36 J2. B. Biggs — Composition of Tourmaline. 



Methods of Analysis. 



A few words on the methods of analysis may not be out of 

 place here, that the character of the work may be the better 

 judged. 



Water. — The water was directly determined by igniting a 

 mixture of the mineral and carbonate of soda in a Gooch tubu- 

 lated crucible, the sodium carbonate being used to hold back 

 any fluorine that might otherwise be driven off. The carbonate 

 of soda used was first fused, and then, in order to ensure perfect 

 dryness, the mixture of mineral and reagent was again dried in 

 an air bath at 105° C. for two or three hours. This estimation, 

 as well as all the other more important ones, was made in dupli- 

 cate. 



Boric acid. — Where the tourmaline contained fluorine, the 

 same portion was used for determining both the boric acid and 

 the fluorine, the filtrate, from the mixed carbonate and fluoride 

 precipitates, being used for the estimation of the boric acid. 

 The borate of lime, which may be formed, is sufficiently soluble 

 in hot water so that no difficulty is experienced in bringing it 

 quantitatively into the filtrate. After evaporating this filtrate 

 to a conveniently small volume, it is brought into a retort and 

 acidified with nitric acid. The boric acid is then volatilized as 

 methyl borate, according to the Gooch method, and weighed as 

 borate of lime. It is scarcely necessary to say that throughout 

 this treatment, nitric acid should be used as the neutralizing 

 reagent, and that, if care be taken in its use, in no case need the 

 amount of salt, which is to be brought into the retort, become 

 inconveniently great. Where no fluorine is present, the soda 

 fusion may be digested with water at once, the solution, con- 

 taining the borate of soda, Altered off, neutralized and treated 

 as above indicated. One might save himself even this filtration, 

 but for the fact that, in using the whole fusion, a quantity of 

 bases would thus be brought into the retort, which, even at the 

 low heat required by the distillation, give up their nitric acid, 

 thereby rendering the determination, in its after stages, both 

 more difficult and possibly less exact. 



Fluorine. — The fluorine was estimated by the Berzelian 

 method. Though it is far from a good method, care and ex- 

 perience enable one to obtain fairly reliable results. The ten- 

 dency is toward too high results, because of the difficulty of 

 freeing the calcium fluoride from last traces of alumina and 

 silica ; and a better method will probably show even less 

 fluorine to be present in the tourmaline than the insignificant 

 quantity now found. 



Ferrous oxide. — Tourmaline, especially the varieties contain- 

 ing lithia and iron, are decomposed by acids with extreme 



