Previous work by the writer has shown that normal development of the tympanic membrane in R. pipiens is due to inductive influences of the annular tympanic cartilage. The present work concerns the nature of the developmental influences, specifically as regards whether or not non-living cartilage still possesses inductive properties. The experiments performed and results obtained were as follows:
Controls consisted of autoplastic transplantation of annular tympanic cartilage beneath back integument of metamorphosed and involuting R. palustris larvae. In general, integumentary thinning, coupled with partial to total glandular degeneration, occurred. This is typical of partial membrane formation.
Similar transplantations were made involving cartilage killed in ethyl alcohol, saturated chloretone solution, and boiling water, or by desiccation in strong sunlight. The results differed considerably depending on the type of killing treatment employed and the degree of subsequent transplant degeneration. Partial membrane formation was usually induced, however, in the overlying back integument. Similar series were run with half-metamorphosed larvae. In general, membrane induction was even more pronounced.
The same experiments were performed on R. catesbeiana larvae. Transplantation, however, was made beneath the skin of the upper jaw, posterior-ventral to the external nares. Wide variations in results were observed. No membrane changes occurred in some cases, while in others considerable transformation was induced.
It is concluded that dead annular tympanic cartilage still retains inductive influences capable of initiating varying degrees of tympanic membrane formation in integument. In general, the greater the degree of cartilage degeneration, the less the extent of membrane formation.
The results are discussed in general and the conclusion reached that the inductive influences arising in the developing annular tympanic cartilage of the anuran are quite probably chemical secretions or excretions. These, when transmitted to the integument of the ear region during metamorphosis, serve to induce a dedifferentiation of the larval skin into that type characteristic of adult tympanic membrane.
The tympanic membrane of the adult frog is divisible into three histologically distinct parts, viz. an inner epithelial layer, continuous with the mucous membrane of the tympanic cavity, the former being closely applied to a thin lamina propria conforming in shape and area to that of the annular tympanic cartilage to which it is attached at its peripheral borders. The third and outer portion of the membrane consists of modified integument which, externally, appears as a conspicuous brownish coloured area almost circular in shape and slightly indented. The histology of the various layers has been described for only a few anurans. In this regard the reader is referred to Moldenhauer’s (1878) work on R. escalenta and R. temporaria, to that of Tokura (1925) on R. nigromaculata, and of the writer’s (1928, 1934 a) on R. pipiens, R. sylvatica and R. clamitans.
The writer (1928) has previously shown that the formation of the integumentary portion of the membrane, during larval metamorphosis, involves a definite sequence of histological changes in the skin covering the presumptive ear region. In general, this consists of two distinct phases in which certain degenerative changes first occur in the various integumentary layers, particularly the stratum compactum, followed by a developmental stage during which the characteristic structure of the adult membrane1 is gradually formed. Thus, the entire process may be considered as a dedifferentiation of normal integument into that type peculiar to the tympanic membrane.
The skin of the presumptive ear region cannot be considered as determined to develop into tympanic membrane as contrasted with other integument of the larva, for when it is transplanted, heterotopically, it fails to undergo membrane transformation during metamorphosis. Conversely, however, any integument transplanted to the ear region will transform into tympanic membrane at the proper metamorphic stage. It was finally found that the dedifferentiation of larval skin into that of the tympanic membrane is the result of inductive influences emanating from the developing and underlying annular tympanic cartilage. This conclusion was based on the results of two types of experiments. In the first, the developing annular tympanic cartilage was extirpated at various stages of larval metamorphosis. If removed prior to the initiation of membrane development, no histological membrane changes occurred during metamorphosis. When extirpated after the onset of membrane transformation, the membrane ceased to develop further and in fact tended, in certain cases, to regress towards normal integument. The second type of experiment involved the transplantation of developing annular tympanic cartilage beneath the skin of the side and back. During larval metamorphosis the integument overlying the transplants underwent normal tympanic membrane transformation. It may be noted here that the development of the thickened elastic fibre region of the lamina propria is similarly induced by influences arising from the tip of the developing columella (Helff, 1931).
