A STUDY OF FACTORS INFLUENCING THE YIELD OF POTATOES 1249 



and 76 exhibits a very obvious similarity. It is evident in both tables 

 that more than ten cultivations were efficient only when the greater 

 amounts of seed and of manure and fertilizer 'were used. This may 

 indicate either that a greater amount of tillage was necessary to control 

 the greater weed growth produced by the increased fertility, or that 

 more tillage was necessary to make available sufficient plant food to 

 support the increased stand of potatoes. It may be concluded that, on 

 the average, it did not pay to cultivate potatoes more than ten times in 

 this region in 1912. The coefficient of correlation between this factor 

 and yield, as shown in figure 145, is - 0.087 0.037. This slightly 



Yield per acre, in bushels 



1 6 27 38 81 43 60 27 23 13 83 330 

 r = - 0.087 0.037 



FlG. 145. CORRELATION OF FREQUENCY OF CULTIVATION AND YIELD ON 330 LONG ISLAND 



FARMS IN 1912 



negative value indicates that the average frequency was a little too high 

 for maximum production. However, the relatively high probable error 

 renders the coefficient insignificant. 



Frequency of cultivation in Steuben County 



Cultivation was not so thoroly practiced in Steuben County in 1912, as 

 on Long Island, the average frequency being 7.6 cultivations. Weed 

 control is much more of a problem here than elsewhere because of the 

 fact that potatoes are usually grown on sod land of several years standing 

 and on land containing a considerable quantity of weed seed or stubble. 

 Furthermore; the seed bed is here more poorly prepared than in most 

 other regions because of the susceptibility of the soil to extreme puddling. 

 Under these conditions, the factor of frequency of cultivation would be 



