Analysis of Heterogeneity in Complex Characters 



259 



and 4; all gave progeny, and in all the expected results were obtained. Each 

 test strain behaved as a homozygous line, giving in F^ an intermediate between 

 biotypes 1 and 4, and these in F, segregated into the three well-defined groups 

 shown in figure 43. F^ mating in 8 pairs gave exactly the expected results of 

 segregation and permanency and in close approximation to expected ratios. 



The extreme male in this experiment in F^ was kept a day or two with the 

 female of biotype 1, and then an equal period with the female of type 4, and 

 this repeated as long as the pairs continued breeding. In this way any error 

 from an assumed influence of age, strength, environment, or the determination 

 of anything by these factors is eliminated, as both lines A and B were served 

 alike by the divergent male. In the cross with the biotype 1 female and F^ 

 population were black, or black with a narrow margin of light color, 19 of the 

 former to 37 of the latter. Six matings from each of these groups, all giving 

 progeny, gave in all identical results, a very ordinary behavior in Fj, with 

 segregation into blacks of biotype 1, clearly marked biotype 4, and intermediate 

 types which were heterozygous, as shown by the breeding of Fj and shown in 

 figure 44. In Fg the totals from the twelve pairs is shown in table 34, which was 

 a fair approximation to a normal ratio, although biotype 1 in all sets of progeny 

 was slightly in the lead in point of numbers. 



Table 34. 



The mating B with the biotype 4 female gave quite different results. In F^ 

 a variable progeny, largely like type 4 in character, resulted, and of these 10 

 pairs were mated, and all gave in Fg like results, and all gave a rather invariable 

 biotype 4 progeny in Fg with low amounts of pigmentation ; of these another 10 

 pairs of Fg mated at random from the population gave in F^ another population 

 of the biotype 4. This result is also shown in figure 44. 



The result of this test of this divergent black male is in every respect to show 

 in an indubitable manner that, irrespective of its superficial aspects, it was not 

 modified at all as far as its homozygous biotype 4 nature was concerned. It was 

 black over the entire surface, but its gametes were not of biotype 1, but were still 

 unmodified biotype 4. No discoverable effect of the quantitative accumulation 

 of pigment was found and every chance was given this extreme male to show its 

 capacities. 



A second series of tests was made of this selected race in F^, where matings of 

 extreme individuals were made in reciprocal crosses with a pure-breeding, 

 homozygous-acting strain of biotype 12. In this cross other characters of alter- 

 native behavior are involved — i. e., body-form, elytral color, etc. These are, 

 however, not considered here — only the array presented of the particular pair 

 of contrasted characters — i. e., selectively modified biotype 4 and homozygous- 



