R. B. Riggs — Composition of Tourmaline. 35 



Aet. III. — The Analysis and Composition of Tourmaline ; 

 by E. B. Riggs.* 



Apaet from the work by Rammelsberg (Pogg., lxxx, 449, 

 lxxxi, 1, cxxxix, 379, 547), very little has been done toward 

 solving the question of the composition and constitution of the 

 varieties of tourmaline. Their apparent complexity and the 

 difficulties attending the determination of certain constituents 

 have possibly turned many aside, possibly also the impression 

 that with Rammelsberg's investigations the matter was settled. 

 While Rammelsberg's work was comprehensive and was good 

 for the times, his analyses are so seriously faulty in certain 

 important respects, as to justify a new investigation. Though 

 the direct estimation of both water and boric acid would seem 

 to be of the highest importance, before any satisfactory conclu- 

 sions could be reached with reference to the constitution of 

 tourmaline, we find that, having failed in one single attempt to 

 determine the water directly, he falls back on the loss on igni- 

 tion, deducts therefrom an amount equal to the amount of sili- 

 con tetrafluoride, representing the fluorine found in the mineral, 

 and calls the balance water. He takes it for granted that the 

 fluorine is driven off quantitatively. But while this supposition 

 is questionable, it is not the ground of objection. In the re- 

 vision (Pogg., cxxxix, 379) of his earlier work, Rammelsberg 

 comes to the conclusion that the iron contained in tourmaline 

 is all there in the ferrous condition, yet wholly ignores the fact 

 of its possible oxidation on ignition, especially an ignition such 

 as would be necessary to expel the fluorine. In a few cases 

 boric acid is determined directly, but by a method (Stromeyer's) 

 which has ever been counted one of the most unsatisfactory. 

 In the majority of the analyses it is estimated by difference. 

 But if the results called water are incorrect and low, as they 

 surely are, the boric acids ought to be correspondingly high or 

 the analysis must be elsewhere at fault. 



The direct estimation of water being possible, and a satisfac- 

 tory method for determining boric acid having lately been de- 

 vised by Dr. F. A. Gooch, (Am. Chem. Jour., Feb., 1887), 

 new tourmaline analyses seemed desirable. Through the kind- 

 ness of many, abundant and varied material has been at my 

 disposal. 



* Ad abstract of a paper which is to appear in a forthcoming Bulletin of the U. 

 S. Geol. Survey. 



