318 J\ T . jy. Evans — Chry sober yl from Canada. 



powder was fused in a platinum crucible for several hours 

 with a large quantity of potassium disulphate, the fusion was 

 extracted with dilute sulphuric acid, and the insoluble material, 

 apparently silica, was filtered out and weighed. It amounted 

 to a few tenths of a per cent and was deducted from the weight 

 of mineral originally dissolved, the percentage of the other 

 constituents being calculated to the remainder. The iron, 

 alumina and beryllia were then precipitated with ammonia, 

 the precipitate washed and redissolved in hydrochloric acid, 

 the solution being then evaporated in a porcelain dish until 

 crystallization just commenced. To this an excess of a strong 

 solution of sodium hydrate, prepared from the metal in plati- 

 num, was added ; the porcelain dish was used here, as several 

 experiments showed that platinum dishes were attacked to a 

 considerable extent. The mixture was diluted somewhat with 

 cold water, and filtered ; the precipitate w r as redissolved in 

 hydrochloric acid and subjected to the same treatment as 

 before, the nitrate obtained being added to that previously 

 obtained. The precipitate was redissolved, reprecipitated with 

 ammonia and treated in the usual way, giving the iron. The 

 combined filtrates from the two precipitations by sodium 

 hydrate were poured into a liter of boiling water, the boiling 

 continued one hour, and the perfectly white beryllia filtered 

 out, washed, ignited and weighed. The filtrate from the 

 beryllia was boiled again for a short time to make sure that 

 the precipitation of beryllia was complete, was then slightly 

 acidified with hydrochloric acid, evaporated to a small bulk, 

 and the alumina precipitated with ammonia and determined in 

 the usual way. Tests were applied for calcium and magnesium, 

 but neither was found. 



The two analyses agreed very closely and the mean was as 

 follows : 



BeO... 17'78$ 



A1 2 3 „ 76-76 



FeO 6-07 



100-61 



The specific gravity of the mineral was 3*52. 



The iron was all calculated as ferric, and this probably 

 accounts in part for the results being high, but no method 

 could be devised for determining the ferrous iron present. 

 No doubt also small quantities of material were dissolved from 

 the dishes and beakers used ; their use, however, seemed 

 unavoidable, as platinum was attacked to such a marked extent. 



Macdonald Chemical Department, 

 McGill University, Montreal. 



