134 EDWIN CARLETON MACDOWELL 



environment (which was just as potent in the early generations) 

 nor to depend upon a supposed lack of any genetic control at 

 all of the numbers of extra bristles; it does depend upon a unifi- 

 cation of the germ plasm brought about by the high selection. 



EXTRACTED LOW SELECTED RACE 



After a race of high grade germ plasm has been isolated by selec- 

 tion, the return selections are unsuccessful. This favors the 

 hypothesis previously raised, that the germinal differences in- 

 volved do not necessarily occur during the process of selection, 

 nor does selection hasten or retard the occurrence of such dif- 

 ferences. Selection has accomplished a sorting out of differences 

 already present. Were it possible to regain the differences lost 

 by the selection process, it would be possible to re-establish a 

 low grade race. According to the theory formerly explained, 

 such a regaining of germinal differences could be accomplished 

 by crosses with normal flies. The modification of the distri- 

 butions of extra bristled flies resulting from such crosses has 

 been considered evidence of the regaining of germinal differ- 

 ences. With this fact in mind, the attempt was made to start 

 a low grade race from the high selected race by first making a 

 cross between the high selected and normal flies, and then 

 selecting low grades from the extra bristled flies appearing as 

 extracted extras in the F.,. 



Three flies of the high selected race, generation 16, all double 

 first cousins, were mated with normal wild flies from a stock 

 that had been bred in the laboratory for two years or more. 

 This stock had been inbred for many generations, and for a 

 few generations single brother by sister matings were made by 

 pairs. Four matings were made between the extra flies (grade 

 2) that appeared in the Fo of this cross. The race started in this 

 way was continued by using the lowest grade flies available as 

 parents. Following the fourth generation of the race all the 

 flies came from a single one of the four matings made in the 

 first generation of this race. The data are to be found in tables 

 2, A and B, and plotted in figures 9, A and B. 



