4(50 



MESSRS. R. S. BUTTON AND J. E. PETAVEL ON ELECTRIC 



TABLE IX. (1). Typical Experiment on the Reduction of Alumina by Carbon (G 33). 



The large furnace enclosure was used in the vertical position. 



Iron liner filled as shown in fig. 11, A. 



Lower electrode, bed of 2 kilogs. retort carbon. 



Upper electrode, a 41-millim. carbon rod 32 centims. long ; weight with holder, 1650 grammes. 



Charge: about 5500 grammes calcined alumina (Brit. Al. Co.), mixed with 1130 grammes petroleum 

 coke (60-mesh). 



Initial pressure, 20 atmospheres coal gas. 



Gaseous products of reaction washed out with coal gas (see p. 442), the gas being let off about every 

 10 minutes until the pressure had fallen to 15 atmospheres, a fresh supply of coal gas being at once forced 

 in, raising the pressure to 25 or 30 atmospheres. 



TABLE IX. (2). 



TABLE IX. (3). 



The total volume of coal gas used during the experiment was 800 litres, and about the same volume 

 during the subsequent cooling period. 



Product : central fused lump surrounding the vertical electrode, to which it adhered ; had a weight of 

 about 1560 grammes, and had the appearance of fused alumina containing some aluminium carbide. 



The fused lump contained 228 grammes of aluminium carbide and 36 grammes of aluminium, some 

 of which was found in the form of plates after crushing. 



The outside fritted material, about 1125 grammes in weight, contained a further 9 grammes of aluminium 

 in large pieces. No reduced material was found in the outer layers of the charge. 



Total yield was therefore 228 grammes aluminium carbide 45 7 grammes aluminium. 



