234 



MUTATION AND PLANT BREEDING 



LU 



Z 



or 



UJ 



m 



98 



100 



102 



10* 



106 



108 



110 



112 



114 



116 



RELATIVE 1000-KERNEL WEIGHT 

 Figure 3. — Effect of recurrent selection for higher 1 ,000 -kernel weight 

 in progenies of X-rayed winter barley (variety Breustedts Atlas), accord- 

 ing to the procedure of Table 9, results of 1959. Control is taken as 100. 

 Only 4 lines out of 31 have a lower 1 ,000-kernel weight than the control. 



In Figure 4 a scatter diagram of the results in 1958 and 1959 is 

 represented. There is a highly significant correlation between both 

 years (r = 0.745, P > 0.1 per cent). A similar correlation was obtained 

 between 1959 and 1960, as is shown in Figure 5 (r = 0.797, P > 0.1 

 per cent) and also between 1958 and 1960 (r = 0.790, P > 0.1 per cent). 

 Because of severe lodging in 1960, only marginal plants could be sam- 

 pled from the drilled plots. Thus, the sampling has been done in 

 1958 from hills, in 1959 from drilled plots, and in 1960 from marginal 

 plants of drilled plots. Because of the high correlation between the 

 different years and growing conditions, there can be no doubt that 

 the differences between most of the lines are genetically controlled. 



At present we are not completely sure, however, whether all these 

 variants of kernel size are true mutants though this is very probable 

 for the majority at least. The seeds, which were irradiated in 1955 

 were "Zuchtgarten-Elitegemisch" and not a pure line obtained by 

 propagation of a single plant for only a few generations. According to 

 the breeding history, the propagation methods, and his experience, 

 the breeder of this barley believes it is not possible that the variants 

 selected in the irradiated material were already present in the starting 

 material (Breustedt, personal communication). 



