length in other places the errors that may arise in these observation, and need not enter largely into the question in this place, but it appears to me a sufficient argument that two sets of rocks are not consecutive if those which correspond in one place do not also correspond throughout. I should expect to find that the same bed of red sandstone exists between Kyleaken and Broadford, where I had no opportunity of looking for it; but beyond that it is not likely to occur, as the limestone which lies at some distance above it in the order of superposition occupies the lowest situation on the shores of Broadford. I mentioned that the same bed is to be found at Loch Scavig, where it is plainly connected with the very large mass of similar sandstone which constitutes the main part if not the whole of the island of Soa; it occupies here the lowest visible situation, being washed by the sea and exposing but an inconsiderable thickness, since it is shortly followed by the trap, when it finally disappears, none of the subsequent stratified rocks which accompany and follow it near Loch Eishort being here found. It is here interstratified with beds of a schist which has sometimes the character of fine clay slate, while at other times, the laminæ being separated by coarser particles of sand and of mica, it puts on the character of that which has been called fine graywacke slate. I have on other occasions shown[1] that these rocks are not necessarily distinguished in position, and that near Loch Ard in Perthshire, and Crinan in Argyleshire, the several sorts of clay slate and graywacke are irregularly intermixed: the same phenomenon is to be seen in many other parts of Scotland, which I cannot here pretend to enumerate; although the question is of importance to geology, this is not the place in which it must be discussed. The fact abovementioned however extends our objections to the commonly received
- ↑ Geological Transactions, vol. 2d.
