354 BOTANICAL GAZETTE [NOVEMBER 



as discussed in case XII, were sound, and that there was not simply 

 the temporary disappearance of the hermaphrodite character 

 through some thinkable vagary of dominance in the F x , 5 males 

 in 08125 were tested in crosses with 2 different females. The 

 resultant progenies consisted of 333 females and 168 males. Not 

 a single hermaphrodite appeared, thus convincingly supporting 



333 168 



FIG. 15. Model pedigree for case XIV 



the view that the appearance of hermaphroditism in C and D was 

 illusive, and that they were therefore only superficially like the 

 genetic hermaphrodites A and B. These results fully justify my 

 conclusion that the hermaphrodites of Lychnis dioica belong to 

 two genotypes, one of which is the same as the normal male, the 

 other different from it. 



Discussion and conclusions 



Although these data from the breeding of hermaphrodites of 

 Lychnis dioica are presented in fourteen sections, each representing 

 a somewhat different direction of attack upon the genetic problems 

 involved, the results under the various sections are remarkably 

 consistent. The hermaphrodites are clearly of two kinds. Those 



