72 
CONTRIBUTIONS TO CHEMISTRY AND MINERALOGY. 
[bull. 167. 
lite, since they were found by Geutli in nearly the same amounts and 
no recognizable iron or magnesium minerals were noticed in the purified 
powder. 
For comparison, the mean of Roscoe's analyses and that one of 
Genth's considered by himself to be his best are also given in the table 
below. 
Yery marked differences are apparent in the three analyses by dif- 
ferent chemists. If titanium was present in the material analyzed by 
Genth and Roscoe, as is very probable, their high results for alumina 
are in great part at least accounted for. It is inconceivable how Genth 
obtained his value for water by ignition, since the mineral oxidizes 
when heated in air. In fact the oxidation in one of my own analyses, 
after allowing for loss of water as ascertained by direct weight, was 
almost what theory requires for the oxidation of V 2 3 to Y 2 5 and of 
FeO to Fe 2 3 , or 5.14 per cent instead of 5.27 per cent. It may fairlybe 
assumed that his water was weighed directly after expulsion by ignition 
of the powder. Roscoe's figures for water, if not for moisture, must be 
affected by error, probably arising from the unsuspected oxidation of 
vanadium. 
Analyses of roscoelite. 
Amount used 
Si0 2 . 
Ti0 2 . 
V 2 3 . 
A1 2 3 
FeO . 
MgO. 
K 2 . 
Na 2 
Li 2 0. 
H 2 below 105° C. 
H 2 105-280° C . . . 
H 2 above 280°C. 
Fl 
Ilillebrand. 
0.3000gr. 0.2531#. 0.2635r/. aO.1560<7 0.2038#. Mean 
45.30 
.77 
23.90 
1.59 
1.G4 
24.00 
11.74 
10.41 
.13 
ft. tr. 
.77 
11.34 
10.32 
none 
4.29 
.80 
24.06 
1.65 1.57 
Gontb. 
45. 17 
.78 
24.01 
11.54 
1.60 
1.64 
10.37 
.06 
ft. tr. 
.40 
.17 
4.12 
none 
47.69 
20.56 
14.10 
1.67 
2.00 
7.59 
.17 
trace 
4.96 
ign. 
86 98. 76 
Roscoe. 
Si0 2 
Y 2 5 
A1 2 3 
Fe 2 3 
Mn 2 3 
CaO 
MgO 
K 2 
Na 2 
Water 
Moisture . 
41.25 
b 28. 60 
14.14 
1.13 
1.15 
.61 
2.01 
8.50 
.82 
1.08 
2.27 
10]. 62 
a 4.94 per cent oxygen used for complete oxidation instead of 5.27 needed for all V as V 2 3 ami Fe 
as FeO. 
b Equivalent to 23.59 per cent V 2 3 . 
