136 



than is lost from above in surface and snbdrainage. It is not yet known to 

 wliat depth we sliould measure our stoclc of i)lant food, but this we know: That 

 the i)r(K'ess of oxidation, ineludins nitrification, and of fermentation and decom- 

 positon, by means of wliicli phuit food is lilierated and made available for 

 plants, is limited almost exclusiv(>ly to a few inches of surface soil. 



While our interest in the physical composition should be measured by feet 

 in depth of soil, our interest in the chemical composition may be measured by 

 inches. (Jiven or 8 feet of soil of perfect physical composition, capable of 

 absorbinj; ;ind retaining moisture to be delivered by capillary action to the 

 throwing plants as needed, with G or 8 inches of surface soil of perfect chemical 

 composition, and large crops are assured; but of little value is a rich subsoil 

 overlain by a worn-out surface. 



The argument is made that immeasurable quantities of plant food are brought 

 from miknown dejiths by the capillary rise of the soil solution. Considering the 

 facts that niinfall far exceeds evaporation in most liumid countries, that sub- 

 drainage waters always carry away some plant food, and that old leached soils 

 contain much lower ]»ercentages of lime and plant food elements than are con- 

 tained in similarly formed soils of more recent origin, we conclude that we must 

 not dc|)end upon the supply of plant food being maintained by the rise of the soil 

 solution. As an average of thirteen years at the Rothamsted Experiment Sta- 

 tion, the annual loss of nitrogen in the drainage water from a 20-inch gauge, 

 containing 20 inches of soil, was 37 pounds per acre, while from a GO-inch gauge, 

 containing GO inches of soil, the loss was oO pounds of nitrogen per acre, or 1 

 pound less from the GO-inch than from the 20-inch gauge, thus showing that 

 there is a loss and that the loss is from the surface soil. 



Another argument which some hold to make ample provision for the indefinite 

 maintenance of the fertility of the soil is the fact that the soil is constantl.v being 

 mixed to miknown depths by ants, earthworms, crawfish, and other agents. 

 The fact that we have soil and subsoil which are markedly different is sufficient 

 evidence that this mixing process is very limited. Indeed, the com])osition of the 

 subsurface soil lying immediately below the plow line is usually appreciably 

 different from that of the iilowed surface, and not infrequently we find at a 

 depth of 8 or 10 inches a stratum differing markedly in character from the sur- 

 face soil. 



With due respect to Darwin and to his more ardent supporters of this theory, 

 we must conclude that these agents could not be depended upon to permanently 

 maintain the fertility of the surface soil, even if their activities extended to 

 inexhaustible supplies of plant food. 



The truly fundamental plant-food questions which probably deserve the most 

 serious consideration are as follows : 



(1) To what depth do plants feed significantly in practical crop rotations? 



(2) What is the total stock of plant food in the soil to this depth? 



(3) How rapidly by practical methods can this plant food be made available? 



(4) When necessary, in order to insure the production of large crop yields, 

 how can we most economically supplement the plant food which can profitably 

 be liberated from the soil? 



Investigations are in progress in Illinois to determine the depths to which 

 different plants feed under varying conditions, and because of the very great 

 importance of this subject I take tlie liberty to lu'ge that other experiment sta- 

 tions assist in solving this problem, especially for the soils and crops of local 

 importance, and to suggest that the assistance of the National Bureau of Soils 

 in this line of investigation would be deeply appreciated and probably of great 

 value. 



As a rule we all prefer to report upon the actual I'esults of investigations 

 rather than upon the iilniis or metliods of investigations in ])rogress, but it seems 

 not in.-ippropriate. considering the needs ;uid possil)ilities of solution, to describe 

 one of the methods by which we are attacking this proitlem. 



We take a given .soil and enrich it to a certain depth, say 7 inches. In another 

 pot or plat we enrich the second 7 inches of soil ; that' is, between 7 and 14 

 inches in depth, and in another pot or i)lat we enrich the entire 14 inches in 

 depth, using twice the quantity of plant food recpiired for a 7-inch stratum. By 

 growing plants in these soils we are thus able to determine the influence ui)ou 

 plant growth of a rich substratum, not only when the surface soil is poor but 

 also when it. too. has liecn enriched. By varying the dejiths and strati to which 

 plant food is applied, the anumnts and kinds of i)lant food added, and the kinds 

 of crops and soils used, the investigation is made to cover a variety of 

 conditions. 



