466 MUTATION AND PLANT BREEDING 



tion has already increased the load of mutations to the optimum 

 for the exigencies of the environment for population size, for loss 

 of mutant alleles through homozygosis, and for balanced mutation 

 rate through modulators, any marked increase in mutation would 

 likely result in a diminution of fitness. Hutchinson (27) and Muller 

 (37) especially have maintained that little is to be expected of muta- 

 tion breeding with this system of mating. These strictures do not 

 preclude positive results from mutation breeding of crossbreds (38) 

 in instances of small breeding populations or where the crossbred 

 population is near or at the limit of its range, either case necessitat- 

 ing a past history of inbreeding. They do suggest that mutation 

 breeding, in a panmictic population of large size and great com- 

 plexity, unaccompanied by a radical departure from customary 

 breeding procedure, would probably be ineffectual. The somewhat 

 negative conclusion here should take into account the selection work 

 reported on irradiated populations of Drosophila (45). 



Lacking the necessary experimental data to prove or to disprove 

 the above conclusion and unable to review the enormous litera- 

 ture allied to this subject, I would like to stop with the suggestion 

 that the effectiveness of selection following mutagenic treatment 

 vs. no treatment, with and without radical change of breeding pro- 

 cedure, be investigated in a panmictic crop plant — shall we say rye 

 or corn. 



Selfbreds 



Mather has emphasized the contrasting organizations of the 

 genetic systems in self- and cross-pollinators. In an extension of 

 Mather's thesis, Gregory (19, 22) postulated that self-pollinators, 

 faced with the alternatives of elimination or fixation of every muta- 

 tion, have evolved genomic systems capable of absorbing relatively 

 large numbers of mutations of small effect without their necessarily 

 reaching the thresholds for phenotypic expression. According to this 

 hypothesis genetic organization favoring resiliency of the pheno- 

 type of the homozygote in the self-pollinator would tend to become 

 established by selection. The genetic factors controlling the ten- 

 dency of mutations to occur in the direction of resiliency would 

 likewise become ensconced in the genome. The genome would 

 become laced with supporting modifier gene complexes with little 



