Chap. XIII.] 
KODURITE SERIES : ALTERATION. 
2ll 
dissolved oxygen. It is interesting to notice that this deduction that 
the ores were formed vmder oxidizing conditions, meaning that they 
were formed comparatively near the surface, agrees with the conclusion, 
discussed later on, that kaolinization takes place comparatively near 
the surface. The calcium bicarbonate and siUca shown in the foregoing 
equation would pass away in solution leaving a mixture of hydrated 
oxide of manganese with ferric oxide. 
The analysis of the Garbham ^ spandite, which is very similar to that 
found in the part of the Kodur mine to which reference is made on 
The removal of alu- P^g^ 269 shows 6-52 per Cent, of AI2O3, whilst 
niina. those of three pieces of manganese-ore, A. 334, 
A.336, and A. 338, from this same part of the Kodur mine show 1*38 
to 6-84 per cent, of AI2O3, the mean value being 3" 71 per cent. 
Now, owing to the removal of sUica, as shown in the foregoing 
equation, we should have expected a small increase in the percentage 
of AI2O3, rather than a decided decrease. As it might be thought 
that the decrease shown above is fortuitous, and dependent on 
the specimens chosen for analysis, it is to be noticed that the 
calculated analysis of the Kotakarra garnet (page 258) shows 18-98 
per cent, of AI2O3, and that the mean percentage of AI2O3 in 10 
Vizagapatam ores of which the analyses are scattered through this 
Memoir is 2-50, the range being from 0-27 to 6-84. Some of the ores 
thus analysed may have been formed entirely by replacement ; but in 
any case the point illustrated is that in whatever way the ores are formed 
there is a tendency for a disappearance of AI2O3. Hence the equation 
given above is not strictly uue, because a portion of the AI2O3 is removed, 
probably by the same solution as decomposed the rock and deposited the 
manganese oxide. (Compare page 268). 
We can regard the ores formed in this part of the Kodur mine as the 
combined product of the alteration of spandite-rock in situ and of the 
deposition of manganese oxide brought in from elsewhere ; the deposition 
of this manganese oxide being probably somewhat of the nature of a 
replacement, in which the manganese contained in the solution attacking 
the garnet is deposited in the place of the constituents being dissolved. 
1 It is a pity that no analysis of Kodur garnet is availahle. The Garbham one 
was chosen because it happened to be easier to pick clean ; but it looked exactly the 
same as the Kodur garnet. I have compared this analysis with those of the Ko^Iur 
manganese-ores noticed above, because the Kodur ores are ones formed in situ from the 
gainet-rock ; whilst this does not apply to the Garbham ores analy.^ed. 
