144 EMIL WITSCHI. 



The result of 29 one-sexed and 27 two-sexed pairs is quite 

 close to expectation (28 2.5). Among the total of 112 indi- 

 viduals we find 59 <? and 53 9 (the latter group including the 

 sex-changing animals). The deviation from the mean is less 

 than the single P.E. It can be considered, therefore, as well 

 established, that there was no sex transformation before the time 

 of sex-differentiation. 



In discussing the fact of the dominance of the male sex in 

 the free-martin, Lillie is inclined to consider the time-factor as 

 an essential one. The observations of his students (Chapin, 

 1917 and Bascom, 1923), relating to the earlier development of 

 interstitial cells in the testes and the later differentiation of 

 the ovaries, seem to favor the theory of dominance of the male 

 twin by earlier hormone production. Lillie admits, however, 

 that other causes may also exist. In the case of our frog twins 

 it is evident that the time-factor has not the suggested sig- 

 nificance. The male-differentiating factor reaches the female 

 gonads after the ovaries have acquired a relatively advanced 

 stage of development. This means, that the female-differ- 

 entiating factor is already in full action. In the resulting 

 competition of both, the male-differentiator acts first as an 

 inhibitor and later on as a suppressor of the female one. This 

 is in strict agreement with the conclusions derived from the 

 author's observations on hermaphroditism (1914; 1921). I 

 think, that there can no longer be any doubt about the antagonistic 

 character of these two factors. 



Lillie (1923) and Bissonnette (1924) report on some earlier 

 stages of the free-martins with rather highly developed ovarial 

 structures. It seems not improbable that the female co-twins 

 started with a typical female sex-differentiation. Such cases 

 would lead to the same conclusions as the parabiotic frog twins. 

 The assumption, that only the male is producing hormones in 

 an early time, would apparently not be sufficient to explain the 

 suppression of the active ovarial development. There seems to 

 be no escape from the postulated sexual antagonism. 



Some years ago Goldschmidt (1920) put forward the idea, 

 that the sex-determining genes act like enzymes, stimulating 

 the production of sex differentiating hormones. The term 



