Analysis of Heterogeneity in Complex Chakacters 245 



The biotype 4 {D') was not mated and one pair of biotype {D"), which gave 

 progeny, gave in F3 biotype 2 (i)"a), biotype 3 {D"b), and an extreme of biotype 

 8 {D"c) . One pair of F3 of the type 4 {D"a) which gave progeny was the basis 

 of a line breeding true and constant. The other pairs gave biotype 4 and biotype 

 3 (not shown). Three matings of biotype 4 {DEa) gave results like similar 

 matings in F3. The biotype 8 extreme {D"c), of which there were only two 

 pairs that gave progeny, gave in one case all biotype 8 extreme {D"ca), and in 

 the other pair (D"ah) two biotypes, 8 and 4, appeared, which, however, were not 

 fully tested. 



From the mating E in F^ the results from two pairs are shown, one (£") 

 breeding true in Fg and subsequently ; the other giving in Fg a population of a 

 biotype 12 {E"a), a biotype 4 {E"b), and an extreme biotype 2 {E"b) . Of these 

 only the extreme biotype of 2 {E"c) were mated and one pair gave in F3 biotype 

 2 again {E"c), a biotype 3 {E"ca) and the other extreme of type 4 {E"cb). 

 Only the biotype 2{E"c) was mated in F3, and this pair gave in F4 biotype 2 

 {E"c) and biotype 3. In F^ five pairs of biotype 2 were mated, one of which 

 gave biotype 2 (E"c) and types 3 and 4, and of this Fg population matings of the 

 biotype 2 {E"c) showed one that was constant in Fg and F^. 



Fig. 38. — Showing results In a second pair of 

 L. muUitCBtiiata from the Chapultepec colony, 

 having a pronotum of biotype 1 ; although ob- 

 viously heterozygous In composition they were 

 apparently much less complex than the similar 

 pair shown In figure 37. 



It is evident that the original pair * were not pure black type 1, but complex 

 hybrid combinations in which different potentialities were combined in such a 

 series of complexes that the chance union of these two gave rise to a striking 

 array of types, out of which several distinct races were extracted and maintained 

 as pure-breeding strains. It is also shown that although the original stocks 

 were exceedingly heterozygous in character, their progeny were, nevertheless, 

 quite fixed in type and the array found was entirely due to diverse germinal 

 constitution. 



Although the pair and their progeny show extremely complicated germinal 

 constitution, not all tested pairs were so complex. Thus in figure 38 is shown 

 a pair obtained at the Chapultepec colony in 1908, which gave a simple behavior, 



^ Both of these were typical " L. melanothorax Stk\ " as found in all museum col- 

 lections and would unless tested by breeding have passed as pure typical specimens 

 of that " species." 



