348 



GENES MODIFYING NOTCH. 



On the other hand, had the sequel shown that the results were due to a 

 number of factors present at the beginning, the mass-culture method 

 would have offered a better chance of collecting the different modi- 

 fiers in the same strain. The object of the selection process here prac- 

 tised was, however, to produce by a rough method results of the kind 

 familiar to the breeder, and then to show, by the refined tests that the 

 Drosophila work has made possible, what had been done to the original 

 stock 



CONDITION OF STOCK BEFORE SELECTION. 



In table 1 there are records of the offspring of 11 pairs of Notch 

 females by normal males. The totals give 577 Notch females, 608 

 normal females, 613 normal males. It is clear that the viability of the 

 Notch females compares favorably with that of the normal females. 

 Very few of the Notch flies could have had normal wings when this 

 class comes so near to the realization of their expected numbers. How- 

 ever, there were other females that had the same origin in which the 

 ratios amongst the offspring were strikingly different. These are 

 given here in table 2. 



Table 1. 



Table 2. 



In these 8 sets the Notch females are not half as numerous as the 

 males and less than half as numerous as the normal females. The 

 normal females are greatly in excess of the males. If we suppose that 

 here a considerable number of the Notch females have normal wings — 

 as was actually shown to be the case later in the offspring of some of 

 these sets — the discrepancies between tables 1 and 2 may be accounted 

 for. Thus, if we add the two classes of females (1429) and divide by 



