260 



The Mechanism of Evolution in Leptinotaksa 



acting biotype 13. In figure 45 I have shown the behavior when biotypes 10 

 and 12, both from homozygous-acting extracted strains, are crossed. In F^, 

 uniformly, type 4 is completely dominant and in F^ there is segregation, 

 usually into three classes, on the average of 1:2:1, as far as this pair of char- 

 acters is concerned. 



Fig. 45. — A second test made with respect to the same modified character as 

 shown in figure 42. 



In the test (fig. 46) 6 reciprocal pairs were mated between Pg of the modified 

 biotype 4 and biotype 12, and gave an F^ population all alike with biotype 4 

 dominant, 247 in all. Of these, 10 pairs mated gave progeny, an Fg generation 

 in three classes, variable biotypes 4, 12, and typical biotype 4 nonvariable. 

 Inbreeding each class in F3 showed the type 12 were true-breeding extractives 



F3 



mm m 



J 



F2 



I 



m ^ ^ 



^- 



"I 



m 



Fig. 46. — Graphic representation of a further test 

 between a selectively modified race crossed with 

 biotype 12, showing that it was not modified but 

 was the normal biotype 4 as far as its gametic com- 

 position was concerned. 



in the 7 pairs tested, and that the biotype 4 nonvariable were pure, while vari- 

 able biotype 4 were heterozygous dominants. At no point in these tests have 

 any indications of gametic alteration in the biotype 4 as the result of the quan- 

 titative accumulation of pigmentation been observed. 



In this entire experiment and its testing great care was exercised to keep out 

 extraneous factors which might vitiate the series. At no time has there been any 

 indication of gametic change. How, therefore, are these changes, often beyond 



