326 PRINCIPLES OF AGRICULTURAL CHEMISTRY 



etc. The fertilizer is applied to plots varying from 1/50 to 1/2 

 an acre, but generally of T/IO to 1/20 acre. The effect of the 

 different applications is measured by the weight of the product 

 on the different plots. The only variable should, of course, be 

 the fertilizer. All other conditions should be the same for all 

 the plots. The crop, however, is subject to other variables, such 

 as differences in soil or subsoil, in stand, damage by insect pests, 

 nearby trees or fences, etc. The best results are secured when 

 the experiment is carried out on the same land for a number of 



A B 



Fi g- 75- Tobacco, fertilized (A), and unfertilized (B). Ohio Station. 



years to eliminate seasonal differences. It is also well for plots 

 to be repeated a number of times in order to eliminate error due 

 to inequalities of the soil. 



In order to study the effect of variable soil on the crop, several 

 experiments have been made in which a field of apparently uni- 

 form soil, bearing a crop under similar conditions in all parts, has 

 been subdivided and harvested in separate small areas. Com- 



