196 G. HARKER. 



transparent rolled pebbles of from one quarter to one eighth of an 

 inch in diameter. The method of analysis first used was that of 

 Wohler, as given in his Mineral Analysis, English edition of 1871. 



" The substance when fused with four times its weight of 

 anhydrous carbonate of soda is decomposed with formation of 

 fluoride of sodium, which is extracted with water. Before filter- 

 ing oft" the residual silicate of alumina, the solution should be 

 digested with carbonate of ammonia to precipitate any small 

 quantities of alumina and silica which may have been dissolved." 

 The residue is evaporated down with hydrochloric acid, and the 

 silica and alumina extracted. The fluorine in solution is removed 

 as calcium fluoride by adding calcium chloride. 



This method when tried on the topaz crystals from Mudgee gave 

 results for fluorine which were low and irregular, viz., from five 

 to eleven per cent. On treating some Brazilian topaz later in the 

 same way, similar unsatisfactory results were obtained, viz. 4*7 

 to 12-9 per cent. 



Second or tetr a- fluoride method. — That of Liversidge as given in 

 Crookes' Select Methods, p. 580, 1884 edition. 



In this method the substance is decomposed with sulphuric acid 

 and silica and the silicon tetra-fluoride passed into solution of 

 ammonia. The mixture is heated in a platinum retort first over 

 the water bath and finally at 160° C, the last traces of gas being 

 removed by a current of air. The ammonia solution is evaporated 

 until the gelatinous silica passes into solution and is then pre- 

 cipitated as protassium silicon fluoride by potassium chloride and 

 alcohol. 



So far this method had only been used for the determination of 

 fluorine in coprolites and apatite. In order to make the method 

 applicable to topaz, which is not decomposable by sulphuric acid, 

 the topaz was first fused with sodium potassium carbonate, silica 

 was then added and the mixture decomposed with sulphuric acid. 



Preliminary trials were made with calcium fluoride, apatite, and 

 cryolite, the substance in each case being fused with two or three 