The writer (1934 a) has furthermore shown that the annular tympanic cartilage continues to exert its inductive influence on membrane formation for some time following complete metamorphosis, but with decreasing intensity. In fact the considerable post-metamorphic enlargement of the tympanic membrane in the growing frog is quite probably due largely, and perhaps entirely, to the corresponding enlargement of the annular tympanic cartilage and its continued inductive effect. Finally, the remarkable intensity of the inductive influence of the developing annular tympanic cartilage has been demonstrated by the writer (1934 b, 1937). The fact that fully differentiated dermal plicae integument (transplanted, hetero-plastically, over the developing annular tympanic cartilage of the metamorphosing host animal) will undergo complete transformation into tympanic membrane, is convincing evidence of the power of the inductive influence at work.
The nature of the inductive influence or influences present in developing annular tympanic cartilage was of considerable interest to the writer. Since this structure consists entirely of hyaline cartilage, it was thought worth while to determine whether other hyaline cartilage possesses the same inductive effect. For this work (1934 c) portions of the quadrate and supra-scapula were removed from newly metamorphosed R. palustris frogs and transplanted, autoplastically beneath the back integument. The results showed that membrane inductive influences of varying intensities were present in such transplants. Usually only slight tympanic membrane transformation occurred in skin over the supra-scapula transplants, while the quadrate implants induced nearly complete membrane development in many cases. This can be readily understood when one considers the fact that the quadrate and the annular tympanic cartilage are quite closely related, embryologically; the latter structure having its origin from the metapterygoid region of the larval quadrate.
Apparently, therefore, tympanic membrane influences are characteristic of hyaline cartilage in general, although the inductive intensity is subject to considerable variation and is particularly strong in the developing annular tympanic cartilage. The fact that actual contact of cartilage and integument is essential before membrane induction will ensue, strongly suggests that the inductive influence is probably chemical in nature. In fact, it was consistently noted that where contact between transplant and skin was strongest, the membrane development was correspondingly accelerated and more pronounced as compared with regions where implant and integument were barely touching one another and the former, therefore, exerting no pressure against the latter.
In the present work an attempt has been made to carry the problem a step further. The question of whether or not non-living annular tympanic cartilage still possesses inductive influences was investigated. Should no membrane induction occur following the transplantation of dead annular tympanic cartilage beneath integument, the results would indicate, although not necessarily so, that the inductive influence is not chemical in nature. However, it is obvious that such a negative result might be explained on the grounds that normally the inductive chemical influence is being constantly produced by the living cartilage in amounts sufficient to initiate and complete membrane development, whereas in the killed cartilage only a limited amount is present which, when used up, cannot be replenished by the dead cartilage cells. If membrane formation was found to be induced by dead annular tympanic cartilage, the probability of the presence of a chemical or secretory influence would be strongly suggested.
MATERIALS AND METHODS
The species utilized for the various transplantations were R. palustris and R. catesbeiana. Specimens of the former species were collected in the vicinity of Lake Winnisquam, New Hampshire, during the month of July 1938, while those of the latter species were obtained from Silver Lake, New Hampshire, during August of the same year. Usually the animals, when collected, were at or near the metamorphic stage desired for operative work, or if not, soon attained the proper degree of transformation in the laboratory.
The technique of annular tympanic cartilage extirpation has been previously dealt with in some detail by the writer (1928, 1934 c) and need not be related here. The methods employed to kill the transplants, however, prior to their implantation beneath back integument, should be described at this time since, as far as the author is aware, the killing of annular tympanic cartilage for subsequent transplantation work has not been done before.
