ccliv 
APPENDIX. 
The principal formation of the island appears to be the fletz sandstone, 
with the subordinate one of coal and ironstone. The structure of the cliffs 
along a considerable extent of the northern shore of Barrow’s Strait, ex- 
hibiting, beside horizontal stratification, numerous buttress-like projections 
and mural precipices, is not of uncommon occurrence in the formations of the 
transition and older fletz limestone ; but still more striking in this respect is 
the appearance of the sandstone formations, especially those of more ancient 
date. Having undergone a peculiar disintegration which acts in a direction 
nearly perpendicular to the horizontal stratification, they exhibit the repre- 
sentations of ruined towers, buttresses, pillars, and similar works raised by 
the hand of men. This structure, so strikingly expressed in the sandstone 
formation of Bohemia, Saxony, and other parts of Germany, at the Cape of 
Good Hope, and particularly in several mountainous tracts of China, appears 
no less characteristic of the sandstone of some parts of the coast of Melville 
Island, especially at Cape Dundas, the westernmost point to which the 
investigation of Captain Parry extended, and the general features of which 
have been so ably described by him in his Journal. 
This sandstone is composed of very fine, flat, confluent grains, with here 
and there the appearance of minute silvery scales, which, when more or less 
aggregate, communicate to the mass a perfectly micaceous appearance. It 
occurs both of a uniform greyish-white colour, and more or less marked 
throughout by small brown ochry spots, which sometimes are confluent into 
large patches. It generally separates into tabular pieces, and is sometimes 
invested on the rifts with thin plates of white carbonate of lime. Some of its 
varieties are not unlike grauwacke slate. It contains secondary fossils. Of 
the specimens which I had an opportunity of examining, two bore the 
impressions of a Trilobite, but too indistinct to admit of being determined 
with precision*. 
* I have since determined it to belong to Brongnart’s genus of Asaphus lately published ; but 
■whether or not it be one of the species described by him. and Wahlenberg, cannot be ascertained 
from the specimen alluded to. 
