on Chemical Equivalence. 171 



phate. The nickel solution is evaporated to dryness (to get 

 rid of the large excess of acid), dissolved in water, and preci- 

 pitated as sulphide with amnionic sulphide. The sulphide is 

 then ignited with excess of pure hydric sulphate in an open 

 porcelain crucible, and weighed as nickelous sulphate. 



In experiments that were tried to ascertain the accuracy of 

 this method, the following results were obtained : — Two solu- 

 tions, each containing 0*2500 grm. manganous sulphate and 

 0-2500 grm. nickelous sulphate, were taken ; the manganese 

 was separated by the method given above, with the following 

 results: — 



(a) Manganous sulphate obtained . = 0-2506 

 Nickelous sulphate 0-2503 



(h) Manganous sulphate obtained . 0'2497 

 Nickelous sulphate 0-2517 



Mean 0-2506 



Having now fixed on the method for separation that was to 

 be used, we made up solutions of the salts required. 



For the manganous and nickelous sulphates, 1-per-cent. 

 solutions were made. The sodic carbonate solution was of 

 5-7151 per cent, strength. It was kept in a stoppered bottle 

 that had been well coated internally with paraffin. 



III. A. Conjoint Precipitability \ 

 A series of nine experiments was made, in which the rela* 

 tive weights of nickelous or manganous sulphate present 

 varied from '1 to '9 grm., — the total weight of nickelous and 

 manganous salt and the volumes of the solution being, how- 

 ever, always the same, viz. 1 grm. and 100 cub. centim. The 

 experiments were conducted as follows : — The bottles contain- 

 ing the solutions of the sulphates and the sodic carbonate were 

 immersed in a tank, through which there was a constant flow 

 of water from the main to bring them to a constant tempera- 

 ture. The necessary quantities of nickelous and manganous 

 solutions were then carefully measured out, mixed, and the 

 temperature observed. 10 cub. centim. of sodic carbonate 

 solution were then added, the mixture stirred, and the tempe- 

 rature again observed. The precipitate was then filtered off 

 as quickly as possible* (an aspirator being used to facilitate the 

 filtration) and washed, first with cold and then with hot water. 



* Continued experience with precipitation experiments has convinced 

 us that oiu- probable error would have been less if we had allowed the 

 precipitate to stand some time. 



P2 



