Dr. Gibbs’s Contributions to Analytical Chemistry. 207 
oxyd of iron also contains manganese which must be separated 
in the usual manner. en however the alkalies are present, we 
may separate the iron and alumina by means of carbonate of 
baryta, and then determine the manganese as above,’ thus avoid- 
ing the presence of ammoniacal salts in the solution. 
_ Separation of manganese from zinc and nickel.—The pro- 
cesses to be followed in this case are the same as in the last. 
rom “zinc, but is less serviceable where manganese is to be sepa- 
rated from nickel, since this metal is usually associated with co- 
alt. When cobalt and nickel are both to be separated from 
Manganese, Ebelman’s method appears preferable to any other. 
From what has already been stated (8 and 9) it will be evident that 
the peroxyd of lead cannot be employed to separate manganese 
from iron and cobalt. On the other hand, however, it is highly 
probable that this reagent will effect the perfect separation of man- 
ganese from glucina, zirconia, and the metals of the cerium 
group. But upon this point the writer cannot speak from actual 
experiment. ° erie 
For the analysis of a mineral which contains silica, alumina 
oxyd of iron, manganese, lime, magnesia, and the alkalies, I 
bonate of baryta to separate alumina and iron. To the filtrate 
from these bases peroxyd of lead is to be added to separate the 
Manganese. The filtrate, after separating the lead by means of 
sulphuretted hydrogen, will then contain baryta, lime, magnesia, 
and the alkalies, which are to be separated in the usual manner. 
By this method we may avoid in a great measure, if not entirely, 
the presence of ammoniacal salts, and obtain more accurate deter- 
Minations of the lime and manganese than by thé ordinary pro- 
ce 3 
. 
Process are not alluded to, it became necessary to make this point 
the subject of special study. The results obtained in this investi- 
Sation are as follows: 
4 aS e 
