THE EFFFXT OF SELECTION UPON MENDELIAN 

 CHARACTERS MANIFESTED IN ONE SEX ONLY^ 



W. E. CASTLE 



In the Journal of Experimental Zoology, vol 7, No. 4. Miss 

 MCracken reports extensive and important observations upon 

 the inheritance of the alternative race-characters of silk-moths, 

 univoltinism (one generation a year) and bivoltinism (two gene- 

 rations a year). The inheritance she characterizes as non-Men- 

 delian on the ground (i) that when a cross is made involving the 

 contrasted race-characters neither condition appears to be uni- 

 formly dominant and (2) that in subsequent generations neither 

 condition can be wholly freed from the other, that is neither be- 

 haves as an extracted Mendelian recessive, and (3) that the 

 proportions of univoltins to bivoltins following a cross does not 

 approximate an ordinary Mendelian ratio, but changes from 

 generation to generation according as selection is made for one 

 condition or the other. 



It seems to me, however, that these reasons are not sufficient 

 to establish the non-Mendelian character of the inheritance, but on 

 the contrary are entirely consistent with a Mendelian interpre- 

 tation. In the first place it is to be observed that the inheritance 

 is strictly alternative. All broods are either bivoltin or univoltin 

 in character. This is prima facie evidence in favor of a Mende- 

 lian interpretation. The two essential features of Mendelian in- 

 heritance, dominance and segregation, are both strongly in evi- 

 dence throughout the entire experiment. The only obgcure points 

 from a Mendelian standpoint are these: (i) Is dominance re- 

 versed within the series, and (2) are the ratios obtained Mendelian 

 ratios. 



Confusion in the interpretation arises fron the fact that the 

 univoltin or bivoltin condition is manifested only in the female 

 line, though transmitted through both sexes. The female silk- 

 moth hatched in the sprmg of the year lays a batch of eggs and 



^ Contributions from the Laboratory of Genetics, Bussey Institution, Harvard University, No. 5. 



