620 



SCIENCE 



[N. S. Vol. XXVIII. No. 723 



more often indicated by a comparative defi- 

 ciency instead of abundance in the ash of 

 the plant. Such a theory is, however, not 

 borne out by more general experiments; 

 many plants do not exhibit such idiosyn- 

 crasies as are shown by wheat and swedes 

 but require a general fertilizer the com- 

 position of which is determined more by the 

 soil than by the plant. Indeed, no theory 

 of manuring can be based upon the plant 

 alone, but must also take the soil into ac- 

 count, so that a fertilizer may be regarded 

 as rectifying the deficiencies of the soil as 

 far as regards the requirements of the crop 

 in question. What those special require- 

 ments are can only be decided by experi- 

 ment, just as the soil conditions are ascer- 

 tainable by trial rather than from a priori 

 considerations of analysis. If an analysis 

 be made of any soil in cultivation it will 

 be found to contain sufficient plant food 

 for the nutriment of a hundred or more 

 full crops: the soil of the unmanured plot 

 on the Rothamsted wheat field contained in 

 1893, after 54 years' cropping without fer- 

 tilizer, 2,570 pounds per acre of nitrogen, 

 2,950 pounds of phosphoric acid and 5,700 

 pounds of potash. Of course much of this 

 material is in a highly insoluble condition, 

 but though attempts have been made by 

 the use of weak acid solvents to discrim- 

 inate between the total plant food in the 

 soil and that portion of it which may be 

 regarded as available for the plant, no 

 proper dividing line can be thus drawn. 

 The availability of a given constituent, say 

 of phosphoric acid, will depend, as has 

 already been seen, upon the nature of the 

 crop; a given soil may contain sufficient 

 easily soluble phosphoric acid for the needs 

 of the wheat plant and yet fail to supply 

 swede turnips with what they require. 

 Again, the mechanical texture of the soil 

 may be such as to limit the root range of 

 the plant, so that a richer soil is necessary 

 to produce the same result as is obtained 



in a poorer soil of more open structure; 

 the state of the microflora of the soil may 

 also have much to do with the amount of a 

 given nutrient which can reach the plant. 

 Perhaps the best general point of 

 view of the action of fertilizers is 

 obtained by extending the "law of 

 the minimum" originally enunciated by 

 Liebig, according to which the yield 

 of a given crop will be limited by the 

 amount of the one particular constituent 

 which may happen to be deficient; if the 

 soil, for example, is lacking in nitrogen the 

 yield will be proportional to the supply of 

 nitrogen in the fertilizer, and no excess of 

 other constituents will make up for the 

 shortage of nitrogen. To take an example 

 from the Rothamsted experiments, the 

 following table shows the yield of wheat, 



EXPEEIMENTS ON WHEAT, BEOADBALK FIEID, 

 ROTHAMSTED 



grain and straw, from the unmanured plot 

 and from a series of plots, all of which 

 receive an excess of phosphoric acid, pot- 

 ash, etc., but varying amounts of nitrogen, 

 ranging from 43 pounds to 172 pounds 

 per acre. That the nitrogen was deficient 

 is shown by the almost negligible increase 



