PROCESSES FOR THE CONVERSION OF CARBON INTO SILICON. 



555 



for two hours, at a full white heat, was found by him to yield silica, when treated 

 with muriatic acid in the manner already referred to. " The weight of silicon was 

 never less than an llth, and never more than a 12th, under the calculable weight 

 of the constituent carbon."* 



We have repeated this experiment many times, both with crude and purified 

 paracyanogen, and in the greater number of cases have obtained silica. In several 

 trials, however, and even with paracyanogen prepared by that gentleman himself, 

 we obtained no silica ; in others, the quantity from paracyanogen of our own pre- 

 paring was very small ; and, in all, it was much more than an llth under the weight 

 of the constituent carbon. The exact quantities of silicon we procured are given 

 in the subjoined table, from which it will be seen that, whereas the proportion of 

 silicon obtained by Dr BROWN was between 91 and 92 per cent, of the weight of 

 the carbon, the proportion we obtained, even from purified paracyanogen, was 

 never more than 15, and sometimes less than 1 per cent. 



We endeavoured to make this a quantitative process, by collecting and exa- 

 mining the gases evolved during fusion. According to Dr BROWN'S views, para- 

 cyanogen, when fused with carbonate of potash, should give off all its nitrogen in 

 the gaseous form, and have all its carbon transmuted into silicon. The latter 

 should then become silica at the expense of the carbonic acid, reduced thereby to 

 carbonic oxide; while the silica and caustic potash contemporaneously evolved 

 should unite to form silicate of potash. There should thus be calculable volumes 

 of carbonic oxide and nitrogen, as well as a calculable weight of silica produced. 

 And all three should agree (within the limits of error in experiment) with the anti- 

 cipated numbers, if a perfect transformation of paracyanogen into silicon and nitro- 

 gen occurred. 



We made this experiment several times, but always found carbonic oxide and 



* Transactions of the Royal Society, vol. xv. p. 236. 

 accidentally inverted, an llth being more than a 12th. 

 VOL. XV. PART IV. 



The numbers, we suppose, have here been 

 7L 



