Section 9 — Population Genetics 



of certain genes with equal frequencies, 0.5. As 

 expected, natural selection led in all cases to 

 decreases in the relative frequencies of the mut- 

 ant and to increases of those of the wild type 

 alleles. The selection rates proved, however, to 

 be in some cases very different depending on 

 whether the foundation stocks contained only 

 one or many kinds of wild type chromosomes. 

 The "single-strain" populations were derived 

 from crosses of mutant flies to flies from a highly 

 inbred wild-type stock. The "multiple strain" 

 population's non-mutant chromosomes were 

 derived from many (about a dozen) wild strains 

 of different origin. With the third-chromosome 

 mutants ebony, scarlet and stubble, and ap- 

 parently also with the second-chromosome 

 brown-75, the selection rates against the mutants 

 were initially much greater in multiple-strain 

 than in single-strain populations; in later 

 generations this difference tends to disappear. 

 With the sex-linked forked and the fourth- 

 chromosome polished, the single strain and the 

 multiple strain populations showed similar 

 behavior within the limits of the experimental 

 errors. The data are, on the whole, consistent 

 with the assumption that the selection rates 

 against certain mutant genes depend not only 

 on the effects of these genes themselves but also 

 on those of the polygenic complexes associated 

 with them in the same chromosome. 



in two strains of Drosophila nielanogaster. 

 One strain, S, had been subjected to 40 genera- 

 tions of stabilizing selection on development 

 time. The other strain, D, had been subjected 

 to 40 generations of disruptive selection for the 

 same trait. 



The primary effects of selection (previously 

 published) show a decrease in the within culture 

 variance of the S line and an increase in the 

 D line and further, that the increased variance 

 in the latter was probably due to a loss of buf- 

 fering. 



The present report concerns further studies of 

 the comparative sensitivity (buffering) of the 

 two lines to (1) environmental variables which 

 operate among cultures, (2) temperature varia- 

 tion, and (3) variations in an utritional variable 

 known to effect development time. 



The results of these studies indicate that the 

 two lines do not differ in their buffering against 

 these three factors. 



It is concluded, therefore, that the genotypes 

 involved in the differentiation of these lines are 

 rather specific in limiting their effects to buffer- 

 ing only against the microenvironmental varia- 

 tion within cultures. If loci for "generalized 

 buffering" exist, apparently they were not segre- 

 gating in these lines. In general these findings 

 contrast somewhat with studies of the effects of 

 inbreeding on buffering. 



9.65. Chromosomal Inversion Polymorphism and 

 Size in Drosophila subobscura. Antonio 

 Prevosti (Barcelona, Spain). 



Data are presented showing that some chromo- 

 somal orders of Drosophila subobscura present a 

 clear frequency cline north — south. The results of 

 experiments of artificial selection for size car- 

 ried on in the laboratory are presented. It is 

 found that as a correlated response, a tendency 

 exists to fixation in homozygous condition of 

 complex orders, when selecting for small size. 

 However, selection for larger size increases the 

 frequency of the heterozygosity between the 

 standard and the complex inversion types of the 

 chromosomes. Experiments planned to correlate 

 the chromosomal structure of individuals and 

 its size are in preparation. The preliminary results 

 will be given. 



9.66. The Effects of a History of Stabilizing Selec- 

 tion on Sensitivity to Foreign Environments. 



T. Prout (Riverside, U.S.A.). 



The sensitivity of the time of development to 

 certain environmental variables was investigated 



9.67. Towards a Synthesis of Population and Physio- 

 logical Genetics. Forbes W. Robertson 

 (Edinburgh, Great Britain). 



The genetic properties of components of fit- 

 ness, their inter-relations and importance in 

 determining the stability of other characters must 

 be related both to development and ecology. 

 Such information will enable us to specify the 

 conditions, genetic and environmental, which 

 favour particular kinds of adaptive change 

 and thereby bridge the gap between population 

 and physiological genetics. This approach has 

 been applied to the growth of Drosophila nie- 

 lanogaster by selecting for body size, develop- 

 ment time and egg production on different 

 controlled diets and also by analysing the 

 changes which have accompanied adaptation 

 by a number of populations to novel nutritional 

 conditions. Growth can be divided into an 

 exponential and a later, slower phase and changes 

 in either appear to be genetically independent, 

 within limits. Both contribute to the total variance 

 of adult size, but their genetic properties dif- 



165 



