298 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 1920. 



matter of considerable interest and importance from the mining en- 

 gineer's point of view, several attempts have been made to formulate 

 the relations, and it will be pertinent to give a very brief account of 

 the subject. 



Among the earliest of the more modern workers to investigate this 

 problem are Vogt, Kemp, and De Launay, 24 who confined their atten- 

 tion chiefly to whether the various elements considered were most 

 abundant in the more or the less siliceous rocks. 



The writer pointed out 25 that " the relations are more complex 

 and are dependent, not so much on the relative amount of silica, as 

 on the relative amounts of other constituents, notably soda, potash, 

 iron, magnesia, or lime." Such relations of common association are 

 shown, in part among the most abundant constituents of rocks and 

 minerals, and in part among the rarer ones, generally in connection 

 with the more abundant. For the most part, the relations so far ob- 

 served, which may be considered as best established, are confined to 

 the petrogenic elements, as would be expected, but there seem to be 

 similar relations, not yet quite clear, between some of the metallo- 

 genic and the petrogenic elements. 



Broadly speaking, silica, alumina, soda, and potash tend to go 

 together ; thus the rocks that are highest in silica have, in nearly all 

 cases, alumina and the alkali metals as the next most abundant con- 

 stituents. At the same time, the alkali metals, and lime (not iron 

 or magnesia), tend to go with alumina; so that a very large number, 

 and among these the most common, of the silicate minerals are sili- 

 cates of alumina and (or aluminosilicates of) soda, potash, and lime. 

 The iron oxides and magnesia do not show nearly so strong a tend- 

 ency to combine with silica or with alumina. In this connection 

 may be mentioned a tendency toward combination with (or affinity 

 for) silica, which may be expressed thus : 



K 2 0>Na 2 0>CaO>MgO>FeO. 



That is, potash will endeavor to take all the silica that it can, so far 

 as is compatible with certain physical conditions, soda next, and so 

 on; iron being the only very abundant element (except silicon) that 

 commonly forms an oxide alone, that is to say, uncombined with 

 silica. This general law or rule, which is based on the most gener- 

 ally observed relations among rock- forming minerals, is the basis 

 of a recently introduced classification of igneous rocks, and it gives 



24 Vogt, J. H. L., Zeits. Trakt. Geol., 1S98, p. 326; Kemp, J. F., Ore Deposits, 3d edition, 

 pp. 34-37, 1900 ; De Launay, L., La Science Geologique, p. 637, 1906. Cf. also Hillebrand, 

 W. P., U. S. Geol. Survey, Bull. 700, p. 25, 1919 ; and Clarke, F. W., U. S. Geol. Survey, 

 Bull. 695, p. 13, 1920. 



25 Washington, H. S., Trans. Amer. Inst. Min. Eng., p. 751, 1908; Cf. Washington, 

 Manual of the Chemical Analysis of Rocks, 1st edition, p. 14, 1904 ; 3d edition, p. 17, 1919. 



