26 PROCEEDINGS OF THE AMERICAN ACADEMY 



witli impure salts of xanthocobalt. Krok* has described a salt with 

 the formula Co2(NH3)ioCl(N02)5 + 3OH2. There is no theoretical 

 reason why such a compound should not exist, but Krok's analyses do 

 not appear to me sufficient, as the cobalt, chlorine, and ammonia only 

 were determined, and not the whole quantity of nitrogen in the salt. 

 Moreover, it is not proved that the salt can be recrystallized without 

 decomposition, or that it forms deliuite compounds with metallic 

 chlorides. ' » 



As the chloride and nitrate of xanthocobalt are capable of uniting 

 directly to form the chloro-nitrate above described, it might be supposed 

 that the two salts are isomorphous, and, therefore, crystallize together 

 m all proportions. According to Professor Dana's measurements, cited 

 in the first part of this memoir, nitrate of xanthocobalt crystallizes in 

 forms belonging to the dimetric or square prismatic system. Professor 

 Cooke has kindly determined the form of the corresponding chloride, 

 and finds that the crystals are either trimetric or monoclinic. The 

 chloro-nitrate cannot, therefore, be regarded as a mixture of two iso- 

 morphous salts. 



11. Finally, salts of xanthocobalt are formed by the action of 

 Fischer's salt, Co., (N 02)12 Kg, upon salts of purpureocobalt and roseo- 

 cobalt. When, for instance, chloride of purpureocobalt is dissolved 

 in boiling water, with a little free acetic or chlorhydric acid, and 

 Co2(NO.,)j2Kg is added, in small portions at a time, the violet color of 

 the salt gradually disappears as the boiling continues, and the solution 

 finally assumes a fine orange-brown tint. The filtered solution gives 

 on cooling fine crystals of chloride of xanthocobalt, the reaction being 

 probably expressed by the equation 



Co2(N02)i2Kg + 3Co2(NH3),,Cl, = 3Co2(NH3),„(N02)2Cl,+ 

 6KC1 + 2Co(N02)2 + 2NO2. 



During tliC boiling red vapors are given otF. In one experiment the 

 chloride of xanthocobalt formed was analyzed, with the following 

 results: — 



0-5027 gr. gave 0-2987 gr. SO^Co=: 22-62 per cent cobalt. 

 0-7616 gr. gave 0-63ol gr. silver =27-35 per cent chlorine. 



The formula Co2(NH,)i„(N02)2Cl4 requires 22-52 per cent cobalt, 

 and 27*09 per cent chlorine. The salt gave ail the reactions of the 

 chloride. 



* Acta Univers. Lund, 1870. 



