AMMONIA-COBALT BASES. 9 



Co 2 Cl 3 ) must be formed, which satisfactorily explains the precipitation of the brick- 

 red chloride by the acid. 



Fremy assigns to the brown substance, which is the first product of the oxidation, 

 the formula 4NH 3 .Co 2 3 . "We have not yet been able to obtain this substance in a 

 condition fit for analysis, and Fremy does not consider the formula, which he pro- 

 poses, as by any means established. 



It will be seen from the above that, in the formation of the chloride of Roseo- 

 cobalt, the elements of ammonia unite with sesquioxide or sesquichloride of cobalt 

 at the instant that these are formed by the absorption of oxygen from the air. 

 Claus has recently shown that the sesquichloride of rhodium 1 unites directly with 

 five equivalents of ammonia to form a chloride exactly analogous to the chloride 

 of Roseocobalt, and having the formula 5NH 3 .Rh»Cl 3 . We have made various 

 experiments to determine whether sesquioxide of cobalt once formed could unite 

 directly with ammonia. A solution of chloride of ammonium was poured upon freshly 

 prepared sesquioxide of cobalt, strong ammonia-water added, and the whole allowed 

 to stand for some time in a closed bottle and in a rather dark closet. Even after 

 many weeks, however, only traces of chloride of Roseocobalt could be detected. 

 A quantity of sesquioxide of cobalt was dissolved in strong acetic acid, and to the 

 solution chloride of ammonium and ammonia-water added. In this case chloride 

 of Roseocobalt was formed after a few days, but it is doubtful whether its forma- 

 tion was not due to the oxidation of a small quantity of protoxide of cobalt in the 

 sesquioxide employed. In another experiment, strong ammonia was added to 

 freshly prepared sesquioxide of cobalt, and the whole allowed to stand for several 

 weeks, after which time it was boiled with chlorhydric acid, and considerable quan- 

 tities of chloride of Purpureocobalt, Luteocobalt, and Praseocobalt, were obtained. 

 This experiment leaves no doubt that the Ammonia-cobalt bases can be prepared 

 by the direct action of ammonia upon sesquioxide of cobalt, though this mode of 

 preparation is not economical. 



The chloride of Roseocobalt may also be prepared by adding cold and strong 

 chlorhydric acid to a completely oxidized solution of the ammoniacal nitrate or 

 sulphate of cobalt. A brick-red precipitate is formed in either case, which must 

 be purified by repeated washing with chlorhydric acid. Strong chlorhydric acid also 

 precipitates the chloride from solutions of the sulphate and nitrate of Roseocobalt. 

 In all these cases, however, it is difficult to obtain the chloride in a perfectly pure 

 state. 



The chloride of Roseocobalt is usually precipitated as a brick-red powder, which, 

 under the microscope, appears to be composed of indistinct granular crystals. It 

 may be purified, though with difficulty, by solution in ice-cold water and sponta- 

 neous evaporation in the cold. The salt is soluble in cold, as well as in hot water, 

 with a dark-red but not violet-red color, the portions still undissolved becoming 

 lilac or purple before dissolving. The most remarkable property of this salt is 

 the facility with which it is converted into chloride of Purpureocobalt. The hot 



1 Since the above was written, Claus has extended his observation to the sesquichloride of iridium, 

 which forms a similar base with five equivalents of ammonia. 

 2 



