78 Intelligence and Miscellaneous Articles. 



ON THE CRYSTALLIZATION OF SILICA IN THE DRY WAY. 

 BY T. HAUTEFEUILLE. 



In 1SGS M. \ oin Rath described, in a trachyte, some small lamelli- 

 form crystals, oL ! which he made a new species, tridymite. The 

 observations of that accomplished mineralogist prove that tridymite 

 is silica crystallized under a form different from that of quartz, and 

 possessing a lower density than that species. Like sulphur, arse- 

 nious acid, flowers of antimony, &c, silica crystallizes under two 

 incompatible forms. It can be prepared, as I shall prove, in the 

 dry way, under both these forms. 



The only known process for crystallizing silica in the dry way we 

 owe to M. GJ-. Rose ; it is based on the employment of the phos- 

 phorus-salt, and permits only tridymite to be prepared. The alka- 

 line tungstates can, with advantage, be substituted for the phos- 

 phates ; for they permit crystallized silica to be obtained, at pleasure, 

 either under the form of tridymite or under that of quartz. 



Reproduction of Tridymite. — Amorphous silica, kept at the melt- 

 ing temperature of silver in tungstate of sodium, crystallizes in a 

 few hours. After the cooling, treatment with water dissolves the 

 alkaline tungstate and lays bare a crystalline sand, the weight of 

 which is, within a few thousandths, that of the silica employed. 



The principal crystallographic and optical characters of tridymite 

 are easily verified in the crystals obtained by this process. They 

 are thin hexagonal scales, mostly piled one upon another to the 

 number of three or four*. Upon the most regular scales one or two 

 half-scales are frequently lodged. 



By the long-continued action of tungstate of sodium at a nearly 

 constant temperature of 1000° C, the tridymite can be obtained in 

 thick scales. These crystals, mixed with large lamella? grouped 

 according to one of the two laws alluded to, are hexagonal tables 

 with faces free from strisef. 



A pencil of parallel-polarized light is not depolarized when it 

 passes quite perpendicularly through these hexagonal scales, what- 

 ever the thickness. 



The density of the crystals prepared in platinum vessels, with 

 pure tungstate of sodium, and with silica containing neither alu- 

 mina, nor oxide of iron, nor magnesia, is 2-30 at 16° C. ; Vom 

 E-ath, operating on some crystals containing about 2 per cent, of 

 oxides, found 2-326, 2-312, and 2-296 for the density at 16°. The 

 determination made upon an absolutely pure product proves that 

 silica in the form of tridymite possesses a density certainly inter- 

 mediate between those of quartz (2-65) and fused silica (2-20). 



Tridymite is more readily attacked than quartz by reagents both 

 in the wet and the dry wa} r . Tungstate of sodium itself can 

 destroy the tridymite. Thus, at a temperature much above 1000°, 



* These scales are often partly corroded, like those met with in the tra- 

 chytes of Mont Dore. 



t The ratio - appears to be constant and equal to -. 



