596 MINERALS AND VEINS OF ^EGIKINE-SYENITE, 



undetected by a careful qualitative test, was proved to be present 

 in notable quantities in several different samples. The estimates 

 were carefully checked byconversion into sulphate and reweigh- 

 ing; two determinations gave 140 % and 0-80 %. Finally, 

 specially fresh material was selected and subjected to a pre- 

 liminary treatment with strong sulphuric acid to decompose any 

 fluorspar which might be present as impurity; on analysis only 

 018 % fluorine was then returned. Water above 110°C was very 

 constant. Due care was taken to eliminate any possibility of 

 contamination with fluorine by the introduction of a granulated 

 fused lithage plug before the calcium chloride weighing tube. 



Development of a formula. — Grouping the vicarious constituents 

 into like sets offers no special case .for consideration, it being 

 evident that titania replaces silica and that the combined water 

 is to be taken with the alkalies. Fluorine might be correctly 

 referred to a combination with the monoxide bases, but as this 

 is somewhat uncertain, the fluorine contents will be overlooked 

 in deducing a formula, its small bulk having very little appreci- 

 able effect on the results. 



Referring to the molecular ratios calculated from the analysis, 

 a fairly simple empirical formula can be deduced, namely 

 i n in 

 K 5 R i H(Si0 3 ) 8 



i ii in 



in which R is chiefly Na, R mainly Fe, and R practically all Fe. 



For a theoretically pure arfvedsonite this formula would read 



5Na 2 0. 8FeO. Fe 2 3 . 16Si0 2 



or stated as a semiconstitutional formula 



in 

 4NaoSiO :3 . 8FeSi0. 3 . 2NaFe(Si0. 3 ) 2 



or 2(NaFe) 2 (SiO :3 ) ;3 . NaFe(Si0 3 ) 3 



Comparison with former Analyses. — The physical properties of 

 the mineral here described appear to answer exactly to those of 

 arfvedsonite; notwithstanding the slight discrepancies in analyses 

 by comparison with that from Kangerdluarsuk, there can be no 

 doubt as to its species. The crystals analysed from Bowral were 

 most of them perfectly formed, and, being smaller than the Green- 



