50 
G : 1 alcohol, absolute alcohol, and ether to remove water and ensure 
rapid drying at a low temperature. Finally the mineral was dried at 
50° C., crushed to pass a 90-mesh sieve, and bottled. The material 
so prepared was found to contain 0 04 to 0.08 per cent, of hygro- 
scopic water, and from nil to 0.20 per cent, of insoluble matter. 
The results obtained on three separate specimens were : — 
Natroj a rosite , K unclip . 
Theory. 
°/ 
/o 
A 
0/ 
/o 
B 
% 
C 
% 
mols. 
h 2 0 
11*15 
10-93 
606 
Na 2 G 
K 2 0 
6-39 
6-32 
•68 
l0 * | 109 
Fe A 
49-42 
49-99 
49-98 
49-86 
312 
FeO 
•16 
2 
S0 3 
3304 
33-08 
3213 
32-30 
403 
PA 
•20 
1 
100 00 
100-45 
All results in the case of “C” are the means of two concordant esti- 
mations, using only vessels of quartz and platinum, and reagents 
of “Guaranteed Reagent’’ quality, whose purity had been confirmed. 
None of the water was removed at temperatures up to 150° C., 
fro.m which it would appear that none of it is water of crystallisa- 
tion. 
The empiric formula for minerals of the Alunite Group is 
6H 2 0.R 2 0.3R 2 0 3 .4S0 3 . 
For Natroj arosite this becomes 
6H 2 0.Na 2 0.3Fe 2 0 3 .4S0 3 . 
This may be written (Hillebrand and Penfield) 
Fa 2 [Fe(0H) 2 L(S0J 4 
The results obtained for the Kundip mineral give slightly high 
results for alkalies and ferric oxide and correspondingly low results 
for water and sulphuric oxide. The departure from the theoretical 
ratios is, however, quite small, viz.: — 
Theory 
mols. 
Found 
mols. 
H a o 
6-00 
5-94 
Na.,0 
1-00 
1-07 
FeA 
3-00 
3-06 
so 3 
400 
3-95 
