Payne: Variations and Selection 



37 



was for a few generations only and then the race became 

 stabile. What selection did according to these experiments 

 was to isolate a homozygous strain. 



Opposed to the multiple factor interpretation is the work 

 of Castle and Philhps ('14) and Castle ('15, '16, and '17). 

 In the hooded rat they have selected plus and minus strains 

 for 16 and 17 generations respectively, and have produced 

 marked differences in the two lines. Instead of assuming the 

 presence of several modifying factors, Castle believes that the 

 single gene or factor which stands for the hooded pattern 

 varies. MacDowell, Morgan, and Pearl, in particular, have 

 attacked this interpretation of Castle's and have attempted 

 to show that Castle's data can be interpreted in favor of the 

 multiple factor hypothesis. These criticisms are well known 

 and need not be repeated here. 



Goldschmidt ('18) extends his ideas of the heredity of 

 sex to all characters. Fluctuating variations, as with Castle, 

 are expressions of the genes which stand for those charac- 

 ters. In addition, however, he maintains this variation to be 

 a quantitative variation. The modifying factor, according to 

 Goldschmidt, is superfluous and hence put into the scrap heap. 

 I am willing to admit that a gene may vary, possibly quanti- 

 tatively and certainly qualitatively, but I also wish to make it 

 emphatic that the modifying factor cannot be so easily buried, 

 as we have demonstrated its action in the case of extra bristle 

 production. 



Three interesting experiments have recently been made 

 which show that selection, even in pure lines, may produce 

 results. One of these is by Middleton ('15) on Stylonychia, 

 in which he is able to isolate lines diff'ering in division rate; 

 another by Stocking ('15) on Paramecium, where she sepa- 

 rated strains differing in hereditary abnormalties ; the third 

 by Jennings ('16) on Difflugia in which he finds a marked 

 effect of selection on the number of spines on the shell, the 

 depth of the shell, the number of the teeth surrounding the 

 mouth, and the diameter of the mouth. Jennings does not 

 state in plain terms that he rejects his former conclusions 

 based on Paramecium and accepts those obtained from his 

 study of Difflugia. He gives one the impression that he does, 

 however, when he says : ''As set forth in our introduction, the 

 present work was designated as a test for the adequacy of the 



