RICHARDS AND BAXTER. — ATOMIC WEIGHT OP COBALT. 119 



usual care in preparing the material used in this investigation. Pure 

 cobalt was prepared by two entirely different methods ; and a third 

 sample was purified by both methods. The agreement of the results 

 obtained from all of these samples leaves no doubt as to the uniformity 

 of the material used. 



In the first place about three hundred grams of commercial cobalt 

 chloride were dissolved in a litre of water. The removal of the copper 

 group as sulphides was of course the first step ; but on account of the 

 difficulty of filtering out small quantities of these sulphides, the solution 

 was not acidified before passing in sulphuretted hydrogen. In this way 

 enough cobalt sulphide was precipitated to sweep the solution free from 

 small amounts of a precipitate which could not be removed by filtration. 

 After long standing the supernatant liquid was decanted through a filter, 

 and, as the precipitate was found to contain copper, the solution was 

 again fractionated with sulphuietted hydrogen. The sulphides from this 

 precipitation gave no test for copper, hence it is fair to conclude that the 

 filtrate was free from this metal. After the addition of a large amount 

 of hydrociiloric acid to this filtrate, and then amnionic hydrate until the 

 mixture was strongly alkaline, pure sulphuretted hydrogen in excess was 

 passed into the solution. The precipitated sulphides were washed several 

 times with water, and then digested with cold dilute hydrochloric acid. 

 The precipitate was then washed free from iron by decantation. As the 

 succeeding purification by means of potassic nitrite was in itself a separa- 

 tion from many other metals as well as from nickel, it was not thought 

 necessary to reprecipitate the cobalt sulphide at this stage. Accordingly 

 the precipitate was dissolved in aqua regia, and after dilution and filtra- 

 tion the solution was evaporated until the greater part of the excess of 

 acid was driven off. Sodic hydrate was then added to the diluted solu- 

 tion in slight excess, and subsequently a large amount of acetic acid. 

 Upon the addition of potassic nitrite all the cobalt was precipitated as the 

 crystalline yellow double nitrite of potassium and cobalt. This precipi- 

 tate was washed by decantation several times, and after its solution in 

 strong hot hydrochloric acid the precipitation was repeated. The re- 

 agents used in this purification were especially tested for calcium ; for 

 only in the absence of calcium is the separation of cobalt from nickel 

 complete by this method. The efficiency of the separation is shown by 

 the fact that the filtrate from the first precipitation was distinctly green, 

 owing to the nickel salts present, while the filtrate from the second pre- 

 cipitation after evaporation to small volume merely turned faintly brown 

 on the addition of ammonic sulphide. The double nitrite was again dis- 



