246 Dr. E. J. Mills on certain Cobaltamines. 



greatest amount of work performed. That the three chlorides 

 are generated in this order is generally admitted ; and I think 

 the above dynamical view of the process will be acknowledged 

 as correct. It is supported, at any rate, by a remarkable result, 

 first obtained by the distinguished American chemist already 

 mentioned. Gibbs found* that the /3-disulphate, on treatment 

 with water, splits into the a-sulphate and free hydric sulphate. 

 Work done in decomposition here effects, as might have been 

 foreseen, a return to the point at which the opposite kind of 

 work had been attended with its first consequence. 



(1) Cobaltic hexammoniotrichloride is readily obtained when 

 cobaltic chloride is heated with sal-ammoniac and aqueous am- 

 monia in presence of an energetic oxidizing agent. It is best to 

 carry out the operation under pressure ; and for this purpose a 

 soda-water bottle with an india-rubber stopper, which can be 

 closed tightly by a simple screw-arrangement, will be found of 

 great convenience. The most suitable temperature of digestion 

 is about 70° C. for a period of not less than twenty hours ; and 

 the number of oxidizing agents which may be employed is pro- 

 bably very considerable. 



One part (15 grms.) of sal-ammoniac, potassic permanganate, 

 and crystallized cobaltic chloride f respectively, together with 

 6| parts of strong aqueous ammonia, yield, after digestion as 

 above mentioned, an orange-coloured liquid and warty aggre- 

 gates of the desired chloride. There is a copious deposit of 

 manganic, containing cobaltic, peroxide. The ammoniacal liquid 

 is transferred to a beaker containing an excess of aqueous hydric 

 chloride; the crystalline aggregates are dissolved out by hot 

 water containing a little hydric chloride, and their solution is 

 mixed with the preceding ; a double volume of strong aqueous 

 hydric chloride is added to the mixture, which is then left to itself 

 for twenty-four hours; at the end of that time the yellow preci- 

 pitate is removed to a filter and washed, first with aqueous hydric 

 chloride, and finally with spirit. In this manner, about 71 per 

 cent, of the theoretical quantity of the hexammoniochloride is 

 obtained. 



(2) If 1 part of potassic dichromate be used instead of the 

 potassic permanganate in the above process, the amount of chlo- 

 ride obtainable corresponds to very nearly 80 per cent, of the cal- 

 culated yield. 



(3) One part of crystallized cobaltic chloride, 1 part of native 

 manganic binoxide J in fine powder, 2 parts of sal-ammoniac, and 



* Silliman's American Journal, vol. xxiii. p. 322. 

 f Vide infra. 



% Since these experiments were performed, Braun has published two 

 processes in which manganic and plumbic peroxide respectively are em- 



