DR BROWN ON THE PRODUCTION OF SILICON FROM PARACYANOGEN. 241 



by fusing carbonate of potassa in them, and discovering not a trace of silicic 

 acid in the salt ; and, even with the former, it is easy to remove the possibility 

 of error by producing, in successive operations, more siliciuret in each than its 

 own weight. 



It is evident that the ferrocyanide may be substituted for the paracyanide in 

 this formula, the cyanide being sublimed away by the heat of the furnace. In 

 one crucible of the capacity of an ounce, and weighing 500 grs., one operation 

 produced 165 grs. of the semicrystalline siliciuret of iron, and the third repetition 

 raised the weight above that of the crucible itself. By working a strong iron 

 tube, like that Avhich is figured in the introduction, several times, and then re- 

 moving the iron, partly by oxidation and partly by solution in acid, there was ob- 

 tained 234 grs. 



The analytical method by which this product was examined, consisted in re- 

 ducing it to a state of fine division in a steel mortello, rubbing it up with a large ex- 

 cess of pure carbonate of potassa, and fusing the mixture in a platinum crucible ; the 

 silicate of potassa, having been dissolved away and filtered, was then decomposed 

 by an excess of hydrochloric acid, and the silicic acid was separated by desicca- 

 tion, ignition, and elutriation ; and the constituent silicon was calculated for. 

 The very same operation, performed in an iron crucible, separated the iron in a 

 peculiar condition which will be explained at the end of the next section. The 

 average result of the application of this double process to the following products 

 was, that they are all of the same composition, and contain 28.5 per cent, of silicon, 

 the remaining weight being iron ; and the inference is that they are disiliciurets, 

 as might have been divined from the composition of the paracyanide and the 

 conclusion regarding the relation of silicon to paracyanogen, which has been al- 

 ready stated. 



2. The best way to procure the semicrystalline product of the foregoing pro- 

 cess is this : Mix ferrocyanide of potassium with its own bulk of cyanide of potas- 

 sium, and treat it exactly in the same manner, only for thrice as long a time ; the 

 supernumerary cyanide acts as a non-reactive flux, and is ultimately driven off 

 by the fire. In this case there is no amorphous siliciuret, but the crucible, or iron 

 tube, is lined with a fine smooth cake of the semicrystalline substance, to the 

 depth of a line and a quarter for every operation, and it may be repeated till the 

 vessel be nearly full. It adheres to the porcelain ; but, if the crucible be broken 

 into fragments, it may be picked off with a knife. It is stratified, and readily 

 separates into layers. Jetty as it is in mass, when pulverized it is a light-coloured 

 powder, of a greenish-grey hue, being the whiter the more finely it is triturated ; 

 but, whenever it is immersed in water, or moistened by any other liquid, it re- 

 sumes the dark appearance of the mass. When broken down and examined by 

 the microscope, it is seen to be perfectly transparent ; and in fine powder, the 



VOL. XV. PART I. 3 T 



