450 RADIATION BIOLOGY 



those with one dose of a normal gene, and hence, too, between homo- 

 zygous and heterozygous normals, to affect their genetic survival sig- 

 nificantly^ — so significantly, in fact, as to have led to the establishment of 

 systems of dosage-compensating genes. This then demonstrated the 

 importance for the organism of shades of difference so minute as to be 

 below the threshold for ordinary detection, and showed that these sub- 

 liminal effects have been accumulated in the course of natural selection 

 until a remarkably high degree of precision of genetic adaptation has been 

 attained by the normal type. 



The conclusion, thus doubly arrived at through radiation studies, that 

 the dominance of normal genes is not actually complete, was later veri- 

 fied, by the direct measurements referred to in Sect. 14, of the viabihty of 

 indi\'iduals heterozygous for lethal and sublethal genes that had been 

 produced in radiation experiments. These showed the amount of expres- 

 sion of the "recessive" mutant genes in the heterozygote to be sufficient 

 to result in mutant genes being eliminated while in the heterozygous 

 condition, in the great majority of cases. Since the evidence of Levit 

 derived from spontaneous mutations in man fitted in with this, an entire 

 reordering of ideas and recalculation of results pertaining to rates and 

 curves of elimination, types of expression, and equilibrium frequencies of 

 mutant genes — whether spontaneous or induced — in populations became 

 necessary. The application of these methods to the actual situation also 

 required the use of another contribution of radiation genetics, in which 

 light had been thrown on the relative frequencies of mutations having 

 different types and degrees of expression: visibles, detrimentals, steriles, 

 and lethals, and in which estimates had thereby been arrived at of the 

 total frequency of mutations. In the process of combining the results 

 from the two fields of investigation, on the degree of dominance and on 

 the frequencies of mutations, respectively, the concept of genetic load 

 had to be introduced. It proved a fruitful one in assessing the effects of 

 radiation and of selection on populations and on the individuals com- 

 posing them. 



Studies of chromosome changes produced by radiation threw light from 

 still different angles on the properties of genes. For example, it was the 

 finding of the regularity with which, in Drosophila, structural changes are 

 accompanied by detectable phenotypic effects, such as lethality, sterility, 

 and morphological abnormalities, that suggested the conception of posi- 

 tion effect as a general, fundamental phenomenon (even though not 

 evident in most organisms), rather than one confined to special cases. 

 Numerous subseciuent studies, employing chromosomes changed struc- 

 turally by radiation in various ways, verified this idea and disclosed 

 important additional features, such as the peculiarities of the position 

 effects resulting from the juxtaposition of eu- and heterochromatic 



