1879.] 



Researches in Chemical Equivalence. 



183 



In the experiments with this process, the nickel was invariably 

 found too high. This was at first thought to be due to alkali which 

 might cling to the precipitated oxide ; and, on testing, some potash 

 was actually found in "the precipitate, which had been thoroughly 

 washed. To remove this source of error, the nickelous sulphate, 

 instead of being weighed as such, was dissolved in water, excess of 

 ammonia added, and the nickel deposited on a platinum crucible, by 

 means of two Grove's cells. 



But the nickel was, with this method, still found too high, and we 

 then thought that the separation had not been perfect. On testing the 

 deposit, a considerable amount of cobalt was, in fact, found in it, and 

 the operation required to be repeated three times before the separation 

 was complete. 



The next process we tried was devised by Rose.* It consists in 

 saturating with chlorine a very dilute solution of the mixed salts, 

 adding baric carbonate in excess, and allowing to stand from twelve to 

 eighteen hours, with occasional shaking. The cobalt then falls as 

 sesquioxide, the nickel remaining in solution. 



But the results obtained by this method, as was also the case with 

 Henry's modification of it (which consists in substituting bromine 

 for chlorine) were very variable. In the former case the cobalt was 

 generally found too low, even after standing over eighteen hours ; the 

 result appearing to depend a good deal on the liquid being shaken up 

 at regular intervals, which cannot be very conveniently done, during 

 an entire period of eighteen hours. In the latter case, the cobalt was 

 generally found too high, nickel being precipitated along with it. 



The last method tried, and the one finally adopted, is due to Gribbs.f 

 In this, the neutral solution of the sulphates is boiled with plumbic 

 peroxide. The cobalt is then precipitated as a higher oxide, while the 

 nickel remains in solution, along with a small quantity of lead. (The 

 author of this process does not claim very great accuracy for it, but 

 we have found it to be of adequate accuracy.) 



The manner in which we operated was as follows : — The perfectly 

 neutral solution of the mixed sulphates, was" boiled for half an hour 

 with plumbic peroxide, about 7 grms. of the peroxide being taken to 

 1 grm. of cobaltous sulphate, and the volume of the liquid being 

 about 100 cub. centims. The liquid was then filtered, and the filtrate 

 evaporated to about 20 cub. centims. Some aqueous hydric sulphide 

 was then added, and the small quantity of plumbic sulphide formed 

 filtered off. The solution of nickelous sulphate was then evaporated 

 to dryness in a weighed crucible, ignited, treated with a little hydric 

 sulphate, heated to very dull redness and weighed. The precipitate 

 containing the cobalt, with excess of plumbic peroxide and some sul- 



* " Pogg. Ann.," lxxi, 545. 

 f " Sill. Am. J.," xiv, 205. 



2 



