1826.] M. Berxelius's Amlj/sis of some Minerals, 119 



temperature of melting tin ; after which dissolve out the nitrate 

 of yttria by water. If the quantity of water be small, the solu- 

 tion will not contain any manganese ; but when we wash the 

 oxide of manganese, a portion dissolves, and the solution as* 

 sumes a dark colour, which it loses by exposure to light." 



Levi/ine. 



M. Berzelius found a specimen of levyine, (sent him from a 

 quarter that precluded all suspicion of error,) so extremely simi- 

 lar to his mesoline^ a mineral which accompanies the tessellite 

 from Ferroe, that he was desirous of ascertaining its exact com- 

 position. The texture of the mineral was crystalline, ^nd 

 wherever the angles could be measured, they were found to cor- 

 respond with those of chabasie. 



Its analysis gave him, 



Silica 48*00, containing oxygen . .24'06 



Alumina 20*00, 9*34 



Lime 8-35^ 



Magnesia.... 0-40! onr 



Potassa 0-41 f "^ "^^ 



Soda 2-7oJ 



Water 19-30 17-16 



99-21 



** This result,'* M. Berzelius observes, " accords perfectly 

 with the proportions of the component parts of chabasie," 

 namely, 



S^ + 3 AS^ + 6 Aq. 



Kj 



as may be more distinctly seen by the following table : 



" With respect to the chabasies, I consider it certain," adds 

 Berzelius, " that the trifling differences which occur in these 

 analyses are occasioned by their not having been perfectly freed 

 from the minerals which accompany them ; hence mesoline 

 ought to be considered merely as a species of chabasie, as is 

 further observable in its granular texture," 

 * Analysed by Arfwedson. 



