COMPOSITION OF N.S.W. LABRADORITE AND TOPAZES. 199 



Third method, fusion with alkaline carbonates and silica. — A. 

 third method for the determination of fluorine was then tried, viz. 

 that of Berzelius as modified by Rose. The ground topaz is fused 

 with one half its weight of pure silica ; the fused mass extracted 

 with water ; five to ten grams of ammonium carbonate added to pre- 

 cipitate silica, and the solution allowed to stand twelve hours. The 

 solution is filtered and evaporated to get rid of excess of ammonia, 

 and a solution of zinc carbonate in ammonia added to precipitate 

 any silica in solution. The liquid is then further evaporated to 

 get rid of ammonia, filtered from zinc oxide and silicate and the 

 fluorine precipitated by calcium chloride in the usual manner. 



The results from the Mudgee topaz were much higher than those 

 obtained by the two previous methods and were constant ; the 

 same constant results were obtained in the case of the Brazilian 

 topaz and the green topaz from New England. 



The calcium fluoride obtained by precipitation in this method 

 from the Mudgee topaz was tested and found to be free from 

 calcium silicate. Berzelius' method therefore gave the best results 

 for fluorine in topaz, but as there seemed to be no reason why the 

 fluorine should be lost in Wohler's method which did not apply 

 equally to that of Berzelius', an attempt was made to find where 

 the fluorine was lost in the former case. 



It was found that no fluorine is left in the residue after extract- 

 ing the fused mass with water. The decomposition in Wohler's 

 method is therefore complete. It was noticed that the precipitate 

 of aluminium hydroxide obtained on addition of ammonium 

 carbonate to the solution was more granular than the usual 

 gelatinous precipitate, and on trial this precipitate was found to 

 contain no silica but always more or less fluorine. (Hydrofluoric 

 acid was liberated from the precipitate by sulphuric acid and 

 etched glass. Mixed with silica and sulphuric acid, silicon tetra- 

 fluoride was given off which was passed into water and the 

 potassium silicon fluoride precipitated.) On one occasion 6-6 per 

 cent, of fluorine was obtained in this precipitate by Berzelius' 

 method, but this was exceptionally large, the usual amount not 



