346 On Coledl and Kkkel. 



; The dry major oxide of nickel is black ; when in a mass 

 its fracture is vitreous. 



This oxide, preserved in ammonia, gives out bubbles, re- 

 turns to the state of gray oxide, and is dissolved in the alkali. 



It gives a considerable quantity of oxygenated acid, with 

 a muriatic acid at 15°. The solution is greenish yellow : 

 crystals are formed upon cooling. 



The oxides of nickel are reduced like those of cobalt. 

 They are melted in the same way, with this difference only, 

 that the cobalt gives a larger globule. 



This metal hns taken a surcharge of sulphur from 46 to 

 100 ; but the author has still his doubts on this subject. 



Arsentie and Arseniate. — They are formed like those of 

 cobalt, and are of a fine apple-green colour. The arsenite 

 healed in the tube loses its colour with water, sets at liberty 

 some white oxide, and passes to the olive-green. Charcoal 

 is necessary in order to take away all the arsenic. 



When heated in a platina spoon, the arsenic is speedily 

 dissipated. An oxide at the minimum remains. 



The arseniate heated in a gun-barrel loses its colour with 

 water ; becoming of a hyacmth and transparent appearance: 

 but at a red heat it passes to the clear yellow, and remains 

 unalterable. 



In the spoon the arseniate becomes white, reddens with- 

 out melting, or emitting the smallest arsenical fumes ; we 

 must avoid flame in order to decompose it. 



Recapitulation. 

 M. Proust concludes from the foregoing facts, and from 

 those he has published in other memoirs, that cobalt, nickel, 

 and most of the other known melals, have only two degrees 

 of oxidation distinctly marked : he has not asserted, how- 

 ever, that a metal can only absorb two proportions of oxygen : 

 he only says that it is not yet time to admit all the oxides 

 hitherto spoken of, and in which we have neither seen the 

 quantity of oxygen ascertained, nor the combinations which 

 they are susceptible of forming with the acids ; and he adds 

 that colour is not a sufficient character by which to distin- 

 guish them. 



8 There 



