514 ALUMINUM. 



deliquesces quickly in the air. When it is attempted to make 

 this salt anhydrous by heat, the chlorine goes off in the form 

 of hydrochloric acid, and pure alumina is left. 



The anhydrous chloride was discovered by Oersted, who 

 made known a method of preparing it, which has since had 

 numerous applications. Pure alumina, free from potash, is in- 

 timately mixed with carbon, in the form of lampblack, and 

 strongly calcined in a crucible. The alumina is thus made 

 anhydrous, without being otherwise altered. It is then intro- 

 duced into a porcelain tube, which is placed across a furnace 

 and exposed to a red heat. Chlorine gas, carefully dried, is 

 conducted over the materials in the tube, when, under the con- 

 joint influence of carbon and chlorine the alumina is decom- 

 posed; its oxygen is carried off by the carbon as carbonic 

 oxide gas, and chlorine unites with the aluminum itself, in the 

 place of oxygen. The chloride of aluminum, being volatile, sub- 

 limes and condenses in the cool part of the porcelain tube. A 

 glass tube, a little smaller than the porcelain tube, should be 

 introduced into this part of the latter, which may afterwards be 

 drawn out, containing the condensed chloride. The salt is 

 partly in the state of long crystalline needles, and partly in 

 the form of a firm and solid mass which is easily detached 

 from the glass. 



Chloride of aluminum is of a pale greenish yellow colour, 

 and to a certain degree translucent. In air it fumes slightly, 

 diffuses an odour of hydrochloric acid, and runs into a liquid by 

 the absorption of moisture. It is very soluble in water, but 

 cannot again be recovered in the anhydrous condition. It is 

 equally soluble in alcohol. Chloride of aluminum combines 

 with sulphuretted hydrogen, phosphuretted hydrogen, and also 

 with ammonia. The latter compound Persoz finds to contain 

 27.61 per cent of ammonia,= A1 2 C1 3 + 3NH 3 . 



The fluoride of aluminum can only be obtained by dissolving 

 pure alumina in hydrofluoric acid ; it does not crystallize. This 

 fluoride unites in two proportions with fluoride of potassium, 

 for which it has a strong affinity. Both the compounds are 

 gelatinous precipitates, which become white and pulverulent 

 after being washed and dried. Berzelius assigns to them the 

 formulae, 3KF + A1 2 F 3 and 2KF+AL, F 3 . Fluoride of alumi- 

 num exists in two crystalline minerals, one of which, on account 



