HEREDITY 343 



Effect of Selection. Would it be possible to develop a race of 

 supermen by carefully selected matings? For illustration let us select 

 one character, height. The height of any human is the sum total 

 of a number of influences. Unquestionably one of these factors 

 is included in the inheritance he derived from his ancestral line, but 

 in addition nutritive conditions and disease during growth, and the 

 activity of the pituitary hormone act on this genetic factor so that 

 the eventual height may be influenced in both directions and 

 amount to the algebraic sum of the effects of these influences. In 

 plants it has been shown that selection is effective in increasing the 

 frequency of occurrence of tall individuals in a population, so that 

 the general average of height increases for a number of generations 

 as selection is continued. But selection soon ceases to be effec- 

 tive; the maximum has been reached. Fluctuations around this 

 maximum are due to non-genetic conditions. This maximum height 

 represents a pure line for all the allelomorphic pairs of genes that 

 have to do with height of the offspring. When all these are homo- 

 zygous no increase in height is accomplished by further selection. 

 A new height character can arise only as a mutation and so far as 

 is known, selection in breeding does not induce a mutation in the 

 character selected. Going back to the original question, a human 

 race might possibly be produced by selection, the average height of 

 which is distinctly greater than that of the present races, but there 

 is no reason to regard it probable that a race of giants could be so 

 developed (Fig. 211). 



The Nature of the Gene. In the preceding discussion of the 

 principles of heredity we have used the term gene freely. It is en- 

 dowed with powers of continuing itself unchanged through genera- 



FiG. 211. — Graphs showing the effects of selection and breeding for size through 

 ten generations of potato beetles. Black areas indicate the sizes selected for mating in 

 each generation. Thus in the second generation a pair smaller than the average and 

 one larger than the average were mated. This was practiced throughout. Note that 

 after the sixth generation the average size did not increase further. In the smaller line 

 in the seventh generation a random mating was made. Note that its descendants 

 returned to the normal for the species in the tenth generation. (After Tower.) 



