Portal de Periódicos da CAPES

The Development of Grafted Embryonic Fragments of the Chick

1925; The Company of Biologists; Volume: 3; Issue: 1 Linguagem: Inglês

10.1242/jeb.3.1.9

1477-9145

P. D. F. Murray, J. S. Huxley,

Animal Genetics and Reproduction

The method of embryonic grafting has proved of considerable value. Murphy (1913), Danchakoff (1916), and others have extended the method to the chick, by utilising the chorio-allantois of other embryos as host-tissue for the grafted pieces.In the following pages some account is given of experiments designed especially to test the capacity for differentiation of whole portions of the body of early chick embryos. Some other experiments with organ-rudiments are also recorded.The methods used were essentially those of previous investigators. 6- to 12-day embryos were used as hosts, the most frequent age being 7 to 9 days. The eggs were "candled "and the position of the embryo marked. The shell was swabbed with 90 per cent, alcohol over the site of operation. A triangular cut was made in the shell a short distance away from the position of the embryo. The cut was made with a small triangular file or else a fret-saw blade. Care was taken to avoid injury to the shell-membrane.The graft had been meanwhile prepared. The embryo was dissected out aseptically and placed in warm sterile Ringer. With fine instruments the required dissection was performed under a binocular microscope.The host egg was then placed close below a strong light. With an iridectomy needle a V-shaped incision was made in the shell-membrane and the resulting tongue turned back; when this is successfully done, the chorio-allantois sinks down a little, but is undamaged. The piece to be grafted is then taken up, either in a pipette or else on a needle (a convenient way is to lift it out of the fluid on a closed forceps, and pick off with a needle). It is then squirted on to or laid on the chorio-allantois, if possible near a large blood-vessel, and not immediately below the site of operation. The shell-membrane flap is then turned back, the piece of shell replaced, and sealed with shell-membrane moistened with albumen from another egg. Sometimes an additional sealing of the edges of the piece of shell-membrane was undertaken later with collodion solution. The egg is then replaced in the incubator, and removed for examination after a suitable time.The best method of examination is to hold the egg with the site of operation downwards, open the egg on the opposite side, and carefully remove the embryo and yolk-sac into a small vessel. The chorio-allantoic membrane should remain in place over the lower half of the egg. If there is a reaction at the site of operation, this is readily visible. A small piece of shell and membrane is cut out surrounding the reaction, and the whole preserved, the chorio-allantois being later stripped off before clearing. Most of the specimens were examined as toto mounts in clove oil before embedding: this often gave valuable information. Bouin's fluid was chiefly used as fixative. 10-μ sections were cut and stained in various ways. Reconstruction models were in some cases made by the wax and blotting-paper method. In other cases graphic reconstructions were plotted on co-ordinate paper. This method serves admirably for comparatively simple structures, and takes far less time than the construction of models.A large number of operations produced no visible result, or the hosts died. The visible result in the apparently successful ones was often a mere reaction of the host-tissue (Huxley and Murray, 1924). A number of successful grafts, however, were obtained. Those described in this paper are as follows:—The Head (No. 21).—In order to arrive at a correct understanding of the anatomy of this specimen, it is necessary to describe the distortions of form which it has undergone (figs. 2, A, B, C). The brain, which is the predominant feature in the sections, is divided clearly into telencephalon, diencephalon, mesencephalon, and rhombencephalon. The entire brain, in all parts, is expanded laterally so as to be far broader than in the normal chick, presumably on account of pressure. This expansion is everywhere markedly greater on the left side. Partly as a result of this, the eyes have been forced into a ventral position. They, of course, start normally with the optic stalk attached ventro-laterally to the diencephalon, so that the position of the eye is at first ventro-lateral. In this specimen, owing to the great extension towards the left of the telencephalon and diencephalon, the left eye is very far ventral, so far that it is really on the right side of the brain and ventral. In other words, the more dorsal and left latero-dorsal regions of the brain have grown out laterally and have left the left eye behind, so that it has come to occupy this ventral position.