Buchner found that sulphide of aluminium, Al_{2}S_{3}, is more suitable for the preparation of Al by the electrolytic method than Al_{2}O_{3}, but since the formation of Al_{2}S_{3} by heating a mixture of Al_{2}O_{3}, and charcoal in sulphur vapour proceeds with difficulty, Gray (1894) proposed to prepare Al_{2}S_{3} by heating a mixture of charcoal, sulphate of aluminium, and sodium fluoride. The resultant molten mixture of NaF and Al_{2}S_{3} gives aluminium directly under the action of an electric current. Aluminium has a white colour resembling that of tin--that is, it is greyer than silver and has the feebly dull lustre of tin, but compared to tin and pure silver, aluminium is very hard. Its density is 2·67--that is, it is nearly four times lighter than silver and three times lighter than copper. It melts at an incipient red heat (600°), and in so doing is but slightly oxidised. At the ordinary temperature it does not alter in the air, and in a compact mass it burns with great difficulty at a white heat, but in thin sheets, into which it may be rolled, or as a very fine wire, it burns with a brilliant white light, since it forms an infusible and non-volatile oxide. Aluminium itself is non-volatile at a furnace heat. These properties render Al a very good reducing agent, and N. N. Beketoff showed that it reduces the oxides of the alkali metals (Chapter XIII., Note 42 bis). Dilute sulphuric acid has scarcely any action on it, but the strong acid dissolves it, especially with the aid of heat. Nitric acid, dilute or strong, has no action whatever on it. On the other hand, hydrochloric acid dissolves aluminium with great ease, as do also solutions of caustic soda and potash. In the latter cases hydrogen is evolved.[38]