openings at Killicrankie, has hitherto proved insufficient to undermine and remove it; in consequence as it would seem of the lateral direction of its course produced by the fall of those portions of alluvium which have formed the small holm on which Fascally is situated. Thus it is for the present protected from change, and perhaps destined to remain, for a longer period than that in Glen Tilt, a monument of those revolutions which point out an extensive but transitory action of water on the surface of our globe.
I have remarked in the preceding paper, that the granite which forms the right boundary of Glen Tilt has the property of affecting the magnetic needle. This influence is far from inconsiderable, and in many cases produces not only remarkable local variations, but a disturbance of that instrument so great as to render it useless for the purposes of ascertaining any meridian whatever. It is not limited to the syenitic varieties, where from the predominance of hornblende we might reasonably expect to find it more active, but is equally to be found existing in those granites which do not contain this ingredient. Neither does it depend apparently on the micaceous ingredient, since in many of the rocks which shew it strongly the mica is in very small proportion. It seems to be equally independent of a state of decomposition in the granite, since it is here, as elsewhere, to be found inherent in fresh specimens, although on the summit of Goatfell in Arran, as I have remarked on a former occasion, it appears most conspicuously in those specimens where the iron is becoming carbonated and the rock is tending to disintegration. In the paper to which I here allude,[1] I noticed this fact as being nearly a solitary one at that time, but I have had occasion to observe it since on various occasions. In a paper on Cruachan, drawn up
- ↑ Geological Transactions, vol. ii.
