LECTURES IN BIOLOGICAL SCIENCES 



emphasized, for instance, by Lerner (1954), that the heterozy- 

 goses for these minor allelic differences are usually fitter than 

 either of the two homozygous types. In so far as this is so, 

 natural selection will not lead to a state of genetic purity in 

 a population, but will preserve a heterozygous condition, that 

 is to say, one which contains much genetic variability. Further, 

 the studies, particularly those of Dobzhansky (1955), have shown 

 that each local population of a species contains a collection of 

 allelic differences which are such that they interact with one 

 another to give favourable heterozygotes or gene combinations, 

 whereas genes taken from one population often give rather bad 

 results when combined, in hybrids with genes from a different 

 population. It is usual to speak of the whole collection of genes 

 which are found in a population as the "gene pool." Dob- 

 zhansky's work, just mentioned, leads to the conclusion that 

 within the gene pool of a single population the genes are co- 

 adapted to one another. 



These developments in the field of population genetics are 

 making possible a much more profound understanding of the 

 processes of natural selection. Before the complexity of the 

 genetic structure of wild populations was fully understood, 

 work had already been carried out on the natural selection of 

 characters controlled by single genes with marked effects. There 

 are comparatively few cases in which evolutionary changes in 

 such characters have been observed to proceed at rates fast 

 enough for detailed scientific study. One of these is the ap- 

 pearance of melanic forms in Lepidoptera in industrial areas. 

 The existence of natural selection, favouring the dark forms in 

 areas contaminated by industrial smoke and the light forms in 

 uncontaminated areas, was demonstrated by Heslop-Harrison 

 (1920), and is still being studied in detail, for instance, by Kettle- 

 well (1955). In quite recent years the appearance of popula- 

 tions of insects resistant to insecticides, such as DDT, has posed 

 both an important practical problem and a wonderful oppor- 

 tunity for the study of evolution (Crow, 1957; Oshima, 1958). 

 In many cases, however, these resistant forms have turned out 

 to depend on the cooperative action of many genes, rather 

 than on a single one with large effect. 



From what has been said above about the genetic structure 



24 



