OP ARTS AND SCIENCES. 5 



the formation of oxalate of roseocobalt, — a result possibly due to the 

 presence of a little free ammonia. The chloro-oxalate does not appear 

 to unite with metallic chlorides. It reduces gold from auro-chloride 

 of sodium, AuCl^Na, while with platinic chloride it forms only the 

 ordinary chlorplatiuate, Co^(NH3)^(,Cly-}-2PtCl^, and with mercuric 

 chloride the anhydrous 6-atom salt, Co2(NH3)jDCly-|-6HgCl^. In this 

 respect it differs remarkably from the oxalate of roseocobalt, which, as 

 I shall show, forms a well-detined salt with platinic chloride. The 

 chloro-oxalate dissolves readily in a hot solution of oxalic acid, and 

 crystallizes in violet needles which contain chlorine, and which appear 

 to be an acid chloro-oxalate. 



Neutral Sulphate. — Schiff has described a violet-colored very 

 slightly soluble sulphate which he obtained, together with sulphate 

 of luteocobalt, by the action of an alcoholic solution of ammonia upon 

 Fremy's sulphate of fusco-cobalt. To this salt Schiff gives a formula 

 corresponding with 



Co,(NH3),,0.(SOj,+30H,. 



This would be the basic sulphate corresponding to the chromate which 

 1 have described, and which has the formula Co2(NH3)ii,.0.(CrO^)2. 

 By the action of baric chloride upon this sulphate, Schiff obtained, as 

 he asserts, a new chloride with the formula : — 



Co2(NH3),,.O.Cl,+30H2, 



which he regards as the true chloride of the purpureocobalt series. 

 From what I have already said, the existence of such a basic sulphate 

 and chloride may be regarded as not merely possible but probable. 

 On the other hand, Braun, on repeating Schiff's experiments, obtained 

 wholly different results. The action of ammonia upon salts of fusco- 

 cobalt yielded him only the ordinary salts of purpureocobalt, together 

 with salts of luteocobalt. Further researches are therefore needed to 

 establish the existence of Schiff's salts. In a single experiment I 

 obtained a small quantity of a salt which appears to be the neutral 

 sulphate of purpureocobalt. A mixture of cobaltic sulphate with 

 excess of ammonia water was allowed to stand some weeks with fre- 

 quent agitation. The deep-red solution was filtered and preci|)itated 

 by alcohol. The heavy red liquid thrown down became solid after a 

 time. Cold water dissolved the larger portion of this mass, but left a 

 beautiful violet crystalline powder, which was washed on a filter till 

 the washings had a fine clear violet tint, and dried iri vacuo over 

 S0^H2 for some days. Of this salt, 



