EYE COLOR IN DROSOPHILA MELANOGASTER 365 



PINKISH 



In the fall of 1913 a stock of eosin black had been made up 

 with which to test the chromosome group of cream II. In the 

 following summer (July 27, 1914) I noticed that a few of the 

 males were somewhat lighter in eye color than the others, but 

 seemed chiefly noticeable because of the weakness of the yellow 

 component of the eosin eye color. The color of the regular 

 eosin male is a pinkish yellow; the color of creams a, II, III, 

 and b is nearly a pure yellow with Uttle of the pinkish tinge, 

 while this new color was somewhat the converse of this and was 

 pale pink with relatively little yellow. 



One of these males mated to a sister gave all of the sons of 

 this pinkish color and all the daughters of a similar color, which 

 is, however, much harder to distinguish from standard eosin. 

 It seems that this character is somewhat sex-limited in the same 

 direction as eosin. Pure stock of the mutation had been obtained 

 at once through the happy selection of a pure pinkish female 

 which had been taken to be simply an eosin female of somewhat 

 lighter eye color because of being freshly hatched. 



Since pinkish appeared in a stock of eosin black, material was 

 on hand to test the chromosome group at once. Accordingly, 

 black pinkish females were outcrossed to eosin males and the Fi 

 eosin females, standard eosin in color, were backcrossed to black 

 pinkish males. In the B.C. cultures half of the flies were not- 

 black, and the not-black pinkish flies were seen to be less marked- 

 ly 'pinkish' in tone than the blacks. In the absence of black 

 the eye color was more nearly hke that of the other creams, 

 though the amount of dilution is less than in the case of any of 

 the other creams. In the first of these B.C. cultures (table 22) 

 males and females were both classified together. Some question 

 having been raised as to the accuracy of the separation of pinkish 

 from eosin among females, the cross was repeated, and the more 

 readily classifiable males (last three cultures) gave the same re- 

 sult as before. It was seen that the new or crossover combina- 

 tions were as numerous (51.4 per cent) as the original classes, 

 and this independent inheritance was taken to mean that the 