Four killing methods were used, viz. immersion in alcohol, prolonged chloretonization, boiling in water, and desiccation. For the alcohol treatment, the extirpated annular tympanic cartilages were placed in 95% ethyl alcohol for 5 min. in some instances and for 30 min. in others. Cartilages, so treated, become more firm and tough and are consequently easy to transplant. There is no apparent shrinkage in size. Subsequent observation showed that while the transplants had been killed when immersed for but 5 min., they apparently had not been preserved against degeneration as adequately as had those treated for 30 min. Killing by chloretonization involved simply the immersion of transplants in a saturated solution of chloretone (trichloro-tert-butyl alcohol) for approximately 10 min. There is no immediate physical effect on the cartilages and they appear the same as when extirpated. When killed in water, the cartilages were boiled for about 5 min. in either tap or distilled water. This treatment, although not shrinking the transplant, caused it to become very soft and white, making its transplantation somewhat difficult. The desiccation treatment involved more care and difficulty as compared with the other methods of killing. Desiccation was quickly, although no doubt not entirely accomplished, by placing the rings of cartilage on a piece of filter paper after which they were exposed to strong direct sunlight for at least an hour. The transplants by the end of this period become hard and brittle and have a tendency to curl up at the periphery. They are also considerably reduced in size. Since they tend to adhere rather firmly to the filter paper, care had to be exercised in removing them to prevent breakage or other injury. It is interesting to note here, however, that following transplantation beneath integument, they imbibe lymph rapidly and return to their normal size and general physical consistency within 5-10 min. Cartilages killed with alcohol or chloretone were always carefully washed in water for at least 10 min. prior to their transplantation. Those killed by means of desiccation or hot water, however, were transplanted without further treatment
All transplantation was autoplastic and the cartilage was always obtained from the right side. In the case of R. palustris animals, the transplants were placed beneath the back integument in the mid-dorsal line just posterior to the eyes. To accomplish this a small transverse incision in the integument was first made quite far posterior on the back. The cut was just large enough to permit of the entrance of the transplant. A small spear-headed scalpel was next inserted through this incision and the back integument carefully loosened anteriorly by cutting the subcutaneous septa or connecting fibres which bind the skin to the underlying musculature. The transplant is now inserted through the cut in the skin and carefully shoved anteriorly to its desired position. Care must be exercised in so doing since the transplant can be easily seriously injured or doubled over on itself. When properly implanted, however, it produces a moderate to a strong bulging of the integument above it by means of which the outline of the transplant can be readily detected externally. Transplantation on R. catesbeiana animals was made on the right side beneath the skin of the upper jaw to a point just posterior-ventral to the external nares. The transplants were first inserted through a skin incision ventral to the eyes and then moved anteriorly to the desired position. Even more care is necessary here in loosening the integument under which the cartilage transplant must pass since it is even more firmly attached than is the case of back skin. Transplants in this location, however, maintain an even stronger contact with the overlying integument and usually produce a very pronounced bulging of the latter. This condition was considered of advantage for the purpose of the experiment.
All operated animals were maintained in separate aquaria, the water changed daily, and frequent observations made. Representative specimens were killed in 70% alcohol for photographic purposes. All specimens killed for subsequent histological study were fixed in an aqueous solution of Bouin’s fixative. The killing and fixation of R. palustris animals was made on the seventeenth to the twenty-ninth day following transplantation, although the great majority were fixed between the twenty-third and twenty-ninth days. R. catesbeiana specimens were fixed between the eighteenth and twenty-third days following annular tympanic cartilage transplantation.
The animals,, at the time of cartilage transplantation, were in approximately one of two stages of larval involution. Stage A larvae were newly metamorphosed young frogs in which the tail had been entirely resorbed or was represented by a black stump not over 8 mm. (R. palustris) or 15 mm. (R. catesbeiana) in length. In such individuals the normal tympanic membrane was about half developed, histologically, and had begun to assume a brownish tint as viewed externally (Pl. I, fig. 1). The annular tympanic cartilages extirpated from such individuals were of good size and quite circular in shape. Stage B animals were characterized by a much earlier stage of larval involution in which the tail was little shortened in length but definitely antrophied dorso-ventrally. The forelegs had emerged, however, and the animal was in general undergoing rapid metamorphosis. The annular tympanic cartilage in such individuals was usually little more than semicircular and about half size as compared with those of stage A animals. Since the cartilage had not as yet reached the definitive tympanic membrane region, there were of course no integumentary changes indicative of membrane development.
The following series of experiments were performed :
Series P (A). R. palustris control series (22 cases): Annular tympanic cartilage extirpated from stage A animals and transplanted, autoplastically, beneath integument of the back.
Series P (B). R. palustris control series (11 cases): Same as series P (A) except stage B animals used.
Series PAB (14 cases): Same as series P (A) except transplants killed by immersion in alcohol for 5 min.
Series PAB (B) (8 cases): Same as series P (B) except transplants killed by immersion in alcohol for 5 min.
Series PLAB (15 cases) : Same as series P (A) except transplants killed by immersion in alcohol for 30 min.
Series PLAB (B) (7 cases) : Same as series P (B) except transplants killed by immersion in alcohol for 30 min.
Series PCB (14 cases) : Same as series P (A) except transplants killed by immersion in saturated chlor etone solution for 10 min.
Series PCB (B) (8 cases): Same as series P (B) except transplants killed by immersion in saturated chloretone solution for 10 min.
Series PDB (16 cases): Same as series P (A) except transplants killed by exposure to strong direct sunlight for 1 hr.
