Manchester Memoirs, Vol. xlvi. (1902), No. II. 7 



bromine water added.* It is now allowed to stand for five 

 minutes, with frequent shaking. If cobalt is present, a 

 black precipitate very soon appears. At the end of five 

 minutes the liquid is filtered and the filtrate tested for 

 nickel by the addition of a drop or two of ammonia and 

 ammonium sulphide. The presence of cobalt in the pre- 

 cipitate may be confirmed by the borax bead test. In 

 this way nickel may be detected even when present in 

 very small quantity, and when the cobalt is largely in 



excess. 



Instead of boiling off the free acid, it may be neutral- 

 ised with sodium or potassium hydrate before adding the 

 carbonate and bromine water ; or the carbonate of barium 

 or calcium may be added in excess to the acid liquid, and 

 then the liquid boiled for a short time to expel the free 

 carbon dioxide. It must then be cooled to the ordinary 

 temperature before adding the bromine water. 



The carbonates of barium, strontium and calcium are 

 not the only ones which act in the manner described ; the 

 carbonates of magnesium and zinc apparently act in the 

 same way. Thomas Moore (67^^///. Nezvs, Vol 82 (iqcxd), 

 p. yi), describes the use of zinc oxide and bromine water 

 for the purpose of estimating small quantities of cobalt in 

 presence of nickel. Mr. Moore finds that those reagents 

 precipitate cobalt as sesquioxide, but not nickel, even on 

 boiling. He insists that bromine 7?mst be used, and not 

 chlorine. I have tried the experiment, and fail to find 

 any difference at all between the action of chlorine and 

 bromine. 



* Barium and calcium carbonates appear to act equally well. The dry 

 substances may be employed. Barium carbonate is preferable if the subse- 

 quent removal of the added metal is desired. 



Central School, 



Manchester. 



