156 
MANGANESE DEPOSITS OF INDIA: MINERALOGY. [ PART I: 
think they can have, appreciably, affected the composition, not even to 
the extent of the small amount of manganese that the analysis shows 
to be present. From the above it will be seen that the specific gravity 
of the pure mineral analysed lies between 2 '959 and 3 "024 ; it can 
g ^.^ therefore be taken as 3'0._ Doubtless there 
^ ^' are some few grains that lie outside the above 
limits, but the majority of such owe their greater specific gravity 
to the inclusions of manganese-ore. Perhaps, allowing for variations in 
composition, we can give the specific gravity of the mineral when 
free from inclusions as ranging from 2'95 to 3'05. The colour of the 
pure specimen of the mineral thus obtained was a bright azure blue 
as viewed in bulk in a glass tube, whilst under the microscope each tiny 
prism showed its distinctive pleochroism. 
The material thus obtained was analysed by Mr. T. R. Blyth, of the 
Geological Survey of India, with the following 
result 
Chemical coinposition. 
Si02 
A'203 
FeO 
MnO 
MgO 
CaO 
NagO 
K2O 
Loss 01) ignition 
Moisture at 100° C. 
55-64 
1- 08 
G-35 
0-77 
22-09 
7-64 
2- 89 
0-98 
2-95 
0-14 
100- 53 
He recorded the absence of titanium, chromium, zinc, barium, phos- 
phorus, and flourine. The amount of material available was not suffi- 
cient for the determination of the state of oxidation of the iron and man- 
ganese. In the first place I assumed the above to be the correct state of 
oxidation and converted all the protoxides into metasilicates of the gene- 
ral formula RSiOs, taking for the purpose the requisite amount of Si02. 
In the same way the AI2O3 was converted into a silicate of the 
formula Al203.3Si02. After doing this 2-65 per cent, of silica was left 
over. If the loss on ignition be assumed to be due to combined water, 
9-89 per cent. Si02 would be required for the formation of H2Si03. I 
then supposed the iron to be present as Fe203 and converted this into 
