NATURE OF THE GENETIC EFFECTS 437 



derived from one or from a mixture of a few more or less inbred lines, 

 or from relatively few progenitors; in that case, moreover, the population 

 will also start out with unusually restricted genetic variability, which the 

 apphcation of radiation will tend to remedy. 



All the above conditions are fulfilled in certain recent experiments 

 which have been carried out with Drosophila populations in independent 

 work of Wallace (1951) and of Buzatti-Traverso (1951). In these experi- 

 ments it was found that application of radiation to considerable labora- 

 tory populations, even for many successive generations, in doses that 

 were rather heavy for this organism, resulted over a long period in a 

 greater improvement of productivity of the flies, as measured under the 

 laboratory conditions, than occurred in the nonirradiated control popu- 

 lations. These important but not surprising results are in harmony with 

 the fact that a single pair of Drosophila will if given an opportunity pro- 

 duce hundreds of offspring, so that lines of descent derived from a rela- 

 tively few flies of higher productivity are under some circumstances 

 capable of displacing in a rather short time those derived from numerous 

 others, of lower productivity. The free opportunity for breeding in the 

 relatively large population cages used allowed intense natural selection 

 of this sort, even though not so marked as this extreme example would 

 suggest. Thus the high proportion of deleterious mutants could be kept 

 from swamping the population, while those rare types which possessed 

 any features that were advantageous under the given conditions outbred 

 the original type. 



Even so, it is very unlikely that a comparable improvement could have 

 been brought about by these means in a natural population of Drosophila 

 living under the conditions of nature. For one thing, these experimental 

 populations must have had a restricted genetic background, as compared 

 with natural populations, and this made a rise in the frequency of muta- 

 tions more advantageous than it ordinarily would be. Second, there are 

 many important features of the natural environment missing in even the 

 best laboratory population cages (e.g., the relative inaccessibility of food 

 in nature, and sometimes of mates; the incidence of drought, wind, preda- 

 tors and parasites; the existence of competing species and of other natural 

 dangers), while on the other hand other features are present in the labo- 

 ratory containers in more marked degree than usual. Under these 

 changed conditions, opening new evolutionary pathways, not a few 

 mutations would now be helpful which would have been disadvantageous 

 in a state of nature, and the accumulation of these mutations might 

 easily more than cancel the effect of an otherwise increased genetic load. 

 In the course of the genetic reorganization process, retrogression in adap- 

 tation to the many features no longer of importance could occur with 

 relative impunity, and those mutations furthering adaptation to the spe- 

 cial laboratory conditions would still be advantageous even if they had 



