Chap. XVII.] 
GONDITE SERIES : ALTERATION. 
359 
seen that the quartz is being bodily replaced by manganese-ore, with- 
out the spessartite being appreciably affected. In other sections, in 
which the change has gone further, it is seen that the spessartite has also 
suffered replacement, with the production of solid masses of manganese- 
ore containing residual patches of spessartite. When it comes to the 
replacement ot the garnet the solution seems to have taken advantage 
of any cracks in the mineral to percolate into its inside and there deposit 
its manganese oxide. And from these cracks the replacement seems to 
have spread out in all directions, so that the final result is a mass of 
manganese-ore with very few or no residual patches of spessartite. 
' When there were no cracks the garnet has been attacked from its ex- 
terior, ragged projection of manganese-ore growing into it. It is evident 
that this replacement-conversion of the garnet into manganese-ore 
must be a combined decomposition and replacement in which the 
attacking solutions break up the garnet somewhat on the lines represent- 
ed in the equation on page 357, and at the same time deposit manganese 
oxide in the interspaces that would be left if the reaction only proceed- 
ed according to that equation ; the manganese oxide deposited being 
thus added to that already in the garnet. This formation of the ores 
by a combined replacement and breaking up of the garnet means of course 
that the incoming waters must contain manganese salts in solution. 
Laci' of examples ^uch manganese was derived, in all probability, 
showing solution of from Other parts of the deposit where solution was 
manganese. going on instead of deposition. Now in the masses 
of kodurite many examples are seen of rocks from which the 
manganese has passed into solution ; but in the case of the manganese- 
ore deposits of the gondite series it is very difficult to point to any 
particular example or case of gondite or other allied rock from which the 
manganese has been removed. In figure 1, Plate 12, is shown a 
photo-micrograph of gondite in which the garnets are each of them 
suffering alteration with the deposition or liberation of manganese oxide 
in the interior of each crystal, without the periphery of the crystal 
being affected. This is possibly a case in which replacement is not 
going on, for the quartz is practically unaffected. Hence it is possible 
that it is an example of gondite being attacked under conditions that 
will permit of the passage of the manganese into solution. 
The next point to be considered is the character of the solutions 
Carbon dioxide as ^^^^ ^^^^ dissolve the manganese of the 
the reagent producing garnet Or rhodonite in one part of a mass of rock 
alteiation. deposit it in another part of the same mass 
