Chap. XIII.] kodurite series : alteration. 
269 
rock that undergoes replacement, when the garnet is attacked the mangan- 
ese it contains must be left behind with the manganese oxide deposited 
from the attacking solution. So that the resultant manganese-ore 
consists of the original manganese contained in the rock together with 
that brought in by the replacing solutions. 
Not only can manganese-ores be formed by replacement in the way 
described above, but they can be formed by 
Formation of raangan- ,^ , , , . . ., « , , i • i i 
ese-ores by alteration tlie alteration m situ of the very highly man- 
in sUu of manganese- ganiferous varieties of the kodurite series. 
This was to be well seen at the time of my 
visit to the Kodur deposit. In the north-east corner of the mine 
at the point A in fig. 82 on page 1060 there was exposed a 
mass of spandite-rock in every stage of alteration into manganese-ore, 
which was being actively quarried. The final product of the alteration 
is a vesicular psilomelane with numerous vesicles averaging about to 
|- inch in diameter. Some of these vesicles are found to contain water 
when freshly broken open, whilst others contain a brownish black powder, 
and some are filled with shining black braunite usually traversed by very 
thin veinlets of psilomelane. Since a common type of ore in this mine is 
massive grey psilomelane with small specks and patches of braimite, 
it seems probable to me that the vesicular ore mentioned above must be 
an intermediate stage between the garnet-rock and this massive ore, 
and that later the vesicles have become filled with braunite by percolating 
solutions of which we have evidence in the water contained in the cavities 
of the ore. I have alluded to this ore as having been formed by the 
decomposition of the spandite-rock in situ ; but it is evident that, if this 
rock were decomposed and most of its constituents removed in solution 
except the manganese oxides and iron, the residue would be a porous mass 
of oxides and not solid manganese-ore. Hence it is necessary to suppose 
that a further portion of manganese was brought in in solution and added 
to that derived from the decomposition of the rock in situ. The question 
now arises as to whether manganese oxide was brought in after the span- 
dite-rock had been decomposed with removal of most of its constituents 
except the oxides of manganese and iron ; or whether the decomposition 
of the garnet and removal of certain of its constituents were effected by 
the same solutions as brought in the further supply of manganese oxide. 
For two reasons it is unlikely that the former is the case. In the first 
place, as carbonated alkaline solutions are supposed to account for the 
changes noticed in other parts of the deposit, it is necessary to suppose 
