214 ANNUAL OF SCIENTIFIC DISCOVEEY. 



When sodium is treated with chloride or fluoride of silicium in a tray 

 placed in a porcelain tube heated to redness, the last traces of the metal may 

 be removed ; and all that is then necessary is to wash the residue, in order 

 to obtain siiicium with all the characters attributed to it by Berzelius. But 

 if the portions which do not adhere to the tray be selected, put into a crucible, 

 surrounded and covered with pure fused chloride of sodium, and heated to a 

 sufficiently high temperature for the volatilization of the greater part of the 

 alkaline chloride, two kinds of products are obtained, which vary according 

 to the temperature and the nature of the flux. In the first place the graphit- 

 oid silicium may be produced: fused silicium is also obtained in the midst of 

 a gangue which resists the action of heat ; it is then frequently crystallized. 



Crystallized silicium has much resemblance in color with specular iron ore 

 when a little iridescent. Its form can not be exactly measured, the faces of 

 crystals being always curved ; but the form presents so close a resemblance 

 to those of the diamond, that this comparison has been made immediately by 

 all the mineralogists to whom I have shown it. In this state siticium cuts 

 glass. 



The analysis of the crystals which accompanied the specimen exhibited 

 furnished the following results : 100 silicium gave 205 of silica; calculation 

 requires 209. The small quantity of matter which was wanting also contained 

 silica and iron, but in proportions which might be neglected. Thus silicium, 

 like carbon, beside, which it has been placed in the series of metalloids, is 

 capable of assuming three distinct forms : 



1. The silicium of Berzelius, which represents ordinary carbon. 



2. Graphitoid silicium, which corresponds with graphite, and is obtained 

 under the same circumstances as artificial graphite. 



3. Crystallized sihcium, which is the analogue of the diamond. 

 Silicium, consequently, differs from the metals in every respect. 



I also exhibit some fused silicium; which has been extracted from different 

 gangues. I can not, however, state exactly either the temperature, which 

 was very high, employed in tins new experiment, or the mode of preparation 

 which is most proper for attaining a certain result. I must observe only that 

 silicium takes up iron, wherever it exists, even in vessels of common porce- 

 lain, which it corrodes in a singular manner.* In preparing sihcium. it is 

 necessary, therefore, to exhaust every precaution in the purification of the 

 original materials, particularly the sodium ; to analyze it, it is put with a few 

 drops of nitric acid into a small crucible of Sevres porcelain, and a very small 

 quantity of pure hydrofluoric acid is added (silicium, when strongly heated, 

 resists the action of hydrofluoric acid and nitromuriatic acid) ; it should dis- 

 solve entirely, and the liquid, when evaporated to dryness, should have no 

 trace of ferruginous matter. 



Silicium alloys metals, especially copper, to which it communicates a hard- 



* It reacts upon alumina, at least in the presence of bases, furnishing vitreous products 

 which appear to be new, and which I am at present engaged in analyzing. The vessels 

 which I prefer are crucibles of coke, calcined, and immersed while still hot in boiling 

 muriatic acid. After remaining for some time in the acid, and being repeatedly washed, 

 these crucibles are very good. 



