3o8 Journal of Agricultural Research voi. xii, no. 6 



CONCLUSIONS 



The evidence presented in the early part of this paper indicates that 

 there is always present in soils, in a condition permitting ready solution 

 in water, enough of the more important chemical elements to supply 

 the immediate needs of crops. It is hardly conceivable that substances 

 in this condition do not represent a potentially available supply. Inas- 

 much as this supply never entirely disappears, even in the case of 

 nitrates, it would seem that there is no such thing as a lack of available 

 nutrients in soils which are at all productive, but that a plant may still 

 be unable to satisfy its requirements if the concentration with reference 

 to the individual ions falls below certain minima. It is furthermore 

 highly probable that the optimum concentration varies with every soil 

 in accordance with the physicochemical system present in the soil solution. 

 Slight differences in the character of this system may modify in a marked 

 degree the power of a plant to absorb solutes, so that even if we were 

 able to obtain and analyze the true soil solution, we would not necessarily 

 be able to say that any figure for individual ions constituted inadequacy. 



Attention has been called to certain characters as reflecting the com- 

 position of the soil and its power to produce crops. Inasmuch as these 

 involve three variables to which it is impossible to assign definite rela- 

 tive weights,anexactcorrelation of productivity with the figures presented 

 is not to be expected. Nevertheless, the correlation between the general 

 magnitudes of the figures presented and the crop-producing powers of 

 the soils studied is sufficiently close to justify the belief that they give 

 expression to the relative power of soils to produce crops, although they 

 are not an exact measure of that power. 



We believe that the evidence obtained is sufficient to justify the hope 

 that we may be able to predict, within reasonable limits, thte relative 

 crop-producing powers of soils by comparing their figures expressing 

 these characters with similar data derived from soils whose productive 

 power is known. Before such a method is generally applicable, how- 

 ever, it will be necessary to study the behavior of many soils with numer- 

 ous type crops. This is quite feasible if the various characters can be 

 developed without the enormous number of analytical determinations 

 involved in the present experiments. We believe that this can be 

 accomplished by substituting for our figures, representing the sums of 

 the basic ions, figures obtained for total soluble salts, or preferably 

 direct determinations of the concentration of the soil solution by some 

 such method as that presented in the preceding paper.^ It is quite 

 certain that we shall never have a precise measure of soil fertility until 

 soils are studied concurrently in the cropped and the uncropped con- 

 dition and under strict control. The reasons are obvious in that data 

 from the soil under crop can not indicate its latent power, and data 



'HOAGLAND, D. R. Op. cit. 



