386 II. S. WASHINGTON ISOSTASY AND ROCK DENSITY 



of deep oceans and elevated continents is striking.'' He expressed the 

 opinion that the correspondence "is in accord with reasonable expecta- 

 tions." Iddings calculated the densities for a few small areas or comag- 

 matic regions: California (2.85), Colorado (2.80), Yellowstone Park 

 (2.85), the Atlantic coast of the United States (2.89), Great Britain 

 (2.91), Fenno-Scandia (2.85), Germany-Austria (2.81), and Italy 

 (2.86), the average compositions of which had been calculated by 

 Clarke, 11 and made an estimate of the average density of some oceanic 

 islands, using analyses of rocks from Eeunion (3.12) and the Hawaiian 

 (3.12) and Society Islands (3.20). His results are uniformly higher 

 than those arrived at by me, the difference being ascribable to the differ- 

 ence in the method of computation. 



In calculating the average densities or specific volumes which are given 

 in the following pages, I have used the normative molecules of wollas- 

 tonite (CaSi0 3 ), enstatite (MgSi0 3 ), and ferrosilite (FeSi0 3 ), as well 

 as forsterite (Mg 2 Si0 4 ) and fayalite (Fe 2 Si0 4 ), instead of taking them 

 combined into diopside, hypersthene, and olivine, as in the regular state- 

 ment of the norm. This was done because the densities of these mixed 

 mineral molecules are additive and vary with the relative amounts of 

 their components; so that more exact and mutually comparable results are 

 obtained by the method adopted here. The normative feldspars present 

 were, of course, calculated from their respective amounts of orthoclase, 

 albite, and anorthite, as in the norm. Especial care was taken to select 

 the most reliable data for the mineral densities, those of pure artificial 

 crystals being adopted when available. Most of the determinations of 

 these were made in the Geophysical Laboratory and are to be found 

 among the various papers which have been published by the Laboratory. 

 Most of the original figures were for specific gravity; -these have been 

 corrected to density, as noted below. 



In Table I, which follows, are given the densities and specific volumes 

 which were adopted for the present calculations, together with those for 

 other normative minerals, which may be of use in some calculations. 



Table I. — Densities and Specific Volumes of normative Minerals 



<5 r Remarks 



Quartz 2 . G4S . 3776 Natural. Johnston and L. H. 



Adams. Journal of the Amer- 

 ican Chemical Society, volume 

 34. 1912. page 572. 

 Orthoclase 2.555 .3914 Calculated from analyses of nat- 

 ural adular. H. E. Merwin. 

 private communication. 



11 F. W. Clarke: Proc. Amer. Phil. Soc, vol. 51, 1912, p. 217. 



