MENDELISM 447 



possesses the characters of the opposite sex to that to which 

 it belongs in a latent state. It is well known that the removal 

 of the generative organs of an animal often results in the 

 appearance of the secondary sexual characters of the opposite 

 sex ; for example, the human eunuch has the high voice and 

 the distribution of fatty tissue which is characteristic of the 

 female. The most decisive case is perhaps that which is cited 

 by Darwin, who relates that when the domestic cock is crossed 

 with the hen pheasant, those of the offspring which happen to 

 be males have the secondary sexual characters of the male pheasant. 

 There is abundant evidence over and above that cited by 

 Prof. Castle for the truth of his contention that members of 

 both sexes are potentially hermaphrodite. So high an authority 

 as Dr. Curt Herbst concludes, after an exhaustive scrutiny of 

 a great mass of evidence, first, that the presence in a functional 

 state of the generative organs of a particular sex are an indis- 

 pensable condition for the elaboration of the secondary sexual 

 characters of that sex ; and, secondly, that the functional 

 generative organ exercises a negative influence — preventing the 

 appearance of the secondary sexual characters of the opposite sex} 



To return therefore to our argument. Prof. Castle contends — 

 and we have seen that his contention is supported by inde- 

 pendent evidence — that the individuals of a normally bi-sexual 

 species are not sexually pure, but potentially hermaphrodite. 

 From this he concludes that the homozygous individuals ? ? 

 and cfc? do not exist. He accounts for their non-existence by 

 supposing that the unions between ova and spermatozoa 

 bearing similar sexes are infertile. And he adduces evidence, 

 other than the a priori considerations we have spoken of, for 

 this view. Only half the unions are fertile, namely, those 

 between ova and spermatozoa bearing dissimilar sexes. And, 

 in this way, only heterozygotes are produced which are 

 capable of producing both sexes. Furthermore, to account 

 for the fact that half of these heterozygotes display the male 

 and half the female primary and sexual characters, we have 

 to suppose that in the case of the allelomorphic pair "male" 

 and "female," we have alternative dominance of either of the 

 pair, as opposed to the fixed dominance of one of the pair 

 which obtains in the ordinary Mendelian pairs. 



We have stated that it has been shown that the characters 



* Formative Retse, p. 76. 



