The salts of sodium are abundantly and universally distributed, the most common and familiar being the chloride-common salt; the nitrate, carbonate, and sulphate also form considerable geological deposits, and the silicate occurs in many minerals. The metal, like potassium, can be prepared • by electrolysis, but less easily. A second method is by decomposing caustic soda with metallic iron at a white heat; and a third, which is that adopted on an industrial scale, consists in igniting a mixture of soda carbonate and charcoal,' the operation giving rise to no risk of explosions as is the case with potassium. In practice, 66 lb. common soda-ash is well ground up with 28 1/2 lb. slack or small coal and 6 1/2 lb. chalk, and the mixture is put into an iron cylinder 3 ft. 9 in. long, and 5 in. in diameter, coated with fire-clay; this is introduced into a reverberatory furnace, and heated to whiteness; its ends are closed by iron plates, one being traversed by a 1-in. iron gas-pipe, through which the gases and sodium vapour escape, the latter being condensed by passage through a cooling receiver exactly as in the case of potassium (p. 405, Fig. 159), and falling into a dry iron pot placed beneath, while the escaping carbon monoxide burns with a yellow flame and forms no explosive compound.
The operation is made nearly continuous by arranging the cylinders in sets in the furnace, and discharging and recharging them in turn. The actual product is only about 1/3 of the theoretical yield, owing to losses incurred by part being volatilized and burned, part adhering to the receiver, and part being imperfectly reduced. The metal thus obtained is pure enough for general use, and only needs to be remelted and cast into rods 1 ft. long and 1 in. thick; these will keep in dry air in closed vessels for a long time, becoming covered with a thin coating of oxide which preserves them from further attack; but small pieces should be stored under petroleum.
"W. P. Thompson has proposed a novel method of preparing all the alkali metals which, if successful,would greatly reduce their cost. The reducing agent used is liquid iron, either alone or in conjunction with hydrogen or carbon, the operation being performed in an apparatus resembling a Bessemer converter. In the preparation of sodium, iron mixed with an equal quantity of carbon is treated with caustic soda in the converter, and the sodium said to be formed under these circumstances is simply distilled off." (Analyst.) This does not appear to offer much advantage over Gay-Lussac and Thenard's process, already mentioned (the second method spoken of).
Metallic sodium has a silver-white colour and lustre; it is hard at -4° F. (-20° C), very ductile at 32° F. (0° C), of a waxy consistence at ordinary temperatures, semi-fluid at 122° F. (50° C), and melts into a mercury-like liquid at 204° F. (95 1/2°,C.); it oxidizes in moist air, volatilizes at a red heat, and has a sp. gr. of 0.9735 at 56° F. (13 1/2° C.); in conductivity of heat and electricity it ranks after gold, and in electro-positiveness after silver, copper, and gold; it forms with potassium an alloy which remains liquid below 32° F. (0° C), if more than 16 parts potassium are combined with 10 of sodium. It is commercially employed as a reducing agent in the preparation of other metals (aluminium, boron, magnesium); and its amalgam with mercury (see p. 12) is largely used in place of mercury alone for catching fine and dirty gold in the apparatus employed for treating auriferous ores (see Lock's'Gold').