55S 



THE AMERICAN NATURALIST [Vol. LII 



behaved quite normally in ¥ t and F 2 generations. One forked 

 son bred and gave offspring which behaved normally. The other 

 two forked sons failed to produce any offspring, even though 

 transferred to new bottles. The two eosin ruby sons were mated 

 to bar females and afterwards rebottled with three females, but 

 no offspring resulted. The eosin male also seemed to be sterile, 

 as he was rebottled and remated without producing offspring. 



The three daughters, which should have been heterozygous for 

 eosin ruby and forked, and which have produced sons correspond- 

 ing to that constitution, were crossed to brothers and gave the fol- 

 lowing unexpected results: 



One in culture 88, mated to an eosin ruby forked brother, pro- 

 duced a total of 41' normal females, 74 forked females, 73 forked 

 males, 37 normal males, and 13 eosin ruby forked males. A 

 second in culture 93, mated to one of the wild-type brothers, 

 produced 50 normal females and 28 forked females, 51 normal 

 males and 73 forked males. The third daughter, mated to an 

 eosin ruby forked brother, produced 2 normal, 35 forked, and 10 

 eosin ruby forked males ; and 26 females all forked. 



It is possible that the females were not virgin in these cases 

 but that could not affect the sex-linked characters of the sons 

 according to the normal mode of inheritance. 



The two large classes of sons should have been the normal and 

 eosin ruby forked classes, while the forked class of sons, which is 

 the largest in all cases, should be no larger than the eosin ruby 

 class, which does not occur even once. 



Efforts to determine what was causing this abnormal inberi- 

 tance were unsuccessful, because further breeding experiments 

 showed the offspring of all classes to behave quite normally in all 

 respects. 



D. E. Lancefield 