* Like the eyes the epiphysis is far to the right of the mid-line of the specimen, and the same applies to the hypophysis. The mid-brain shares in this same distortion ; it is expanded to some extent in both lateral directions, as though forming optic lobes, but especially and far more to the left. The extension of the mid-brain to the left is shown by the point of continuity with the hind-brain, which is far over on the right side.This lateral left extension of all regions of the brain is probably due to the position of the graft on the chorio-allantoic membrane. The graft was found with its dorsal surface facing inwards towards the centre of the egg, but with a tilt over to the right. Thus the right side would have had less room to grow in and so the left side expanded at its expense.Another distortion of form has involved the establishment of a new and abnormal brain flexure. In a normal chick, the head, as a consequence of the cranial flexure, is in such a position that it faces backwards. In this specimen the graft was made before the cranial flexure appeared, when the brain was a straightforward continuation of the spinal cord, without any flexures. The graft was made in such a way that the ventral surface of the fore-brain region became closely attached to the chorio-allantoic membrane, as shown in figs. 2, A2, etc. It was thus impossible for the cranial flexure to influence the telencephalon, on account of its firm attachment: it could, however, influence the diencephalon and mesencephalon, and has done so, because they were not in such close and direct connection with the chorio-allantoic membrane, but separated from it by a considerable space of more or less loose mesenchyme, and also by the gut region. The result of this has been that, while there is a cranial flexure having the apex of its arch in the mid-brain, the telencephalon, instead of constituting one of the limbs of the arch, has maintained its original antero-posterior position, and so it has introduced a new upward flexure as seen in the figures.Before proceeding to the more detailed description of the brain of this specimen, it is necessary to mention the main features of the brain of a normal 24-hour chick. In such a chick the neural groove is just closing, but the closure is not complete until thirty-three hours. The outlines of the three primary brain-vesicles are barely indicated, but the fore-brain has large lateral expansions, the optic cups, and a constriction is appearing separating it from the mid-brain. A little later, at about twenty-six hours, the mid-brain becomes marked off from the hind-brain. There are no brain-flexures at this time, the cranial flexure beginning at twenty-nine hours.In the graft, fixed four days after grafting, the neural tube is completely closed, and the brain has increased greatly in size and histological differentiation, and is clearly divided into telencephalon, diencephalon, mesencephalon, and rhombencephalon.The telencephalon is large and roughly approximates to the normal form at a stage prior to the formation of cerebral hemispheres, except for the two features mentioned above: the lateral expansion to the left, and the fact that it points forward, as it did before the formation of any cranial flexure in the brain, instead of downwards and backwards as is normally the case. Cerebral hemispheres are entirely absent, there not being any trace of them at all. As these are strongly indicated in a normal chick even as early as seventy-two hours, it is clear that there has been considerable retardation in the development of the telencephalon. Retardation appears to be a general feature of grafts such as these. A peculiar abnormality in the floor of the telencephalon will be described later.The diencephalon has also advanced considerably. It is expanded laterally to the left like the rest of the brain, but somewhat less so than the other regions. This expansion is, however, sufficient to put the epiphysis well on to the left side. The two eyes are both present. Their position is ventral, as described above.The mid-brain has a flat floor and longer arched roof (figs. 2, A 2) and, as described, is much expanded to the left, and also, though less, to the right, as though it had been trying to form optic lobes (o.l.). It is marked off posteriorly from the rhombencephalon by a constriction, the isthmus.The hind-brain is present in part, the posterior part of the medulla oblongata being absent. The roof of the anterior end is thinner than elsewhere, being in fact epithelial, and marking the position of the myelencephalon.As regards the flexures of the brain, the cervical flexure does not come into the discussion, because none of the spinal cord was included in the graft. The pontine flexure is not well marked, but is present. It would have been clearer but for the fact that at the posterior end of the graft where the brain was cut it has closed again and taken a downward curve (figs. 2, A 2), thus obscuring the pontine flexure a little farther forward. The cranial flexure, on the other hand, is marked, the brain forming an arch of which the mesencephalon forms the apex and the diencephalon and rhombencephalon the two limbs. The relations of the telencephalon to this flexure have already been described.It should be pointed out that the normal turning of the embryo on to its side has been prevented by the fixation of the brain to the chorio-allantoic