y RESEARCHES ON THE 



sesquioxide of cobalt, which is sometimes almost inappreciable, sometimes in com- 

 paratively large amount. In one experiment, in which we employed perfectly pure 

 chloride of cobalt and pure ammonia, there was no deposit whatever of oxide. In 

 this case, however, no chloride of Roseocobalt, but only chloride of Purpureo- 

 cobalt was formed. When impure materials are used the precipitate is abundant, 

 and contains many of the impurities of the substances employed, as well as much 

 sesquioxide of cobalt. The rate at which oxygen is absorbed varies much with the 

 degree of concentration of the solution, with the temperature, with the quantity of 

 ammonia present, and with the extent of liquid surface exposed to the air. Fre- 

 quent agitation of the solution materially shortens the time required for complete 

 oxidation, and the same effect is produced by passing a current of oxygen directly 

 through the liquid, which soon becomes brown and subsequently red. As a general 

 rule, the first effect of the oxidizing action is to give the liquid a brown color, the 

 layer next the surface being the first to change its tint. The brown color then 

 passes gradually into a deep red, and the oxidation is complete, when the whole 

 mass of liquid has the color of red Burgundy wine. 



The presence of chloride of ammonium is not necessary in this process ; a large 

 quantity of this salt in the solution often gives a lilac or purple precipitate as the 

 oxidation advances, but this is composed principally of the chloride of Purpureo- 

 cobalt. As will be seen from the above, the chloride of Roseocobalt is not always 

 formed during the oxidation of an ammoniacal solution of chloride of cobalt. On 

 the contrary, it often happens that not a trace of this salt can be obtained from the 

 oxidized solution, which contains only the chloride of Purpureocobalt. We have 

 observed the absence of the chloride of Roseocobalt only in solutions which had 

 been oxidized in a warm room, or during the summer season. This fact, taken in 

 connection with the facility with which heat transforms solutions of Roseocobalt 

 into those of Purpureocobalt, renders it, to say the least, extremely probable, 

 either that a comparatively high temperature prevents the formation of the chlo- 

 ride of Roseocobalt entirely, or else that this salt is converted into chloride of 

 Purpureocobalt as fast as it is formed in the solution. 



To obtain the chloride of Roseocobalt from the oxidized solution, cold and 

 strong chlorhydric acid is to be added to it, the slightest elevation of temperature 

 being carefully avoided. A brick-red precipitate is thrown clown, which is to be 

 washed with strong chlorhydric acid and then with ice-cold water, thrown upon a 

 filter, and dried by pressure, great care being taken to operate at as low a tempera- 

 ture as possible. 



As the formula of the chloride of Roseocobalt is, 5NH 3 .Co 2 Cl 3 +2HO, its forma- 

 tion by the oxidation of the ammoniacal solution of chloride of cobalt may be 

 explained by the equation 



6CoCl+10NH 3 +3O=2(5NH 3 .Co 2 Cl 3 )+Co 2 O 3 . 



In those cases in which no sesquioxide of cobalt is precipitated, we may suppose 

 that the sesquioxide unites directly with ammonia, as represented by the equation 



6CoCl+15NH3+30=2(5NIl3.Co 2 Cl 3 ) + 5NH 3 .CoA- 

 On adding an excess of chlorhydric acid to such an oxidized solution, 3(5NH 3 . 



