44 CONTRIBUTIONS TO MINERALOGY. [bull. 202. 
If the mineral is to be regarded as anhydrous, the comparatively 
simple formula R" (Mn 3 7 )" satisfies the above ratio, and it may be 
written structurally 
= Mn""- 
-0 
A 
\ 
1 
\ 
= Mn"" 
R 
i 
/ 
/ 
V 
0=Mn""- 
in which R" — Pb" or Mn". This is to be regarded as a saturated salt 
of one of the numerous possible derivatives of ortho-manganous acid 
that may be derived from it by removal of water, in the present case 
as follows: 
3H 4 Mn0 4 -5H 2 0=H 2 Mn 3 7 
An acid of the same empirical formula would result by removal of i 
two molecules of water from three of metamanganous acid, H 2 MnOl 
It is probably best to rest for the present content with the above-r 
relatively simple formula and to regard the water found as due tm 
incipient alteration. But if the water is to be considered as wholly or 
in part essential, and furthermore constitutional — and this may veil 
well be the proper view to take — then the formula becomes much 1 
more complex, namely, R/'H 2 (Mn 12 2<J ), when none of the water is 
allotted to the foreign matter. This formula is still referable graph- 
ically to a more highly condensed manganous acid, and a number of 
isomers would be possible. 
Such intricate formulas as this should not cause the least surprise, \ 
however unlikely they may at first appear to be. The great number 1 
of manganites in varying degrees of saturation and hydration observed* 
in nature and prepared artificially, some of them of even greater com-il 
plexity than the above, are certainly not all mixtures of only a few! 
simply constituted molecules. A very short study of the graphic i> 
formula corresponding to the above empirical formula R" 4 H 2 (Mn 12 29; '.,| 
will show what a vast number of closely related bodies are theoretic 
ally producible by hydrating the molecule step by step or by adding tc 
or reducing 1 the number of bivalent atoms or substituting for their 
those of another valence. Similar varieties in great number would b( 
derivable from other condensed manganous acids of both higher anc 
lower orders, and it is plain that because of the very slight difference; I 
in percentage composition between many of them it is almost as hope | 
less to expect analysis to reveal the exact empirical formula in th< i 
majorit} 7 of cases as it is for the enormously complex album inott I 
