crags above. Hence the precipice would shrink backward into the mountain. To this unequal weathering I believe we should ascribe the singular extremes in the scenery of granite mountains, as well as the picturesque forms which are often assumed by groups of granite boulders.1

Throughout the Highlands, where the older palæozoic rocks adjoin formations of later date, there is usually a well marked contrast of scenery. Thus along the shores of the Moray Firth the brown rough mountains of the interior are fringed with a border of fertile ground, marking where the Old Red Sandstone takes the place of the schists. Further north a similar contrast shows where the Oolitic sandstones and shales of Sutherlandshire run as a strip along the coast, at the base of the line of rounded bare conglomerate hills which rise above them. On the west side of the island also, the Liassic and Oolitic strata, owing to the comparative richness of their soil and their low level, are sharply marked off from the Cambrian and Lower Silurian mountains which surround them.

1 I must confess, however, that the Highland precipices present not a few difficulties to the geologist who would explain their origin and persistence. I should much like to know to what extent they and the deep glens from which they rise were modified during the glacial period. One may feel sure that they are not due to subterranean convulsion, though he may not be able to follow back to the first beginning the various stages of their formation during the slow excavation of the rock.

Q

As a final example of the close relation between the contour of the Highland hills and the nature of the rocks that compose them, let me refer to the Inner Hebrides. These islands, from the north end of Skye to the south of Mull, are formed in great part, of greenstone, basalt and other masses of volcanic origin, piled above each other in horizontal or gently inclined beds, to a height of many hundred feet. Some of these rocks weather away more rapidly than others. Hence the profile of a hill in these districts

FIG. 17.-VIEW OF A TERRACED TRAP HILL, ISLE OF EIGG.

usually shews a succession of terraces, each of the harder beds standing out sharply against the sky like one of the steps of a staircase. Moreover, the general decay of the rocks gives rise to a rich loam, and thus these trap-islands are marked by the verdure of their valleys and the close bright sward which is found in patches even on the tops of their hills. No one who, in a sunny day, has sailed through the Sound of Mull can fail to remember how strikingly these features

are there brought out. The green mountains of Mull rise from the water's edge, sometimes in steep rocky banks seamed with glistening waterfalls, and sweep inland, terrace above terrace and hill after hill, till their minor features are lost in the distance. The profile of one of these terraced slopes sometimes shows twenty or thirty distinct steps with gentle green slopes between them. The effect is often heightened where a soft shale or yellow sandstone is banded with the greenstones and basalts; for then, as the soft stratified intercalation moulders away, the igneous rock above it may be traced as a dark crag, running far along the side of the hill, and keeping parallel with the other terraces overhead and below. The landscapes in the interior of Mull and of the northern half of Skye owe the singularly artificial look of many of their lower hills, and the general terraced character of the whole surface, to the manner in which the bedded igneous rocks have yielded to denudation. The penning of these lines recalls to the writer many a pleasant reverie among the wilds of the Inner Hebrides, when, sitting in the light of the long autumn evenings, he used to mark the sinking beams strike along the sides of those truncated pyramidal hills, revealing terrace over terrace in alternate bars of dark crag and green slope-features that were but faintly seen in the glare of noonday—to look over the wild heathy uplands that stretched around

to right and left in utter solitude and stillness; to watch how hill-top after hill-top would lose its blush of sunset, as if the dying day were slowly climbing the steps cut along the flanks of the western hills, and how the chill shadows, struggling upward from dark and lonely glens, crept up the same gigantic staircase until the whole landscape melted into grey gloom, and the night began to fall.

These and other local but characteristic parts of Highland scenery are to be traced, then, to the varying nature of the rocks of the country. I have already said that before the great erosion of the Ice Age began, these peculiarities were in all likelihood more strongly marked than they are now, and that, slowly recovering from the effects of the glaciation, they are returning to their former condition. I have been often struck with the progress of this change along the shores of Loch Fyne. The hard quartzose rocks opposite Tarbert are beautifully ice-worn and smoothed; their striæ, still fresh and clear, may be seen running out to sea under the waves. The lower parts have been protected from decay, owing partly to the recentness of their upheaval into dry land, and partly to their having been shielded by a coating of boulderclay, not yet worn away from the bays. But above high-water mark, though the track of the old ice-sheet is still strikingly shewn, the rocks have begun to split up along their joints. Hence the low cliff that rises

along the shore is rent into endless chinks and clefts, large angular blocks have been detached from it, and its base is cumbered with the ruins. Some of the islands show well the union of the glaciated outlines with this subsequent weathering. They still rise out of the water in long flat curves, like so many whales,the form that was impressed upon them by the ice; yet they are split across along the joints into open cracks which one might fancifully compare to deep paralle gashes made across the whale's back.

FIG. 18.-ICE-WORN ISLETS IN LOCH FYNE, WEATHERING ALONG THE PARALLEL JOINTS OF THE ROCK.

The examples already cited from Ben Nevis and elsewhere, of the waste of cliffs and ravines, illustrate this obliteration of the marks of the ancient glaciers. It is unnecessary to multiply instances of a feature of Highland scenery which may be seen more or less. distinctly on almost every hill-side and valley. Nor need I again point to the numerous ravines and riverchannels which have been excavated or deepened since the ice disappeared, and which show that the