ILLINOIS FIELD EXPERIMENTS 469 



Plots i (especially) and 2 in this series were naturally more 

 productive than the other plots, it being the regular custom of the 

 Illinois Station to use the most productive land for the untreated 

 check plots if any such differences are apparent when the field is 

 established, as was the case in this instance. Plot i serves only as a 

 check against the lime treatment, and the average of plots 2, 4, 5, 

 and 8 gives a more reliable basis of comparison for ascertaining 

 the effect of nitrogen. 



Potassium is evidently the second limiting element in this soil 

 where decaying organic matter is not provided, but the limit of 

 potassium is very far above the nitrogen limit. 



During the six years plot 7, receiving nitrogen and potassium, 

 produced 291.3 bushels of corn (averaging 72.5 bushels a year), 

 54.7 bushels of oats, and 36.5 bushels of wheat, per acre. To pro- 

 duce the increase of plot 7 over plot 5 would require about 75 per 

 cent of the total nitrogen applied. Thus, there has been a loss of 

 25 per cent of the nitrogen applied, which is a smaller loss than 

 usually occurs where commercial nitrogen is used. Without doubt, 

 larger yields would have been produced, especially of corn, if 150 

 or 200 pounds of nitrogen per acre per annum had been used, which 

 would have increased the cost of nitrogen to $22.50 or $30, re- 

 spectively, per acre each year. 



It need scarcely be mentioned that commercial nitrogen is used 

 in these and other experiments in Illinois only to help discover what 

 elements are limiting the crop yields. It should never be purchased 

 for use in general farming, but, if needed, secured from the atmos- 

 phere by legume crops to be returned to the soil directly or in ma- 

 nure. 



On three other series of plots on the Green Valley soil experi- 

 ment field, a three-year rotation of corn, oats, and cowpeas is prac- 

 ticed, every crop being represented every year. On plots receiv- 

 ing lime and phosphorus and legume crops, as green manure, the 

 yield of corn was 45.6 bushels in 1906 and 67.8 bushels in 1907, 

 compared with 70.8 bushels and 64.7 bushels with lime, phosphorus, 

 and nitrogen on plot 6 (see Table 90) and with 10.4 bushels and 

 13.1 bushels with no nitrogen on plot 4, for the respective years. 

 On other plots receiving comparable treatment, where lime, phos- 

 phorus, and potassium were used with nitrogen-gathering legume 