Series PDB (B) (9 cases): Same as series P (B) except transplants killed by exposure to strong direct sunlight for 1 hr.
Series PWB (15 cases): Same as series P (A) except transplants killed by boiling in water for 5 min.
Series C (A). R. catesbeiana control series (12 cases) : Annular tympanic cartilage extirpated from stage A animals and transplanted, autoplastically, beneath integument of upper jaw.
Series CAN (7 cases) : Same as series C (A) except transplants killed by immersion in alcohol for 5 min.
Series CLAN (8 cases): Same as series C (A) except transplants killed by immersion in alcohol for 30 min.
Series CCN (6 cases) : Same as series C (A) except transplants killed by immersion in saturated chloretone solution for 10 min.
Series CDN (9 cases) : Same as series C (A) except transplants killed by exposure to strong direct sunlight for 1 hr.
Series CWN (8 cases) : Same as series C (A) except transplants killed by boiling in water for 5 min.
R. palustris transplantations
(1) Series P (A) and P (B)
In both these series the transplants remained large and of normal circular outline, as could be readily ascertained by the shape and degree of bulging produced in the back integument under which they had previously been transplanted (Pl. I, fig. 2). There were no indications, however, that any enlargement or growth of the cartilage implants had occurred even in those cases where fixation of the animal was not made until the twenty-eighth day. The integument above the transplants was usually under strong pressure (series P (A)) due to the presence of the latter, and under slight to moderate extension in series P (B) where the size of the cartilage implants was much less. In a few series P (A) animals, a brownish coloration, indicative of late membrane formation, was attained. Usually, however, the same greenish, black or grey tint was maintained characteristic of the back integument of newly metamorphosed frogs. Parenthetically, it may be stated that the back integument of metamorphosed R. palustris larvae exhibits a wide variation in coloration between different individuals. In all cases where a brownish tint was attained and in many where it was not, a lack of pigmentation developed in those areas of the skin directly over the peripheral portions of the transplant. To the eye, such areas appeared more or less grey and translucent. Although the translucency was usually confined to isolated spots over the transplant rim, several instances were noted in which a complete translucent ring developed. The developnient of such translucency constitutes one of the normal external manifestations of late membrane formation.
A histological study of the various representative cases sectioned, corroborated the external indications as to the integrity of the transplants. There were no signs of cartilage degeneration and this was true not only of the annular tympanic cartilage but also of the columella which had frequently been included in the original implant (Pl. 2, figs. 14, 17). The areas of integument overlying the more central regions of the transplants gave evidence of pronounced glandular degeneration and correlated reduction in thickness of the stratum spongiosum. The stratum compactum was also undergoing rapid dissociation and fibril degeneration. In several cases there was evidence of the fusion of fibrils within a dissociated bundle. The epidermis in the more central regions was, however, practically unchanged. The integument above and in strong contact with the periphery of the transplant exhibited an even more advanced stage of membrane formation. Here, the epidermis was thinner and corresponded closely to that of normal fully developed tympanic membrane, while the stratum spongiosum was frequently entirely lacking or much reduced in thickness coupled with total glandular degeneration. Of particular interest, in many cases, was the development of the double set of fibrils in the stratum compactum which is characteristic of fully formed tympanic membrane. As previously described by Tokura (1925) for R. nigromaculata and by the writer (1928, 1934 a) for R. pipiens, R. sylvatica and R. clamitans, these consisted of an inner thick region of horizontal fibrils and an outer region of smaller fibres running vertically, which cross with those of the inner region to form a complex network. When it is added that the subcutaneous connective tissue was frequently entirely absent in this region, the conclusion can be drawn that complete membrane development had been induced (Pl. II, fig. 17). It should also be noted that where the columella had been included in the original transplant, the lamina propria overlying the former was usually greatly thickened and composed chiefly of yellow elastic fibres; demonstrated by staining with Weigert’s resorcin-fuchsin (Pl. II, fig. 14). In general, it can be stated that membrane induction was somewhat more pronounced in series P (B) than in series P (A) animals.
