COLOURS FOUND AT POMPEII. 303 



dows of some houses. It is produced by the calcination of 

 the yellow ochre just mentioned. 



No. 4 is a pumice stone, very light, and very white. It is Pumice stone. 

 of a fine and close grain. 



The other three are compound colours, which I have 

 been obliged to analyse, in order to know their constituent 

 principles. 



The first of these, No. 5, is a fine, deep, and mellow A bine coin- 

 blue. It is in small pieces of similar form. The outside P omu • 

 of each piece is a paler blue than the inside, the colour of 

 which is more bright and lively than that of the finest ver- 

 diter. 



Muriatic, nitric, and sulphuric acids, produce a slight Treated with. 

 effervescence with this colour. The}- appear to brighten aci s > 

 it, even with long boiling. Oximuriatic acid has no action 

 on it. It differs therefore from ultramarine, which is de- 

 stroyed by these four acids, as Clement and Desormes ob- 

 served. 



Ammonia has no action on it. ammonia, 



Exposed to the flame of the blowpipe it grows blackish, and the blow- 

 and the continued action of the flame converts it into a ! ;1 l' e - 

 reddish brown frit. With borax it fuses before the blow- 

 pipe into a greenish blue glass. Treated with potash' on a 

 stand of platina it produces a greenish frit, which after- 

 ward becomes brown, and at length assumes the metallic 

 colour of copper. This frit is partly soluble in water. Mu- 

 riatic acid poured into this solution produces a copious 

 flocculent precipitate ; and the liquor decanted from this 

 precipitate yields another in considerable quantity with oxa- 

 late of ammonia. 



Nitric acid dissolves with effervescence the residuum, 

 which the alkali could not dissolve, and the solution is 

 green. Ammonia produces in this solution a precipitate, 

 which it redissolves when added in excess, and then the so- 

 lution becomes blue. 



This colour then appears to be composed of oxide of cop- Its composi- 

 per, lime, and aluinine. It approaches to verditer in the tl0n " 

 nature of its principles, but differs from it in its chemical 

 properties. It appears to be the result not of a precipita- 

 tion, 



