64 M. Wohler on an Ailotropic Modification of Silicon. 



4. The anhydrous fluorides are remarkable for their stability ; 

 the hydrated fluorides, on the contrary, are only slightly stable, 

 and are decomposed even by being dried in racMO, hydrofluoric acid 

 being disengaged, and a residue of oxyfluoride or oxide being left. 



5. The auhydi'ous fluorides have a strong tendency to form 

 double fluorides; this property is seen even with insoluble 

 fluorides. These last compounds cannot be obtained by double 

 decomposition, as they always retain, as a double salt, part of 

 the soluble salt used in their preparation. 



6. Hydrogen does not decompose all the fluorides, even at a 

 high temperature ; it does not act on fluoride of calcium, but it 

 reduces the fluorides of lead, of tin, &c. with the greatest ease. 

 The reduction by hydrogen of metallic fluorides such as those 

 of lead, tin, &c., which resist the reducing action of carbon, 

 appears to demonstrate in a positive manner that these com- 

 pounds contain no oxygen, and are really binary compounds. 



7. All the fluorides, even those of potassium, sodium, and 

 calcium, are rapidly decomposed by the vapour of water. 



8. Oxygen and chlorine decompose fluoride of calcium at 

 furnace heat, and liberate a gas which appears to be fluorine. 



9. Vapour of sulphur does not act on fluoride of calcium, but 

 this body is completely decomposed by the vapour of sulphide 

 of carbon ; there is formed sulphide of calcium, and probably 

 fluoride of carbon : the presence of silicious substances facihtates 

 the reaction. 



10. The analyses of the principal fluorides made, such as those 

 of potassium, sodium, calcium, tin, lead, and silver, show that the 

 equivalent of fluorine, as determined by Berzelius, is correct. 



11. All the fusible fluorides are decomposed by the voltaic 

 current, and disengage a gas which appears to be the radical of 

 the fluorides. 



The author is still engaged on the subject, and promises still 

 more extensive contributions. 



In the preparation of aluminium, Deville often accidentally ob« 

 tained a dark gray crystalline substance, which was found to be 

 silicon in a form analogous to that of graphite. Wohler com^ 

 municates, in Liebig's Annalen for March, a method by which 

 this ailotropic modification of silicon may be produced at pleasure. 



Aluminium is fused in a Hessian crucible with about twenty 

 times its weight of perfectly dry silico-fluoride of potassium. On 

 cooling, the slag in the crucible is found to contain a well-formed 

 dark-coloured regulus, from which, by treatment with hydro- 

 chloric acid, a crystalline mass is obtained, consisting of about 

 70 per cent, of the aluminium used. In this case the alumi- 

 nium appears to determine the crystallization of the silicon, as 

 melting cast iron determines that of graphite. 



