108 CHARLES ZELENY. 



perature is not as great as that produced by two ultra-bar factors. 



The analysis makes it evident that there are two distinct proc- 

 esses involved in the mutation from full to ultra-bar. One of 

 these consists in an essential change in the type of reaction as 

 shown by the change in the temperature coefficient. The other 

 process involves a change in the general level of the rate of the 

 reaction without affecting its specific character. The first is fully 

 accomplished by a single ultra-bar factor. The second is influ- 

 enced quantitatively by the number of ultra-bar factors. 



The Effect of Temperature upon Dominance. — Since the hetero- 

 zygote has the same temperature coefficient as homozygous ultra- 

 bar, and one that is much greater than that of homozygous full 

 eye, it becomes a matter of interest to consider the effect of tem- 

 perature upon dominance. Elsewhere (1920, p. 308) I have dis- 

 cussed a method of determination of the coefficient of dominance 

 by the use of a factorial scale in which the effect of a degree of 

 temperature is taken as the measure of an unit factor. It is 

 obvious that on this basis there can be no change in dominance 

 with temperature, because the ommatidial value of the unit varies 

 with change in effect of temperature. 



If, however, it is desired to get an expression for dominance 

 which is based directly upon the somatic expression — i.e., upon the 

 ommatidial number — such a value changes with the temperature. 

 Suppose that complete or 100 per cent, dominance of full is a con- 

 dition in which the heterozygote has the same ommatidial number 

 as full eye and complete recessiveness or zero per cent, dominance 

 of full a condition in which the heterozygote has the same omma- 

 tidial number as ultra-bar. Likewise suppose that the ommatidial 

 count is the scale of values. Then the coefficient of dominance of 

 full as expressed on a percentage basis is 



H — B'u 

 C - D - F = F — B 'u Xl0 °> 



in which H is the ommatidial count of the heterozygote, B'u that 

 of ultra-bar, and >F that of full. Correspondingly 



