Nature of Inherited Variations 207 



is a mere quantitative alteration, in the extent or intensity of 

 pigmentation or other characteristics. In many cases such 

 alterations have occurred in characters that in other re- 

 spects behave like "single unit characters." Such work was 

 done on the rat by Castle and his associates ; on Drosophila 

 by MacDowell, Zeleny and Mattoon, Reeves, Morgan, Stur- 

 tevant, and others. 



Two views have been held by investigators as to the nature 

 of the change in such cases. Castle and a number of others 

 have long held that there was occurring a gradual change, 

 perhaps merely quantitative in nature, in the single unit 

 factor on which the adult character depends. On the other 

 side, many have maintained that these gradual alterations 

 are due to the fact that the adult character depends on 

 many distinct genes or unit factors, each affecting the adult 

 character but little. By selective breeding many of these 

 factors are gradually accumulated in one set of progeny, 

 few in the other ; so that the adult features become slowly 

 very diverse. That is, it is maintained that the apparent 

 changes in the hereditary characters are really due, like all 

 Mendelian inheritance, to recombinations of the existing fac- 

 tors. 



This explanation, commonly called the hypothesis of mul- 

 tiple modifying factors, has recently been accepted, on the 

 basis of crucial experiments, by Castle himself. 2 There can 

 hardly be doubt that it is correct for at least most cases of 

 this kind. 



Let us, therefore, accept this explanation, and proceed to 

 an examination of its relation to the questions in which we 

 are interested. What bearing have the facts, so interpreted, 

 on the nature of hereditary variations and on the method of 

 evolution ? 



3 Castle, W. E., Proc. Nat. Acad., April, 1919. 



