Nos. 622-623] SHORTER ARTICLES AXD DISCUSSION 557 



bred to determine whether the missing vein represented a genetic 

 characteristic. This peculiarity probably had no relation to the 

 exceptional nature of the offspring produced. The bottle was 

 kept on the desk in the laboratory and the temperature was 

 rather low most of the time, so that the larvae developed slowly 

 and the bottle became moldy before the flies finished hatching. 

 It yielded exceedingly few flies, probably on this account as well 

 as owing to the fact that nearly all the females were eliminated. 



A count of the flies, as they hatched, gave 38 males and 3 fe- 

 males, a ratio which is inexplicable. The classes obtained were 

 also as surprising. Owing to the cold the flies developed very 

 slowly, so that the first offspring were removed on March 1 and 

 comprised 18 eosin ruby forked sons. Four more hatched on the 

 second and third, making a total of 22 eosin ruby forked sons 

 which are of the expected class, since the three characteristics 

 are sex-linked and the mother was homozygous for them. The 

 count, continued until the thirteenth of March, gave a total of 

 3 normal females, which were expected in equal numbers with 

 the males; 30 eosin ruby forked males; 2 eosin ruby males; 1 

 eosin male ; 3 forked males ; 2 normal or wild-type males. 



The count in this ease was kept up for more than the usual 

 10 days, but that could not have had any effect on the result in 

 this case as no F 1 females were found until March ninth and 

 could not have produced offspring, even had the temperature 

 not been so low as to lengthen the incubation period beyond 13 

 days. The exceptional males, which are the cross-over el asses 

 ordinarily obtained in the F, generation from such a cross, could 

 not be the result of a back-cross of the original mother to a son, 

 as the only sons with which she could have come in contact were 

 eosin ruby forked. In cases of primary non-disjunction, where 

 sons inherit the sex-linked characters of the father, they inherit 

 all his sex-linked characters, so that this can not be a case ex- 

 plainable by that means. 



Contamination can hardly account for the results as the early 

 males were of the expected class and later males always carried 

 characters used in the cross, and the females were normal in ap- 

 pearance. Moreover, there is no known source of contamination 

 that would give such a sex-ratio as this. 



Results from the offspring were interesting but have not sug- 

 gested any possible explanation of what occurred in the first 

 generation. The eosin ruby forked sons were crossed out and 



