EYE COLOR IN DROSOPHILA MELANOGASTER 



357 



light of subsequent experiments these are known to be the cul- 

 tures in which the mother was heterozygous for the crossover 

 variation (Cm Cm). The other three cultures (578, 579, 581), 

 however, gave a much higher percentage of crossing over, and 

 for this one of two explanations seemed possible. Either the 

 cream III ebony stock was not pure for C^ as supposed, in 

 which case this high value would be comparable with the regular 

 values obtained in experiments free from Cm; or the eosin stock 

 used as Pi (in the mating of cream III ebony male by eosin fe- 

 male) itself carried Cm (in heterozygous form, since she gave 

 a ratio of 1:1), in which case the high crossover value is com- 

 parable with the crossover values obtained from mothers homo- 



TABLE 15 



The coupling backcross offspring from the Pi mating of an cream III ebony male to 

 an eosin female. The Fi eosin females back-crossed singly to cream III 



ebony males 



zygous for C. While crossing over in females heterozygous for 

 Cjii is very much lower than in females free from Cm, yet in 

 females homozygous for Cm it is the highest of all (Muller, '16). 

 I was not able to decide without testing whether the high 

 values were due to homozygous Cm on the one hand or to homo- 

 zygous not-Cm on the other. 



This question was not tested until December, 1915. Mean- 

 while the number of stocks of primary mutations had increased 

 to such an extent that it had become necessary to eliminate as 

 many as possible of those stocks which were decreed less generally 

 useful. Among those discarded (or lost) at various times were 

 cream a, cream II, whiting, and cream III. However, the stock 

 of cream III ebony had been spared for this special test. 



