OF MUTANT CHARACTERS. 



173 



The linkage results of these back-crosses were somewhat unexpected, 

 for in four of the lines no cross-overs at all were obtained, and in 

 another only a few. In the original F2 culture several cross-oxers had 

 been noted, and five F2 cultures raised from the brothers and sisters of 

 these back-crossed males seemed to be giving in the neighborhood of 

 15 per cent of cross-overs (table 28). A suspicion was aroused that 

 the results of the back-cross and of the F2 were of a different order, 

 but this idea did not develop beyond a suspicion because of the con- 

 tradictory result given by the different back-cross cultures. It was not 

 at first clearly realized that these aberrant cultures were descended 

 from a single set of parents, so that the difference was disprojiortion- 

 ately blurred. Whatever difference was recognized w-as attributed to a 

 possible difference in the linkage results given by back-crosses as con- 

 trasted with F2's. Since this w'as the first back-cross invohing auto- 

 somal linkage that had been tried, there was no corrective evidence to 

 show that these two kinds of results might not be different. 



Table 28. — F2 offspring given by the Fi (vermilion) sons and daughters from 

 the out-cross of (vermilion) purple vestigial males to vermilion females. 



BACK-CROSS TEST OF FEMALES. PURPLE VESTIGIAL "COUPLING." 



A second back-cross experiment using the simple purple vestigial 

 stock instead of the vermilion purple vestigial was started (June 25, 

 1912) a month later than the first and before the results of the first 

 were fully known. A purple vestigial male out-crossed to a wild 

 female produced wild-type sons and daughters (B 39; -f- 9 15, -}- cf 

 10). Four of the Fi females were back-crossed each by two or three 

 purple- vestigial males from stock. In this case Fi females happened 

 to be chosen because, as is usually the case, they hatched somewhat 

 earlier than their brothers in the same culture. 



These back-cross cultures (table 29), in common with the previous 

 r2 cultures (table 28), showed a fair amount of crossing-over between 

 purple and vestigial. A calculation showed that the percentage of 

 crossing-over was 9.1. This was recognized as being of a different 

 degree from the apparent percentage of 1.8 calculated from the first 

 back-cross (table 27). It was now reahzed for the first time tliat the 

 two back-crosses had differed in the sex of the Fi flies tested by the 



