NATURAL SELECTION: I 455 



small, crumpled wings, elongated, cylindrical abdomens, and sometimes 

 short and crooked legs. Obviously these flies were far below their "adap- 

 tive peak." Stocks of flies homozygous for this chromosome were estab- 

 lished in culture bottles and raised generation after generation. The cul- 

 ture bottles became overcrowded and were deliberately kept that way to 

 provide a restrictive factor making for natural selection. In the later gen- 

 erations of the experiment natural selection was further abetted by keep- 

 ing the cultures at a temperature high enough to be deleterious. From time 

 to time, as the generations passed, flies were removed and the PA748 chro- 

 mosomes they carried were tested to determine whether there had been 

 any improvement in the genes present. It was found that by the tenth gen- 

 eration viability had clearly improved and that by the fiftieth generation 

 it was almost normal. Improvement in the speed of development from egg 

 to adult occurred more slowly, but at some time between the thirty-eighth 

 and fiftieth generations it had become normal. (We recall that rate of 

 development may constitute an important factor in the relative success of 

 a species. See p. 357.) By the fiftieth generation, also, the wings and legs 

 had become normal, though the elongated, cylindrical abdomen had re- 

 mained unchanged. Evidently appropriate mutations for rectifying the ab- 

 normalities of the abdomen had not occurred. 



All told, Dobzhansky and Spassky performed this experiment with 

 strains of flies homozygous for seven different chromosomes having dele- 

 terious effects. Of each strain two stocks were established: one received 

 X-ray treatment; the other was left untreated. It was thought that the 

 radiation might increase the rate at which mutations occurred and hence 

 the rate of evolutionary change. The X-ray treatments seemed to have 

 no particular advantageous effect, however. Ten of the fourteen experi- 

 mental stocks showed improvements in the genetic contents of the re- 

 spective chromosomes concerned. Three stocks remained unchanged as 

 the generations passed, and one actually deteriorated. This is exactly the 

 sort of result which would be expected if improvement depended upon the 

 chance occurrence of suitable mutations for natural selection to act upon. 



The results in terms of viability alone are summarized in Fig. 20.2. It 

 will be noted that eleven of the stocks improved in viability — some very 

 markedly, some only slightly. Two stocks declined slightly in viability; one 

 remained unchanged. 



Parallel with the stocks just mentioned the authors kept "balanced 

 strains" in which the chromosomes under study were "protected" from the 

 action of natural selection by the presence of normal genes in the other 

 chromosome of the pair. Of the chromosomes possessed by these stocks six 



