22 



Indiana University Studies 



mean bristle number of the eosin flies is 4.416, very near to 

 that of the other non-cross-over class, eosin miniature, which 

 is 4.329. This demonstrates quite clearly, I think, that there 

 is a factor for extra bristle number in the X-chromosome 

 somewhere near the eosin end. The cross of the selection 

 line to eosin ruby forked brings out the same fact. The 

 cross was made in the same W2iy as with eosin miniature. 

 The results are given in Table VI. Since eosin and ruby are 

 only about five units apart, there is very little crossing-over 

 between them. Of the non-cross-over classes, normal and 

 eosin ruby forked, the mean bristle number of the normal 

 is much higher, being 5.79 while it is only 4.55 in the eosin 

 ruby forked flies. Of the cross-over classes, eosin ruby and 

 forked are of most interest. The number of flies in the other 

 classes is too small to be of much significance. The mean 

 bristle number of the eosin ruby flies is 4.65, which is very 

 close to the mean of the non-cross-over type, eosin ruby 

 forked. On the other hand, the mean of the forked flies is 

 5.58, approaching that of the normal. Here again, of the 

 cross-overs, one has a much higher bristle number than the 

 other. The evideice in this case corroborates that of the 

 former in demonstrating the presence of a factor for bristle 

 number in the X-chromosome near the eosin end. All the 

 facts then point to the presence of a factor for extra bristle 

 number in the X-chromosome, and my conclusion can be none 

 other than that there is such a factor present. 



The experiment in which the X-chromosome was eliminated 

 by crossing the selection line to bar demonstrated the presence 

 of a factor or factors outside the X-chromosome. The next 

 problem was to locate them if possible. In an attempt to 

 do this, the selection line was first crossed to black, pink, bent 

 stock. The gene for black is in the second chromosome, that 

 for pink in the third chromosome, and that for bent in the 

 fourth. The easiest method would be to mate the F^^ females 

 from this cross back to black, pink, bent males. I have mated 



to F^. It is a little more laborious this way, but the results 

 show the same thing. The data are given in Table VII, a 

 study of which shows that the pink files when compared to the 

 normal, to black, and to bent have fewer extra bristles. Fur- 

 ther, 85 per cent of them have the normal bristle number, 

 while in the case of the black it is only 37 per cent, in the 

 normal 33 per cent, and in the bent 26 per cent. The highest 