(2) Series PAB, PAB (B), PLAB, PCB and PWB
Although the original cartilage transplants of these series were of the same size as those of the comparable control series P (A) and P (B), considerable degeneration ensued following their transplantation. This was evidenced by the slight bulging of the overlying integument at the time the animals were fixed and also by a study of the histological sections. The degeneration was particularly marked in series PAB, PAB (B) and PWB, in which many cases occurred involving no external signs of integumentary bulging indicative of the persistence of an underlying transplant. Sections revealed, however, that a remnant of the cartilage implant always remained. In a few cases the transplants underwent only moderate degeneration and the integument above consequently remained under considerable pressure (Pl. I, fig. 3). The transplants of series PLAB and PCB, although undergoing post-operative degeneration to varying degrees, were in general better preserved. In fact several cases were observed (Pl. I, fig. 4) in which, apparently, no diminution in size had taken place. There were only a few instances in which distinct macroscopic signs of tympanic membrane formation could be detected in the integument overlying the transplants. These occurred in series PLAB and consisted chiefly of coloration changes resulting in the attainment of a brownish tint in two instances. Other than these two cases, there were no other examples of macroscopic signs of membrane formation in any of the five series under discussion.
A histological examination of the transplants revealed decided cartilage degeneration, especially in series PAB and PWB. In series PAB the cartilage cells appeared to be greatly enlarged and vacuolated, while the matrix had decreased in amount (Pl. II, fig. 13). Series PWB transplant cartilage resembled that of series PAB except that the cells usually appeared to be polyhedral in shape and apparently somewhat collapsed. This may very well be, however, simply a later stage of degeneration as represented by the enlarged vacuolated condition. In series PLAB the matrix was very little decreased in amount, while the cartilage cells were only slightly enlarged. Series PCB transplants appeared to be particularly resistant to degeneration and there were many cases in which the cartilage had remained practically the same as when originally implanted.
The histological changes in the integument covering and in contact with the transplants were of that type characteristic of the first or degenerative phase of tympanic membrane formation. These included a general thinning of all layers together with the partial to total degeneration of glandular elements in the stratum spongiosum. The most pronounced changes were observed in series PAB (B), while least degeneration was recorded for series PWB. In all five series, however, the integumentary changes over the periphery of the cartilage transplants were more accentuated as compared with those prevalent in the more central areas (Pl. II, fig. 18). Finally, it should be noted that there were no instances in which the later developmental phase of membrane formation was exhibited.
(3) Series PLAB (B), PCB (B), PDB and PDB (B)
The degree of integumentary elevation, indicative roughly of transplant integrity in these series, may be briefly summarized as follows. There were no instances of bulging in series PCB (B), only three in series PDB (B), and but two in series PLAB (B). Individuals of series PDB, however, always showed some degree of integumentary bulging, which in some cases was very pronounced. Usually, the size and shape of the elevation closely approximated that which was produced at the time of the original transplantation, suggesting little if any degeneration of the transplant. Although two cases of integumentary translucency developed in series PDB, there were no other external signs of membrane formation in evidence in these individuals or of any description in the remaining thirty-eight cases of the four series in question.
Histological sections revealed a variety of pictures as regards the condition of the transplants. All cartilages of series PDB (Pl. II, fig. 15) and five of series PDB (B) appeared quite normal. The matrix was of average amount and had taken a natural light blue stain while the, cartilage cells were very little, if any, enlarged. Four of series PDB (B) were of unusual interest in that the peripheral portions of the transplants were very evidently undergoing rapid degeneration and dissociation. In this same location were present scattered, deep blue staining, solid masses having the appearance of early bone formation. Occasional irregularly shaped structures, which may have been degenerating cartilage cells, were found imbedded in these masses. The condition of series PLAB (B) and PCB (B) transplants was even more striking for here usually the entire transplant had been obliterated, at least as far as their original cartilaginous make-up was concerned. In their place occurred large irregularly shaped masses of the same texture and staining reaction as previously noted for the four series PDB (B) transplants cited above. The mass, although usually maintaining continuity between its various parts, contained large cavities, while the peripheral portions were characterized by trabecular-like projections and deep pockets ; the latter filled with loose connective tissue and connective tissue cells (Pl. II, fig. 16). Somewhat similar cases were encountered in series PWB except that the degenerating cartilage cells, were more numerous (Pl. II, fig-18).
The microscopic appearance of the integument overlying the transplant region in these four series gave even more evidence of tympanic membrane transformation than previously described for the other R. palustris experimental series. In addition to a generalized thinning of all integumentary layers, considerable glandular degeneration occurred resulting frequently in their complete absence where the transplant rim was in strong contact with the skin (Pl. II, fig. 15). It was furthermore noted that certain areas of the stratum compactum were dissociating and undergoing degeneration. The development of a double set of stratum compactum fibres (typical of fully formed tympanic membrane and as noted in certain cases of the control series P (A) and P (B)) was not observed.
