[405] SCIENTIFIC MANUAL. 119 



but detailed directions, embracing a clear statement of the 

 questions to be asked, and the precautions to be used, to 

 make the results reliable, and the correct interpretation of 

 the answer possible. 



Perhaps the most important of the list of experiments 

 enumerated, is that designed to ascertain \vhat elements cf 

 plant-food required by plants, are not supplied by the soil. 



On soils in good "condition," the answer to this ques- 

 tion would not be very definite, since such have an accu- 

 mulated stock of available plant-food ; but on soils that are 

 exhausted in the agricultural sense, the answer would be 

 readily interpreted, since, in such, there is little more than 

 their "natural strength,' 1 and the effects of different fertil- 

 izing substances, either singly or in combination, will be 

 marked, and the results readily interpreted. 



Soils that have become so far exhausted as to refuse to 

 return remunerative crops for the labor of the husbandman, 

 are defective in one or more of the necessary elements of 

 plant-food in available form, which must be restored before 

 such soils can again become productive. 



They may have stiil remaining a sufficient supply of some 

 of the elements in an available form, or they may be defi- 

 cient in all. The problem, therefore, is to ascertain what 

 substances are deficient in the soil. The object of experi- 

 ment No. 1 is to ask nature just this question. 



The farmer, in order to do this, applies to one part of the 

 soil under investigation potash alone, to another phosphoric 

 acid, to another nitrogen, to another potash and super- 

 phosphate, to another nitrogen and potash, to another ni- 

 trogen and phosphoric acid, to another potash, phosphoric 

 acid and nitrogen together, to another lime, etc., with un- 

 fertilized plats between to detect any want of uniformity in 

 the strength of the soil. 



To make the experiment accurate and reliable, the pre- 

 paration of the soil and the cultivation of the crop should 

 be identical on every plat ; the distribution of each fertili- 



