Mr Magnus's Analysis of Picrosmine. 109 



taken from the same specimen, part of which I subjected to 

 analysis. 



When exposed alone to the heat of the blow-pipe, it is in- 

 fusible, but it assumes a much higher degree of hardness. In 

 the mattrass, its colour first becomes black, then white again, 

 and it gives off water. With solution of cobalt, it gives a 

 rose-red colour, indicative of magnesia. Salt of phosphorus 

 and borax dissolve it ; a skeleton of silica being visible in the 

 former. If the mineral has been previously heated to red- 

 ness, it is almost insoluble. Treated with soda or charcoal, 

 it forms a half-vitrified mass, which is not transparent. 



In order to ascertain the composition of picrosmine, 2.144 

 grammes, reduced to an impalpable powder, and carefully 

 washed, were exposed in a platina crucible to the action of 

 fuming dilute fluoric acid. The mineral was decomposed, and 

 so much of caloric disengaged, that the fluid mass began to 

 boil. While it cooled, it was now and then stirred with a 

 small platina spoon, then some distilled sulphuric acid was 

 added, and the whole carefully evaporated to dryness, and 

 exposed to a slight red heat, in order to remove the fluo- 

 silicic and the superfluous sulphuric acids. 



What remained was dissolved in water ; a small insoluble 

 residue was left, probably of silica, but it could not be ascer- 

 tained, whether it still contained a portion of the undecom- 

 posed substance, becanse there was too little of it. It might 

 have been supposed that it was sulphate of lime, but some 

 other experiments, afterwards to be mentioned, leave no 

 doubt that the mineral does not contain a trace of lime ; 

 and I am therefore inclined to consider it as produced by the 

 decomposition, at a higher temperature, of some of the sul- 

 phates of alumina or iron. 



Pure ammonia, added to the clear solution, produced a 

 precipitate of alumina, oxide of iron, and a little manganese 

 and magnesia. The filtered fluid, which still contained the 

 greater part of the magnesia and of the manganese, yielded 

 no precipitate with oxalate of ammonia, and therefore did not 

 contain any lime. It was evaporated to dryness, and then 

 ignited, to drive oft' the remaining sulphate of ammonia, and 

 rediBSolfed in a small quantity of water. 



