156 ADRIANO A. BUZZATI-TRAVERSO 



light-eyed gene of about .63. Since in both parallel populations the gene fre- 

 quency was similar, one could conclude that selection had favored the mutant 

 type, shifting its frequency from .5 to .63 in the course of about twenty 

 generations. 



Such an experiment did prove that the mutant gene had a positive selec- 

 tive value. It was impossible to know whether in the long run it would have 

 eventually eliminated its normal allele from the population. At this stage, I 



0.875^ 



Fig. 9.1 — Variation in the frequency of the Hght-eyed gene in selection experiments. In the 



abscissae is the number of generations, in the ordinates the gene frequency. Each line 



represents a single experiment on an artificial population. 



have begun a new experiment along the same lines, but with different gene 

 frequencies to start with. Two populations were started with 2 males and 2 

 females of the mutant type, plus 14 males and 14 females of the normal type. 

 Two populations were started with 16 mutant and 16 wild flies, and two 

 populations with 28 mutant and 4 wild type flies. 



I had, therefore, at the beginning of the experiment six populations. Two 

 had a gene frequency of the light-eyed mutant approximately equal to .125. 

 Two had a gene frequency of .5, and two had a gene frequency of .875. 

 Figure 9.1 shows the result of such an experiment after about fifteen genera- 

 tions. Crossings of wild type males, taken from the populations, with homo- 

 zygous recessive females showed that there was no significant departure from 

 random mating within the population. The gene frequencies indicated on the 



