GAUL: INDUCED MUTANTS IN SEED-PROPAGATED SPECIES 239 



has to be done in the generations succeeding the F 2 . The structure of 

 total work to be done by a breeder may be compared with a pyramid 

 having the top in the Fi generation and becoming broader with suc- 

 ceeding generations. Another example of efficiency in cross-breeding 

 was recently given by Williams (127) in tomatoes. The frequency of 

 successful hybrid combinations appears to be around 2 per cent in 

 that crop. Two such Fj hybrids which are used commercially in 

 England were analysed in order to fix heterosis and Williams isolated 

 1 desirable recombinant in every 1,000 to 1,500 F 2 individuals. 



Gregory (42), on the other hand, in his extensive work with 

 small mutations in X-rayed peanuts, indicates that the frequency of 

 mutants which are superior in yield may be of the order of 1 among 

 500 to 5,000 Mo population plants. Gustafsson (45, cf. also 43, 44, 48) 

 stated that a higher productive mutant is formed "once in 500 to 

 1,000 genotypical changes". It should be emphasized that this estimate 

 is based mainly on the selection of superior large mutations. From 

 our preliminary studies in barley it appears that out of 5 to 10 small 

 or fairly small earliness mutations, 1 may outyield the mother line. 

 Earliness mutations are frequent and are easy to detect; the small 

 ones, however, only in a group of plants. 



Consequently, if one critically tries to compare both methods, 

 breeding with mutations and traditional cross-breeding, there is no 

 evidence that the first is inferior. Much more experimental evidence 

 and practical experience are necessary for a sound evaluation of the 

 significance of breeding with mutations. 



The significance of mutations in the future depends largely on 

 the question of whether a higher total mutation rate can be obtained, 

 and particularly, whether the output of useful mutants can be 

 increased. The final output of mutants depends on the technique of 

 original induction as well as of selection. In both these fields theoreti- 

 cal progress has been made, and there is hope that the efficiency of 

 mutation production will be considerably increased. 



It would be useful, if, in various parts of the world, mutant 

 collections of the main crops could be established which are based 

 on adapted varieties with high performance. Collections of large 

 mutations are easy to establish even with present methods. The 

 expenditure is not too great as compared, e.g., with collection trips to 

 centers of genetic diversity. These large mutations imply a valuable 



