THE MOSAIC STAGE OF DIFFERENTIATION 259 



tance during ontogeny. In frogs, on the other hand, the inductive 

 substances must be capable of moderate diffusion. The effect here 

 recalls the graded distribution of limb-forming potencies in the 

 limb-field of Urodeles (p. 223), and the probable diffusion of in- 

 ductive substances from the presumptive dorsal lip region during 

 cleavage (pp. 134, 311). 



In Urodeles, on the other hand, diffusion is so complete that there 

 is no evidence of any concentration-gradient. It is possible, though 

 not demonstrated, that here the substance diffuses into the blood- 

 stream and is transported by it. We have thus within the boun- 

 daries of one class of vertebrates either a complete or a nearly com- 

 plete gradation between sharply-localised morphogenetic substances 

 and freely-circulating hormones. It has indeed been suggested that 

 the sex-hormones of the adult gonads are identical or homologous 

 with these morphogenetic substances produced by the embryonic 

 cortex and medulla, merely differing in being secreted into the 

 blood-stream instead of soaking through the tissues. 



In support of it we find indications in cases of hermaphroditism 

 or asymmetrical development of gonads that the accessory sex 

 characters (male and female ducts), whose differentiation is known 

 to be dependent upon sex-hormones, are locally better developed 

 in regions of greater development of the gonad of corresponding 

 sex.i While no certainty can yet be reached on this point, it is a 

 valuable suggestion to guide further research. In any event, it is 

 clear that the physico-chemical conditions regulating diffusibility 

 of morphogenetic substances are of great importance in ontogeny. 



In Urodele parabiosis the failure of the medulla of genetic female 

 gonads to differentiate in the male direction after regression of the 

 cortex under the influence of the male partner is in marked contrast 

 with the results in frogs. It appears to be general in the subclass. ^ 

 No adequate explanation is yet forthcoming : in general, it appears 

 to link up with the subject of metaplasia (p. 211). The medulla of 

 all female Amphibia differentiates in a specifically female direction ; 

 that of the Anura retains its bisexual potency, and is capable of 

 metaplasia and male histo-differentiation ; that of the Urodela loses 

 the original bisexual potency and is capable only of continued 

 development or of regression within the limits of female-type 

 1 Witschi, 1933 B. 2 Witschi, I933 A. 



17-2 



