Manchester Memoirs, Vol. xlvi. (1902), No. 11. 5 



materials I used were the crystallised chlorides of the two 

 metals, each guaranteed by the makers as free from the 

 other. 



A solution of chloride of nickel was prepared, of no 

 definite strength, and the amount of nickel in it was 

 determined by electrolytic deposition, from an ammoniacal 

 solution, on a weighed platinum basin, — a process of 

 estimating nickel which appears to be most satisfactory. 

 In this way, in two separate experiments 50 c.c. of the 

 solution, gave O" 1 3 50 and OT355 gramme of metallic nickel 

 respectively. In two further experiments, 50 c.c. of the 

 same solution were mixed, in each case, with an 

 indefinite amount of a solution of cobalt. The two metals 

 were then separated by adding excess of precipitated 

 barium carbonate, made into a paste with water, and 

 bromine water. Ten minutes were allowed for the 

 precipitation of the cobalt. After filtering and washing, 

 the barium was removed from the filtrate by precipitation 

 with sulphuric acid, the liquid evaporated to a small bulk, 

 and the nickel determined by electrolytic deposition. 

 The two experiments gave respectively 0'i345 and 01347 

 gramme of nickel. 



In a similar way a solution of chloride of cobalt 

 was prepared, and the amount of cobalt in it determined 

 by electrolytic deposition from a solution of the double 

 oxalate of cobalt and ammonium, this being one of the 

 solutions of cobalt which is said to give the best results. 

 But, in my hands at any rate, the process of electrolytic 

 deposition is not so satisfactory with cobalt as with nickel. 

 The deposited metal is not so clean, and it is difficult to 

 know when the process is complete. Thus, in one experi- 

 ment 50 c.c. of the solution of cobalt, electrolysed for 5 

 hours (when the deposition seemed to be complete), gave 

 0'I5I7 gramme of metallic cobalt. In another experiment, 



