MUTATIONS IN EYE-COLOR OF DROSOPHILA. 337 



The analysis is as follows: 



vPOX vPOX Pink 9 



vpOX vpo Orange <? I 



vPOX — vpOX Pink 9 (F x ) 



VPOX Vpo Red d" (pure strain) 



vPOX VPOX Red 9 



vpOXVPOX Red 9 lET 



vPOXVpo Red cf T 2 



vpOXVpo Vermilion <j 



The numbers are small, but they agree fairly with the expectation. It will 

 be noticed that the vermilion males are actually ahead of the red males. 



When P 9 's (from P 9 by & ) are mated to vermilion males of pure strain 

 the results were: 



Red 9 57 Red tf 45 Vermilion 9 57 Vermilion <? 41 



The analysis follows: 



vPOX vPOX Pink 9 



vpOX vpo Orange d* 



vPOX — vpOX Pink 9" (F,) 



VpOX Vpo Vermilion cf (pure strain) 



v POX VpOX Red 9 



vpOXVpOX Vermilion cf 



vPOX Vpo Red 9 



vpOXVpo Vermilion d" 



In this experiment the vermilion hold their own well 





The Vermilion-Pink Ratio. 



In the cross between vermilion female and pink male recorded in my former 

 paper there appeared in the F 2 generation twice as many pink 9 as pink <? and 

 twice as many orange 9 as orange <?. The numbers however were small, and 

 it was not clear whether they had any significance. I have repeated the experi- 

 ment on a larger scale, and have found that the numbers were not significant. 

 A number of back-crosses were also carried out and counter-crosses which may 

 be given since they also furnish a further and desirable confirmation of the general 

 explanation of the heredity of eye color. The F 2 records are given separately 

 in the next table : 



There is nothing unusual in the results and even the separate records run 

 fairly evenly. As usual the pink and orange classes fall below the expectation 

 (3tol). 



22 JOURN. ACAD. NAT. SCI. PHILA.. VOL.XV. 