R. catesbeiana transplantations
(1) Series C (A)
The transplants of all series C (A) animals maintained moderate to strong pressure against the skin of the upper jaw until such time as the young frogs were fixed for histological study. This resulted in a pronounced elevation of the integument, the height and diameter of which served to indicate that while the cartilage had not increased in size, it had obviously undergone no decided degeneration (Pl. I, fig. 5). Histological examination, moreover, showed that the normal cytological condition of both the annular tympanic cartilage (Pl. I, fig. 9) and the columella (Pl. I, fig. 10) had remained unchanged.
The external appearance of the integument overlying the transplants underwent changes indicative of membrane formation. In three cases this consisted of a coloration change in which a circular band of skin in contact with the rim of the transplant had turned to a brownish shade. The more central areas of the integument, however, still remained greenish in coloration similar to that of the adjacent skin of the jaw. This condition is not abnormal, since in the normal development of the tympanic membrane the integumentary transformations involved always begin at the periphery and progress towards the centre. In the remaining nine cases, however, development had proceeded to the extent where the entire circular area over the transplant was of a nearly uniform brownish tint. Considerable translucency also developed in these individuals. The general appearance compared quite favourably with that of the normal left-side tympanic membrane of the same animal (compare Pl. I, figs. 5, 7).
Sections made of the integument covering the transplants showed that although no decrease in thickness usually occurred in the more central areas, that associated with the transplant rim was invariably much thinner as the result of degenerative changes (Pl. I, fig. 9). The epidermis of this region was slightly decreased in thickness, which was also true of the stratum spongiosum. The failure of the latter layer to degenerate completely was probably due to the presence of many large mucous glands, which are characteristic of the jaw integument in 7?. catesbeiana much more so than is the skin of the normal tympanic membrane region. It was noted, however, that the nearer to the rim of the transplant one approached, the smaller the glands became, while many were apparently undergoing degeneration. The stratum compactum gave more clear-cut evidence of membrane transformation. In addition to the loss of its convoluted appearance coupled with the partial degeneration of its constituent fibres, there was evidence of the development of the double set of fibres typical of fully formed tympanic membrane. Mention should also be made of the fact that where the columella had been incorporated in the transplant the lamina propria covering its tip was greatly thickened and composed almost entirely of yellow elastic fibres (Pl. I, fig. 10).
(2) Series CAN, CLAN, CCN, CDN and CWN
The integrity of the transplants in the R. catesbeiana experimental series, as evidenced by the degree of integumentary bulging, was subject to considerable variation. There were but four cases in which the cartilage implants had degenerated to the extent of no longer producing an elevation of the adjacent skin. There were two each of these in series CAN and CDN. In the remaining thirty-four individuals, however, slight to strong bulging of the integument was maintained, the majority producing a moderate elevation of the skin. Frequently, the elevated area was not completely circular in outline, indicating either a mechanical doubling over of the transplant, due to the difficulty of proper implantation, or (as later shown by histological examination) to a more rapid degeneration of one region of the cartilage as compared with another (Pl. I, fig. 6). It is of interest to compare this condition with the appearance of the half-formed membrane in the ear region on the same side of the animal (Pl. I, fig. 8). At the time of annular tympanic cartilage extirpation for transplantation further anterior, the integument of this region had evidenced distinct macroscopic signs of membrane formation which have persisted for 19 days in this case. In the absence of the cartilage, however, the area affected has not increased in size as can be determined by comparison with the left-side normal membrane of the same individual (Pl. I, fig. 7). There were only two series in which macroscopic changes occurred in the integument, overlying the transplants, indicative of membrane formation. In series CCN, two instances occurred in which a slight translucency was evident, particularly in those areas in contact with the periphery of the transplant. In series CWN, four cases were noted in which not only translucency developed but a distinct brownish coloration was apparent.
Sections made through the transplant region showed that in all instances varying degrees of cartilage degeneration had taken place. Since the histological picture differed, considerably, between the various experimental series, a brief description for each is necessary. Series CAN transplants usually showed a normal amount of matrix but with the cartilage cells collapsed and misshapen. In series CLAN, the matrix was somewhat reduced in proportion, while the cellular elements were generally larger, ovoidal in shape, and definitely vacuolated (Pl. II, fig. 20). Series CCN transplants were usually degenerated more peripherally than centrally with cartilage cells slightly vacuolated (Pl. I, fig. 11). The transplants of series CDN were always undergoing rapid degeneration and as a result were greatly reduced in size. Furthermore, they were characterized by a minimum of matrix and large, vacuolated cells (Pl. I, fig. 12). Lastly, series CWN transplants were peculiar in that although the cartilage cells were almost invariably misshapen with large nuclei, they were nevertheless usually small in size and not vacuolated (Pl. II, fig. 19).
