352 



CALVIN B. BRIDGES 



be either unaffected or less affected by the action of the whiting 

 gene. Should the white-eosin whiting flies be colorless, then 

 the colorless classes should be 3 white + 1 whiting white + 1 eosin 

 whiting (eosin whiting if males, white-eosin whiting if females), 

 and these classes should total |, or 62.5 per cent. The colorless 

 flies did in fact constitute 61.7 per cent. \^Tiat is even more 

 significant, the proportion of colorless flies among the females 

 was just as great as among the males (table 9), The colorless 

 males are known to include the (eosin) whiting flies, and the 

 conclusion is justified that the like class of colorless females 

 included white-eosin whiting flies. 



TABLE 10 

 The Fi and F-z offspring from the outer oss of (eosin) whiting males to cherry female 



The most interesting of these three experiments was that in 

 which (eosin) whiting males were outcrossed to cherry females. 

 The Fi females w^ere eosin-cherry compounds, and the males 

 were cherry which is of nearly the same color (table 10). Fo of 

 the cross of eosin male by cherry female gives in equal numbers 

 the F2 classes eosin-cherry compound 9, cherry 9, cherry cT, 

 and eosin cf. The eosin-cherry female and the pure cherry 

 female are so nearly the same color that they form one phenotype, 

 but the eosin males can be easily separated from the cherry 

 males. The cross of (eosin) whiting male by cherry female will 

 give further subdivisions of each of these classes into 3 not- 

 whiting: 1 homozygous for whiting. 



