Analysis of Heterogenetty in Complex Chaeacters 251 



ing a larger biotypic group. Thus No. 1 is in nature usually associated with the 

 melanothorax form, broad and rounded, while the biotypes 9, 10, 11, and 12 are 

 always present in and correlated with the appearance of the muUilineata form- 

 factor. Biotypes 2 to 8 are found associated in the species with the form and 

 body characters conmion to multitceniata. In any population there are thus 

 complicating factors in any character, so that a complete analysis would be an 

 analysis of all characters and their interrelations. The pronotal pattern, how- 

 ever, is sufficiently independent and so accurately a gametically determined char- 

 acter that it serves well for certain purposes of investigation. 



To recognize these biotypes is of little or no use unless they be tested fully with 

 regard to their properties and capacities for existence. Can such groups exist in 

 freely-breeding cultures without change? Can they be selectively changed by 

 quantitative selection without adding something to the complex ? What is the 

 relation of the different groups when crossed? These are some of the deter- 

 minations that must be made concerning the nature of these biotypes. 



THE BREEDING OF PRIMARY BIOTYPES SELECTION. 



When one of these biotypes has been isolated by any process, can it stand 

 alone, isolated from intercrossing, without constant selection? This I have 

 tested as follows : Biotypes extracted as shown in this section were started from 

 1, 2, 3, or more pairs of like, homozygous-extracted types. These were either the 

 progeny from a single pair or from one or more pairs, and were allowed to inbreed 

 and continued to test the stability of the type. 



The simplest to deal with is the type 1, and this I have reared in such cultures 

 many times. A pair of F^ A' primary biotype 1 were taken from the culture 

 shown in figure 37 and allowed to breed. They gave 97 progeny, 47 males and 50 

 females, all of the black type, with no variation in pattern. These are not 

 melanothorax, but the black thorax transferred onto the multitceniata form. Of 

 these, 20 males and 20 females taken at random from the population were allowed 

 to breed freely and gave a progeny of 224 males and 217 females, all of the par- 

 ent type. From this population, 20 males and 20 females taken at random were 

 again allowed to breed, and these gave in the third generation 196 males and 194 

 females, all black. This operation I continued until the sixth generation with 

 exactly the same result, when it was discontinued. At no time was there trace 

 of other type or of variation in this line. 



An example of the breeding of a primary biotype as a group culture is shown 

 in figure 40, which represents a group culture of biotype 2. The original pair 

 were an extracted stock bred pure in pedigree culture for four generations and 

 came from parents obtained at the Chapultepec colony. 4 males and 4 females, 

 2 of each of the kind shown, were mated and these gave a progeny (FJ of 64 

 males and 80 females, distributed as shown in figure 40, on the scale of the range 

 in this biotype shown in figure 35, 2. From these 3 males and 3 females from 

 each class represented were taken at random and again inbred and gave a progeny 

 (Fj) of 183 males and 195 females, distributed between the two main classes. 3 

 males and 3 females from each of these classes were taken at random, mated, and 

 gave a progeny (F3) of 200 males and 197 females, distributed in four classes, 

 as shown. From these, 3 of each class were taken at random, as far as possible, 

 mated as the parents of generation 4, and gave a progeny of 369 males and 360 



