Chap. IV. ] 
PSILOMELANE. 
Ill 
not always conform to the formula given in Dana's System of Mineralogy, 
namely 3Mn20;}.MnSi03 ; and in the second place because the difference 
between the amount of oxygen required with this method of calculation 
and that actually found is usually not very large. In cases where this 
error is large, say over 1%, then it might be found that the assumption 
of one of the other formulfB for braunite would reduce the difference 
between the oxygen required and that actually found. But, as in the 
case of the analyses of specimens of pure psilomelane given on the 
preceding pages, this difference can be more easily accounted for by 
assuming a portion of the manganese to be in the form of manganic 
manganate instead of manganous manganate, in the case of an excess of 
oxygen ; or by assuming a portion of the iron to be in the form of ferrous 
manganate instead of ferric manganate, in the case of an oxygen deficit. 
In the following table are given 12 analyses of psilomelane extracted, in 
the manner explained above, from the analyses of specimens of manganese- 
ore scattered through the descriptive portion of this work. In these no 
attempt has been made to allow for the oxygen difference, but all the iron 
has been entered up as ferric manganate and all the basic manganese as 
manganous manganate. 
