CLA 
STE 
KK I \M> I 
KIGEK. J 
ON 
CAL1FORNITE. 
73 
These analyses may be reduced to uniform type as follows: Throw 
out hygroscopic water, P 2 5 , and C0 2 , with the equivalent amount of 
lime representing calcite. Recalculate Ti0 2 to equivalent Si(),; Fe 2 3 
to Al 2 (),; FeO, MnO, and MgO to CaO, and F to H 2 (). Then recom- 
pute to 100 per cent. The reduced analyses are as follows: 
Analyses. 
Molecular ratios. 
A. 
B. 
c. 
A. 
B. 
C. 
Si0 2 
Al„(), 
35. 58 
19. 24 
41.03 
4.15 
37. 28 
19.74 
39. 42 
3. 56 
39. 09 
22. 82 
37. 22 
.87 
0. 589 
.188 
.731 
.231 
0.617 
.193 
. 703 
.197 
0. (547 
. 223 
CaO 
.064 
H\,0 
.048 
100. 00 
100. 00 
100. 00 
Analysis C gives the garnet ratios quite sharply, but as the mineral 
was white and massive, somewhat resembling chalcedony, its nature 
was at first unsuspected. Its density is that of garnet, however, its 
usibility is the same, and in thin section under the microscope it is 
optically isotropic. It is undoubted^ garnet containing as an 
mpurity about 1 per cent of calcium carbonate. 
Analyses A and B are less easily interpreted. B gives approx- 
mately the empirical formula: 2H 2 0, 2A1 2 3 , TCaO, 6Si0 2 ; or, con- 
titutionally, 
Al 2 Ca 7 (Si0 4 ) 6 (AlOH) (A10 2 H 2 ) H. 
If part of the ivater is extraneous, the following- much simpler 
•rmula may hold: 
Al 2 Ca 7 (Si0 4 ) fi (A10H) 2 ; 
id this fairly represents a good many other analyses of vesuvianite, 
jut not all. Vesuvianite varies in composition, and its variations are 
Lther troublesome to interpret, for they may signify either actual 
jifferences in the mineral itself, alterations, or impurities. The 
•nnula proposed by one of us" some years ago. 
Al 2 Ca r , (Si0 4 ) 5 (AlOH), 
a good expression for the average composition of the species, but it 
:>es not fit the extremes. 
fl Clarke, Bull. U. S. Geol. Survey No. 125, \>. 25. 
