MUTATIONS IN EYE-COLOR OF DROSOPHILA. 



333 



males, because in the reciprocal cross equal numbers of pink males and orange 

 males were present. In order to study this exceptional case I have repeated t he 

 experiment and at the same time have mated pairs of pYs, and made back-crowe 

 and counter-crosses. The repetition has shown that the peculiar ratio was the 

 result of some error in the experiment, since the new results conform to expec- 

 tation. The other crosses have also shown that there is nothing unusual in this 



Incidentally the new data give a searching test of the correctness of th< 

 general explanation of eye-color, and give a welcome addition to the facts pre- 

 viously recorded. 



When orange females were crossed to pink males all the female offspring were 

 Pink and all the male offspring Orange. The new numerical results for th< first 

 and second generations were as follows: 



case 



Pink 9 969 f 



Orange 9 by Pink d 1 



Pink 

 Pink 



9 1,738 

 <* 1,641 



Orange c? 941 



Orange 9 1,846 

 L Orange & 1,621 



The results of each bottle given separately show how evenly the results 



9 byPtfPi 



P 9 



<?Fi 



1 

 2 



3 



4 



5 



6 



7 

 8 

 9 



M 



1 P<? 

 80 



09 



0<? 



91 



87 



75 



85 



79 



75 



66 



76 



43 



33 



37 



73 



67 



63 



55 



102 



70 



90 



73 



44 



41 



45 



44 



30 



30 



25 



24 



23 



20 



25 



23 



10 1 



1 5 



4 



2 



The result gives a close approximation to expectation. There is nothing un- 

 usual in the ratios. Some error must have arisen in the course of the former experi- 

 ment that is most regrettable. There can be no doubt that the counts were accu- 

 rately made, for, at the time we commented on the peculiar ratio that was appear- 

 ing. The error must have arisen in the mating. If by chance one or more parent 

 flies ( 9 's) (t. e., orange 9 by pink d") had been accidentally left over in the 

 culture bottle the error can be explained; for, as seen above, the presence of such 

 flies would increase the ratio of pink 9 's and of orange & 's which added to the 

 Ft count would give the recorded results. I am inclined to account for the error 

 as due to this oversight, because without the utmost care F x flies may be left 

 in the old bottle when the parents are supposed to be removed. We are on our 

 guard against this danger, and usually take the precaution of removing all the 

 old food and looking over it carefully to see that no flies have been left sticking 



