1824.] Chrysoberylsjrom Haddam and Brazil. 429 



posed, that as barytes could be brought into contact with this 

 substance more conveniently than potash at a high temperature, 

 it might decompose it. With this view, a portion of the inso- 

 luble matter was exposed to a strong heat, during one hour, with 

 six parts of nitrate of barytes in aplatina crucible. The calcined 

 mass was boiled in nitric acid. In this way nearly two-thirds of 

 the matter that could not be entirely attacked in any other way, 

 were dissolved. The same treatment was repeated, until nearly 

 the whole of it was taken up, which happened after the fourth 

 calcination. It was then no further acted on. 



After making numerous experiments on the matter that 

 resisted nitrate of barytes and nitric acid, I ascertained that it 

 was not acted on by alkalies nor acids when used separately, 

 but after having been previously calcined with caustic potash, 

 it readily dissolved in muriatic acid, yielding a solution of a pale- 

 yellow colour, which gave a reddish precipitate with an infusion 

 of galls, a deep-green precipitate with the hydrosulphate of 

 potash, and a white precipitate with alkalies. Hence it was 

 oxide of titanium. 



After the barytes was separated with sulphuric acid, the nitric 

 solutions were united, and treated with an excess of subcarbonate 

 of ammonia. An abundant precipitate ensued, which entirely 

 redissolved in the excess of subcarbonate. By ebullition it was 

 again precipitated ; and when calcined, it was in the form of a 

 light white powder, possessing all the properties that characte- 

 rise glucina. With the sulphuric and muriatic acids it formed 

 very sweet astringent deliquescent salts. By caustic potash it 

 was precipitated from its solutions, and the precipitate redis- 

 solved in the excess of the alkali. Klaproth and Arfvvedson, in 

 their analyses of the chrysoberyl from Brazil, considered the 

 insoluble matter remaining, after they had treated the mineral 

 with potash and muriatic acid, to be silica. This will explain 

 why their results differ so essentially from mine. 



After having thus satisfied myself of the composition of the 

 residue above-mentioned, I resumed my preliminary experi- 

 ments, and proceeded to examine the muriatic solution obtained 

 from the treatment of the mineral with potash and muriatic acid. 

 From this solution some silica was separated. A portion of the 

 liquid was treated with caustic ammonia, and then tested for 

 lime with oxalate of potash, but none of it could be detected. 

 To the remaining liquor a considerable excess of subcarbonate 

 of ammonia was added, and the precipitated matter was digested 

 twenty- four hours. It was then separated by filtration, and the 

 fluid was boiled till all the ammonia was expelled. No glucina 

 was thus precipitated. Hence we conclude that the very small 

 portion ol titanium above-mentioned, rendered the whole of the 

 glucina so refractory. The alumina precipitated by the sub- 

 carbonate of ammonia was mixed with a small quantity of oxide 

 of iron. It was soluble in caustic potash, and with this alkali 



