296 The Genetics of the Dutch Babbit — A Reply 



as suggested by Punnett, we should expect segregation ' of more than 

 two groups., or the formation of multimodal groups, with extended 

 ranges resulting from recombination of factors, none of which are in 

 evidence. 



Punnett attempts a detailed application of his three-factor hypothesis 

 to only one cross, that in which the parent races differ most widely, one 

 race being non-Dutch and entirely devoid of white (the Self race), the 

 other being the whitest of all the Dutch races, the so-called White race. 

 F^ from this cross consisted of animals bearing a very small amount of 

 white. In j^ the Self type reappeared in 21*4 per cent, of the young 

 (expected on a monohybrid basis 25 per cent.). The F^ type (of grades 

 1 — 3) was found in 46"6 per cent, of the F^ young (expectation on a 

 monohybrid basis 50 per cent.). The remaining 32 per cent, of the 

 individuals are found in the range of the White race (where, on a mono- 

 hybrid basis, we should expect 25 per cent, to occur, if segregation is 

 perfect) or in the region between the ranges of Fi and uncrossed White. 

 The excess of 7 per cent, over expectation in this group is readily 

 accounted for, if we explain the individuals in the intermediate region 

 between the range of uncrossed White and of ^i, as an assemblage of 

 modified Whites and of modified F-^s respectively. If this is so, a 

 1:2:1 ratio is approximated. The data present no serious difficulty to 

 a simple monohybrid explanation, but to make sure of the soundness of 

 such an interpretation a back-cross was made with the White race. 

 This shows a beautiful segregation into two sharply separated groups, 

 each of 58 individuals distributed in a monomodal and substantially 

 symmetrical figure. The range of the lower group is exactly that of 

 the original F^ group plus the intermediate region in which the doubtful 

 individuals of the F^ distribution occurred. The range of the upper 

 group is exactly that of uncrossed White plus two additional lower 

 grades (13 and 14) of the intermediate region. Three unoccupied 

 grades (10 — 12) separate the two groups. It would be difficult to 

 imagine a clearer demonstration of simple monohybrid segregation, if it 

 be granted either that the segregating factors are mutually modified 

 (an hypothesis which I no longer entertain for reasons explained in my 

 former paper) or if it be granted that slight modification of the extracted 

 types has resulted from recombination of mij^r (modifying) genetic 

 factors in association with the single pair of major factors. 



As an alternative to this explanation Punnett has worked out in 

 detail a three-factor hypothesis, which regards the Self race as homo- 

 zygous for three dominant factors, P, T, and S, the White race being 



