SELECTION IN BAR-EYED RACE OF DROSOPHILA. 365 



became hard, and larvae did not seem to thrive well in it. This 

 defect, however, was remedied after one or two generations had 

 been reared from the vestigeal-winged stock and long before the 

 experiments on the long-winged stock were undertaken. 



In some preliminary work the number of facets was estimated 

 in the eyes of the parents selected and the actual counts were 

 made at the time the parents were killed. This method was 

 employed by Zeleny and Mattoon. The author, however, 

 found that his estimates w r ere not close enough and that in 

 some cases parents were lost because they died and were destroyed 

 by the larvae before their death was discovered. The facets of 

 dead flies also disintegrate when they remain in contact with 

 food for a short time. For those reasons all selections during 

 the experiments were made from actual counts. 



During all of the work on the vestigial-winged flies and during 

 part of the work on the long-winged flies, selections were made 

 at noon and night after all the flies had been removed in the 

 morning. In that way the specimens were never more than six 

 hours old at the time of selection. But observations during that 

 time showed that males seldom or never mated before they 

 w r ere twenty-four hours old, and for that reason later selections 

 were made every twelve hours. All desirable specimens were 

 saved and when no mates were present they were kept in vials 

 with food until matings could be made. 



All flies were etherized and the facets were counted while they 

 were quiescent. The larvae were never subjected to ether, as 

 the flies to be examined were first transferred to empty vials. 

 Selected flies recovered from the effect of the ether in fifteen 

 minutes to half an hour. The other flies were preserved in 85 

 per cent alcohol. 



For counting facets the flies were placed in a little pit on a 

 paraffin block and illuminated by means of a 25-watt tungsten 

 lamp. This gave sufficient light but only a moderate amount 

 of heat. A Leitz microscope with a number four ocular and a 

 number three objective and the tube drawn out to its full length 

 was found to be most convenient. This gave a relatively high 

 magnification with a suitable depth and size of field. 



Errors in counting can not be avoided ; it is merely a question 



