96 
MANGANESE DEPOSITS OF INDIA : MINERALOGY. [ PaRT I : 
In any particular case a more definite formula can, of course, be 
given. Thus the three analyses given above correspond roughly to the 
following formulae : — 
Balaghat rn(jVIn,Ba,K2,H2)2Mn05 + K(Fe,Al)4(Mn05)3. 
Gowari Warhona ??i(Mn:Ba)2Mn05 + wFe4(]Mn05)3. 
Junawani »?i(Mn,Ba,C'a)2 ]Mn05 + ?!{re,Al)4(]Mn05)3. 
As will be seen from these formulae these manganates correspond to 
the hypothetical acid H4Mn05, and it seems as if there is a whole series 
of compounds derived from this acid by isomorphous replacement. 
Laspeyres ^ has already suggested that psilomelane may conform to this 
formula, and we see that the existence of compounds corresponding to 
this formula is put beyond doubt by the discovery of the definitely 
crystallized minerals here described. As will be shown under the heading 
of psilomelane this latter mineral also conforms to the formula suggested 
by Laspeyres, at least as far as can be judged from a large number of 
analyses of Indian psilomelanes. Hence it seems probable that hoUan- 
dite is to be regarded as the crystalline form of psilomelane, there being 
no essential difference in chemical composition between the two minerals. 
An interesting problem for future solution is to find out what deter- 
mines whether the mineral takes the crystalline or amorphous form, 
and why it is that crystalline manganates, which are by no means un- 
common in India, are not found wherever psilomelane occurs in all parts 
of the world.2 It is interesting to note that W. Lindgren and W. F. 
Hillebrand [Amer. Jour. Set., XVIII, pp. 448460, (1904)] describe a 
^ ^.^ new mineral from Arizona under the name of corona- 
dife. They regard their analysis as indicating the 
mineral to be a salt of a derivative of ortho-manganous acid, H^Mn^O^. 
I find, however, that their analysis corresponds very closely with 
RgMnOg, where R = Mn and Pb chiefly. The mineral is fibrous, and 
1 Jour, furprak. Chemie, XIII, p. 215, (1876) [Dana]. 
2 It is interesting to note that M. Al. Gorgeu in a paper called ' Sur les oxydes de 
manganese natiirels'. Bull, de la Soc. fr. de Mincrnlogiqxie, III, (1890), says, on page 22, 
that the interior of a specimen of psilomelane from Romaneche, in Saone et Loire, 
France, was crystalline. He gives analyses of both the exterior non-crystalline portion, 
and interior crystalline portion. He calcidates out formulae for these specimens, and 
also for specimens of p-ilomelane from Thiuiugia in Germany, and Lorsa in Spain, 
according to which psilomelr.ncs arc hydratcd manganites with " bases multiples et vai ices'. 
I find, however, that both tlie Komaneche analyses can be converted into a seiies of 
n.anganates of the form IJoMnOs, the oxygen errors, however, being rather large, 
namely deficits of r44: and 100 ; but these would have been less had I taken account of 
Ihc arsenic oxide of which amounts of 1-50 and 0 70 % are shown in the two analyses, 
and converted it into arsenates. I think it is probable that this crystalline psilomelane 
is identical with h jlh.ndite. The non-crystalline and crystalline portions contain 1C'20 
and 14'45 per cent of BaO, respectively. 
