376 Inbreeding, Selection, and Heterosis 



parents of the next generation in the one case, and the smallest 

 plants in each generation in the other. Such a regular choice, 

 year after year, of a plant of a certain type as the parent of the 

 next generation is known as selection. It is probably the oldest 

 method of plant and animal breeding and is certainly a very 

 important one. In some of our generations above, selection was 

 an important factor in increasing or decreasing the height of our 

 plant, but in other generations it was not. Why is there this 

 difference? When we study the genotypes of these two groups 

 of plants, we find that the plants that were not effective in 

 selection w^ere homozygotes, whereas those that were effective 

 were heterozygous for at least one locus. It is a w^ell-established 

 rule, specifically established by Johannsen in 1903, that when 

 we are dealing with homozygotes we can not vary our strain 

 by selection. 



In discussing the previous case, we simplified our problem by 

 the implied assumption that all the variation was hereditary. 

 Obviously, it would not be so in an actual experiment, although 

 the extent to which variation is the result of environmental 

 factors will differ with the organism, the particular part of the 

 organism, and the nature of the environment. Even though we 

 attempt to eliminate the effect of environment by raising all the 

 plants or animals of a given experiment under essentially the 

 same conditions, we can never control external conditions so 

 rigidly that their effects cannot be noticed at least to a small 

 extent. 



Environment may sometimes have such a large effect that 

 organisms of one genotype may appear phenotypically like those 

 of a class larger or of a class smaller. Thus the action of en- 

 vironment on quantitative characters may result in the over- 

 lapping of phenotypes. In our hypothetical problem, some T^t^ 

 T2T2 T3T3 plants which should be theoretically 50 cm tall might 

 grow in such rich, moist soil that they would attain a height of 

 53 or 54 cm. On the other hand, some T^Ti T2T2 T^Ts plants 

 might grow in such an unfavorable location that they would be 

 only 51 or 52 cm. Because of environmental differences, these 

 two groups of plants would be confused with plants of different 

 genotypes. In trying to establish a giant strain, we should 

 probably choose the 53- or 54-cm plant as the parent of the next 

 generation because of its large size, but we would find that its 