membrane by its ventral side.Histologically the brain has reached a high stage of differentiation, as shown by the development of grey and white matter and of cranial nerves. As there are a number of mitoses among the germinal cells lining the ventricles, it is clear that development had not by any means stopped when the specimen was fixed.Turning to the organs of special sense, we find both eyes developed, but very abnormal. Both are very small as compared with the eyes of a normal 5-day chick. Each has an optic cup and a lens. The optic stalks are narrow, the process of narrowing down the original wide aperture from brain to optic cups having been complete. The lenses show signs of considerable retardation, and one of them of abnormality of development. In the more normal of the two (the left—figs. 1, E) the lens is well formed, except that the thinning out of the external layer of cells has not proceeded nearly as far as it should have done in a 5-day chick. The lens of the right eye, however (fig. D), is of abnormal form, being in places triangular in section, with but little difference between the two layers. The optic cups are both very abnormal. Instead of being large, approximately spherical sacs, they are without any regular shape, and the walls are thrown into a number of large folds, somewhat as though an unsuccessful attempt had been made to grow into a large sphere of normal size. They are also very much smaller than they should be in a 5-day chick. The explanation of this may perhaps lie in the position of the eyes relative to the other structures of the graft. It is possible that the eyes, growing, as a consequence of the expansion of the lateral expansion of the brain, under pressure, may have attempted to reach the normal size and form of the eyes of a 5-day chick, but, there being insufficient space for their expansion, were unable to do so, so their continued growth led to folding of the walls of the optic cup. Apart from the question of size and form of the cup, there are two other signs of considerable retardation. The difference between the thickness of the inner and outer walls of the optic cup is much less than in a normal chick of this age. There is, in fact, hardly any difference at all, while in a normal specimen the inner wall is considerably thicker than the outer. The optic nerve seems not to have begun developing in the graft, while in a normal chick it reaches nearly or quite to the brain during the sixth day. A small amount of pigment has been developed in the optic cup, but not as much as normally. Barfurth and Dragendorff (1902) also found abnormal folding in the eyes of operated chick embryos, as did Hoadley (1924) in the eyes arising from grafted rudiments. In all recorded operative cases, the lens is much more normal than the optic cup.Fig. F shows an organ which may be the otic sac. As, however, some of the sections in this part of the series have unfortunately been lost from the slide in staining, and as part of this organ is involved in the loss, it is impossible to be certain. The structure in question is a sac occupying approximately the position expected of the left otic sac. It has walls in part thick, in part thinner, and a cavity with an irregular shape. This cavity gives off two tubes which run dorsally and end close to the mid-brain. One of these may be the endolymphatic duct, but not both. Neither can either of them be the lagena or the sacculus, as the former does not develop till the seventh day and the latter is by then barely indicated. The matter must therefore remain in doubt. Neither can they be semicircular canals as both are blind ducts, not re-entering into the sac. This is the only structure that has been found in any way resembling an otic vesicle labyrinth, and if it is that, its fellow is completely absent. It could hardly be among the missing sections as there are too few of these, and all are close to the structure just described.It is doubtful if olfactory pits are present. Just anterior and lateral to the mouth on each side is a small pit which may possibly represent the pits. As, however, one of these is very small and the other rather a groove than a pit the matter must remain in doubt, although since they are in the correct position, the probability is that they do represent retarded olfactory organs. Hoadley (1924) has shown the capacity for differentiation of olfactory organs in grafts consisting of the front end of the head grafted prior to the appearance of visible olfactory rudiments.The epiphysis (fig. G) is present in its usual position and is a long tube. This tube normally lies flat over the roof of the telencephalon, thus lying forward from its base. In the graft, however, it has grown in a sideways direction, and lies stretching towards the right. It is possible that this occurred through the base being pulled to the left through the marked growth of the brain to that side.The hypophysis is represented by a small mass of tissue applied to the diencephalon floor (fig. H), and still attached (in another section) to the gut by a Stalk. The infundibulum is a deep depression reaching out posteriorly