446 MESSRS. R S. BUTTON AND J. E. PETAVKF. OX KLECTRIC 



tube fuses within the first few minutes ; the power can then be rapidly increased, arid 

 the temperature of formation of carborundum is soon attained. 



There is little doubt that the production of carborundum is the result of the 

 interaction of the vapour of silica and the highly heated granular carbon. From the 

 previous work on quartz we were therefore led to anticipate that under pressure the 

 reaction would not occur very readily. 



Several experiments confirmed this impression. An examination of the furnace 

 after the run showed, in every case, that, though the mixture of sand and carbon had 

 attained a sufficiently high temperature to effect the fusion of the quartz to a 

 considerable depth, thus agglomerating the mixture, only a small quantity of 

 carborundum was formed and that immediately around the central core. 



Another distinctive feature of the pressure experiments is the almost entire absence 

 of the amorphous variety of carborundum. The reaction progresses at a slow rate, 

 but apparently uniformly, the resulting pressure being, as in the case of calcium 

 carbide, a linear function of the time (see fig. 12). 



In order to study more in detail the progress of the reaction, an experiment was 

 planned in which the whole of the gas generated was retained in the furnace. 



As is shown in Table VII., analyses were made at intervals both during the run 

 and in the subsequent cooling period. From these it is clear that in this case there 

 is no inverse reaction. The high absorbing power of carbon for carbon monoxide as 

 compared with hydrogen fully accounts for the slight decrease in the percentage of 

 the former.* 



ON THE DIRECT REDUCTION OF ALUMINA BY CARBON. 



The methods used in practice for obtaining aluminium from its ores are indirect and 

 inefficient. 



The preparation involves a lengthy and complicated purification of the oxide, 

 followed by its electrolysis in a bath of cryolite. Early work showed that where 

 aluminium alloys are required they could be obtained by a simple method involving 

 the reduction of alumina by carbon, but the process has never been successful for 

 the production of the pure metal. Up to the present time opinion seems to be 

 divided as to the effect of heating alumina and carbon together in the electric 

 furnace. 



Several authorities definitely state that alumina is irreducible by carbon.t whilst 

 others affirm that it is quite easily reduced.^ 



* DEWAK, 'Roy. Soc, Proc.,' 1904, vol. 74, pp. 122-127. 



t W. HAMPE, ' Chemiker-Zeitung,' 1888, 12, 391. S. A. TUCKER and H. R. MOODY, 'Journ. Soc. 

 Chem. Ind.,' 1901, 20, 970. 



J COWLKS see W .P. THOMPSON, 'Journ. Soc. Chem. Ind.,' 1886, vol. 5, p. -'06; W. BOKCHERS, 'Elektro- 

 Metallurgie,' 3tc Aufl., 1903, p. 102. 



