147 



by feldspar, calcite, etc. Less frequently it is found associated 

 with neptunite and epididymite. In colour it is sometimes 

 white or yellowish -white, sometimes a light brick -red. When 

 brick- red the mineral is not quite pure. The hardness is 

 somewhat varying ; the red mineral has the hardness of 

 quartz or a little more; the white mineral is scratched by 

 quartz. The specific gravity of the former is 2,.594, that of 

 the latter 2,524. Before the blowpipe the mineral easily fuses; 

 in the closed tube it yields abundance of water. In the borax 

 bead it is dissolved with difticulty into a clear glass ; the salt 

 of phosphorus bead, on the other hand, remains opaque. It is 

 decomposed by hydrofluoric acid, but not acted upon by other 

 acids. All the above characters are quoted from Lindström. 



Apart from the rock-forming minerals feldspar and aegirine, 

 no mineral probably occurs so abundantly on Narsarsuk as 

 elpidite. Detached pieces of it lie scattered all over the plateau, 

 and at most of the localities worked by me it was met with 

 in situ in larger or smaller quantities. It varies considerably 

 both in form, structure and colour. As a rule it may be 

 said to present rather an unattractive appearance. It has a 

 columnar structure , and is mostly of a greyish colour. Thus 

 at first sight it looks like wooden slips or sticks in a state 

 of decay. 



The separate columns of elpidite vary in size from the thinness 

 of a hair to the thickness of a finger or more. In the longi- 

 tudinal zone they are chiefly bounded by faces which make with 

 each other about the same angles as occur on amphibole, viz. 124° 

 and 56°. Crystalline terminations are extremely rare and occur 

 only on very small, clear and colourless or enamel-white indi- 

 viduals. The common individuals are either terminated by 

 irregular fractures or rounded as if they had been acted upon 

 by some solvent. Occasionally the larger individuals are split up 

 at the ends into a number of pointed fibres. Individuals of 

 some length are seldom quite straight, but mostly bent like a 



10* 



