dr - 
Analyses of the Chrysoberyls from Haddam and Brazil. 107 
mineral par sa dureté, par sa pesanteur specifique, et 
méme par le resultat de son Analyse, qui a donné environ 
72 parties d’ alumine sur 100, avec 18 de silice, et 6 de 
chaux.”* Iwas anxious to examine the cymophane found 
nd. 
Three grammes of the mineral were examined under 
the impression that Professor Klaproth’s analysis was ae- 
curately made. It was decomposed in the usual manner 
with four parts of caustic potash, and subsequently treated 
with diluted muriatic acid ; but the solution was imperfect. 
The insoluble matter was collected on a filter, and it 
amounted to 25 or 30 per 100. It was repeatedly acted 
on in the same way, and each time it diminished in quan- 
tity, until the fourth experiment. It then weighed about 
een-hundredths, and thereafter resisted all further efforts 
to render it soluble by these means. This residue was then 
boiled in coneentrated sulphuric and muriatic acids, but 
neither of them dissolved more than one-third of it. These 
solutions were tested by different re-agents, and greatly to 
my surprise, the addition of subcarbonate of ammonia oc- 
casioned a floculent precipitate, which entirely re-dissolved 
in an excess of the alkaline subcarbonate. I immediately 
suspected the presence of Glucina, but was much at a loss 
to explain its insolubility, until I observed Berzelius’s analy- 
sis of the Euclase,t in which he met with a com und of 
glucina and oxide of tin that obstinately resisted acids. 
He also met with refractory combinations of this earth and 
* Traité de Mineralogie, 2me Edition, vol. ii. p. 309. 
+ Nouveau Systeme de Mineralogie, p. 2€9. 
