458 Sir Henry JRosfoe [May 3, 



iron if it be present as ferric salt, or a very large percentage if it is 

 in the ferrous condition. Even when exercising all possible precau- 

 tions, the average analysis of the crude double chloride shows about 

 • 4 per cent, of iron. The metal subsequently made from this chloride 

 therefore never contained much less than about 5 per cent, of iron, 

 and, as this quantity greatly injures the capacity of aluminium for 

 drawing into wire, rolling, &c., the metal thus obtained required to 

 be refined. This was successfully accomplished by Mr. Castner 

 and his able assistant Mr. Cullen, and for some time all the metal 

 made was refined, the iron being lowered to about 2 per cent. 



The process, however, was difficult to carry out, and required 

 careful manipulation, but as it then seemed the only remedy for 

 efioctively removing the iron, it was adopted and carried on for some 

 time quite successfully, until another invention of Mr. Castner 

 rendered it totally unnecessary. This consisted in purifying the 

 double chloride before reduction. I cannot now explain this process, 

 but I am able to show some of the product. This purified chloride, 

 or pure double chloride, is, as you see, quite white, and is far less 

 deliquescent than the crude, so that it is quite reasonable to infer 

 that this most undesirable property is greatly due to the former 

 presence of iron chlorides. I have seen large quantities containing 

 upwards of Ih per cent, of iron, or 150 lbs. to 10,000 of the chloride, 

 completely purified from iron in a few minutes, so that, whilst the 

 substance before treatment was wholly unfit for the preparation of 

 aluminium, owing to the i^resence of iron, the result was, like the 

 sample exhibited, a mass containing only 1 lb. of iron in 10,000, 

 or O'Ol per cent. The process is extremely simple, and adds little or 

 no appreciable cost to the final product. After treatment, this pure 

 chloride is melted in large iron pots and run into drums similar to 

 those used for storing caustic soda. As far as I am aware, it was 

 geneially believed to be an impossibility to remove the iron from 

 anhydrous double chloride of aluminium and sodium, and few if 

 any chemists have ever seen a pure white double chloride. 



Aluminium Manufacture. 



I now come to the final stage of the process, viz., the reduction of 

 the pure double chloride by sodium. This is effected, not in a tube 

 of Bohemian glass, as shown in Mr. Barlow's lecture in 1856, but iu 

 a large reverberatory furnace, having an inclined hearth about 6 feet 

 square, the inclination beinjj towards the front of the furnace, through 

 which are several openings at different heights. The pure chloride 

 is ground together with cryolite in about the proportions of two to 

 one, and is then carried to a staging erected above the reducing 

 furnace. The sodium, in large slabs or blocks, is run through a 

 machine similar to an ordinary tobacco-cutting machine, where it is 

 cut into small thin slices ; it is then also transferred to the staging 

 above the reducing furnace. 



