106 T. L. WATfeON — W'EATII EKING OF GRANITIC KOCKS OF GEORGIA 



wliich he characterizes as " highl}^ plastic " and " dee}), red-brown " in 

 color, yielding on chemical analysis 12.18 per cent of Fc^O.^. 



The writer (Watson) has described elsewhere the changes incident to 

 the weathering of an even grained granite near Greenville, Meriwether 

 county, Georgia, the accompanj^ing residual product from which is a 

 highly ferruginous deep red, dark colored, stiff, plastic chw, slightly 

 gritty, chief!}'' from free quartz, and yielding on chemical analysis only 

 6.33 per cent of Fe.^O:i. 



Still another case in question is the weathering of a similar fine and 

 even grained granite near Newnan, Georgia. The residual i)roduct 

 from this granite is a cla3'e3' mass, less plastic and not so dark in color 

 as the Greenville material, but is a bright cherry-red in color, and 

 yields only 1.91 per cent of Fe.,0^ on anal}' sis. Practically the same 

 results are shown on comparing the recorded results of basic igneous 

 rocks. According to these data, we apparently have residual products 

 of essentially the same character and depth of color, but varying evi- 

 dently in the amount of coloring matter — ferric oxide — present. The 

 two extremes are represented by 12.18 and 1.91 })er cents respectively, 

 the former containing more than six times the amount of ferric oxide 

 shown in the latter, and yet the same results as regards color are seem- 

 ingly produced. 



Results based on assumptively insoluble AL.Oa and Al,,03 -j- Fe^Oj 



It has been shown ''^ that of the essential constituents in silicious crys- 

 talline rocks the iron and aluminum oxides are the most refractory, and 

 accordingly the least likely to suffer loss during the })rocesses of rock 

 degeneration. The work of Dunnington.f Clarke,]; and Smyth § strongly 

 indicates concentration, instead of loss, in titanium, in the residual 

 product derived from the titanium-bearing minerals, upon weathering. 

 This element occurs so sparingly distributed through the acid crystal- 

 line rocks, averaging, as a rule, less than 0.5 of 1 per cent, that it can not 

 be regarded as a safe basis of assumption for calculating the percentage 

 amounts of the more essential constituents saved and lost present in the 

 rock. The same statement, as regards concentration instead of loss, also 

 ap})lies in some cases to several other ones of the accessory minerals 

 sometimes present. 



* Merrill, Geo. P., op. cit. 

 t Amer. Jour. Sci. (:5d ser.), xli, pp. 4!H-49.">. 

 X U. S. Geol. Survej', Bulletin no. 78, pp. .14-42. 

 § Bull. Geo). Soe. Am., 1898, vol. 9, pp. 2r)7-2G9. 



