gaul: induced mutants in seed-propagated species 



235 





E 



I 

 to 



LU 



LU 



2 

 CE 



LU 



O 



o 

 o 



50" 



49-- 



48- 



47 - 



46-- 



45-- 



44- 



43 



46 



47 



-t- 



49 



WEIGHT 



—I 



50 



1958 



51 



52 



48 



1000-KERNEL WEIGHT gm. 



Figure 4. — Correlation of 1 ,000-kernel weight between 1958 and 1959 

 in 31 selected lines from X-rayed winter barley with regression lines. 

 One barley line, marked by a dot xuith arrow, lies outside of the dia- 

 gram (1958, 55.2; 1959, 50.2). r = 0.745. 



We also investigated this question in more detail by statistical 

 means. The irradiated seeds were in the F 2 i generation and practical- 

 ly all genes were, therefore, in a homozygous condition. The irradi- 

 ated material was a mixture of 15 "nursery-lines", which were pheno- 

 typically very similar, if not alike. In 1960, progenies of eight of these 

 (untreated) lines were available for us to determine the 1 ,000-kernel 

 weight. In the meantime, 23 lines had been developed from the eight 

 progenies. The variance of the 1 ,000-kernel weight of the 23 lines was 

 s 2 = 0.571 and the corresponding variance of our 31 selection lines 

 was s 2 = 4.92 (determined as mean square deviation in relation to the 

 mean value of our 31 selection lines). According to that, the variance 

 of the selection lines is 8 to 9 times greater. The difference is highly 

 significant (P > 0.1 per cent). 



A similar selection program as with winter barley was conducted 

 with spring barley, variety Haisa II (Hochzucht). The results of the 

 first drill test in 1959 are demonstrated in Figure 6. It is obvious 

 that a great number of small variants of kernel size have been 



