342 MUTATION AND PLANT BREEDING 



erectoicles mutations, respectively, have been analyzed. As a general 

 rule lor all characteristics, it seems more difficult to induce and 

 observe an improvement in a direction to which the parental variety 

 has already been intensively bred (70). 



It is thus evident that the possibility to be able to induce, 

 observe, and properly evaluate a mutation is highly dependent on 

 parental genotype, and for that reason the breeder should not restrict 

 his basic materal too drastically (15) and only to his very best strain. 

 The choice of material especially suitable for the detection of the 

 desired mutations can sometimes greatly improve the experiment. 

 Thus, Melchers (75) demonstrated the use of haploid plants in 

 Antirrhinum for detection of recessive mutations. Monosomies or 

 other types with known deficiencies could be used in a similar man- 

 ner. Mertens and Burdick (76) used the technique of irradiating 

 autodiploids (2n ex-haploids) and then backcrossing and comparing 

 them with an untreated control in order to detect mutations in quan- 

 titative traits that could exhibit a dominance type of heterosis in 

 tomato to F/s. Genotypes with appropriate marker genes or trans- 

 locations can greatly improve the detection of desirable chromosomal 

 rearrangements in an irradiated material (19, 46, 93). 



The choice of the most suitable genotype in accordance with 

 the intention of the mutation experiment is, however, only one side 

 of the problem of how to render mutation breeding as efficient as 

 possible. The chance to induce the desired mutations also depends 

 much on the metabolic stage of the treated cells, the mutagen, and 

 the dose applied as well as on the handling of the material before, 

 during, and after treatment. These problems are, however, taken up 

 by other speakers on this symposium, and we will here continue to 

 center our interest in pure breeding techniques. 



In the treated Mi generation, the breeder has certain possibili- 

 ties to influence the final yield of induced mutations. As in a hybrid 

 bulk population, competition between genotypes will occur, but this 

 competition will start already on the cellular level if plants or seeds 

 are treated. Such an intrasomatic or diplontic selection is desirable 

 as long as only deleterious mutations are screened away as in the very 

 first cell divisions after treatment. Later on, the risk increases that 

 cells carrying valuable mutations may also be suppressed and lost, 

 and it is, therefore, important to overcome this phenomenon of 

 intraplant selection. 