from the fore-brain. It is much larger than would be expected, both in depth (figs. 1, 2, H) and lateral extension (fig. C). Very probably an explanation of this is to be sought in mechanical causes. Owing to the abnormal upward flexure of the telencephalon, a kink is bound to occur at the juncture of telencephalon and diencephalon on their ventral surface—a kink which will produce a fold of just such a nature as the one described, which would incorporate the normal infundibulum in itself. Details of this somewhat aberrant pituitary complex are being investigated by Mr de Beer.In connection with the brain, there should be mentioned a peculiar bar of tissue (fig. 2) which extends across the cavity of the brain from the ventral wall of the telencephalon to the posterior (morphologically ventral) wall of the diencephalon. This bar consists of neuroblast tissue similar to that of the future grey matter of the brain, but includes no fibres. It is traversed by a blood vessel. Just behind its ventral end is a slight depression in the brain-floor. Its interpretation is exceedingly difficult. Obviously it corresponds to no normal structure in the brain of the chick.Another abnormal feature is the existence in several places in the brain, but especially on the left ventro-lateral region of the wall of the mesencephalon, of groups of small vesicles or tubes (figs. J, K). Most of these small vesicles or short tubes have small cavities. They are usually embedded in the brain wall, external to the neuroblasts. Sometimes, as in fig. K, they protrude slightly from the general brain surface. Their walls are composed of neuroblasts with often several mitotic figures in the central layer, as though these were germinal cells. Usually there is a clear area without nuclei immediately surrounding the central space, and this is suggestive of the narrow clear area internal to the germinal area of the brain-wall itself. It differs, however, in having taken a lighter stain.A still more striking abnormal development of the same general type is illustrated in figs. 3, 4, C, L, M. Two quite large "organs" of brain-tissue have been produced and make prominent features on the ventral surface of the fore-brain. They are completely continuous with the brain proper in certain regions, but elsewhere protrude freely over large areas. Else-where again they are closely applied to the brain-wall in such a way that it is often hard to say whether continuity exists or no. This partial continuity is brought about by the fact that in several places both they and the brain-wall proper are not sharply delimited in the normal manner, but show an irregular margin with cells apparently proliferating outwards in a loose way into the surrounding mesenchyme.In other parts, however, the tissue both of these pseudo-organs and of the brain is dense and sharply delimited against the mesenchyme. Numerous vesicles or tubules are to be seen in these "organs." Often the vesicles are identical with those just described as existing in the brain-wall. Sometimes they are very minute, with only a virtual cavity; or even only an indication of radial cell-arrangement. Some, however, are large and spacious, looking like definite morphological featuresIn some places the "organs" are in contact with the ectoderm ; but no modification of either tissue is visible in these regions.It is clear that these large pseudo-organs result from a further extension of the process which gives rise to the embedded vesicles: as to the particular reason for their production, however, we are quite in the dark. It may be noted that the smaller (left) of the two "organs" is much elongated laterally (fig. C). It looks as if the primordium of this formation had been produced early, and had then become involved in the great lateral extension of the whole of this side of the graft.The undefined margin of the nervous tissue in certain areas is interesting, as is the fact that in such regions the cells appear to be migrating irregularly out into the surrounding tissue, and form a much less compact tissue than usual. We have here something which reminds us of the conditions obtained by Hoadley (1924, 1925) in some of his mesencephalon grafts, although the details are different. Other authors have noted extra proliferation of the embryonic central nervous system of the chick as a result of interference, etc. ; e.g. Alsop (1919) (temperature); Harman (1918) (abnormal specimens found in nature); the folded optic cups above discussed, etc. However, so far as we are aware, such extensive abnormalities as the bar and the pseudo-organs have not hitherto been described.(figs. N, O).