118 THE NORTH AMERICAN SPECIES OF DROSOPHILA. 



produce, by D. simulans males, only regular daughters and exceptional 

 sons. The exceptional daughters and regular sons die. In every case 

 it has been possible to determine, by means of sex-linked genes, that 

 the female hybrids carry two X chromosomes, the males only one X, 

 as is the case in both parent species. 



The hybrid females from the two crosses first mentioned must have 

 the same chromosomal constitution; yet in one case they live, while 

 in the other they usually die. This can only mean that the result is 

 due to an interaction between the chromosome complex and the egg 

 cytoplasm in which it finds itself, unless selective fertilization occurs. 

 It seems probable that in this case, as in other instances in which the 

 cytoplasm plays a part in the result, the nature of the cytoplasm is 

 itself detemimed by the chromosome complex of the female that pro- 

 duced the egg. 



These data indicate that the two species differ in a number of genes. 

 Not only are they slightly different in appearance in a number of 

 waj's, but the same end-results (viabilitj'', fertility, bristle-number, 

 abdominal banding, wing-shape, venation) must be brought about by 

 somewhat different sets of genes in the two species, since when both 

 sets are present in the same animal the end result is different from 

 that which either one alone would produce. 



As stated above (p. 14) mutations have been found in D. simulans 

 that appear to be identical with previously known ones in D. melano- 

 gaster, and in five cases crosses have shown that the mutant genes in 

 the two species are allelomorphic. Four of these are sex-linked, 

 and have the same linear sequence in the two species, though they 

 do not show quite the same frequencies of crossing-over. This fur- 

 nishes a demonstration that the germ-plasms of two distinct species 

 contain some identical genes that are subject to identical mutations. 



