FERTILITY AND STERILITY IN DROSOPHILA 209 



but that the productivity of the long-winged individuals is greater 

 than their truncate brothers and sisters I can testify from my 

 experience with them in mass culture, and it would seem as though 

 this increased productivity is due to the increased fertilizing power 

 of the egg and sperm. I shall return to this question in Part IV. 



BEHAVIOR OF THE TRUNCATE WING IN HEREDITY 



Attention has already been called to the fact that many genera- 

 tions of inbreeding the truncate brothers and sisters has failed to 

 purify this stock. The truncate always throw long-winged males 

 and females. The long-winged flies in turn throw some truncates. 

 Table 1 gives the number of truncates and longs that appeared in 

 the different generations by continually selecting the truncates. 

 From a total of 6356 there are 5469 flies with truncate wings and 

 887 flies with long wings; a ratio of 1 long to 6.2 truncate. Dur- 

 ing the period under investigation this ratio remained fairly 

 constant. 



When the long-winged flies thrown by the truncates were bred 

 together there appeared in a total of 962 flies, 825 with long wings 

 and 137 with truncate wings; a ratio of 6.2 longs to 1 truncate. 

 AVhen the Fo generation of long-winged flies were bred together 

 the different classes appear as shown in table 24. In a total of 

 1924 there are 1686 long-winged flies and 238 truncate flies. The 

 males and females are about in equal proportions. This gives a 

 ratio of 1 truncate to 7.1 long. This ratio is practically the oppo- 

 site of the condition found in the truncate wing which threw 1 

 long to 6.2 truncates. 



When the truncates are paired with the long winged wild stocks 

 such as the inbred or Woods Hole, long wings are dominant to 

 short, but not completely so; for while no truncate wings appear a 

 new class of wing arises. This is a long wing with small crescent 

 shaped piece cut out of the inner margin at the tip of each wing. 

 These will be referred to as ' crescents. ' 



In the cross as given in table 3 there appeared among a total 

 of 1784 hybrids, 41 crescent males and 35 crescent females, a 

 ratio of 1 crescent to 21.2 longs. In the reciprocal cross from a 

 total of 4711 offspring there were 30 crescent males and 84 crescent 



