452 INVESTIGATION BY CULTURE EXPERIMENTS 



30 pounds of phosphorus during the five years, or 50 per cent more 

 than was applied. 



TABLE 84. EXPERIMENTS AT STRONGSVILLE, OHIO 

 Data per Acre for Five-year Rotation: Increase Only, except as Noted 



'On this plot 114 Ib. of nitrogen; otherwise the applications were: nitrogen, 

 76 Ib. ; phosphorus, 20 Ib. ; and potassium, 108 Ib. ; per acre, in five years. 



In general, it may be stated that the plant food applied in the 

 Ohio experiments produced small effects the first few years. Thus, 

 in the 5-year rotation at Wooster the largest increase in corn in 

 1904 was 3.5 bushels per acre, on plot 21, and in 8 cases out of 20 

 the treated plots produced lower yields than the untreated control 

 plots. As an average of the first five years, phosphorus (on the 

 No. 2 plots) produced increases per acre as follows: corn, 4 bushels; 

 oats, 3.5 bushels; wheat, 3.1 bushels; clover, 390 pounds ; timothy, 

 186 pounds. These are about one half the effect for the i5~year 

 average, as will be seen from Table 82. 



From a consideration of the Rothamsted and Pennsylvania 

 data, it seems probable that more or less of this apparent average 

 increase is due to a comparative decrease on the unfertilized plots, 

 although the time is too short since the Ohio rotations and soil 

 treatment have been fully under way to make trustworthy averages, 

 because of seasonal variations; for, while land may be decreasing 

 in normal productive power, a few favorable seasons following 

 unfavorable years may furnish data that indicate increasing yields, 

 as in the Minnesota experiments'ref erred to on a former page. 



