TEMPERATURE COEFFICIENT OF A HETEROZYGOTE. 107 



Heterozygotes Exhibit a Complete Clw/nge in Reaction from the 

 Full to the Ultra-bar Type. — In the case of heterozygous females, 

 containing one full and one ultra-bar factor, the mean ommatidial 

 values as determined by the present experiments are given in the 

 second line of Table I. There are 37.1 ommatidia at 30 , 

 40.5 at 27 °, and 112. 1 at 15 . The average increase per degree 

 determined by the method previously described is J.J per cent, 

 between 30 ° and 15 and 8.9 per cent, between 27 and 15 . 

 These values are strikingly different from 2.5 per cent., the value 

 for full eye, and strikingly similar to 8.0 and J.6, the values for 

 homozygous ultra-bar. 



The temperature coefficient of the heterozygotes must, therefore, 

 be considered as essentially like that of the ultra-bar parent and 

 wholly different from that of the full parent. While a single 

 ultra-bar factor is not sufficient to bring about the complete effect 

 in reducing ommatidial number as produced by two ultra-bar fac- 

 tors, it is sufficient to produce the complete change to a physiologi- 

 cal system of the ultra-bar type. 



The change in ommatidial number with change in temperature 

 can be explained most satisfactorily by assuming a differential 

 effect of temperature upon the physiological processes involved in 

 ommatidial production as opposed to other physiological processes. 

 In view of the fact that temperature is effective only during a few 

 hours of larval life, it may be considered that the initial steps in 

 the formation of ommatidia are confined to a definite embryologi- 

 cal period. The length of this period is determined by the general 

 physiological processes of the larva, while the rate of formation of 

 ommatidia during the period is a function of special processes 

 which have a different coefficient. It is evident that under these 

 circumstances two different temperatures must give two different 

 ommatidial numbers. The difference in the temperature coeffi- 

 cients is slight in full eye and the mutation to bar or ultra-bar 

 involves a marked increase of this difference. Further analysis 

 awaits a more accurate knowledge of the nature of the embryo- 

 logical processes involved. Whatever the character of these proc- 

 esses, however, it is clear that the reaction system produced by the 

 introduction of a single ultra-bar factor is of the ultra-bar type, 

 even though the reduction in ommatidial count at any specific tern- 



