28 
CONTRIBUTIONS TO MINERALOGY. 
[bull. •202. 
If it be assumed that the bivalent elements offer the most accurate 
determinations, the above ratios may be reduced to the following 
simpler terms on that basis: 
l-a . 
l-b . 
l-c . 
ll-a 
ll-b 
III . 
R'. 
R". 
2.41 
1 
2.80 
1 
2. 54 
1 
2. 21 
1 
1.88 
1 
.99 
1 
3.56 
4.12 
3.17 
3.60 
3.27 
2.39 
u. 
0. 
3.19 
20.68 
3. 45 
23. 03 
2.93 
18.98 
3. 41 
21. 32 
3.12 
19.50 
2.16 
13.98 
HoO. 
1.95 
:;. 23 
2.75 
2.23 
3. 33 
3.09 
The results, however, show a great lack of agreement and wide 
variation. It is plain that no probable formula can be calculated fori 
the yellow body. The variations are of such a nature as to indicate 
in the plainest manner that it is a mixture of several substances. 
So detailed a discussion as the foregoing would hardly have been i 
justified, in view of the negative conclusions arrived at, but for thej; 
fact that Messrs. Friedel and Cumenge in their paper announced at 
simple formula for the body examined by them and gave it the specific; 1 
name carnotite. Their published analyses are as given below, from 
which they have excluded considerable sand and traces of barium, 
aluminum, lead, copper, and radio-active bodies as present in excess-] 
ively small quantities. They make no mention of calcium and admit 
that their values for water are open to doubt. The formula deduced 
by them is 2U 8 O s ,° V 2 5 , K ,0, 3H 2 0. 
Found. 
Calculated,! 
1. 
2. 
3. 
4. 
Uo<v 
64.70 
20.31 
10.97 
62.46 
19. 95 
11.09 
63. 54 
20. 12 
10. 37 
5. 95 
v 2 o 5 
K 2 
H 2 
5.29 
4.81 
Fe 9 0, 
.96 
.65 
99. 98 
T4- nw ~^„ n 
a Old n< 
)tation, equiv 
alent to the i 
Qodern U0 3 . 
i ~„a n„* 
great good fortune obtained a variety of samples of the pure potassium ! 
compound, free from calcium and without appreciable admixture of 
