INHERITANCE IN HYDATINA 85 



ments described in this paper, the only two parthenogenetic Hnes 

 (C X D and D X C) that were derived from crosses between 

 other Hnes, were intermediate between their parents in respect 

 to the sex-ratio. This is what would be expected if the sex- 

 ratio were dependent on a number of genes. But in a previous 

 paper (Shull '11) I described a case in which a cross gave a 

 much higher proportion of male-producers than either of its par- 

 ent lines. There is a bare possibility that the two lines crossed 

 in this earlier experiment were old; that their percentage of 

 male-producers had been reduced by long continued partheno- 

 genesis, as I have shown (Shull '12 a) to be a not infrequent 

 occurrence ; and that when these lines were young their sex-ratios 

 were so high as to have made the Fj which was later obtained by 

 crossing, and which was itself a young line, actually intermediate 

 between them. This possibility was discussed in the earlier 

 paper, but the view taken that it was probably not the correct 

 explanation. If my earlier supposition was well grounded, there 

 is still no simple explanation for the inheritance of the propor- 

 tion of male-producers. 



The assumption of numerous genes for the sex-ratio would 

 harmonize with the fact that inbreeding usually does not change 

 the sex-ratio in the resulting parthenogenetic line. Repeated 

 inbreeding showed little alteration in the sex-ratio. Since but 

 one or two eggs were selected from each lot to breed partheno- 

 genetic lines, the chances were many to one that these would 

 be similar to their parent line. The few exceptions, such as the 

 low percentage of male-producers in the high viability line in 

 Li (13.6), and the high percentage in H: from H (35.8), may be 

 regarded as due to selecting the extremes for parents, or, which 

 seems to me more probable, to the fact that the lines were reared 

 only a few generations during a period of few male-producers 

 in the former line, and through a period of many male-producers 

 in the latter line. 



