the mode of its formation. The other I have called mountain trap.
The terraced trap forms by far the greater portion of the surface of Sky; but as it is sufficiently defined in the map and in the general description, I will forbear here to name its boundaries. Its northern and principal tract is every where continuous, but at the southern side of the island there are detached portions, which I shall first notice. The southernmost in position is a hill of no great extent, which is seen above the secondary strata already described at Swenish point, between Loch Slapin and Loch Eishort. This mass is connected with two large bodies like roots, (inasmuch as they have neither the parallelism nor the independence of trap veins) which cut through the whole mass of strata, and disappear below the sea. I consider this place as of great importance in the history of trap, as it shows plainly how a particular mass of that rock may appear fairly incumbent on a given stratum, while it is in fact connected with a much deeper set of rocks.
It is plain that partial views of this mass at any one point of all the various substances which it traverses, would assign to the same rock all the several hypothetical characters according to which trap has sometimes been divided. It would be called in one place primitive, in another transition, in a third flœtz, and so forth. As I do not intend to enter into the well-known questions respecting the origin of trap, I forbear to point out how this appearance bears on any of the hypotheses which regard its formation.
The next of the stratified masses of trap which are seen toward the south are the hills which decline from Blaven to Strathaird, and appear to be similarly incumbent on the stratified rocks of that promontory, but I had no success in my attempts to discover their
