on the Ammonia-cobalt Bases. 241 
sulphate of baryta appears to have been completely washed. 
é have, therefore, in all cases preferred to decompose the salt 
be analyzed, by boiling it with a little ammonia. After com- 
plete precipitation of the sesquioxyd of cobalt, chlorhydrie acid 
i8 to be added to reduce me dissolve the oxyd, when the sul- 
phuric acid may be directly thrown down by chlorid of barium, 
Even with these precautions, our results are not unfrequently 
0g aga or three-tenths of one per cent too high, almost never 
Ww 
Oxalic acid.—The ordinary methods for the quantitative esti- 
mation of this acid fuil entirely with the class of salts under 
consideration. A solution of terchlorid of gold is reduced only 
after very long and tedious boiling, and then incompletely. 
Even after previous reduction of the cobalt to the form of pro- 
toxyd, the method is found to be very inconvenient and inaceu- 
Tate. The conversion of the oxalic into carbonic acid by oxyd- 
ation, and its determination from the weight of this last, gave 
no better results, inasmuch as the oxydation is effected wit 
difficulty. We have therefore in all cases had recourse to the 
ultimate organic analysis, which alone gives reliable results. 
The methods employed in the determination of other substan- 
Ces will be described, when necessary, in treating of particular 
compounds. 
ROSEOCOBALT. 
The description of the salts of Roseocobalt forms, upon the 
Whole, the most convenient starting point in a statement of 
the results of our investigation. These salts are in general 
easily obtained, and the products of their decomposition include 
Several of the other bases, which we shall have occasion to 
describe. They are almost all well crystallized, and are in 
general nearly insoluble in cold water, soluble without decompo- 
sition in warin water slightly acidulated, but easily decomposed 
when the neutral solutions are boiled, a hydrated hyper-oxyd of 
cobalt being thrown down, while free ammonia 1s given 
The salts of Roseocobalt have a purely saline, not metallic taste; 
their color varies, being sometimes dull or brick-red, - 
tines cherry-red. They are usually dichrous, though a few of 
them do not exhibit this property in a marked degree. Heat 
decomposes the dry salts Hes the final products of the de- 
Composition being usually ammonia, a salt of ammonium and a 
salt of protoxyd of cobalt. Intermediate products are, how- 
ever, sometimes formed, as we shall hereafter see. Thus in 
teocobalt, which then, bv continued boiling, are completely de- 
composed. The salts of Roseocobalt may almost always be pre- 
Pared by the direct oxydation of ammoniacal solutions of salts of 
SECOND SERIES, VOL. XXIII, NO. 68,—MARCH, 1887, 
31 
