GREGORY! EFFICACY OF MUTATION BREEDING 



477 



crosses A and D in 1960. In 1960 the plant progenies of the high and 

 low selections were randomized together by treatment permitting a 

 significance test for high versus low. This was not possible in 1959. 



The individual means and variances were estimated for each 

 treatment. The treatment means, together with the indications of 

 significance of differences among progenies in treatments, are shown 

 in Table 7. 



The comparisons of s G by F 2 generation family for the high 

 and low selections from the varions treatments are given in Figure 

 3 for crosses B and C. The cross mean is shown as 100 per cent of 

 all treatments. The bases of the bar graphs represent the F 2 fam- 



Taele 7. — F 2 Family Mean Yields cf F5 Plant Progenies in F 6 Generation by 

 Treatment in High and Low F z Families of Crosses B and C, 1959; 

 Crosses A and D, 1960. 



"Field design did not permit a high vs. low test in crosses B and C 



''Omitted because found to be a hybrid irradiated population selected and tested by accident. 



NS = Nonsignificant. 



•Significant at the 5 per cent level of probability. 



•'Significant at the 1 per cent level of probability. 



Superscript to the right of each mean refers to significance of differences between Fo progenies in 



families. 



