332 



Morgan. 



a much smaller degree and in practically all cases the two kinds of 

 wings in males may be distinguished; but in the females the wings 

 of these two kinds are often crinkled and the ends often turned 

 under, so that their length can not be readily determined. Of course, 

 the difficulty can be met by testing, with single pairs, which reveal 

 in their offspring which type had been used, but as the flies having 

 these characters are often sterile, or often die in the cultures without 

 laying, I have resorted almost entirely to mass cultures. 



When rudimentary-miniature males (rm) are mated to long winged 

 females (RM) there are produced long-winged males and females. These 

 inbred have given the following results: 



Long 9 9637 



RM 9 by rm (5 



Long 9 I 068 



,. c5 996 



.. c5 3594 

 Miniature c5 683 

 Rudimentary (J 79 

 Rudimentary- miniature cj 931 



The analysis is as follows: 



Long 9 RMX — RMX 



Rudimentary-miniature 3 . rmX — rm 



Gametes of Fj 



When rudimentary-miniature males are crossed to miniature 

 females both the female and the male offspring are Miniature. These 

 inbred give the following F2 classes: 



,, „ t. ^ f Miniature 9 



rM 9 by rm cJ = I I 



Miniature 9 41 21 

 C? 2061 

 Rud.-min. (J 479 



The analysis foUows: 



Miniature 9 • ■ • 

 Rud.-min. (5 . . . 



rMX — rMX 

 rmX — rm 



Fl 



rMXrmX Miniature 9 

 rmXrm ,. cj 



