IO6 THE SOIIv SOLUTION 



evidence favors the view that all the factors are dependent 

 variables, although numerous attempts have been made from time 

 to time to show that some one factor, such as the rainfall for 

 instance, or the mean annual temperature, or available plant- 

 food, is practically an independent factor. Although it should be 

 rather easy to determine experimentally the nature of the func- 

 tion, if any of these various factors were independent, this has 

 never been done, and this fact is itself a strong argument that 

 all the factors in crop production are dependent on one another. 



When there is introduced into the equation a factor for any 

 one of the methods of soil control, i. e., tillage, crop rotation, 

 or fertilizers, it becomes even more apparent that the function 

 is determined by dependent variables, for the new factor always 

 more or less affects several if not all of those already cited. 

 For instance, fertilizers certainly affect the chemical properties 

 of the soil, its physical properties, the soil bacteria, perhaps the 

 plant- food supply, the oxidation of plant effluvia and other 

 factors. It is obvious, therefore, that a satisfactory theory of 

 fertilizer action can not be a simple one but must of necessity be 

 complex ; and the same statement is no less true as regards tillage 

 and crop rotation. 



The recognition of the fact that the action of fertilizers is a 

 complex function depending upon many factors and groups of 

 factors which vary among themselves and with each individual 

 soil, carries with it the conviction that an exact or quantitative 

 fertilizer practice, while theoretically possible, is probably 

 unattainable since methods for the solution of such complex 

 functions are generally wanting. It is not surprising, therefore, 

 that the empirical experience of the past has failed to develop a 

 quantitative practice. However disappointing this may seem at 

 first sight, the prospect is not altogether hopeless, for this point 

 of view indicates a systematic scheme for experimentally deter- 

 mining a qualitative, but nevertheless rational, fertilizer practice. 

 The dominance of the plant-food theory of fertilizers in the 

 past, shutting off, as it has, a rational attack of the problem, is 

 causing the annual waste of millions of dollars in misapplied 

 fertilizers, and it is of scarcely less economic than scientific 



