Chap. XIV.] 
DHARWAR SERIES : ALTERATION. 
295 
earth. There is no evidence that this alteration is going on at the present 
day, except perhaps to a very insignificant extent. 
Coming now to the way in which this alteration was effected, the evid- 
Method of alf^-- ^^^^ obtained by the microscopic examination 
ation of the manganpse of a large number of thin sections of spessartite- 
sihcates. rhodonite-bearing rocks in all stages of alter- 
ation to manganese-ore indicates that, in the case of spessartite, 
the mineral did not as a rule decompose in situ with the solution 
and removal of the alumina, silica, lime, and magnesia, and 
the resultant residual concentration of the oxides of manganese 
and iron. On the contrary, the garnet was attacked in one place 
with the solution of tlie manganese, which was carried in 
solution to another part of the deposit, where it replaced not only 
the spessartite already there, the manganese and iron contents of this 
replaced garnet being no doubt added to the oxides of manganese and 
iron brought in by the replacing solutions ; but also the quartz, when the 
spessartite occurred as spessartite-quartz-rock. If the formation of 
the manganese-ore had been brought about by the decomposition in situ 
of the garnet and the accumulation of the manganese oxides, then the 
resultant ores would have been porous, imless further supplies of man- 
ganese had been subsequently brought in to fill up the interspaces. 
On the other hand, if the formation of the ore took place by replacement, 
as is suggested by the microscopic aspect of these rocks, the resultant 
ores should be compact, as is the case. Hence the evidence 
of the physical aspect of those ores that can fairly be supposed to have 
been derived from the spessartite-bearing rocks supports the theory that 
such ores were largely formed by replacement. But it must be noticed 
that this feature of the ores, namely, whether porous or compact, is not 
necessarily good evidence of the way in which they were formed, because 
their formation probably took place at such a depth that the pressure 
there existing may have been sufficient to close up any small open spaces 
and render the ores more compact than they would otherwise have 
been, though probably not as compact as an ore formed by replacement. 
In the case of rhodonite the ores seem to have much more frequently 
formed by the alteration of the mineral in situ, the alteration of the 
mineral starting along cracks and gradually spreading throughout 
the whole crystal. In cases where the rock contained both spessartite 
and rhodonite the latter mineral is frequently found nearly or completely 
altered to manganese oxides whilst the garnet remains perfectly 
fresh 
