100 T. L. WATSON — WEATHERING OF GRANITIC ROCKS OF (GEORGIA 



tion of the more soluble constituents, as liberated in the free-nascent- 

 state. It has been shown,^' however, that rocks composed of silicate 

 minerals, almost or entirely lacking in alkalies, lost a portion of their 

 silica with equal readiness. This has been explained on the supposition 

 that as the silica was liberated in the nascent state it was soluble either 

 in pure or carbonated water. 



The iron is mostl}?- present in these rocks in the form of ferrous iron. 

 The stability of this constituent characterizing all the analyses of the 

 fresh and decayed rocks ma}^ be readil}^ accounted for on the supposi- 

 tion that the decomposition was promoted in a sufficient supply of 

 oxygen, whereby all of the iron was converted into the form of insoluble 

 hydrated sesquioxide and retained wdth the residue. The analyses in- 

 dicate onl}^ a slight loss in most cases in the alumina, while in others 

 it has all been retained in tlie transition from the fresh to the decaj^ed 

 rock. 



The lime, magnesia, and alkalies obey the usual laws in the propor- 

 tional amounts lost and saved for such rocks. The lime has invariably 

 disappeared in larger percentage amounts than the magnesia, and like- 

 wise the sodium salts have been removed, as a rule, in larger quantities 

 than those of potassium. The felds})ars are, of course, the principal 

 source of the soda and potash in the rocks, and the removal of soda in 

 larger amounts than the j)otash indicates that the potash felds])ar varieties 

 are more refractory toward normal atmospheric agencies than the soda- 

 lime or plagioclase varieties.!* 



The accessories — tourmaline, magnetite, zircon, and garnet — all occur 

 to some extent in the residual granite sands scarcel}'' affected at all by 

 the atmospheric agents. 



In the most advanced stages of decay the residual product is a higlil}'' 

 plastic ferruginous clay, rendered gritty from the presence of free quartz 

 granules mainly, along with a small percentage amount of the unde- 

 composed silicate minerals. In such cases the transition from fresh to 

 decomposed rock has been accompanied by an unusuall}^ large propor- 

 tional loss in the constituents, greatl}^ exceeding in several instances 50 

 })er cent.+ In these cases the change is not far fmm complete, as shown, 

 not only in the chemical analyses, but in the relatively small amounts 

 of undecomposed minerals present in the clay as well. Where the change 



* Ebelmen : Ann. des Mines, 1845, vol. vii. 



Merrill, Geo. P. : Op. cit., pp. 226, 227. 



t Roth's Allegemeine u. Cheniische Geologie, 3d edition, 2d lieft. Geldmacher M., Beitrage 

 zur Verwittening der Porphyre, Leipzig, 1889. Lemberg, Zeit. der Deut. Geol. Gesellschaft, 1876, 

 p. 28. 



J Merrill, Geo. P.: Rocks, Rock Weathering, and Soils, New York, 1897, p. 234. 



