Mr Connell ati the Chemical Constitution ofBrewsterite. 37 

 chiefly small octahedrons, and mixed with some deliquescen 

 matter. 



The deliquescent matter was taken up by alcohol, which 

 left the crystals undissolved. The alcoholic solution was de- 

 canted, and converted to a watery solution, to which oxalate of 

 ammonia was added. A white precipitate fell, and the whole 

 was evaporated to dryness. The dry mass was then ignited, a 

 little solution of carbonate of ammonia added, and heat again 

 applied. Some carbonate of hme was thus got, having a trace 

 of iron. 



The nitric crystals left by the alcohol were now ignited in a 

 silver crucible, to drive off the acid. The caustic residue was 

 dissolved in very dilute muriatic acid. The muriatic solution 

 was evaporated to dryness, and the dry mass ignited, and 

 weighed. Redissolved in water, a little residue was left, the 

 weight of which was subtracted from the ignited mass. The 

 solution by evaporation gave a mixture of long prismatic, and 

 tabular crystals, having all the appearance of muriate of strontia 

 and muriate of baryta. The former were taken up by hot al- 

 cohol, to which a few drops of muriatic acid were added. The 

 latter were left undissolved ; and after being separated from the 

 alcoholic solution, and washed with alcohol, were ignited and 

 weighed, and their weight subtracted from the joint weights of 

 the ignited chlorides, by which means the amount of the chlo- 

 ride of strontium was determined. The amount of the chlorides 

 gave by computation that of the strontia and baryta respective- 

 ly. From the small quantity of matter left undissolved on the 

 solution of the chlorides in water, after their joint ignition, a 

 little silica, and also a little carbonate of strontia, were separated, 

 which were both duly taken into account. 



On recrystallizing the chlorides from watery solutions, the 

 salts obtained had the crystalline form, and all the properties of 

 muriate of strontia and muriate of baryta. The former gave to 

 flame a fine red colour, and the latter a slight greenish tinge. 

 The solution of both afibrded white precipitates with sulphuric 

 acid. 



Another analysis of 47*37 grains of the mineral was executed 

 in a manner not materially differing from the preceding process, 

 the principal distinction being, that the proportions of the alka- 



