484 ANNUAL. KEPORT SMITHSONIAN INSTITUTION, 1910. 



One of the points of difficulty about tlie theory that one sex is 

 homozygous and the other heterozygous in respect of the sex deter- 

 minants is that it appears inconsistent with Wilson's theory based on 

 the study of " idiochromosomes." But phenomena such as he de- 

 scribes have hardly been observed outside most orders of insects and 

 possibly Arachnids, and are probably not of universal occurrence. 

 And if all individuals of one sex are heterozj-gous, those of the other 

 homozygous in sex, it may be imagined that in the homozygous sex 

 two sex determinants would not be necessary; one of them might 

 become vestigial, as Wilson describes, if at the same time spermatozoa 

 bearing such a vestigial determinant can only conjugate with eggs of 

 one kind. But it must be" admitted that any suggestion of selective 

 fertilization interferes with the extreme simplicity of the theory out- 

 lined above.^ 



The hypothesis here described not only explains the cases "which 

 led up to it and such facts as the effects of castration, but also ac- 

 counts for the phenomena of sexual dimorphism and the inheritance 

 of some structures by one sex only. But at present the more complex 

 cases of sexual polymorphism, such, for example, as are known in the 

 African butterflies of the genus Papilio, still remain obscure, although 

 it is probable that when we have more extensive records of breeding 

 experiments, these also will be found to fall into line. And it should 

 be explained that some forms of sex-limited inheritance are of quite 

 a different nature — e. g., color blindness and the disease ha?mophilia in 

 man. In these diseases the abnormal condition is dominant in one 

 sex (male) and recessive in the other and may appear in the female 

 if both parents are tainted. Possibly a combination of some con- 

 dition of this kind with sex relations, such as we find in A. grossu- 

 lariata and the Cinnamon Canary, may ultimately be found to ac- 

 count for the complex sexual polymorphism found in the African 

 Papilios. 



We have now sketched the principal lines of evidence which have 

 been collected in recent years, pointing to the conclusion that the 

 sex determinant is present in the germ cell and is probably com- 

 parable in nature with a Mendelian unit. In a paper of this kind 

 it is clearly out of place to attempt even to mention a tithe of the 

 numerous hypothesis concerning sex which have been advanced even in 



1 Wilson has recently (Science, vol. 29, Jan., 1909, p. 53) put forward a fresh sugges- 

 tion, viz, that the idiochromosomes do not bear the determinants for maleness or fe- 

 maleness as such, but that one idiochromosome in the fertilized egg causes it to develop 

 into a male, two into a female, so that the difference is rather quantitative than quali- 

 tative. Castle (Science, vol. 29, Mar., 1909, p. 395) has taken up this idea with the fur- 

 ther suggestion that while some species are as Wilson suggests, in others the presence 

 of one idiochromosome determines femaleness and absence of any idiochromosome at 

 all brings about maleness. If this last condition should be found to exist in Atraxas 

 grossulartata, it would then fall into agreement. In this connection it is of interest 

 that cases such as Wilson describes have beteti observed in most of the chief ordters of 

 insects tfut not in the Lepidoptera. 



