1822.] Method of analyzing the Ores of Nickel. 213 



and separates the potash ; and when the liquor which passes 

 through the filter is evaporated to dryness, and leaves no resi- 

 duum, the washed oxide is an hydrate of nickel which contains 

 no trace of potash ; if, on the contrary, there was an alkaline 

 earth, it would remain combined with the oxide. 



There is considerable difl&culty in ascertaining whether the 

 oxide thus obtained contains alkali or not. Thinking at first 

 that the washed oxide still retained the potash precipitated with 

 it, I endeavoured to analyze it in order to determine the quantity 

 of potash retained ; but all methods in the humid way completely 

 failed. There remained then only to reduce the oxide by means 

 of hydrogen gas, as I had done with the oxides of lead and of 

 copper. One hundred parts of nickel, heated to redness before 

 the experiment, gave me 78'8 parts of nickel, which, digested for 

 a long time in water, did not impart to it the property of acting 

 like an alkaH ; or, if there was any action at all, it was scarcely 

 perceptible : then, as we know from other experiments that lOQ 

 parts of oxide of nickel contain 78*70 of nickel, it is evident that 

 the oxide contained no potash. In another experiment I 

 obtained from 100 parts of oxide of nickel, 79*7 of nickel; but 

 this nickel imparted to water the property of acting as an alkali. 

 This water became turbid by exposure to the air and on the 

 addition of oxalic acid, it, therefore, contained lime. This oxide 

 was obtained by an analysis, in which 1 had not separated from the 

 ore the traces of carbonate of lime that it frequently contains. 



Those bases which are soluble in ammonia possess the other 

 properties of oxide of nickel to such an extent, that it is often 

 difficult to discover them, especially when they are in small quan- 

 tity : these bases are the oxides of cobalt, copper, and zinc. 



Oxides of Nickel and Cobalt. — I have already observed that 

 Mr. Phillips has given us a method of separating these two 

 oxides by means of caustic potash. In order that it may succeed, 

 it is requisite that the ammoniacal solution should be very 

 dilute, and that the water with which it is diluted should be freed 

 from atmospheric air by long ebullition ; for the addition of 

 potash gives the oxide of cobalt, which is dissolved, a strong 

 disposition to become peroxide, which appears to depend upon 

 the potash combining with the acids, the oxide of cobalt is held, 

 therefore, in solution by the ammonia alone ; whereas, before it 

 was dissolved in the state of a double subsalt, consisting of 

 ammonia and oxide of cobalt. 



If the solution contains atmospheric air, the oxide of cobalt 

 combines with its oxygen, and the oxide of nickel comes down 

 with the peroxide formed. The more concentrated the solution 

 is, the greater is the tendency of the oxide of cobalt to peroxi- 

 dize ; and it then frequently deposits during filtration. It is, 

 therefore, easier to separate a large quantity of nickel from a 

 «mall quantity of cobalt, than the converse, although in every case 



