OP ARTS AND SCIENCES. 21 



pro})ortion of the octamin nitrate, by the action of a mixture of potas- 

 sic nitrite and ammonia upon cobaltic nitrate in presence of air; but 

 that xautliocobalt is exckisively formed by the action of the same 

 mixture upon a solution of ammonie and cobaltic sulphates. I am 

 unable to offer any plausible explanation for the difference of the 

 products in the two cases. 



When cobaltic nitrate, amnionic nitrite, and ammonia are mixed and 

 placed in a tightly corked bottle, no action whatever appears to take 

 place, even after the mixture has stood some days. But if plumbic 

 hyperoxide, PbO^, is added, the mixture soon becomes yellow, and 

 after a few hours large crystals of nitrate of xanthocobalt are formed, 

 with distinct reduction of the plumbic hyperoxide. The reaction in 

 this case may be represented by the equation : — 



2Co(N03), + IONH3 + 2NH, . NO., -f PbO. = 

 Co,(NH3),„(NO,),(N03), + PbO -f (NHJ.O. 



Potassic hypermanganate may also be employed as an oxidizing agent, 

 but is less convenient. The experiment just detailed appears to me to 

 render it most probable that in the action of the red gases upon salts 

 of cobalt in presence of ammonia, the resulting salts of xanthocobalt 

 are not formed by the direct union of the cobaltic salt with ammonia 

 and nitroxyl, but that amnionic nitrite is first formed, and that the 

 oxygen necessary for the completion of the reaction is derived from 

 tlie decomposition of some element of the complex mixture of NO, 

 NO2, NjOg, and NO3H, which make up the red vapors. 



The formation of salts of xanthocobalt by the action of the red gas 

 upon salts of purpureocobalt and roseocobalt in the presence of free 

 ammonia is easily explained. We have here simple cases of double 

 decomposition, a particular instance of which, covering in substance 

 the whole ground, may be expressed by the equation : — 



Co,(NH3),„(N03)e + 2NH, . NO, = Co,(NH3),„(NO,),(N03), + 



2NH, . NO3. 



Salts of xanthocobalt are always formed when salts of purpureocobalt 

 and roseocobalt are heated or even digested in the cold with alkaline 

 nitrites. I have made a si^ecial study of the action of potassic and 

 sodic nitrites upon chloride of purpureocobalt, the details of which are 

 as follows : 



10. Action of Sodic and Potassic Nitrites wpon CMoride of Purpu- 

 reocobalt. — A quantity of chloride of purpureocobalt was dissolved in 

 boiling water, with a little free acetic acid to prevent decomposition, 



