EICHARDS AND BAXTER. — ATOMIC WEIGHT OP COBALT. 121 



acid solutions. At first this compound was made by passing a current 

 of air for some time through a solution containing cobaltous chloride, 

 ammonic chloride, and ammonia. A considerable quantity of purpureo- 

 chloride was formed by this process ; and from the solution, by acidifica- 

 tion with hydrochloric acid and cooling, fully as much more was obtained. 

 Later another method, capable of giving a much lai-ger yield, was used. 

 Into the strongly ammoniacal solution bromine was dropped slowly. A 

 heavy precipitate resulted, and the solution upon acidification and cooling 

 as above yielded the greater part of the cobalt it contained. Several lots 

 of amines which had been prepared in one or the other of these ways 

 were mixed and dissolved in hot ammonia. After filtering the solution 

 was acidified with hydrochloric acid, with the result that on cooling 

 nearly all the cobalt was reprecipitated. The precipitate was washed by 

 decantation with cold dilute hydrochloric acid, and the whole process of 

 recrystallization repeated. This time the precipitate was collected on a 

 pure filter and dried. After separation from the filter paper it was con- 

 verted into the sulphates of cobalt and ammonium by heating on a sand 

 bath with sulphuric acid until all chlorine or hydrochloric acid fumes had 

 ceased to come off. From the solution of these sulphates metallic cobalt 

 was prepared by electrolysis, solution in nitric acid, precipitation as hy- 

 drate, and reduction, in the manner already described. This preparation 

 was designated as Sample II. 



For the final analyses the cobalt was purified by a combination of all 

 the before mentioned methods. The carefully purified cobaltous nitrate 

 from Sample I. was converted into a cobaltamine compound, and this salt 

 was six times recrystallized by solution in ammonia and precipitation 

 with hydrochloric acid. Each time the precipitate was washed twice by 

 decantation with liydrochloric acid, and after the second, fourth, and sixth 

 dissolving the ammoniacal solution was filtered. The final precipitate 

 was converted into spongy metallic cobalt in the usual manner. After 

 sublimation part of the bromide prepared from this cobalt was resublimed 

 to see if this process had any influence on the salt. That which had been 

 resublimed was labelled IV., the other III. The apparatus used for the 

 sublimation has been already described (page 117). 



Silver was prepared with great care, according to a method similar to 

 that described in previous papers from this Laboratory.* The only im- 

 provement introduced was the removal by means of nitric acid of the 

 surface of the buttons, which had been fused in a vacuum upon a boat of 



* These Procee(lin<TS, XXXIT. R2. 



