GREGORY: EFFICACY OF MUTATION BREEDING 481 



ological mutants. Variable F 2 individuals of the same mutant pheno- 

 types had arisen from a cross of two highly uniform mutant par- 

 ents. The most plausible explanation for the new variability observed, 

 lay in the presumption of the genetic segregation of modifiers 

 located in the chromosomes of the background genotype. This pre- 

 sumption was further confirmed when Emery, Gregory, and Loesch 

 (unpublished) made a study of the control type segregates from F 2 

 generation hybrids of various morphological mutants. They dis- 

 covered highly significant genotypic variances among nonsegregat- 

 ing control-type F 2 progenies in F 4 generation for several quantita- 

 tive characters. Finally, Emery (unpublished) showed that even the 

 quantitative variation of the double recessives of these morphologi- 

 cal mutants could be attributed to effects of genetic background. 

 Furthermore, Emery showed in the double recessives that the back- 

 ground effect was significantly associated with the parental source 

 of the background, suggesting at least a multi-chromosomal effect 

 if not a polygenic one. 



These observations support the thesis that the mutagenic action 

 of radiation may be used for the induction of variability in quanti- 

 tative characters. They also lend support to the hypothesis that 

 certain species of plants are capable of absorbing relatively large 

 doses of mutation without crossing the threshold of obvious pheno- 

 typic expression and to the further hypothesis that such mutation per- 

 mits a response to selection sufficient to exceed the reduction in the 

 mean fitness occasioned by the treatment. 



These observations do not lend much support to the general 

 thesis that high mutation tolerance in the selfbreds has provided 

 for the function performed by balanced heterozygosity in the cross- 

 breds. Credence here can come only with relatively large accumu- 

 lations of data on a number of species of different breeding system 

 and from experiments specifically designed to furnish information 

 on this subject. Nevertheless it is our working hypothesis. 



The data brought forward in the present paper give only a 

 foretaste of the solution of the problem of the relative efficacy of 

 conventional and mutational breeding in selfbreds. Much more 

 must be learned concerning additional environmental agents and 

 breeding procedures which selectively eliminate undesirable changes 

 induced by mutagens while permitting the successful exploitation 



