Prof. How on the Mineralogy of Nova Scotia. 131 





Oentrallassite, 1859. 



,„, A 





Found. 



Calculated. 



Lime . . 



. 27-91 



29-20 



Alumina . 



. 1-14 





Magnesia 



. 0-16 





Potassa . 



. 0-59 



* * 



Water 



. 1P4I 



11-74 



Silica . . 



. 58-85 



59-06 



100-07 100-00 



establishes beyond a doubt the identity of the mineral; but 

 the recurrence of an almost absolutely similar quantity of water 

 is so far counterbalanced by dissimilar amounts of the other 

 essential constituents that I must own my old formula is not 

 supported as I should like. Since one of the minerals asso- 

 ciated with the specimen first analyzed contained much silica 

 (cyanolite ; it gave 74 per cent.), and was not easily separated ; 

 and as it is not present with the mineral whose analysis is now 

 published, I think it most probable this last is the more correct. 

 The oxygen ratio is seen to be nearly 1:3:1 for lime, silica, 

 and water : now, taking the accepted constitution of silica and 

 the equivalent notation still most commonly employed in mine- 

 ralogical writings, this ratio leads to the formula 



2CaO, 3SiO a + 2HO, " 

 of which the calculated percentages — 



2CaO = 56 34-14 



2 HO = 18 10-98 



3Si0 2 = _90 54-88 



164 100-00 



agree well enough with the analytical results brought forward, 

 allowing for replacements. The ratio for the same constituents 

 given by Anderson for gyrolite (" Description and Analysis of 

 Gurolite, a new Mineral Species," Phil. Mag. Feb. 1851), was 

 2:6; 3 ; and the percentages calculated from the formula he 

 gave (changing Si0 3 to Si0 2 ), 



2CaO, 3Si0 2 + 3HO, 



2CaO = 56 32-37 



3HO = 27 15-61 



3Si0 2 = _90 52-02 



173 100-00 



agree tolerably well with the numbers found by him, and 

 better with those found by myself (" On Gyrolite occurring 

 with Calcite in Apophyllite in Trap of Bay of Fundy," Eclin. 

 New Phil. Journ. and Silliman's Journal, 1861) in the only 



