38 Mr Connell on the Chemical Constitution of Brewsterite. 



line earths were determined on a separate quantity of the mine- 

 ral of ^5'62 grains, a portion of them from the original quantity 

 having been accidentally lost. 



The manner in which the analytic process was conducted af- 

 forded sufficient evidence that the alkahne earths did not exist 

 in the mineral in combination with sulphuric acid. But to as- 

 certain with still greater precision whether the mineral contained 

 any traces of sulphates, a little of it, in powder, was fused with 

 carbonated alkalis, the mass treated with hot water, the solution 

 deprived of silica and alumina by carbonate of ammonia, and 

 saturated with muriatic acid. No precipitate was then got with 

 muriate of baryta. 



On treating a similar solution, to which ammonia had been 

 added, to take off the excess of muriatic acid, with muriate of 

 lime, and keeping it for a few days in a close vessel, some pre- 

 cipitate which had formed seemed to be carbonate of lime, with 

 traces perhaps of silica or alumina. There was thus no evidence 

 of phosphoric or fluoric acids. 



To ascertain whether the mineral contained any alkali, 33-11 

 grains in fine powder were strongly ignited with six times their 

 weight of carbonate of baryta. The usual steps were taken to 

 separate silica; and the earthy contents of the muriatic solution 

 were thrown down by carbonate of ammonia, its action being 

 aided by heat ; the whole evaporated to dryness ; the residue re- 

 peatedly redissolved ; and new additions made of carbonate of 

 ammonia. The muriate of ammonia was then driven off by heat. 

 No trace of an alkaline chloride could be found. 



The two analyses which have been mentioned corresponded 

 perfectly with one another, in respect of the nature of the con- 

 stituents of the mineral ; and, although in the one, some loss, 

 and in the other, some excess, was got *, yet they did not differ 

 essentially in the proportions of the constituents in relation to 

 one another. By taking a mean of the two analyses, the excess 



• The excess was, I believe, owing to the circumstance, that, in heating 

 the mineral, after being reduced to powder, and before weighing, with the 

 view of expelling any hygrometric moisture, the heat was raised a little too 

 high, by which means, some of the chemically combined water was expelled. 

 The loss in the other analysis, was partly owing to the circumstance that se- 

 parate portions of the mineral were used in the course of it. 



