106 PHYSIOLOGICAL GENETICS 



over Wild; it decreases or increases in reciprocal crosses for 

 dominance of Ultrabar over Bar. Hersh draws the general 

 conclusion from these tacts that in its capacity as a modifier of 

 growth during the course of development, the more dominanl 

 gene has relatively greater or less effect as the growth changes 

 are affected by the temperature. He then seems to think thai 

 the activity or the potency of the two allelomorphic genes (which 

 are thus complements! to each other) has been measured. I 

 should prefer to assume that the temperature coefficients of the 

 rates of the developmental processes controlling the beginning 

 or end, etc. (see Fig. 23), of the facet forming (or iacet-A nlagc 

 destroying) process have been measured or of such processes as 

 control the underlying numerical system of cell divisions or of 

 both. We shall again consider this same set of facts. 



Hersh (19346, c) returned later to the same problem in connec- 

 tion with the remarkable difference of temperature action upon 

 Bar and Infrabar (Luce, see page 143). It will be shown in 

 detail (page 143) that in experiments involving Bar, Infrabar, 

 and the heterozygote, Bar is dominant at all temperatures. As 

 Bar produces fewer facets with increasing temperature, but 

 Infrabar more facets, the dominance of the heterozygote must 

 in cicase in favor of Infrabar with a decrease of temperature. 

 Near 17° this leads actually to a reversal of dominance, as 

 illustrated in Fig. 29 (page 143). From this the conclusion may 

 be drawn "that any theory of dominance . . . based on the 

 quantitative serial order of the 1 phenotypic effects at 25° would be 

 inconsistent with a theory based on the serial order of the data 

 at 17° or below." Such a conclusion does not, however, apply to 

 the type of theory of dominance represented in Fig. 23; here only 

 the difference of reactivity to temperature between Bar and 

 Infrabar would have to be added as a further variable Involved. 



In the chapter on phenocopies, we mentioned that Friesen 

 (1936) succeeded in producing the same types of phenocopies in 

 Drosophila by the action of X rays as Goldschmidt (1929a, 

 1935a) had done by temperature shocks. Both shocks are sup- 

 posed to act upon rates of reactions connected with morphogene- 

 sis. As we do not doubt that the shift of dominance is primarily 

 also concerned with rates of developmental processes, we might 

 expect that X rays will also be able to perform a shift in 

 dominance. 



