4.8 Analyjis of the Spinel Ruby. 



and expofed to heat depofited a white powder, which, waflied 

 and dried, weighed 17 parts, that were reduced to 8*3 by 

 calcination. They confifted of magnefia, Mill mixed with a 

 few particles of oxyd. The precipitate formed in the mo- 

 ther water by* the carbonat of potafli was treated with a fo- 

 lution of cauftic potafh. By thefe means the alumine was 

 diflblved, and the oxyd of chrome remained without under- 

 going any fenfible alteration : it weighed 5 parts. 



This analyfis perfectly confirms the former; for, fome 

 flight differences in the proportions excepted, it absolutely 

 gave the fame refults. It proves above all that the ruby 

 does not contain filiceous earth, and that the fmall quantity 

 found in the products was furniflied by the mortar in which 

 the ftone was pounded. 



I muft here obferve, that it does not always happen that 

 the whole mafs of the ruby is diflblved during the firft ope- 

 ration. There remains fometimes feveral parts which have 

 undergone no alteration. This depends on the time em- 

 ployed for ebullition, and ft ill more on the degree of finenefs 

 to which the ftone is reduced before it is fubjefted to trial. 

 But when that happens, after having waflied the refiduum, 

 which is infoluble in the acid, it muft be dried, pounded 

 again, and treated in the fame manner as before. Nothing 

 then ought abfolutely to remain but the filiceous earth be- 

 longino- to the mortar. 



XIV. The muriatic acid diflblves the ruby alfo : but a 

 confiderable portion of it is required ; becaufe it volatilifes in 

 a moderate heat, and without a ftrong degree of heat it 

 makes no fenfible impreflion on that fubftance. I obferved 

 that this acid diflblves the principles of the ruby in the fame 

 proportions as they hold between themfelves in the ftone ; 

 for the portion not attacked abfolutely preferves the fame 

 {hade of colour as that which it had before fubjected to the 

 aftion of the acid. This proves that it does not diflblve one 

 principle in preference to another, and that they are in a 

 (late of real combination in the ruby. 



3 From 



