THE ANALYSIS OF VIRGIN SOILS. 347 



a valuable indication as to the one of these that will first be 

 required to be added when production slackens. 



What arc " Adequate " Percentages of Potash, Lime, Phos- 

 phoric Acid and Nitrogen ? It is evident that a very critical 

 discussion of cultural experience can alone answer this ques- 

 tion ; and at first sight such experience often appears very con- 

 tradictory when compared with the results of analysis. 



One of the chief causes of such apparent discrepancies is readily in- 

 telligible when we consider the differences in root-development of the 

 same plant in different soils. In "light" or sandy lands the roots may 

 penetrate to several times the depth attained by them in heavy clay 

 soils. Having thus within their reach a soil-mass several times larger, 

 and aerated to a much greater depth, it is but reasonable to expect that 

 in deep, sandy lands plants would do equally well with correspondingly 

 smaller percentages of plant-food than would suffice in clay soils, in 

 which the root-range is very much more restricted. The well-known 

 fact that the production of heavy clay lands may be increased by their 

 intermixture with mere sand, adding nothing to their store of plant- 

 food, emphasizes this expectation and elevates it into a maxim. On 

 this ground alone, therefore, it is evident that the mere consideration 

 of plant-food percentages found, can be a true measure of productive- 

 ness only in the case of virgin soils with high percentages. 



Soil Dilution Experiments. The extent to which dilution 

 with mere " lightening " materials can be carried without im- 

 pairing production, can of course be determined for concrete 

 cases only; but the following experiment made at the Cali- 

 fornia Station is a case in point: 



One kilogram of the heavy but highly productive black clay 

 soil of the experimental grounds of the University of Cali- 

 fornia was used in each of five experimental cultures, each 

 made in duplicate, in cylindrical vessels of zinc-covered ( " gal- 

 vanized ") sheet iron, all proportioned alike in height and 

 diameter, but containing respectively one. two. four, five and 

 six volumes of total soil. Tn the snrillest was placed one kilo- 

 gram of the undiluted, original soil, in the others successively 

 the same amount of the soil thoroughly mixed with one, three, 

 four, and five volumes of a dune sand fully extracted with 

 chlorhydric acid, and washed with distilled water. The water- 



