NATURE OF THE GENETIC EFFECTS 405 



If the theory of dominance which has been presented is correct, neo- 

 morphic mutant genes, since they have not had an opportunity to be 

 exposed to a prolonged natural selection for high potency, or for modify- 

 ing genes in other loci that would tend to stabilize their effects, would usu- 

 ally find themselves in the near-linear region of the gene activity-effect 

 curve. Hence marked phenotypic differences would usually result (1) 

 when the dosage was changed, (2) when there were relatively slight 

 changes in environmental conditions and/or (3) in genes of any one of 

 numerous other loci than their own ("modifying genes"). Although the 

 normal allele, not having an effect of the same kind, would not be domi- 

 nant over them, neither would they show dominance over the normal 

 but only the same sliding-scale result as when, in the absence of a normal 

 allele, their dosage was changed. All this has been found to be the case. 

 Thus the heterozygous neomorph is usually approximately intermediate 

 in phenotype between the mutant and the normal homozygote, respec- 

 tively. These findings then serve to confirm the general interpretation 

 given in the foregoing discussion. 



In Drosophila, where the matter has been most studied, the gene muta- 

 tions produced by radiation have not so far appeared to show a different 

 trend from the spontaneous ones in regard to any of these dosage or 

 dominance effects. Moreover, the gene mutations originating from 

 irradiation or spontaneously, when classified according to their types of 

 dosage effects, have not shown obvious differences in their frequency 

 distributions. Nevertheless it is necessary for those dealing with radia- 

 tion mutations to be aware of these relations in order to know the way 

 in which mutations induced in one generation become expressed in sub- 

 sequent generations, and in order thereby to deduce (as shown in Sect. 

 20) the manner and speed with w^hich they become either multipUed or 

 eliminated from the population. 



15. RADIATION AND SPONTANEOUS GENE-MUTATION FREQUENCIES 



IN DROSOPHILA 



The determination of whether and to what extent radiation or any 

 other agent or condition affects the incidence of mutations has depended 

 upon the development of objective genetic techniques for the large-scale 

 detection and counting of representative mutations and for distinguishing 

 between newly arisen mutant genes or chromosomes and those pre- 

 viously present in a hidden heterozygous state in the stocks used. In 

 Drosophila sex-Hnked lethal mutations, i.e., those in the sex-determining 

 or X chromosome, were early found, in work of the present writer and 

 Altenburg, to be especially suitable material for such a study (Muller 

 and Altenburg, 1919, 1921 ; Muller, 1928c) . The chief advantages of using 

 them were that (1) the detection of sex-linked lethals already present in 



