SOILS OF THE ARID AND HUMID REGIONS, 389 



In the analysis of silicates we employ caustic lime for the setting- 

 free of the alkalies and the formation of easily decomposable silicates, 

 by igniting the mixture; but the carbonate will slowly produce a similar 

 change, both in the laboratory and in the soils in which it is constantly 

 present. This is strikingly seen when we contrast the analyses of 

 calcareous clay soils of the humid region with the corresponding non- 

 calcareous ones of the same. In the former the proportions of dis- 

 solved silica and alumina are almost invariably much greater than in the 

 latter, so far as such comparisons are practicable without assured absolute 

 identity of materials. That is, calcareous clays or clay soils are so sure 

 to yield to the analyst large precipitates of alumina, that experience 

 teaches him to employ smaller amounts for analysis than he would of 

 non-calcareous materials, in order to avoid unmanageably large bulks of 

 aluminic hydrate. It is but rarely that even the heaviest non-calcareous 

 soils yield to the acid usually used in soil analysis more than 10 per 

 cent, of alumina ; while heavy calcareous clay (prairie) soils commonly 

 yield between 13 and 20 per cent. 1 It would be interesting to verify 

 this relation by artificial digestions of one and the same clays with 

 calcic carbonate at high temperatures, as it must always be extremely 

 difficult to insure absolute identity of all other conditions in natural 

 materials. 



In most of these cases, what is true of alumina is also true of the 

 soluble silica. But since the latter is constantly liable to be dissolved 

 out by solutions of carbonated alkalies, it is not surprising that this 

 relation is not always shown. 



Hydrate. In numerous cases, the amount of 

 alumina dissolved in analysis is greatly in excess of the soluble 

 silica, so as to force the conclusion that a portion of the latter 

 must be present in a different form from that of clay ( kaoli- 

 nite) ; the only choice being between that of complex hydrous 

 silicates (none of which, however, could contain as large a per- 

 centage of alumina as clay itself) and aluminic hydrate. The 

 latter is alone capable of explaining the presence of more 

 alumina than silica in easily soluble form; 2 and the visible 

 occurrence of " gibhsite " and " bauxite " in modern forma- 



1 Report of the Tenth Census, Vols. 5 & 6 ; see especially the analyses of soils 

 from Mississippi and Alabama. Also the Reports of the California Experiment 

 Station. 



- Excepting the relatively rare minerals of the Allophane, Kollyrite, and 

 Miloshite group. 



