no 
MANGANESE DEPOSITS OF INDIA : MINERALOGY. [ ParT I : 
This analysis can therefore be rewritten in the following way : — 
Apatite 
Braunite 
Psilomelane 
Al4(Mn05)3 
Ba2 MnOs 
Ca2 MnOs 
Mg2 MnOs 
K4 MnOs 
Na4 MnOo 
Cu2 MnOs 
Zn2 MnOs 
H4 MnOs 
Mn2 MnOs 
. 2-99 
, 1-06 
, 1-61 
. 2-14 
. 3 -51 
. 0 42 
. 0 03 
. 0-08 
. 7 -71 
.33 81 
0-846 
4217 
53 36 53 00 
Quartz (free silica ) ......... 2 '70 
Sulphur 0 -021 
Ti02 0-04 
Moisture at 100° C 0 '60 
Oxygen unused .......... 0 '43 
100-167 
Now by taking the above psilomelane and reducing the whole to 100 
Calculated analyses we obtain an analysis of this mineral expressed in 
of psilomelane. terms of manganates ; and by a re-grouping of 
the constituents of the manganates in terms of oxides we obtain an 
analysis in the usual form. One defect of analyses of psilomelane 
so extracted from analyses of mixed specimens of ore is that in extract- 
ing the braimite aU the iron oxide is often assimied to replace Mn203 
in the braunite, leaving none for the psilomelane. On examining the 
analyses of psilomelane given on page 1 12, it will be seen that, although 
this mineral often contains but a small quantity of ferric oxide, yet, on 
the other hand, it sometimes contains a considerable quantity of this 
constituent. Unfortunately there is often no means of telling in any 
particular case whether the iron present is in the braunite or in the 
psilomelane. An other draw-back to this method of extracting analyses 
of psilomelane is the fact that, as shown under the heading of braunite, 
the composition of braunite is not a constant one, but usually corres- 
ponds to one of three formulae, namely • — 
(1) 3Mn203.MnSi03; (2) 7Mn203.2MnSi03 ; (3) 4Mn20a.MnSi03 
and there is usually no means of telling which to employ in any partic- 
ular case. As already mentioned, however, I have assumed the hrst of 
these formulae in making these calculations ; in the fiist place because when 
I first made them I had not discovered that the formula of braunite does 
