358 
MANGANESE DEPOSITS OF INDIA : GEOLOGY. 
[Part II 
extent into the psilomelane. The psilomelane is represented in the equa- 
tion as a manganate of manganese of the formula Mn2Mn05. The 
theoretically pure mineral of this composition has not yet been cer- 
tainly identified : for a portion of the basic manganese is always replaced 
by calcium, magnesium, iron, derived from the spessartite as explained 
above ; by hydrogen, derived no doubt from the waters that brought 
about the alteration of these rocks ; and by barium, which, in the ab- 
sence of any evidence of its existence in the original spessartite or gondite, 
must be supposed to have also been brought in by the attacking waters, 
though possibly only from another part of the deposit ; for as mentioned 
on page 339, some of the rocks of the gondite series contain barytes. 
The amounts of alkalies in the Central Provinces psilomelanes are too 
small to be worth consideration. 
From the equation it will be seen that the chief constituents to be 
removed are the alumina and nearly all the silica. 
One result of the removal of the alumina and sihca would be that 
The ores are not residual mixture of braimite and psilomelane, 
porous. formed in accordance with the equation on page 357, 
would be very porous. For if the following specific gravities be assumed : 
spessartite, 4-2 ; braunite, 4*8 ; and psilomelane, 4-3 ; then according 
to the equation, 100 cc. of spessartite would give rise to 29'2 cc. of brau- 
nite and 17'6 cc. of psilomelane, so that the total volume of the braimite 
and psilomelane would be 46-8 cc, or not quite half the volume of the 
original spessartite. The fact that the masses of ore are no longer porous, 
except for occasional cavities found in almost all the deposits, can be 
explained in two ways. In the first place, if the alteration of the ores 
took place in Archaean times, there may have been subsequent tectonic 
disturbances sufficiently intense to compact the porous and cavernous 
masses of ore. I do not, however, think that this explanation would 
apply to the majority of cases, for the simple reason that I do not think 
the ores ever were in this porous condition. The 
The ores formed by . ,. „ . l- t 
coiubiiicd rephieement exammatiou of numcrous microscope sections of 
and decomposition of gondite in process of alteration has convinced 
rhodonite° me that the alteration is of the nature of 
a replacement, in which the incoming solutions 
add a portion of manganese oxide to that Hberated in accordance 
with the equation given on page 357. The best rock in which to study 
this replacement is gondite, or spessartite-quartz-rock ; because, in 
rocks made up entirely of spessartite, it is difficult to distinguish 
alteration from replacement. In the sections of gondite, it is often 
