CH. x] Nature of Sex 1 65 



analogy. Mendelian experiments have shown that in all 

 cases which can be adequately investigated, a mixture of 

 distinct zygotic types occurring in one family is due to 

 gametic differentiation. Sex is a case of such mixture 

 of zygotes and the presumption is thus created that the 

 case is comparable in causation with those amenable to more 

 direct analysis. This suggestion was made in 1902 (Rep. 

 Evol. Ctee, i. p. 138), as a natural deduction from Mendelian 

 discoveries and it is interesting to know that the same 

 possibility occurred to Mendel himself (197, p. 241). 



The argument from analogy may perhaps be carried 

 a step further. If the distinction between the sexes is the 

 result of gametic differentiation, the fact that in ordinary 

 cases the two sexes are produced in equal numbers must 

 be taken as a strong indication that one sex is heterozygous 

 in respect of sex-character, and the other homozygous. That, 

 as we now well know, is the simplest way by which numerical 

 equality in the production of two types is brought about. 



In his important paper on this subject Castle argued 

 that both sexes are to be regarded as heterozygous in sex. 

 But in order to apply this suggestion two serious assump- 

 tions are required. First, since male and female are each 

 regarded as heterozygous in sex, unions between ova and 

 male cells bearing similar sex-factors must be assumed to 

 be impossible. At the present time no fact can be adduced 

 which negatives this assumption, and there are indeed 

 general considerations which may be appealed to as render- 

 ing it somewhat probable. 



The second assumption involved is more serious. As 

 both sexes are regarded as heterozygotes containing the 

 same factors, the nature of their dissimilarity is still so far 

 unrepresented. In view of this difficulty Castle regards 

 dominance as a matter of chance. Doncaster, in respect 

 of a case to be discussed below, suggested that dominance 

 may belong exclusively to the cells coming from one parent, 

 say, for example, those of the mother. These postulates 

 seem unsatisfactory and do not accord well with anything 

 that we know in regard to the nature of dominance else- 

 where observed ; for the applicability of all schemes hitherto 

 discovered lies in the fact that it has been found possible to 

 represent zygotic composition and structure as determined 



