166 



SCIENCE 



N. S. Vol. XXX. No. 762 



from about 180 lakes and rivers, represent- 

 ing the drainage from almost the entire 

 surface of the country. To illustrate the 

 plan of the work one example may be 

 taken, that of the Mississippi just above 

 New Orleans. A sample of water was taken 

 daily, and at the end of each week the seven 

 samples were combined and analyzed. This 

 procedure was followed during a year, and 

 the average of the 52 composite analyses 

 thus made, gave the composition of the 

 saline matter annually contributed from the 

 entire drainage basin of the Mississippi to 

 the Gulf of Mexico. In other cases ten-day 

 composites were taken instead of weekly 

 ones, and in some instances monthly an- 

 alyses were made. These were combined 

 with careful daily gaugings of the stream 

 flow, so that the contribution of each river 

 system to the ocean could be determined 

 with a considerable degree of accuracy. A 

 preliminary discussion of the data, which is 

 subject to some small corrections yet to be 

 investigated, leads to the conclusion that 

 the surface of the United States loses to the 

 ocean annually 87 tons of dissolved solids 

 per square mile and 166 tons of suspended 

 matter. This is equivalent to a lowering of 

 the surface by aqueous denudation, one 

 inch in 760 years. The usefulness of the 

 investigation, however, is not limited to its 

 geologic bearings; it is of direct value for 

 industrial purposes, giving information as 

 to the availability of the several waters for 

 use in steam boilers, or in a variety of 

 manufacturing processes. In its systematic 

 character and its great extent the work is 

 especially notable. 



In the prosecution of their analytical 

 work the chemists of the survey have not 

 only been obliged to study existing meth- 

 ods, but also to devise new ones. The work 

 of Hillebrand upon rock analysis has al- 

 ready been mentioned, but it does not stand 

 alone. It was in the survey laboratory 



that Gooch developed his methods for de. 

 termining boric acid and titanium, and for 

 separating lithia from the other alkalies. 

 Whitfield studied the indirect determina- 

 tion of chlorine, bromine and iodine ; Chat- 

 ard investigated the separation of titanium, 

 chromium, aluminum, iron, barium and 

 phosphorus; Hillebrand devised methods 

 for the colorimetrie estimation of chro- 

 mium and the volumetric determination of 

 vanadium. A comparison of the wet and 

 crucible methods for the assay of telluride 

 gold ores is due to the joint labors of Hille- 

 brand and Allen ;^^ and a process for the 

 estimation of small quantities of fluorine 

 in rock analyses was worked out by Steiger. 

 All of this work, which I can only mention 

 briefly, grew out of the necessities of the 

 chemists in handling the analytical prob- 

 lems submitted to them by the geologists of 

 the survey, and therefore had good reasons 

 for its prosecution. 



Apart from the routine analytical work 

 of the laboratory, a variety of researches 

 upon mineralogical and geochemical prob- 

 lems have been carried out. Some of the 

 mineral analyses were made with direct 

 reference to purely abstract investigations, 

 such as the series of researches upon the 

 constitution of the natural silicates; a 

 problem which was commonly assumed to 

 be hopelessly complex. The silicates, how- 

 ever, are not remarkably complicated sub- 

 stances, their complications being more 

 apparent than real. The conditions under 

 which they are formed, by deposition from 

 molten magmas, or by alteration of such 

 primary compounds, would seem to pre- 

 clude any great complexity. Only rela- 

 tively stable and therefore presumably 

 simple compounds could exist; a conclusion 

 which is borne out by their comparatively 

 small number. The apparent complica- 

 tions are due to impurities, to alterations 



^Bull. 253, U. S. Geol. Survey. 



