OF MUTANT CHARACTERS. 



249 



elimination of black was to mate together some of the not-hlack 

 pinkish flies of table 95. One-third of the not-black ()fTsj)rinK of .siu-h 

 pairs should be of the desired kind— that is, entirely free fruni bhick. 

 Our task was then to pick out from the mixture of pure grays and 

 grays heterozygous for black some pure gray males. In this si)cciiil 

 case we were aided by the fact that black happens to be slightly 

 dominant— that is, the heterozygous blacks are somewhat darker tlian 

 the pure grays. While this difference is not marked enough to Ije 

 used regularly in classification, it enables us to pick out by insj)e('ti()n 

 a greater proportion of pure grays than we would get by random 

 selection. Four such males were selected as being probaijly free from 

 black and were mated to eosin females. Into the same b(jttle with 

 each pair of these flies was put a ^'irgin (red-eyed) black female. 

 The offspring from these two mothers are easily distinguished, since 

 they are eosin-eyed if from the eosin mother and red-eyed if from the 

 black mother. The offspring from the black mother constitute a test 

 of whether the father were free from black, for in this case none of 

 the red-ej-ed offspring hatching in the double-mating culture should 

 be black, while if the father were heterozygous for black half of the 

 red-eyed offspring should be black. Only one of the four cultures gave 

 black offspring, and this culture was then discarded. . . , The eosin- 

 eyed flies of the other three cultures were all heterozygous for j^inkish, 

 and at the same time free from black. By mating together some of 

 these eosin-eyed flies pure pinkish offspring should be obtained as a 

 quarter of the offspring. A more efficient method, and the one actually 

 followed, was to save the fathers and mate them to their eosin-eyed 

 daughters, since in this case half, rather than a quarter, of the progeny 

 should be pure pinkish. 



Table 96. — B. C. offspring given by the Fi eosin star dichcete sons, from the 

 out-cross of a pinkish female to a star dichcete male, luhen back-crossed to 

 pinkish females. 



In order to show by an actual test that the gene for pinkish is in the 

 third chromosome, it was decided to take advantage of the fact of no 

 crossing-over in the male and to run a back-cross test of a male hetero- 

 zygous for pinkish and for the dominant third-chromosome cliaracter 



