constitute a class of rocks entirely different, have been evidently formed together with the rocks which they accompany, and are, like them, referable to the same epoch, never intruding to their disturbance, but occupying situations and maintaining characters, in every respect conformable to them. It would tend to the accuracy of geological language, if they were to receive designations derived from their simple parent, hornblende rock, and it is to be desired that the occasional mixture of felspar with hornblende, whether that rock be massive or slaty, should be designated by a term, if it requires a distinct one, derived from its true basis, so as to prevent the confusion which their present names introduce into our ideas and descriptions.
The Shiant isles having escaped the notice of mineralogists, I shall make no apology for giving such a cursory description of them as my opportunities of observation allowed me to draw up, for the purpose of explaining the specimens which I have transmitted.
These islands are situated to the north of Ruhunish point, in Sky, from which they are distant about 15 miles, and they lie about 8 miles nearly south-east of the entrance of Loch Brolum, in Lewis. There are three principal islands, forming a nearly equilateral triangle, besides a few subordinate rocks of considerable height, but of small superficial dimensions. The names of the three largest are Gariveilan, Eilanakily, and Eilan Wirrey. Of these the two former are about a mile and half in circumference, and the latter appears to be less than a mile. They constitute one farm, the house appertaining to which is situated on Eilanakily. In ap-