22 CONTRIBUTIONS TO MINERALOGY. [bull. 262. 
CHEMICAL EXAMINATION OF CARNOTITE ORES. 
The chemical problems involved in the analysis of the carnotite ores 
were peculiarly intricate. Ideal material was quite unobtainable and 
mechanical separation was impossible. The first specimen received 
happened to be of higher grade than any of the subsequent ones, con- 
taining about 5 per cent of sand grains and showing only very faint 
delicate reddish tracings, indicative of some foreign iron mineral. In 
mass it was of a beautiful canary-yellow color and easily broken 
down by pressure. Other specimens were more coherent, the degree 
depending altogether on the extent to which the sandstone had been 
impregnated and altered, and they sometimes showed more of the 
peculiar reddish admixture. 
A fact only suspected in making the first analysis was confirmed by 
subsequent work on lower-grade ores, namely, that the vanadium existed 
in two conditions, in entirely distinct minerals, the greater part b}^ far 
as pentavalent vanadium in the easily soluble carnotite, and a smaller, 
much less soluble portion, almost vanishing in the purest ores, in the 
trivalent state as a constituent of a silicate free from uranium. This 
observation explained the statement of Messrs. Poulot and Voilleque' 
that they had found the low-grade ores relatively richer in vanadium,, 
as compared with uranium, than the high-grade ores. 
While the carnotite dissolves at once in cold dilute nitric acid,, 
unfortunately the vanadiferous silicate is not quite insoluble, hencet) 
arose an important difficulty in the way of arriving at the true com-] 
position of the carnotite. It is true that in one case the silicate has] 
been analyzed (p. 30), but it would be unsafe to correct the carnotite* 
analysis on that basis. Moreover, the analyses were not all made oe| 
the same plan; some are less complete than others; they can not icfi 
some respects be rigidly compared with one another. 
As to the carrying out of the analysis, various procedures werei 
tried and no one was found which gave altogether satisfactory resultfl j| 
as to each constituent, though many could be determined with th<i< 
usual degree of accuracy. The difficulty was due mainly to vanadiuni 
and the small amount of phosphoric acid usually present. These twc\ 
constituents were likely to be found in different precipitates and could 
not be separated from them at one stroke. Their complete remova j 
from other bodies was at times impossible, and the weight of the latte 
had then to be corrected for these residual amounts. 
Two entirely different lines of attack were open. One, that o 
Friedel and Cumenge, described in their paper on carnotite, ha 
alread}^ been used with apparent success. It depends on renderin 
the vanadium insoluble in water by evaporating the nitric acid soli i 
tion to dryness. Water extracts the alkalies and uranium withoi I 