—There is present a gut (pharynx), mouth, and anterior intestinal portal. The mouth opens into the gut. No trace can be found of lungs. These should be well marked at forty-eight hours, but it is doubtful if the site of the laryngo-tracheal groove would have been included in the graft. A solid mass of cells is found attached to the floor of the pharynx. This presumably represents the thyroid, although it is considerably retarded.Of visceral pouches (v.p., fig. O) there is one on the left side, two on the right. Those on the right side are large with prominent openings, while that on the left is small and leads to the exterior through a narrow tube. The external apertures of the two right pouches are more or less confluent. In a normal chick both the second and third pouches have double openings, but one is dorsal and the other ventral. There is apparently no trace of this division here. Another peculiarity of the right pouches is that there is given off from their united wall, close to the external opening, a short prominence containing a cavity (fig. O (x)). There is nothing to show to what, if anything, this corresponds in the normal chick. There are no traces of visceral arches, and in the absence of both these and of the other pouches it is impossible to tell to which pouches of the normal chick those present in the graft correspond. The presence of three open visceral pouches is another feature indicating retardation, since the third pouch, the last to close, is normally closed during the fifth day. The absence of the others, indeed the absence of any trace of them, indicates that they never formed. Normally they would close on the third, fourth, and fifth days, and one would expect that some trace of them would remain in a chick showing considerable retardation of at least most of its organs such as the one under consideration. Even in a normal chick the epithelial tubes representing the second, third, and fourth gill-pouches remain till the sixth day. That the left visceral pouch in the graft has already reached the condition of an epithelial tube, argues that it is further advanced than those on the right. So we are forced to the somewhat curious conclusion that one póuch is slightly retarded, two are much retarded, and the remainder have not developed. There is no trace of a tubotympanic cavity, part of which is of course normally derived from the dorsal division of the first visceral pouch.Attached to the dorsal side of the gut just anterior to the anterior intestinal portal is an apparent organ consisting of a number of tubules and vesicles (fig. P). It seems impossible to regard this as having any correspondence with any organ in a normal chick except the liver or pancreas. It can, however, hardly be the former, as it arises from the wrong side of the gut, or rather of the anterior intestinal portal, being on its dorsal side instead of ventral. It is therefore impossible to say what it may be, except just possibly the pancreas, which of course is a dorsal structure. The organ in question, while arising from the dorsal side, lies mostly laterally. In a normal chick both liver and pancreas arise about the end of the second day. Now the graft was made at about twenty-four hours. Further, the liver and pancreas are both very close to the opening of the anterior intestinal portal when they arise. Since a considerable portion of the gut must become closed in during the second day, it is highly improbable that the region destined to form liver and pancreas could be included in the region closed off in the graft, since it is hardly possible that, being affixed to the chorio-allantoic membrane ventrally, any more of the gut became closed off after grafting. It is thus improbable that this body can be either liver or pancreas, and so it is impossible to say what it is, unless one or other of these has been a bnormally formed more anteriorly than usual.The blood vessels of the graft have not been worked out thoroughly. It, of course, receives its blood supply from the vessels of the chorio-allantoic membrane. The heart is present, but in part only (figs. Q, R). The piece which is present consists of a bulbus arteriosus and at least part of the ventricle, but there is no auricle or sinus. Evidently the cut which isolated the head passed through the heart in such a way that the future ventricular region or a large part of it was included in the graft, while the auricular region was not; the part contained in the graft continued its differentiation, but there was no regeneration of the auricle. Thus in the heart rudiment at twenty-four hours the auricular and ventricular regions are predetermined, although not visibly differentiated, as otherwise the part included in the graft should have made at any rate some attempt at the formation of both these regions. In a normal chick the constrictions between the chambers only appear during the third day. The septa between the auricles and ventricles only begin forming on the fifth day in a normal chick, and in this specimen are entirely absent. The cardiac muscle, on the other hand, seems to be well-developed histologically, there being thick muscular walls. As usual, however, with the organs of the graft, the entire organ is small compared to the normal size.The mesenchyme of the surrounding region is continuous with mesenchyme within the heart, there being an open collar of muscle around the large area of entrance (fig. R). From the centre of this area springs a vessel which continues the cavity of the bulbus. It gives off a short blind prolongation, and continues outward to link up with blood vessels in the vicinity. The ventricle