76 



THE PEANUT— THE UNPREDICTABLE LEGUME 



and other lateral branches. The test was planted in a split plot design 

 where the sources of plants were the subplots (main stem, lateral branch, 

 or seed) and the three varieties used were the whole plots. No significant 

 differences in yield were found between the two sources of cuttings. The 

 yields of plants from seed were significantly less than those of plants 

 from main stem cuttings but not less than those from lateral branch 

 cuttings. The seed were planted the same day the rooted cuttings were 

 placed in the field, but the plants from seed may not have been strictly 

 comparable in developmental stage to the cuttings. The following table 

 shows the results obtained : 



Yields of Peanuts in Pounds /Acre 



(c) Field plot size 



Beattie et al. (8) concluded after some years' experience that the 

 optimum plot size fpr testing varietal differences in peanuts was a single 

 100-foot row or five to six 20-foot rows. Robinson et al. (60) investigated 

 optimum plot size from uniformity data. Their conclusions, briefly sum- 

 marized below, are perhaps the most reliable information to be had on 

 this subject. The uniformity data were handled in terms of decreasing co- 

 efficients of variation with increasing plot size. The following diagram 

 gives the pertinent results : 



