220 EOCK DISINTEGRATION AND DECOMPOSITION 



South America is much less favorable to the preservation of 

 shell structures than would be a moist one where the salt would 

 be removed too rapidly for the double decomposition to be 

 brought about. 



Resume. — Making all due allowance for possible sources of 

 error in our methods, there are certain general deductions that 

 may be safely drawn. Not, it may be, from our own analyses 

 alone, but from numerous others as found in existing literature.^ 



Let us briefly review the subject and make the deductions 

 accordingly. 



In glancing over the analyses, it is at once apparent that 

 hydration is an important factor, the amount of water increas- 

 ing rapidly as decomposition advances. In the earlier stages of 

 degeneration it is doubtless the most important factor.^ There 

 is, moreover, among the siliceous crystalline rocks, in every case 

 a loss in silica, a greater proportional loss in lime, magnesia, and 

 the alkalies, and a proportional increase in the amounts of 

 alumina and sometimes of iron oxides, though the apparent 

 gain may in some cases be due to the change in condition from 

 ferrous to ferric oxide. As a whole, however, there is a very 

 decided loss of materials. Among siliceous crystalline rocks, 

 this loss, so far as shown by available analyses and calculations, 

 rarely amounts to more than 60% of the entire rock mass. 

 Among calcareous rocks, on the other hand, it may, in extreme 

 eases, amount to even 99%. 



Of all the ordinary essential mineral constituents the free 

 quartz is the most refractory toward purely chemical agen- 

 cies, and the amount of silica lost from this source must be 

 small, though Sorby^ thinks to have distinguished chemically 

 corroded quartz granules in some of the sands examined by him. 

 It is, however, safe to say that the mineral suffers chiefly from 

 mechanical disruption, — that silica in any rock which is re- 

 moved during the process of decomposition is derived mainly 

 from the silicates, and not from the free quartz. According to 

 Bischof, and as shown by our own work, the silicates that are 



^ See especially Eotli 's Allegemeine u. Chemisehe Geologie, Vol. Ill, and 

 Ebelmen's papers in Ann. des Mines, Vols. VII, 1845, and XII, 1847. 



^ Hydration stands as the most extensive reaction in the belt of weathering. 

 In its importance in this belt, as a geological process, it is second only to 

 carbonation. (Van Hise, Treatise on Metamorphism, p. 481.) 



®Proc. Geol. Soc. of London, 1879. 



