Cyanide of Potassium, and on its Applications. 269 



tative Analysis. — Nickel, cobalt and manganese are so nearly 

 related in their properties, that their separation is attended 

 with great difficulties. In one single form of combination only 

 does nickel differ from cobalt to such an extent that this 

 might be used as an absolute means of separation. 



The oxide, protochloride, or any salt of cobalt warmed 

 with cyanide of potassium and an excess of hydrocyanic 

 acid, is converted into the percyanide of cobalt and potassium 

 (the cobalti-cyanide of potassium), the aqueous solution of 

 which, according to the observation of L. Gmelin, does not 

 undergo the slightest decomposition from boiling with hydro- 

 chloric, sulphuric or nitric acid. 



The oxide of nickel and its salts are thrown down by the 

 cyanide of potassium ; this precipitate dissolves in an excess 

 of the precipitating agent of a yellow colour; and the double 

 compound of cyanide of nickel and cyanide of potassium, 

 although not decomposed by acetic acid, is perfectly so by 

 dilute sulphuric acid, and the cyanide of nickel again pre- 

 cipitated. 



When a mixture of a cobalt and nickel salt containing free 

 acid is treated with an excess of cyanide of potassium, so that 

 the precipitate formed is redissolved, there are in solution free 

 hydrocyanic acid, cyanide of potassium, cyanide of nickel, and 

 the protocyanide of cobalt ; the latter changes immediately, 

 on being slightly warmed, into the cobalti-cyanide of potas- 

 sium; if now dilute sulphuric acid be added in the cold, three 

 cases present themselves. 



If the cobalt and nickel in solution are in the proportion 

 by weight of two cobalt to three nickel (quantities which cor- 

 i-espond to their atomic proportions in the cobalti-cyanide of 

 nickel), the precipitate produced is cobalti-cyanide of nickel, 

 and is of a bluish-white colour. The filtered liquid contains 

 not a trace of cobalt or nickel. 



If the solution contains less nickel than corresponds to the 

 above proportions, there remains in solution a certain quan- 

 tity of cobalti-cyanide of potassium, and the precipitate is still 

 cobalti-cyanide of nickel. 



If there is more nickel present in the solution, the preci- 

 pitate is a mixture of cyanide of nickel and cobalti-cyanide of 

 nickel. 



In the first and second cases, the precipitate produced by 

 dilute sulphuric acid is boiled so long with the acid fluid in a 

 vessel until not a trace of hydrocyanic acid is observed to 

 escape (or it may be evaporated to dryness in a watei'-bath), 

 and then slightly warmed with an excess of carbonated or 

 caustic potash ; the cobalti-cyanide of nickel is decomposed by 



