gaul: induced mutants in seed-propagated species 227 



been investigated in several other plants, e.g. in maize (1), sorghum 

 (64), pea (6), and blue lupin (36). 



In practical breeding work, selection for mutations may perhaps 

 alreardy be started with seedlings of the Mi generation. We have 

 shown that raising X-rayed barley seeds in the greenhouse before 

 transplanting them to the field may result in 4 to 5 times better 

 survival than when seeds are sown directly in the field (33). The 

 mutation frequency of the "greenhouse seedlings" was not essentially 

 lower than that of the "field seedlings". This finding complements 

 interesting results of Caldecott (13), who demonstrated that heavily 

 damaged seedlings (shortest height class of the length of the first leaf) 

 have a 2 to 4 times higher mutation rate than those which belong; 

 to the tallest height class. Thus, with the technique of raising seed- 

 lings in the greenhouse, it might be expected (a) that higher doses 

 can be applied, resulting in higher mutation rates, and (b) that the 

 selection of seedlings with short leaves (but which are still viable) 

 will result in an additional increase in mutation rates. 



Similarly, Blixt, et al. (7) found recently in peas, after treatment 

 with ethylen imine, a correlation between the M 2 mutation rate and 

 the frequency of leaf-spots on the first leaves of Mi plants. In order to 

 obtain a high rate of M 2 mutations, they suggest, therefore, taking 

 offsprings from those plants only which show the highest frequency of 

 leaf-spots at the Mi seedling stage. 



Selection of Mutants — General Results 



In practical breeding work it will often be advisable to com- 

 mence with the selection of mutants only after mutagenic treatments 

 have been applied for several years or generations to the same mate- 

 rial (23, 34). Recurrent induction of mutations leads to an accumula- 

 tion of mutants in the treated population and increases the efficiency 

 in breeding with mutations. 



Selection for fertility can already start in the Mi generation or 

 in case of recurrent induction of mutations, it may be repeated in 

 each succeeding generation. After treatment of barley seeds, we have 

 shown repeatedly that the mutation frequency of fertile Mi spikes is 

 no smaller than that of the Mi spikes with partial sterility if the tiller- 

 ing of the Mi plants is reduced (25, 28, 29, 34). Recent results in rice 

 (4, 5) agree basically with barley. Selection of fertile (or nearly fertile) 



