SOILS — FERTILIZERS. 61 7 



mechanical separation of the constituents of seven soils ac*cording to their 

 specific weights for the determination of humus, colloids, etc., and several years' 

 cropping experiments with six soils to determine the manner in which indi- 

 vidual crops are nourished from the same soil. 



In the first experiments bromoform having a specific weight of 2.83 was 

 diluted with benzol until at a specific weight of 2.G4 the heaviest parts of th(i 

 soils siink in the mixture. By further decreasing the specific weight of the 

 liquid mixture to 2.55, 2.49, and 2.B6 the soils were separated into five fractions, 

 according to specific weight, the lightest of which was suspended on the liquid 

 having a specific weight of 2.86. 



The ease or difiiculty with which a soil was separated into its constituents 

 was found to indicate the uniformity of its composition. An abundant sepa- 

 ration of the specifically lighter constituents indicated a soil rich in colloids, 

 humus, and available plant food, while the separation of an abundance of 

 specifically heavier constituents indicated a less productive soil, poor in humus 

 and available plant food, although not necessarily with a small absolute plant- 

 food content. 



Clay soil contained none of the constituents of specific weight higher than 2.64, 

 and the sand, loamy sand, loam, and limy soils contained only 1 per cent, while 

 (he shale soils contained up to 7.3. The greatest percentages of organic matter, 

 colloids, and available plant food were usually found in the fifth and lightest 

 fraction, and there was usually a steady increase from fraction 1 to fraction 5 

 in this respect. The most silica, iron oxid, and clay were found in the fraction 

 containing the most plant food. It is concluded that the analysis of a soil with 

 fluids of different specific weights indicates the relative quantities of its con- 

 stituents and the distribution of jjlant food in them. 



In pot experiments in the plant house crops yielded more dry matter per 

 unit area of soil than those grown in the open. The assimilation of plant food, 

 especially phosphoric acid, by crops was much greater in warm, dry summers 

 than in cool, wet summers, but the increase in dry matter was not correspond- 

 ingly great. Definite relations existed between the quantity of plant food and 

 the quantity of dry matter obtained, the most marked of which was with 

 nitrogen. It is concluded that if 100 gm. of plant dry matter contain less than 

 1.61 gm. of nitrogen, 0.59 gm. of phosphoric acid, or 1.66 gm. of potash, the soil 

 needs feitilization with the deficient foods, while if the 100 gm. of dry matter 

 contain more than 2 gm. of nitrogen, 1 gm. of phosphoric acid, or 2.14 gm. of 

 potash, no fertilization with these plant foods is necessary. 



Considerable variation was found in the quantities of plant foods absorbed 

 by individual crops, especially the legumes. Red clover contained much nitro- 

 gen and relatively little phosphoric acid, while serradella contained more phos- 

 phoric acid than any other crop. Red clover and potato plants absorbed the 

 most potash, and red clover absorbed the most lime, followed in order by red 

 beets and potatoes. Lime and potash were found in practically the same 

 amounts in plant dry matter. A one-sided fertilization with potash and phos- 

 phoric acid in dry weather when the soil was not deficient in either did not 

 increase the quantity of dry matter. , 



In the six soils, with one exception, definite relations were established 

 between the quantity of dry matter obtained from crops and the following 

 physical and chemical i>roperties of the soils: The absorptive power for 

 ammonia, according to Knop, and for potash and phosphoric acid, according 

 to Fesca, hygroscopicity, according to Mitscherlich (E. S. R., 24, p. 419), the 

 absorptive power for methyl violet, osmotic water absorption, and the electro- 

 lytic conductivity. 



