Chap. IV. ] 
PSILOMELANE. 
105 
aluminium enter into the composition of the manganate^. It will be 
noticed, however, that these specimens cannot be conveniently divided 
into baryta-psilomelanes and potash-psUomelanes. Thus 1157 can be called 
a potash-psilomelane and A. 380 a baryta-psilomelane, but the other three 
all contain moderate proportions of both constituents. 
The general formula 
wR"2lIn05 + nR"'4(Mn05)3. 
to which reference was made above, can be expressed, so as to show che 
constituents corresponding to R" and R'", in the following comprehensive 
way :— 
TO[{Mn,H2,Ba,K2,Ca,Mg,Na2,Zn,Co,Ni,Cu)2Mn05] + n[(Mn,Fe,Al)4{]Mn05)3]. 
Omitting the less important constituents in each case, the composition of 
the five specimens can be stated as follows : — 
Guguldoho(1157) 19[(Mn.H2.K2)2Mn05] + (Mn,re)4(Mn05>3. 
Avagudera (A. 372) 10[(Mii.H2,K2,Ba>2MnO5] + (Fe,Ai;t(]\In05)3. 
Tekrasai (A. 380) 145L(Mn,Ba.H2)2Mn05] + (Al,Fe)4(l\Iii05)3. 
Tekrasai (A. 381) 13[(Mn.H2,Ba,K2)2Mn05] + (Mn,Al)4(Mn05)3. 
Tekrasai (J. 917) ll[(Mn,H2,K2,Ba)2Mii05] + Mn4(Mn05)3. 
In these formulae I have placed the elements constituting the basic 
radicles in the order of their importance considered from the point of 
view of the amount of manganate they form by combination of their 
oxides with the requisite amounts of MnOa. 
If R be entirely Mn, then the formula of the mineral becomes 
Mn2Mn05 or MugOg, which would contain 67-35% Mn and 32-65% 
0. At least two observers have previously suggested the existence of 
this compound in nature. In one case J. A. Jones- gives an analy- 
sis of a ' cryptocrystalline ' mineral from Covadanga in Northern 
Spain, which he thinks corresponds fairly closely to 2Mn02.MnO 
( = Mn305). A calculation of his analysis shows, however, an 
oxygen deficit of 2-81%, if an attempt be made to convert it into 
1 I have already noted — footnote on page 96 — that M. Gorgeu deduces from his 
analyses of specimens of psilomelane that psilonielanes are hydrous manganites, but 
that his analyses of two specimens from Romaneche conform closely to manganates of 
the general formula Rj MnOs He gives the formula corresponding to the Lorca 
analysis as 6(Mn02) RO.2H2O, He regards this, like the other formulae given, as a 
hydrated manganite. The deficiency of the base portion of this formula, as of the others 
given, is noticeable ; whilst it is interesting to note that the formula given for the Lorca 
analyses can be re-arranged as follows : — 
6(Mn05).R0.2H20 = R0.2H20.3Mn0.3Mn03 
= 6R0.3Mn03 
= 3R2]\In03 
in which there is no deficit or surjilus of either basic or acid radicle, 
2 Trans. Intt. Min. Met, III, p. 276, (1894-96). 
