164 



SCIENCE 



N. S. Vol. XXX. No. 762 



figures which can not be very far out of the 

 way. We have here a first approximation 

 to an analysis of the entire crust of the 

 earth, which can be applied to the discus- 

 sion of various large geological problems. 



Upon the basis of the survey analyses, 

 Cross, Iddings, Pirsson and Washington de- 

 veloped their "Quantitative Classification 

 of the Igneous Kocks, "' a classification 

 which became possible only after a large 

 number of complete analyses had been 

 made, and which is applicable only to 

 rocks which have been thoroughly analyzed. 

 In this system, small errors become cumu- 

 lative ; and one effect of its publication has 

 been to encourage better analytical work 

 in several laboratories, and so to increase 

 the accuracy of our knowledge relative to 

 the composition and differentiation of rock 

 magmas. It is interesting to note at this 

 point that the average igneous rock is quite 

 near an andesite in composition, and that 

 it has very close to metasilicate ratios. 

 From the last consideration it may be in- 

 ferred that in the crust of the earth ortho- 

 silicates and trisilicates exist in approxi- 

 mately equivalent molecular proportions. 



The chemists of the survey have also 

 analyzed more than four hundred sedimen- 

 tary rocks; slates, shales, sandstones, lime- 

 stones, etc. They have also made com- 

 posite analyses of many commingled sam- 

 ples in each sedimentary group, in order to 

 determine the average composition of each 

 class. The same method was also applied 

 to the analysis of the well-known "red 

 clay," which forms the floor of the ocean 

 at its greatest depths. Fifty-one samples 

 of the clay, kindly furnished by Sir John 

 Murray, and gathered from all of the great 

 oceans, were ground into one uniform 

 sample and then analyzed. Even here the 

 supposedly rarer elements were found in 

 proportions similar to those of the igneous 



"University of Cliioago Press, 1903. 



rocks, and such substances as titanium, 

 barium, strontium, nickel, chromium, van- 

 adium, copper, lead, zinc and arsenic were 

 easily determined. 



Next in importance to the rock analyses 

 are the analyses of minerals. Of these over 

 600 have been analyzed, covering more than 

 180 species, including eighteen which were 

 described as new. The new species were 

 josephinite, cuprobismutite, warrenite,'' 

 guitermannite, elpasolite, coronadite, zuny- 

 ite, ptilolite, hydronephelite, lucasite, mor- 

 encite, purpurite, antlerite, knoxvillite, 

 redingtonite, plumbojarosite, emmonsite 

 and powellite. Furthermore, the exact 

 composition of many imperfectly described 

 minerals has been well established, as in 

 the cases of tengerite, kleinite, carnotite, 

 roscoelite, patronite, xanthitane, hetaerol- 

 ite, mackintoshite and yttrialite. In Hille- 

 brand's analyses of uraninite its gaseous 

 constituent was first noted; and this gave 

 Ramsay the clue which led to the discovery 

 of helium. Molybdic ochre, previously as- 

 sumed to be molybdic oxide, M0O3, has been 

 shown by Schaller, by the analysis of sam- 

 ples from five localities, to be a hydrous 

 ferric molybdate. The true composition 

 of dumortierite was also established by 

 Schaller. The twenty complete analyses 

 of tourmaline by R. B. Riggs and also his 

 analyses of the lithia micas are important 

 contributions to chemical mineralogy. The 

 analyses of 29 meteorites can be added to 

 this schedule. 



Over 150 analyses of waters have been 

 reported from the central laboratory of the 

 survey. Some of these waters were from 

 springs of commonplace type, but others 

 have been of unusual character. One 

 spring, from southwestern Missoiiri, ac- 

 cording to Hillebrand's analysis, contains 

 zinc sulphate as its principal saline con- 



' According to L. J. Spencer warrenite is iden- 

 tical with janiesonite. 



