i6o THE CHEMICAL ANALYSIS OF SOILS [CHAP. 



special conditions of soil. By pot cultures in the 

 laboratory it can be shown that of the above elements, 

 the carbon, hydrogen, and oxygen are drawn from the 

 atmosphere or the water, and that nitrogen, chlorine, 

 sulphur, phosphorus, among non-metals, and potassium, 

 calcium, magnesium and iron, among metals, are ele- 

 ments indispensable to the plant, and are derived by 

 way of the root from the soil. In view of the above 

 facts it is clearly unnecessary to make an ultimate 

 determination of all the elements present in the 

 soil, which has already been shown to consist largely 

 of sand and various silicates of alumina, etc. These 

 materials constitute the medium in which the plant 

 grows, but do not themselves supply it with any 

 food ; they need not, therefore, be estimated chemi- 

 cally. 



The chemical analysis of a soil, then, resolves itself 

 into determinations of the nitrogen, phosphorus, potas- 

 sium, calcium, and (of less importance) of sodium, mag- 

 nesium, iron, aluminium, chlorine, and sulphur. To 

 these must be added the determination of the carbon 

 compounds of the soil, which have already been touched 

 on under the head of humus, and of the carbonates of 

 calcium and magnesium, which in most soils constitute 

 the bases available for neutralising any acids that may 

 be produced. Having decided upon the elements to be 

 determined it would then be possible to proceed as in 

 an ordinary mineral analysis: the sample of soil would 

 be reduced to such a state of division as would admit 

 of drawing an accurate small sample, and then entirely 

 disintegrated by some such reaction as fusion with 

 ammonium fluoride. But results obtained in this way 

 would give very imperfect information about the soil, for 

 the procedure draws no distinction between material 

 present in the unvveathered interior of the stones and 



