344 BOTANICAL GAZETTE [NOVEMBER 



case IV. The genetic equivalence of the different hermaphrodites 

 again stands out clearly in these results, and when the ratios of 

 the two series are compared, it is found that the percentage of 

 hermaphrodites produced by the hermaphrodite offspring of a 

 self-fertilized hermaphrodite is slightly lower than that produced 

 by the offspring of these cross-bred hermaphrodites, the former 

 producing only 38.2 per cent of hermaphrodites and the latter 

 42.6 per cent. The difference is too small to be of significance, 

 particularly in view of the fact that much wider differences than 

 this are found in families produced from different seed capsules 

 on a single plant when pollinated by a single male. It might have 

 been expected, perhaps, that a self-fertilized hermaphrodite would 

 have produced a larger percentage of hermaphrodites than would 

 be produced by the same hermaphrodite crossed upon a female 

 of a normal family. The fact that such a result does not appear 

 is further proof that, although the hermaphrodite is a heterozygote, 

 its egg cells are of a single type and like those of the normal females. 

 The last family under this section was produced by crossing a 

 hermaphrodite of 08128(16) upon a female in a geno typically 

 distinct strain of Lychnis dioica, received several years ago from the 

 vicinity of Harrisburg, Pennsylvania. The result is quite the 

 same as in the other families, all of which were derived from a 

 common stock secured at Cold Spring Harbor, Long Island. 



CASE VIII 

 WHEN HERMAPHRODITES ARE POLLINATED BY NORMAL MALES 



I have already remarked the difficulties encountered in the use 

 of hermaphrodites as self-fertilized seed parents. The difficulties 

 are still greater when the problem requires the crossing of the 

 hermaphrodites with other males, for nearly all the numerous 

 castrations which have been made have resulted in the dropping 



