EXPERIMENTS IN FIELD PLOT TECHNIC 61 



sixth plot. As a result of the adjustment of yields, the average devia- 

 tion of 3-row plots was reduced from 15.35 per cent to 12.91 per cent, 

 a reduction of 16 per cent, and that of 5-row plots from 15.44 per 

 cent to 12.15 per cent, a reduction of 21 per cent. 



The relative variability of actual and adjusted yields in the oats 

 variety test of 1919 is shown in Table 32. In this field the check, 

 Red Rustproof, was in every ninth plot. When the series affected by 

 the faulty check yields of the border plots is included the variability of 

 the adjusted yields is slightly higher than that of the actual yields, 

 but when this series is discarded the average variability as measured 

 by the mean deviation is reduced 19 per cent. 



It might be expected that the oats strains grown on the same field 

 would show a greater reduction of variability than the varieties, since 

 practically all of them were of the same variety as the check, and since 



TABLE 33. RELATIVE VARIABILITY OF ACTUAL AND ADJUSTED YIELDS. 



Average Deviation in Percentage of Yield. 



Oats Strains Test 1919. 



Planting Accession Average deviation 



number number Actual yields Adjusted yields 



3 interior rows 5 rows 3 interior rows 5 rows 



* Not taxonomically Red Rustproof. 



the check plots were more frequent, being in every sixth plot. The 

 results of adjusting yields in this test, both for protected 3-row plots 

 and for unprotected 5-row plots in four series are shown in Table 33. 

 Contrary to expectation, the variability was not reduced by adjustment 



