230 M. Rose on those Minerals which possess [March, 



obtained from all magnesian minerals, as, for example, ser- 

 pentine. 



It was, however, necessary to ascertain, whether the whole 

 loss in the fire consisted of water. A quantity of another speci- 

 men was, therefore, distilled, which, in different experiments, had 

 lost 3*09, 2*99, and 3*25 per cent, on a red heat, in a porcelain 

 retort with a receiver, to which was adapted a tube with muriate 

 of lime. This experiment, which was made with the greatest 

 care, was decisive, for no gas was given out, and the loss of 

 weight of the retort answered completely to the increase of the 

 receiver, and the tube with muriate of lime.* 



Fragments of the specimen which had served for this experi- 

 ment, were employed for two different analyses. The frao-ments 

 for the first analysis were taken from one end of the large piece ; 

 those which served for the second were from the opposite end. 

 The result of these two analyses was : 



Silica 58*08 containing oxygen 29-21 



Lime...... 11-24 3-16 



Magnesia with manga- 

 nese 22-28 8-62 



Protoxide of iron 5-30 1-20 



Alumina 0-47 



97-37 



Silica 58-30 containing oxygen 29*32 



Lime 9*89 2-78 



Magnesia 24*22 9*37 



Manganese 0*68 



Protoxide of iron 4*24 0*96 



Alumina 0-11 



97*44 



If 3*11 per cent, of water are added, which is the mean quan- 

 tity of the loss that the mineral sustained by a red heat, a 

 small increase is obtained in the analysis. Though the oxygen 

 in the surplus of silica is equal to that of the water obtained, and 

 it, therefore, might appear proper to consider the hydrate of 

 silica as isomorphous with the form of the pyroxene ; yet no true 

 hydrate of silica being either found in nature, or produced by 

 art, there exists no reason why it should occur here. It seems 

 as if the distinguishing property of those sahlites, which, on being 

 exposed to heat in a covered vessel, become black, give out water, 



* The quantity of purified sahlite which was used in this experiment weighed 35*916 

 grammes. The retort had lost I "168 grammes; the receiver had gained 1-151,-f- and 

 the tube with muriate of lime 1-009 gramme. 



■ -_- — — — ■ i i . 



-*- The number is most probably 0-151. — Edit. 



