vertical faces, where large depressions have been formed. As 

 usually, it has proceeded from the first points of attack along 

 the cleavages of the mineral; between the paths thus formed 

 by the proceeding decomposition columns of undecomposed 

 aegirine remain. Between them and parallel to the longitudinal 

 direction of the aegirine individuals, crystals of the new mineral 

 lorenzenite have formed. It cannot, indeed, be directly proved 

 that these crystals have formed simultaneously with the decom- 

 position of the aegirine ; this is, however, very probably the 

 case. A circumstance which particularly speaks in favour of 

 this supposition is the fact that the new crystals are arranged 

 in a position perfectly parallel to the remains of the original 

 mineral. The new crystals, however, occupy only a small part 

 of the space left by the decomposition of the aegirine. The 

 greater part of the constituents of the aegirine has, conse- 

 quently, been carried off. On the other hand, the new mineral 

 contains titanic acid, a constituent which probably does not, 

 to an amount worth mentioning , enter into the composition 

 of aesirine. 



15. Arfvedsonite. 



Arfvedsonite was originally an exclusively Greenlandian mi- 

 neral. Only recently has it also been met with near Lange- 

 sund in Norway, though in extremely small quantity. The original 

 Greenlandian locality was the sodalite- syenite area on the firth 

 of Kangerdluarsuk, where the mineral occurs in abundance and 

 sometimes as very large crystal individuals. On Narsarsuk, on 

 the other hand, arfvedsonite is found rather sparingly, and 

 the crystals are there in general very small. The arfvedsonite 

 material collected by me on Narsarsuk consists of two sorts of 

 crystals, viz. such as are found in the Lützen collection (Type I), 

 and such as show" a different character (Type II). 