is honeycombed with irregular spaces (not shown in fig. R). At one place (see fig. Q) its wall is incomplete, so that here too the external mesenchyme penetrates into the interior of the organ The muscular tissue of the bulbus is much better differentiated than that of the ventricle.As regards skeleton, there is no cartilage in the specimen, and the presence of procartilage is very doubtful. Chondrification begins in the normal chick only on the sixth day, so the absence of cartilage cannot be taken as a sign of retardation. Examination of the notochord (nch., fig. C) reveals a very curious condition. It is very minute. It does not extend as far forwards as the diencephalon as is normally the case, but comes to an abrupt end on the ventral side of the hind-brain, petering out into a small mass of what appears to be degenerate tissue. As this mass of degenerated tissue is very small, it is improbable that any conclusions can be drawn from it as to the fate of the rest of the notochord. Until more is known about the origin of the notochord, and especially of its anterior end, in the normal chick, speculations as to the reasons for its absence in this specimen have little value.The next two specimens to be described, Nos. 45 and 50, do not require detailed description as they both show in essentials the same general features as the previous one, i.e. continued growth and differentiation after grafting; retardation of both growth and differentiation; and distortion. Both were pieces taken from the middle of the body, in the case of No. 45 of a 2-day embryo and in the case of No. 50 of a 24-hour embryo.In both, the spinal cord has undergone differentiation, and there are dorsal and ventral spinal roots. In both, the vertebrae are well developed and chondrified (fig. 6). The notochord, prominent in No. 45 (fig. 6), is altogether absent in No. 50. This is curious, especially as No. 45 represents a 7-day, and No. 50 a 6-day chick. Normally chondrification of the vertebrae begins on the fifth day. In a normal chick, the notochord is prominent in the centre of at least many vertebrae up to eight days. There is, of course, no ossification in either.No. 45, which was taken from the region of the body including the omphalo-mesenteric arteries, shows on each side a region of thickened mesenchyme, resembling procartilage (h.l.b., fig. 6). This is probably the rudiment of the skeleton of the hind limb bud. The four projections, representing the four regions, axis of limb, ilium, pubis, ischium, are not as yet distinguishable. There is no true cartilage in this mass of dense mesenchyme, and even procartilage is doubtful.There is no alimentary canal in either specimen. This is presumably because both specimens, when grafted, were taken from regions in which the gut had not been closed off. When they were grafted they became attached and spread-eagled out on the chorio-allantoic membrane, so that closure of the gut became impossible. There is a gut present in the above described head-region (No. 21) because in that case it had been closed off before grafting.No. 45 (fig. 6) has, on each side, a mesonephros, lying in its usual position. The cœlome is present and the mesonephros lies against it as usual. No. 50 has a mesonephros on one side but none on the other, and has no coelome. As the mesonephros only begins to develop at about forty-eight hours, it is clear that in the case of No. 50, which was grafted at twenty-four hours, the entire mesonephros must has developed since grafting; while in the case of No. 45 it might have been laid down as vesicles before grafting, but certainly has undergone most of its development since then.In No. 50, the embryo has been distorted obliquely; this is especially noticeable in the spinal cord. Myotomes are developed, especially well in No. 50.The chief point of interest lies in the power of independent differentiation of a whole region of the body, such as the head or a section of trunk, when supplied with food by the blood-vessels of the host chorio-allantois. This is combined with an apparently total lack of regeneration. This finding confirms those in many other vertebrate groups, both as regards single organs and large fragments of the embryo.To mention only the work on the chick, the self-differentiating capacity of various organs and regions has been demonstrated by Barfurth and Dragendorff (1902—eye), Lillie (1903, 1904—large anterior regions of embryos), Spurling (1923—limbs and girdles), Minoura (1921—gonads), Atterbury (1923—metanephros), Hoadley (1924—eye, olfactory region, auditory organ, mesencephalon), Agassiz and Danchakoff (1923—spinal cord). In particular Danchakoff (1923) has recently published a preliminary account of experiments essentially similar to ours both in material and results. We may perhaps mention that much of our work had already been accomplished before this note was published.It thus appears that in all vertebrates there comes a period after which the potentialities of each region of the germ are rigidly fixed, as opposed to an earlier period during which complete regulation is possible. Spemann (references in Spemann, 