470 MUTATION AND PLANT BREEDING 



Table 1. — Experimental Design and Model Analysis for Comparing PI, P2, PIX, 

 P2X, F->, and F>X in Three Peanut Hybrids. 



Experimental design Treatment 



Replications 3 



Blocks in reps 4 



Between treatments in blocks 6 



Between trts. in block 1 3 



Between trts. in block 2 3 



Between Fi families in trts. in bl 32 



Between Fi fam. in trts. in bl. 1 16 



Between Fj fam. in trts. in bl. 2 16 



Between F> progenies in Fi fam. in trts. in bl 360 



Reps, x trts. in bl 18 



Reps, x fam. in trts 96 



Reps, x F 2 prog, in Fi fam 1 ,080 



Pooled error 1,194 



Total 1,599 



*Block 1 consisted of 10 F2 progenies of 5 Fi families each of treatments PI, PIX, F2, and F2X; 

 Block 2 consisted of P2, P2X, F2, and F2X. PX and FX refer to P-irradiated and F-irradiated, 

 respectively. 



progenies by treatments are given in Table 3. From this table the 

 effects of radiation vs. hybridization may be observed in the varia- 

 tion among the means as well as in the differences between treat- 

 ment grand means. The average genotypic variance for P was 

 exceedingly small, being zero for four cases out of six. Gregory (20) 

 postulated that the variation induced by radiation might be 

 cumulative with that of hybridization such that o- 2 G P + o- 2 G FX = 

 o- 2 G PX + o- 2 G F. When calculated over all three crosses, this expec- 

 tation was approached in the present experiment with o- 2 G P + o- 2 G FX= 

 76.9% of o- 2 G PX + a 2 G F. 



The above generalities are of importance, but the detailed 

 behavior of the individual Fi generation families holds the greater 



