WASHINGTON AND LARSEN: MAGNETITE BASALT 451 



results. Titanium was determined (in duplicate) by the color- 

 imetric method, as its amount was small, the color effect of the 

 very large amount of iron being removed by the addition of 

 phosphoric acid, a corresponding amount being added to the 

 standard manganese solution.- In a special large portion nickel, 

 chromium and vanadium were tested for. Not a trace of nickel 

 was found with dimethylglyoxime as a reagent. A coloration, 

 possibly due to chromium, was barely perceptible, even when 

 the solution was highly concentrated, and the amount of this 

 cannot be as much as 0.01 per cent, even if present. Vanadium 

 was determined by Hillebrand's method and its presence was 

 verified after the titration. Combined water was determined by 

 Penfield's method. The amount of V2O5 was subtracted from 

 that of P2O5, as it is precipitated and weighed with the phospho- 

 molybdate,^ if the latter is in sufficient excess. This is a point 

 to be borne in mind in the analysis of rocks containing notable 

 amounts of vanadium. 



The analysis shows some remarkable features and, while nearly 

 all its constituents can be matched in the other analyses of iron 

 ore rocks derived from igneous magmas, it does not correspond 

 in all respects with any of them. One of the Adirondack ores 

 (B) resembles it most closely. It is especially remarkable in the 

 very low titanium, the hematite ore of Kiruna alone approaching 

 it in this respect. The phosphorus is notably higher than else- 

 where, while the absence of nickel and chromium is noteworthy. 

 The amount of vanadium is about that of most of the Adirondack 

 and Ontario ores. 



The norm of the Colorado basalt is as follows: 



Or 3.89 Mt 49.88 



Ab 3.67 



An 22.24 



C 0.10 



Hy 7.70 



01 1.12 



/ 99.77 



2 Cf. W. F. Hillebrand, Bull. 422, U. S. G. S., 1910, p. 134. 



3 Cf. J. R. Cain and J. C. Hostetter, Techn. Pap. Bur. Stand., No. 8, 1912. 



