308 ANNUAL OF SCIENTIFIC DISCOVERY. 



carbonate of soda, furnished by springs like those of Carlsbad and Vichy, 

 or contained in the waters of the Loire, the Ottawa, and probably many 

 other rivers that flow from regions of crystalline rocks, to be reminded that 

 the same process of decomposition of alkaliferous silicates is still going on. 



V. A striking and important fact in the history of the sea, and of all 

 alkaline and saline waters, is the small proportion of potash-salts, which 

 they contain. Soda is preeminently the soluble alkali; while the potash in 

 the earth's crust is locked up in the form of insoluble orthoclase, the soda 

 feldspars readily undergo decomposition. Hence we find in the analyses of 

 clays and ai-gillites, that of the alkalies which these rocks still retain, the 

 potash almost always predominates greatly over the soda. At the same 

 time these sediments contain silica in excess, and but small portions of lime 

 and magnesia. These conditions are readily explained when we consider 

 the nature of the soluble matters found in the mineral waters which issue 

 from these argillaceous rocks. I have elsewhere shown that, setting aside 

 the waters charged with soluble lime and magnesia salts, issuing from lime- 

 stones, and from gypsiferous and salifcrous formations, the springs from 

 argillaceous strata are marked by the predominance of bicarbonate of soda, 

 often Avith portions of silicate and borate, besides bicarbonates of lime and 

 magnesia, and occasionally of iron. The atmospheric waters, filtering 

 through such strata, remove soda, lime, and magnesia, leaving behind the 

 silica, alumina, and potash the elements of granitic and trachytic rocks. 

 The more sandy clays and argillites being most permeable, the action of the 

 infiltrating waters will be more or less complete; while finer and more com- 

 pact clays and marls, resisting the penetration of this liquid, will retain their 

 soda, lime, and magnesia, and, by subsequent alteration, will give rise to 

 basic feldspars, containing lime and soda, and, if lime and magnesia pre- 

 dominate, to hornblende or pyroxene. 



The presence or absence of iron in sediments demands especial considera- 

 tion, since its elimination requires the interposition of organic matters, 

 which, by reducing the peroxide to the condition of protoxide, render it sol- 

 uble in water, either as a bicarbonate or combined with some organic acid. 

 This action of waters, holding organic matter upon sediments containing 

 iron oxide, has been described by Bischof and many other writers, particu- 

 larly by Dr. J. AV. Dawson, in a paper on the coloring matters of some sedi- 

 mentary rocks, and is applicable to all cases where iron has been removed 

 from certain strata and accumulated in others. This is seen in the fire-clays 

 and ironrStones of the coal-measures, and in the white clays associated with 

 great beds of greensand (essentially a silicate of iron), in the cretaceous 

 series of New Jersey. Similar alternations of white feldspathic beds, with 

 others of iron ore, occur in the altered Silurian rocks of Canada, and on a 

 still more remarkable scale in those of the Laurentian series. We may 

 probably look upon the formation of beds of iron ore as in all cases due to 

 the intervention of organic matters, so that its presence, not less than that 

 of graphite, affords evidence of the existence of organic life at the time of 

 the deposition of these old crystalline rocks. 



The agency of sulphuric and muriatic acids, from volcanic, and other 

 sources, is not, however, to be excluded in the solution of oxide of iron and 

 other metallic oxides. The oxidation of pyrites, moreover, gives rise to 

 solutions of iron and alumina salts, the subsequent decomposition of which, 

 by alkaline or earth}' carbonates, will yield oxide of iron and alumina ; the 

 absence of the latter element serves to characterize the iron ores of organic 



