CHANGE IN THE BAR GENE OF DROSOPHILA 295 



of Rochester, New York, was used for a continuous record of the 

 temperature. This shows, for the period covered by the present 

 experiments, a departure from 27° not exceeding 0.5°. The 

 accessory tests make it probable that this appUes to the temper- 

 ature of the larvae as well as to that of the thermometer bulb. 



The satisfactory character of the incubator was in part due to 

 its being placed in a room with both refrigeration and heat con- 

 trol and maintained at 24° with an extreme departure of 1°. 

 This constant-temperature room is cooled by air blown over 

 brine coils and warmed by steam coils. Both systems are under 

 automatic control installed by the Johnson Service Company of 

 Milwaukee. 



In choosing a suitable constant temperature for the experi- 

 ments with germinal factors affecting facet number in the bar 

 stocks, the primary consideration was based upon the fact that 

 facet number decreases very strikingly with increase in tempera- 

 ture, and facility and accuracy in counting are of course easier 

 the smaller the number of facets. An added inducement for 

 using a warm temperature is the shorter developmental period. 

 The choice of the exact temperature was more difficult. It was 

 desired to go as high as possible without noticeable disturbance 

 of the viability and without subjecting the work to the criticism 

 that any germinal changes observed are the results of violent 

 reactions to unfavorable conditions. Twenty-seven degrees cent- 

 igrade was chosen after a preliminary trial had shown that at 

 29° viability is considerably lowered. 



Flies raised in dry food frequently are much smaller than those 

 raised in moist food, but preliminary counts under these condi- 

 tions show no consistent difference in eye-facet number between 

 the two. In figure 1 it is noticed that the magnification in each 

 of the five cases is the same and that the size of individual 

 facets is in general larger the larger the individual fly. In indi- 

 viduals of uniform size, however, as found in stocks raised under 

 optimum conditions, the number of facets is proportional to the 

 area of the eye as shown by Seyster ('19). 



In striving for the ideal of basic stocks which will give perfectly 

 constant facet numbers and in which any departure, no matter 