The integument over series CAN, CLAN, CCN and CDN transplants usually remained quite normal, histologically (Pl. I, figs. 11, 12; Pl. II, fig. 20). There was no appreciable thinning of the various skin layers while the large mucous and poison glands in the stratum spongiosum appeared quite normal as compared with those in adjacent integument. A few cases in series CCN were noted, however, in which slight glandular degeneration was apparent. In contrast to the above series, the integument in contact with series CWN transplants showed slight to pronounced signs of degeneration. This usually resulted in a definite thinning of the skin, especially at the expense of the stratum spongiosum and stratum compactum layers. In the former, pronounced glandular degeneration was usually seen, while the latter layer was characterized by dissociation and degeneration of its constituent fibre bundles (Pl. II, fig. 19). The development of fully formed tympanic membrane stratum compactum fibres was not observed, however. Mention should also be made of the fact that scattered ossifying (?) masses were frequently seen in series CCN, CDN and CWN in the loose connective tissue surrounding the transplants (Pl. I, figs. 11, 12; Pl. II, fig. 19). No generalized or central ossification of the cartilage was observable.
Previous work of the writer has demonstrated the potentiality of side and back integument to undergo tympanic membrane transformation when in contact with transplanted annular tympanic cartilage. The results of the present work on R. catesbeiana lead one to conclude that jaw integument must be added to this list. The latter, however, is apparently not as susceptible to such transformation as is true of the former two integuments. This is quite probably due to the presence of numerous and abnormally large glands in the stratum compactum layer, which must first undergo degeneration before the developmental phases of membrane formation can take place.
The results of the present work furthermore demonstrate and emphasize the point that membrane induction is essentially a contact phenomenon. The greater the area of contact with the cartilage and the stronger the pressure of the latter against the integument, the more certain and rapid will be the membrane formation. Just how the pressure factor operates is somewhat problematic. Inasmuch as integument under the strong pressure of an implanted glass bead (Helff, 1926) will undergo complete degeneration and even perforate, it is logical to inquire whether the degenerative phase of membrane formation might be due to this factor alone. Although a few isolated cases of minute perforations occurred in the skin over the transplants, subsequent histological study of these instances and of all other integument sectioned showed that the sequence of degenerative changes that had occurred were typical of tympanic membrane formation and not of the generalized type characteristic of “pressure atrophy”. It seems more likely, therefore, that strong contact involving pressure functions chiefly to insure a more rapid transference of the chemical or other inductive factors in the annular tympanic cartilage to the integument.
It is of interest to inquire whether the degenerative changes which occurred in the integument of the experimental series were due simply to contact with a dead degenerating mass of tissue (the transplant) or represented the normal first phase of tympanic membrane formation. That the latter conception was no doubt the case, is borne out by a consideration of the following points : In the first place, the degeneration which occurred did not involve all integumentary layers at the same time or with a uniform rate, which is apt to be the case where disintegration is caused solely under the influence of adjacent decomposing tissue. As previously stated, the various degenerative changes occurred at varying rates and in a sequence typical of normal membrane formation. Secondly, complete degeneration of the overlying integument did not occur but only to that extent normally characteristic of membrane formation. Finally, there were many instances in the R. catesbeiana series in which no integumentary dissociation occurred at all in spite of the fact that the transplants were undergoing pronounced degeneration.
An analysis and comparison of the results of the fourteen experimental series (exclusive of the control series) reported in the present paper, lead one to conclude that there exists no consistent correlation between the type of killing method employed for the cartilage transplants and the latter’s subsequent average degree of degeneration. For example, transplants immersed in alcohol for 30 min., underwent practically no degeneration in series PLAB, moderate degeneration in series CLAB, and excessive degeneration in series PLAB (B). The only definite correlation noted in this regard concerned those transplants killed by chloretone solutions, in which moderate cartilage degeneration occurred in both stages A and B R. palustris series and the R. catesbeiana series. Within any one series, however, the results were usually more consistent, although instances of rather wide variation were sometimes recorded. A still further analysis was made to determine whether a correlation was evident between the degree of transplant degeneration and the extent of tympanic membrane induction. Although instances of non-correlation were found, it may be said that in general the greater the cartilage degeneration, the less the membrane induction in the adjacent integument. This was true for both stages A and B R. palustris series and the R. catesbeiana series.
