262 THE FARMER OF TO-MORROW 



in a single field. Many of the variants are 

 too subtle to be calculated, yet they make 

 for the widest divergence in results. 



As an illustration of this point, let us take 

 the data of experiments in wheat fertilizers, 

 3,227 in number, conducted by state experi- 

 ment stations, and tabulated by the Bureau of 

 Soils at Washington. It is to be presumed 

 in this connection that the soils used in the 

 experiments, even though they are scattered 

 over the length and breadth of the country, 

 are at least ordinarily suitable for wheat 

 culture, otherwise the work would be useless. 



The usual method of experimenting is to 

 divide a field into parallel plots as nearly uni- 

 form as possible, both as to fertility and as to 

 size. "Test" or "check" plats are always main- 

 tained in connection with an experiment to 

 compare the results with and without fer- 

 tilizers. 



Yet in these official experiments the pro- 

 ductivity of the check plats is found to vary 

 in yield from less than one bushel to more 

 than fifty bushels to the acre. That is, with- 

 out the use of fertilizers of any kind. 



With oats the yield of the check plats varied 

 from less than ten to more than eighty bushels 



