103 



the crystal, a natural prism is obtained, which, in the crystal 

 used, had an angle of 52° 27'. As I had no light of one colour 

 of sufficient intensity, common light has been used and adjusted 

 to the yellow part of the image; by this means we get directly: 



(0 = 1-7640, 

 moreover, according to methods stated by Wolfing^): 



e = 1-5772, 

 but this value must, of course, be rather inaccurate. By using 

 the total reflectometer we get for faces belonging to the base 

 of the largest crystal a quite faint boundary line giving 



£ = 1-5762, 

 while I had no fluid sufficiently refracting for the determination 



of ÛJ. 



It will appear from the above that the cordylite is pos- 

 sessed of an uncommonly strong, negative double refraction. 



Astrophyllite. 



This mineral has not hitherto been known from Narsarsuk; 

 but it is found abundantly in Bernbdrg's collections, and is of 

 a quite peculiar appearance. 



It is always formed as quite thin needles or rather as 

 ruler-shaped bodies. To judge from the optical relations, the 

 longitudinal direction of the needles must be parallel to the 

 crystallographic a-axis, according to Brøggers^) orientation; the 

 needles are flattened after 6{010|, and parallel to this plane is 

 also found a high degree of cleavage; thus the form resembles 

 upon the whole that from El Paso Co., Colorado, described 

 by König and Bucking-^); but while this astrophyllite is imbedded 

 in quartz and feldspar, the greater part of that described here 

 is freely developed and surrounds the other minerals as a 

 golden coat of hair or as a felt-like mass. 



1) Rosenbusch: Mikr. Phys. I, 1, 1905, pp. 86 and 76. 



2) Zeitschr. f. Kryst. 16, 1890, p. 200. 



3) Zeitschr. f. Kryst. 1, 1877, p. 423. 



