OF ARTS AND SCIENCES. 3 



in tlic excess of nitric acid. From tliis salt many otliei: salts of tiie 

 roseo-series may be prepared with facility. 



Chloro-mtrate of Purpnreocobalt. — Acid sulphate of roseocobalt 

 not free from chloride of jiurpureocobalt was treated in the cold with 

 a solution of potassic nitrite and nitrate. After standing some days a 

 red mother licpior was formed, together with a mixture of a red crystal- 

 line salt, and a bright yellow powder. On filtering and washing with 

 hot water, a fine violet liquid was obtained, which, on standing, gav« 

 very well-defined large* octahedral crystals of a deep cherry-red color. 

 The crystals were easily soluble in hot water, and contained only a 

 trace of sulfihuric acid. Qualitative analysis showed the presence of 

 chlorine, nitric teroxide, cobalt, and ammonia. The salt gave the 

 reactions of nitrate or chloride of purpureocobalt with more or less 

 distinctness. Of this salt, 



0*4olo gr. gave 0-1516 gr. silver = 8-30 (y^ chlorine. 

 0-4000 gr. gave 0-1332 gr. silver = 8-23% chlorine. 

 0-4809 gr. gave 0-2397 gr. SO,Co= 18-98% cobalt. 

 0-4448 gr. gave 0-2228 gr. SO.Co — 19-07% cobalt. 

 1-0770 gr. gave 0-4GoO gr. water = 4-79% hydrogen. 



The formula Co,(NIL),oCl3(N03)3+Co2(NH3)ij,(N03)« requires 



Cobalt, 4 



Chlorine, 3 

 Hydrogen, GO 



I can assign no plausible explanation of the formation of the chloro- 

 nitrate under the circumstances, and did not succeed in obtaining the 

 salt by mixing the chloride and nitrate of purpureocobalt in the proper 

 proportions, and allowing the mixed solutions to stand. The crystals 

 formed always consisted principally of the chloride, I was not more 

 successful in the attempt to form a chloro-nitrate with the formula, 

 Co2(NH3)j(,(N03)3Cl3, by mixing solutions of the chloi'ide and nitrate, 

 the chloride crystallizing from the mixtui'e unchanged. I remark, 

 however, that as the acid sulphate employed contained chlorine, prob- 

 ably as undecomposed chloride of purpureocobalt, and as the potassic 

 nitrite contained also nitrate, the reaction must have been between the 

 chloride and nitrate. Two chloro-nitrates of roseocobalt have been 

 observed by Krok, in combinatioD with mercuric and platinic chlorides, 

 the salts having respectively the formulas Co2(NIIg)jy(N03)3Cl3-|- 

 3IIgCl, and Co,(Nll3),,(N03),Cl,+2PtCl,. 



