is an indifferent stage the subsequent differentiation of which into testis or ovary is determined by the type of physiological environment (established by specific inheritance) in which it happens to be situated. In other words, the gonad is moulded by its environment, and does not mould it except as regards the late developing secondary sex characters.

This is best illustrated by the case of the ovary in the fowl. Results from experimental grafting of ovaries in very young chicks (Finlay, 1925) demonstrate that the ovary possesses a certain instability ; the engrafted gonad in its further development may show the presence of definite spermatic tissue. A parallel instance is shown in the cases of naturally occurring intersexes, many of which have been recorded in scientific literature. A change in the physiological environment may cause a concurrent change in the structure and function of the gonad, so that from a typical ovary secreting presumably ovarian hormone, it is modified into a testis secreting a testicular hormone. It is interesting to note, however, that there is no evidence whatsoever that would lead us to suspect a similar instability in the case of the testis in the fowl. There is no case on record in which intersexual gonads were found in the male. It would thus seem that the production of intersexual forms by ovarian grafts in male individuals is highly improbable.

It has to be stated that at a later stage in the development of the gonads they condition the production and maintenance of the secondary sexual characters.

It is shown in the present paper that the evidence derived from a study of gonad implants on the developing chick does not support the contention that in the fowl the processes of sexual differentiation are capable of modification solely through the action of the gonadic implant.