8i 



less aegirine passes here into crocidolite. The line between 

 them is well marked, but jagged. Bundles of crocidolite fibres 

 intrude into clefts in the aegirine. Close to the boundary line 

 the crocidolite fibres are arranged parallel to the c-axis of the 

 other minerals, but at some distance from it they form the 

 usual felt-like aggregate. 



The foregoing shows that, as pointed out by Brogger^), 

 crocidolite is formed from aegirine and, after an intermediate 

 aegirine stage, also from arfvedsonite. But crocidolite may also 

 be derived directly — without an intermediate aegirine stage — 

 from arfvedsonite or riebeckite (?), as has already been pointed out 

 and described by Us sing ^). In cleavage fragments of arfved- 

 sonite from Narsarsuk, which show no trace of the cleavages 

 of aegirine, one sometimes observes inclusions of a blue, flnely 

 fibrous mineral, which evidently is crocidolite. In»microscopical 

 sections of such arfvedsonite, orientated parallel to the plane of 

 symmetry, no aegirine is found. On the other hand, the fibrous 

 portions are easily recognizable as crocidolite. The felt-like 

 texture is here, however, seldom distinct; the fibres are arran- 

 ged more or less parallel to the c-axis of the arfvedsonite. It 

 is evident that the crocidolite has been formed here in the 

 same way as asbestus is derived from common hornblende. It 

 is thus probable enough that all the crocidolite occurring on 

 Narsarsuk is the result of the alteration of aegirine or arfved- 

 sonite. 



17. Spodiophyllite. 



When first observed at the locality, this mineral was sup- 

 posed to be a species of chlorite. This supposition seemed to 

 be corroborated by the determination of its specific gravity and 



') Zeitschr. f. Kryst. Vol. 16, p. 332, 1890. 

 ') This journal, Vol.14, pag. 205, 1894. 



