184 
MANGANESE rEt'O.SITS OF INDIA: MINERALOGY. [ PaRT 1 : 
assumption. It is probable that in carrying out the analysis Piddington 
weighed the iron as Fe203, and the manganese as Mn304, and that he 
assumed the iron to be present in the rock as Fe203 and the manganese 
as MnO. In fact he probably omitted to determine the amount of 
available oxygen in the rock. Had he done this he might have found 
that a portion of the iron '.vas present in the protoxide form, giving the 
correct ratio of peroxide to protoxide for a garnet. On the supposi- 
tion that the analysis is substantially correct, only lacking the 
determination of the available oxygen, it can be re-arranged mineralogi- 
cally as follows : — 
Garnet : — 
MnO 
CaO 
FeO 
AI2O3 
Fe203 
SiO-2 
21 00 
100 
7'50 
0-35 
21-78 
2o-28 
7()!)7 
Quartz 
Oxygen 
Arsenic 
Los.s 
70-97 
21 07 
0-84 
0-20 
0-92 
100-00 
The 0"84% of oxygen left over according to the foregoing interpre- 
tation was not, of course, determined to be piesent. and is to be 
regarded as another part of the undetermined constituents. If this in- 
terpretation be the correct one, then it is evident that we have here a 
garnet corresponding to the formula 3MnO.Fe903.3Si02. None of 
the six type garnets possesses this foimula, so that it seems possible that 
a seventh is to be added to the six garnets at present recognized. It 
will be desirable to use the term calderite to describe this garnet rather 
than the whole rock. Mallet has in fact already used this term as the 
name of a mineral rather than of the rock in ^\'hich it occurs. But he has 
unfortunately used it, in his Mineralogy, page 89, for a garnet that con- 
rains only traces of manganese, as well as for the highly manganiferous 
garnet to which it properly belongs. He says : ' In the metamorphic 
rocks of the Hazaribagh district irregular beds of massive garnet, some- 
