CHEMICAL SCIENCE. 213 



nitrate of potash is added to the fused mixture in the crucible. Is not 

 this, therefore, an indication that the oxide of copper itself has under- 

 gone a change? The chemical action here is somewhat obscure, but 

 it is doubtless one of oxidation. It is evident that the potash is not 

 the effective agent ; for if, after adding the nitrate to the mixed arseni- 

 ates, the heating be long continued after the effervescence has ceased, 

 the compound again takes its original bluish-green color, a change 

 which can only be here traced to the liberation at an increased temper- 

 ature of the oxygen which had, in the early stages of the process, 

 produced the characteristic fine blue color. The question then is, Does 

 an oxide of copper exist containing a larger proportion of oxygen than 

 that forming the base of the ordinary green salts of that metal] Com- 

 bined with arsenic acid, the superoxide seems to be stable at common 

 temperatures, but easily reducible to a lower degree by exposure to a 

 red heat, allowing the excess of oxygen to escape in a free state. 

 The double arseniate of copper and potash, when placed in contact 

 with water, is decomposed, and the arseniate of copper, on account of 

 its insolubility, may then be readily separated. From its beautiful blue 

 color, it can, no doubt, be rendered extensively useful as a pigment. 

 London Mining Journal. 



ON A MINERAL BLUE COLOR EMPLOYED BY THE ANCIENTS, 

 WITH DIRECTIONS FOR ITS PREPARATION. 



M. GIRARDIN has communicated to the French Academy a paper on 

 the analysis and manufacture of a mineral blue paint, found in a Gallo- 

 Roman villa, in the forest of Bretonne, in Normandy. 



Several kilogrammes of this blue color were found in an earthen 

 jar, in the state of friable concretions, which had evidently been a fine 

 powder. It had no taste, w T as insoluble in water, but effervesced vio- 

 lently on contact with acids ; 100 parts gave out 15.50 of carbonate 

 of lime, with traces of oxide of iron, when treated with hydrochloric 

 acid. After this treatment, the insoluble powder remaining had all 

 the appearance of artificial ultramarine ; it resisted the most powerful 

 heat, and was neither fused nor altered in color. The most powerful 

 acids had no action whatever on it ; it was scarcely acted upon by ni- 

 tro-muriatic acid ; but when heated to redness with several times its 

 weight of caustic potash, it fused, and on cooling presented a mass of 

 a sombre green color, for the most part soluble in hydrochloric acid. 

 No trace of cobalt was found. A quantitative analysis gave the fol- 

 lowing results : Silica 49.4, alumina 6.4, lime, with traces of mag- 

 nesia and iron, 19.4, soda 15.5, oxide of copper 9.3 ; total, 100.0. 



This blue substance is, therefore, a glass, colored by oxide of cop- 

 per, in all respects analogous to the cccruleum of Vitruvius, or the 

 Alexandrian glaze which the Roman artists employed for fresco-paint- 

 ing, and the decoration of apartments. 



Chaptal, in 1809, made a qualitative analysis of a color of the same 

 kind, which was found in the shop of a color-dealer in Pompeii, and 

 Descotils has subsequently recognized the same copper color in the 

 hieroglyphical paintings in an ancient Egyptian monument. 



