MR ANDERSON ON THE ANALYSIS OF CAPORCIANITE AND PHAKOLITE. 333 



concerned ; but it differs from them in this much, that, previous to melting, it 

 swells out only to a very inconsiderable degree ; for it melts almost at the same 

 instant that the swelling manifests itself. 



The analysis yielded the following results : 



Silicic acid, . 52.8 oxygen contained 27.43 8. 



Alumina, -. 21.7 10.15 1 



Peroxide of iron, 0.1 >. 0.03 J 



3.651. 



Lime, . 11.3 3.23 



Magnesia, . 0.4 0.15 



Potassa, . 1.1 0.22 



Soda, . 0.2 0.05 



Water, ., 13.1 11.64 3. 



100.7 



If we here express by r the monatomic bases, then the quantities of oxygen in 

 r, A, S, and Aq are to each other as 1 : 3 : 8 : 3, which evidently determine the 

 mineralogical formula to be r S 2 + 3 A S 2 + 3 A q. This when transformed to the 

 chemical formula, becomes r 3 Si 2 + 3 Kl S' 2 + 9 H. 



It thus appears that Caporcianite stands chemically in near relation with the 

 minerals, Analcime, Ledererite, Potash-Harmotome, Chabasie, and Levyne, from 

 which it is separated merely by the difference in the quantity of water which it 

 contains. All these minerals consist of a bisilicate of the first as well as of the 

 second term ; and the quantity of oxygen in the alumina is in all of them three 

 times that contained in the monatomic basis. The formulae of these minerals are 

 as follows : 



Analcime, "| ( r N. 



rS 2 + 3AS2+2Atf{ 

 Ledererite, J I r = C.N. 



Caporcianite, . r & + 3 A. S 2 + 3 A.y r = C. 

 Potash-Harmotome, rS2 + 3AS 2 + 5Ay r= K.C. 



Chabasie, -| . r r - C.N. 



r S 2 + 3 A S 2 + 6 A q \ ^ ... T 



Levyne, J I r = C.K.N. 



The formula r S 2 + 3 A S 2 is thus, then, known to exist in no less than four diffe- 

 rent combinations with water, namely, with 2, 3, 5, and 6 atoms, the second of 

 which results from the foregoing analysis. 



PHAKOLITE. 



This mineral occurs in small crystals in the Bohemian Mittelgebirge, and 

 was from crystallographic investigation believed to be nearly related to Chabasie. 

 But the following analysis shews that this supposition is not confirmed by its 

 chemical constitution. 



