378 CALVIN B. BRIDGES 



modifyiDg gene (cream a, b, etc.), which by itself produces little 

 or no visible effect, and of a particular gene (eosin) that is 

 necessary as a 'base' or 'differentiator.' 



The scale of the modifications of eosin produced by these 

 several modifiers ranges on the one hand to a deep pink darker 

 than eosin, and on the other hand to a pure white. This scale 

 is purely artificial and descriptive, for these modifiers were 

 entirely independent of one another in origin. Furthermore, 

 the order of their occurrence bears only a random relationship 

 to the dark-light seriation. 



By judicious combination of several such simple modifiers a 

 multiple heterozygous stock could be obtained which would be 

 amenable to selection and which would offer upon analysis a 

 satisfactory parallel to such a case as that shown by Castle's 

 rats. The first result of selection in such a heterogeneous stock 

 in the direction of lighter forms would be to pick out individuals 

 homozygous for one or more of the modifiers and probably 

 heterozygous for others. These different individuals of course 

 would not necessarily be homozygous for the same factors, and 

 therefore the population might still remain heterogeneous for 

 these factors. Continued selection would result in a greater 

 and greater degree of homozygosity and homogeneity and a 

 consequent slowing down of the speed of the progression of the 

 population in the direction of selection. The grade of the form 

 reached when the population is homozygous and homogeneous 

 would depend on the number and character of the particular 

 modifiers in the initial population. 



Ihe fact must not be lost sight of that there is not one iota 

 of evidence to show that either the rate or the direction of the 

 mutation processes that are characteristic of the species are 

 altered by such selection. During the course of any selection 

 experiment modifiers and other mutations should arise at the 

 normal rate. In Drosophila roughly 25 per cent of the mutations 

 found were wing or venation characters, 16 per cent body-color 

 characters, and 20 per cent eye colors. Only eye-color mutations 

 would have any effect upon the progress of our selection. The 

 eye-color mutations found in Drosophila have been roughly in 



