NATURAL SELECTION: I 457 



probably have any conception of the magnitude of the investigation we 

 have summarized so briefly. The authors state that 410,784 flies were classi- 

 fied and recorded! And those must have constituted but a small fraction of 

 the total number of flies raised. While we may not be able to "put a fin- 

 ger on" the individual mutations whose occurrence led to the improve- 

 ments cited, we cannot doubt that such favorable mutations occurred, 

 and that natural selection utilized them for the improvement of the strains. 



THE ROLE OF H ET E R OZ YGOT ES 



Slightly dift'ering experiments in natural selection have 

 revealed the importance of heterozygotes in evolution. Pioneer experimenta- 

 tion in this field was that of L'Heritier and Teissier (1937), who designed 

 a type of cage for rearing Drosophila which made possible the maintenance 

 of a large population over an extended period of time. Experiments began 

 with populations of around 4000 individuals, all of which possessed a cer- 

 tain mutation. A few normal, "wild-type" flies were then introduced into 

 the cage. Thus competition was established between normal individuals and 

 individuals possessing a mutation. Nature was allowed to take its course 

 as time passed and generation followed generation. 



In some experiments the flies originally present possessed a malforma- 

 tion of the shape of the eye, called "bar eye." The population was "in- 

 fected" by addition of a few normal flies. At first the population was prac- 

 tically 100 percent bar eyed. But in the ensuing competition the proportion 

 of bar-eyed flies declined (Fig. 20.3). At first the decline was rapid, but 

 as the number of bar-eyed flies decreased the rate of decline lessened. By 

 the end of 235 days, in one experiment, bar-eyed individuals constituted 

 only about 28 percent of the population, the remaining 72 percent being 

 normal eyed. By the end of 426 days the bar-eyed flies constituted only 

 about 1 percent of the population. They continued at or below this very 

 low frequency as long as the experiment lasted. Here is another example of 

 natural selection operating under experimental conditions. In this case the 

 bar-eyed flies were clearly less well adapted than were normal flies and 

 hence lost out almost completely in competition with the latter. 



Another mutation in Drosophila is "ebony," the most evident eff'ect of 

 which is darkening of the body color, from the normal gray. L'Heritier 

 and Teissier performed the same experiment starting with ebony flies. As 

 shown in Fig. 20.3, the proportion of ebony flies declined rapidly at first; 

 by the 235th day it had reached about 28 percent, the remaining 72 percent 

 consisting of normal flies. But the further course of the experiment dift'ered 