1921) has shown in Triton that the time when the first phase passes into the second is about the middle of gastrulation. One of us (Huxley, 1924) has proposed the term chemo-differentiation for the process of differentiation which then sets in, since it appears to make of the germ a mosaic of chemically different regions, each one capable of pursuing its own development independently of the rest (provided suitable conditions are provided) up to the onset of the third period, that of functional or co-ordinated differentiation, when nervous and endocrine co-ordination, as well as functional modification, come to play a part.The three stages, and especially the second and third, overlap considerably as regards different organs and parts of organs. For any one organ or part of an organ, however, they seem to succeed each other invariably in this order. This can be regarded as one of the most important empirical laws discovered in recent times as regards the developmental physiology of vertebrates. It seems valuable to have the existence of the same stages demonstrated for the highest, homothermic vertebrates, and these observations provide some evidence helping to establish this.Grafts like these, however, throw certain new light on the question of early differentiation. They show (1) that retardation and inhibition of development accompany any unfavourable circumstances; (2) that the self-differentiation above described is essentially chemical, leading to the production of specific tissues in definite amounts; but that (3) the assumption of particular morphological form must be largely determined by physical (mechanical) influences.As regards (1) effects of inhibition and retardation due to operation or to depressants, etc., are of course well known. It is worth while stressing the fundamental similarity of result obtained by very different methods. The total absence of certain organs in certain grafts only (see Hoadley, 1924), or of one only of two paired organs (e.g. the auditory vesicle of our No. 21) can only mean that although chemical pre-determination is complete from a certain stage, the realisation of its potentialities will not occur in certain unfavourable conditions. E.g. immersion in various salts will produce cyclopia and failure of certain anterior median regions to develop. Pressure and insufficient blood supply, presumably, have brought about the failure to develop in certain organs in our grafts—e.g. some of the gill-slits, the R. auditory sac, etc. Retardation is thus a partial inhibition.As regards (2), there are two points to be considered: (a) Spemann (1918) has shown that the parts destined to produce retina, tapetum, and optic stalk are not only qualitatively but quantitatively predetermined, within a definite area, in the neural plate stage. However, any part of this area, even though it may possess abnormal proportions of the different substances, will form an optic cup with reasonable approach to normal shape, i.e. the shape of the optic cup is determined quite differently from the histological differentiation of its various parts.(b) Purely mechanical considerations of available space are also of great importance. This was first pointed out by Cotronei (references in Cotronei, 1921) whose work, however, does not seem to be well known. The same point has recently been emphasised by Hoadley (1924, 1925). In many of Hoadley's grafts, the histological appearance was normal or nearly so, while grave distortions of morphological form had been mechanically brought about.This distinction between chemical and physical, between histological and morphological differentiation, is clearly one of some theoretical importance.The extreme asymmetry of our grafted head affords another interesting example of the influence of mechanical causes upon morphological differentiation, as does the reversed flexure of the telencephalon. Clearly the provision of adequate space and growth unhampered by adhesion and adjacent structures is as important in development as the relative timing of chemical processes.Finally, the extraordinary proliferations of the brain tissue, leading in extreme cases to the production of new "organs," is very interesting. Hoadley (1924, 1925) has shown that the absence of the normal limiting membrane on the outside of the brain in a graft leads to abnormal fibre-outgrowths simulating nerves, together with abnormal immigration of blood vessels. (We are reminded of the importance of bounding membranes in the thyroid : higher forms possess a capsule to the organ, which if it enlarges, forms a localised goitre; but e.g. Teleost fish have no capsule, and if a goitrous condition arises in them, the thyroid cells proliferate irregularly, and actually invade muscle and bone—Marine, 1911.)The work was carried out with the aid of grants from the Department for Scientific and Industrial Research and from the Royal Society, to both of which we desire to make grateful acknowledgment.Note.—AU figures except fig. 6, plate I., refer to grafted head No. 21. For the meaning of the abbreviations, consult the list below. The small arrows indicate the morphological axis.