In considering the various cytological changes characteristic of membrane formation which occurred in the integument, the peculiar and transitory hypertrophy of the basal layer of epidermal cells (described by the writer, 1928, for R. pipiens) was not observed in either R. palustris or R. catesbeiana. “ It is possible, of course, that this particular degenerative feature is not an essential feature of membrane formation in all forms and is lacking in the two species in question. Again, it may be that in the present instances, it occupies a different position in the sequence of histological changes and normally occurs at a later stage of membrane formation than was attained in any of the transplant series. This is considered rather unlikely, however, since such a late development as typical membrane stratum compactum fibres was obtained quite frequently in the control series of both species. What seems more probable is that the transforming integument was not fixed for sectioning at the right stage of membrane formation. In this regard it may be stated that the phenomenon of epidermal cellular hypertrophy occupies a very brief period of time and its presence in R. pipiens only detected after sectioning integuments in many closely related stages of membrane transformation.
The development of a double set of fibres in the stratum compactum constitutes one of the last steps in integumentary transformation and hence practically characterizes complete membrane formation. It is of interest and importance, therefore, to inquire why some of the control animals and all of the experimental individuals failed to attain this degree of development. As for the stage A control animals, it can only be suggested that when the annular tympanic cartilage was removed for transplantation, the normal tympanic membrane was approximately already halfformed. Hence, considerable of the total inductive power of the cartilage had already been expended. Furthermore, when transplanted to its new location, the cartilage, although showing no signs of degeneration, failed to increase in size and no doubt became less potent as regards membrane induction. Stage B transplants, although in a more active stage of membrane induction, no doubt failed to induce complete integumentary transformation for the same reason. It will be recalled in this regard, however, that membrane formation in the stage B control series was usually more advanced than was true of the stage A control series.
The failure of the experimental series to develop membranes to as full an extent as the control series, calls for another explanation. In the first place, the killing of the cartilage prior to its implantation, quite probably made impossible the subsequent secretion or formation of additional inductive substances. Thus, as the inductive substances present were gradually utilized, there came a time when the supply was exhausted, resulting in a cessation of membrane induction. Still another contributory factor should be mentioned in this regard. As has previously been mentioned, most transplants of the experimental series underwent disintegration and in general the greater the degree of cartilage degeneration, the less the extent of membrane formation. This would seem to imply, therefore, that during the disintegration of the cartilage, substances are released or formed which are capable of neutralizing the inductive substances or of counteracting their effect on the integument.
The results of experiments described in the present paper lead the writer to believe that the integumentary portion of the anuran tympanic membrane is developed largely, and possibly entirely, through the action of an inductive substance or substances produced by the developing annular tympanic cartilage. Although the exact nature of these substances is still problematic, it seems quite logical to predict that they will ultimately be shown to be either direct secretory products of the cartilage cells, products arising from the ageing of the matrix or other chemical changes occurring within it, or perhaps a combination of both. In any event, a chemical inductor is, obviously, strongly indicated. There can be no question of the fact that the inductive substances cannot become functional until transmitted to the larval integument of the ear region through contact of the annular tympanic cartilage. Furthermore, continued contact with the cartilage during the entire developmental stage of the membrane, is essential if complete integumentary transformation into tympanic membrane is to occur. That this is a case of complete dependence is further emphasized by the fact that even enlargement of the membrane during the post-metamorphic life of the frog is entirely dependent on the growth and corresponding enlargement of the annular tympanic cartilage with which it must maintain contact (Helff, 1934 a). There is some evidence serving to indicate that the gradual thinning of the integumentary layers (particularly the epidermis), during membrane formation, may be partially due to the action of increased tension within these layers ; the latter being due to pressure exerted by the developing annular tympanic cartilage.
EXPLANATION OF PLATES I AND II
The terms “membrane “and “tympanic membrane “aa used henceforth in the present paper, refer only to the integumentary portion of the tympanic membrane.
Inasmuch as two or more series usually exhibited practically the same reactions as regards the degree of tympanic membrane induction, they will be discussed together in order to eliminate, as far as possible, needless repetition of results.