n8 BOTANICAL GAZETTE [FEBRUARY 



independence of the gene whose presence results in the hermaph- 

 rodite modification of the male. If the females may carry the 

 hermaphrodite modification as a latent character, and if the hermaph- 

 rodites of the Fj. differ in their gametic composition from those of 

 the Pj generation, the hypothesis that the female is a positive homo- 

 zygote may be substantiated; but if the hermaphrodites of the F x 

 are identical in their behavior with those of the P 5 , and if the her- 

 maphrodite modification cannot be transmitted through the female, 

 the results will support the hypothesis that the female is a negative 

 homozygote. 



It is not at all improbable that the hermaphrodites discussed 

 above as C and D differed from A and B in having the gene for the 

 hermaphrodite modification in the heterozygous state, while A and 

 B were homozygous with respect to this gene. If this should be 

 correct, then C and D were of the type which will appear in the F 2 

 of crosses between normal females and such hermaphrodites as A and 

 B, in case the hermaphrodite modification is independent of the 

 male germ cells. The occurrence of such hermaphrodites as C 

 and D may be accepted as rather favorable to the view that the female 

 is a positive homozygote. 



There is one fundamental difficulty with any hypothesis which 

 assumes that the hermaphrodites are heterozygous. , The eggs of these 

 hermaphrodites are of a single kind with respect to sex, all carrying 

 the female character. This difficulty is inherent in the material itself, 

 however, and is not opposed in any way to the assumption that the 

 normal females are either positive or negative homozygotes, for the 

 male germ cells of the hermaphrodites are clearly of two types as 

 required by the hypothesis that these hermaphrodite plants are hetero- 

 zygous. Cytological investigations of oogenesis in the hermaphrodites 

 may perhaps give a complete solution to the exceptional situation pre- 

 sented by the egg cells. If the females are positive homozygotes 

 there may be simply a failure of those eggs to develop which do not 

 possess the gene for the female sex. If the female is a negative homo- 

 zygote, there may be a failure of those eggs which possess the male 

 gene, or there may be an exclusion of this gene during oogenesis. 



While I have left out of account the two hermaphrodite individuals 

 in the cross (08116) between a hermaphrodite and a normal male, 



