Chap IV. ] 
PSILOMELANE. 
97 
hence corresponds to the fibrous variety of hollandite. At present I see 
nothing that certainly shows the reason of the formation of crystalline 
manganates at one time and non- crystalline ones at others. An examina- 
tion of the analyses of hollandite shows, however, that the crystalline 
manganates always contain a fair quantity of barium and iron ; 
this may indicate that the manganates can only assume the crystalline 
form in the presence of fair quantities of these two elements, although the 
number of analyses available is not really sufficient to base such a general 
statement upon. Such a hypothesis would explain why ores low in these 
constituents were found as psilomelane and not as hollandite, but it 
would not explain the case of varieties of psilomelane containing as large 
quantities of iron and barium as hollandite. Hence it seems more 
probable that the reason for the formation of crystalline manganates at 
one time and non- crystalline ones at others is to be looked for in the 
conditions of formation rather than in the chemical composition of 
the manganates. 
As might be expected, the great variability in the composition of the 
manganates from different localities is accompanied 
Specific gravity. , . , . -r- 
by a correspondmg variability m the specmc 
gravity of these minerals. Thus a picked specimen of this mineral from 
Sitapar in the Chhindwara district was found to have a specific gravity 
of only 4-70, whilst the specimen analysed by Mr. Winch from 
Kajlidongri had a value of 4*95 for this constant. Determinations of 
the specific gravity of several other picked specimens from this locality 
have given values ranging from 4'93 to 4'99. The massive specimen of 
hollandite of which the analysis is given on page 93, had a specific 
gravity of only 4-59, this low value being in all probability due to the 
fact that the specimen was an aggregate and not a piece of a separate 
crystal. It should be noticed, however, that these values for the 
specific gravity correspond roughly with the differences in the amounts 
of baryta present in each specimen, the values of this constituent 
in the Kajlidongri, Sitapar and Balaghat specimens being 17"59, 6'20. 
and 2 "96, respectively. 
Psilomelane. 
Of all the manganese-ores found in India, this is the most abun- 
Occurr ce ^^'^^ ' ^^^^ braunite it makes up by far the 
larger proportion, probably at least 90 per cent., of 
the manganese- ores exported from India. Its existence in India was 
apparently first recognized by A. J. Scott in association with 
