New Substance discovered in Davidsouile. 427 



Silica 66"59 



Alumina . : . . . 32-12 



Water 1-30 



Calculating the atomic proportions of these constituents, 

 he found it to he, 



Silica 24 atoms 



Alumina .... 1 atom* 



The extreme improbability of the existence of such a 

 compound, induced him to request me to make another 

 analysis of this mineral. 



For this purpose, 30 grains of the mineral were reduced 

 to a fine powder, and fused with 60 grains of carbonate of 

 soda. The fused mass was treated with dilute muriatic 

 acid, and the silica separated in the usual way. It was 

 collected on a filter, washed and dried, and found to weigh 

 13"49 grains, or 67*45 per cent. 



The licpiid which passed through the silica filter, was 

 concentrated on the sand bath, and nearly neutralized with 

 carbonate of ammonia. Caustic ammonia was then added 

 in excess. A white precipitate appeared, which was sepa- 

 rated by a filter, and washed. During the evaporation of 

 the washings, a copious deposition of white flocks took 

 place, as the ammonia was disengaged by the heat. This 

 precipitate was dissolved in muriatic acid, and the evapo- 

 ration continued. When sufficiently concentrated, oxalate 

 of ammonia was added, but no precipitate appearing, the 

 liquid was evaporated to dryness, and the dry mass heated 

 to redness. The residue was dissolved in very dilute mu- 

 riatic acid, and the solution boiled in a flask, with carbo- 

 nate of soda. A white flocky precipitate appeared, the 

 colour of which, during the boiling, changed to brown. — 

 It was separated by a filter, and washed. 



The precipitate which was obtained by caustic ammonia, 

 though quite white when originally thrown down, had, 

 during the washing, gradually acquired a brown colour. 

 This precipitate being supposed to contain peroxide of iron, 

 it was dissolved oft" the filter with weak muriatic acid, and 

 boiled for some time with caustic soda. The soda threw 

 down a white precipitate, which, after a short boiling, al- 

 most entirely disappeared, leaving only behind a small quan- 

 ' Thomson's Mineralogy, i 847. 



