I have referred to the impressive evidence of denudation displayed on the west side of the island of Eigg. The vertical distance from the summit of the Eigg plateau to the bottom of the submarine valley between this island and Rum is about 1500 feet, but as that summit lies below the original surface of the lava-field, the depth of rock which has been removed must exceed 1500 feet. We thus learn that since the close of the volcanic period the hollow between the islands of Eigg and Rum has been eroded to this great depth.

Still more striking is the evidence of enormous waste presented by the Faroe Islands. The cliffs there are loftier and barer, and the fjords have been cut more deeply and precipitously out of the basalt-plateau. I shall never forget the first impression made on my mind when the dense curtain of mist within which I had approached the southern end of the archipelago rapidly cleared away, and the sunlit slopes and precipices of Suderö, the two Dimons, Skuö and Sandö, rose out of a deep blue sea. Each island showed its prolongation of the same long level lines of rock-terrace. The eye at once seized on these features as the dominant element in the geology and the topography, for they revealed at a glance the true structure of the islands, and gave a measure of the amount and irregularity of the erosion of the original basalt-plateau. And this first impression of stupendous degradation only deepened as one advanced further north into the more mountainous group of islands. Probably nowhere else in Europe is the potency of denudation as a factor in the evolution of topographical features so marvellously and instructively displayed as among the north-eastern members of the Faroe group.

Availing ourselves of the datum-lines supplied by the nearly level bars of basalt, we easily perceive that in many parts of the Faroe Isles the amount of volcanic material left behind, stupendous though it be, is less than the amount which has been removed. Thus the island of Kalsö is merely a long narrow ridge separating two broad valleys which are now occupied by fjords. The material carved out of these valleys would make several islands as large as Kalsö. Again, the lofty precipice of Myling Head, 2260 feet high, built up of bedded basalts from the summit to below sea-level, faces the north-western Atlantic, and the sea rapidly deepens in front of it to the surface of the submarine ridge 200 to 300 feet below. The truncated ends of the vast pile of basalt-sheets which form that loftiest sea-wall of Europe bears testimony to the colossal denudation which has swept away all of the volcanic plateau that once extended further towards the west.

Nevertheless, enormous as has been the waste of this plateau of the Faroe Islands, we may still trace some of its terrestrial features that date back probably to the volcanic period. Even more distinctly, perhaps, than among the Western Isles of Scotland, we may recognize the position of the original valleys, and trace some of the main drainage lines of the area when it formed a wide and continuous tract of land.

A line of watershed can be followed in a south-westerly direction from the east side of Viderö, across Borö to the centre of Osterö, and thence by the Sund across Stromö and Vaagö. From this line the fjords and valleys diverge towards the north-west and south-east. There can hardly be any doubt that on the whole this line corresponds with the general trend of the water-parting at the time when the Tertiary streams were flowing over the still continuous volcanic plain. Considerable depression of the whole region has since then sent the sea up the lower and wider valleys, converting them into fjords, and isolating their intervening ridges into islands.

The topography of the Faroe Islands seems to me eminently deserving of careful study in the light of its geological origin. There is assuredly no other region in Europe where the interesting problems presented by this subject could be studied so easily, where the geological structure is throughout so simple, where the combined influences of the atmosphere and of the sea could be so admirably worked out and distinguished, and where the imagination, kindled to enthusiasm by the contemplation of noble scenery, could be so constantly and imperiously controlled by the accurate observation of ascertainable fact.

2. Impressive and easily comprehended as are the proofs of denudation supplied by the basalts of the plateaux, they are perhaps to a geological eye less overwhelming than those furnished by the eruptive rocks which have been injected into these plateaux. In the case of at least the basic intrusions, we may reasonably infer that they assumed their present position under a greater or less depth of overlying rock which has since been removed. When, therefore, they are found at or above the summits of the plateaux, they demonstrate that a vast amount of material has been removed from these summits.

The argument from the position of the dykes has already been enforced. It is absolutely certain that valleys several thousand feet deep must have been excavated since these dykes were erupted, for had such valleys existed at the time when the dykes were injected across their site, the molten rock, instead of ascending to the tops of the surrounding mountains, would obviously have rushed forth over the valley-bottoms. I have shown that this reasoning applies not merely to the volcanic districts, but to the whole surface of the country within the region of dykes. Thus the uplands of Southern Scotland, and wide areas in the Southern and Western Highlands, can be proved to have had glens cut out of their mass to a depth of hundreds of feet since the Tertiary volcanic period.

Not less convincing is the evidence afforded by the great eruptive masses of gabbro. We have seen that these complex accumulations of sills, dykes, and bosses include rocks so coarse in grain as to show that they must have consolidated at some considerable depth, but that they now appear in hill-groups 2000 to 3000 feet in height, the whole of the original basaltic cover having been stripped off from them. But these gabbro hills have been in turn traversed up to the very crests by later basalt-dykes, which thus supply additional proof that the erosion here has been stupendous.

The granophyre bosses tell the same tale. Though, like the domite Puys of Auvergne, they may still retain, in their conical forms, indications of the original shapes which their component material assumed at the time of its protrusion, we may be confident that their existing surfaces have been reached after the removal of much rock which once lay above them. This inference is confirmed by the fact that these eruptive bosses have been invaded by a younger system of dykes. The black ribs of basalt which may be traced along their pale declivities, which cross the glens that have been eroded in them and which mount up to their very crests, prove that since the latest manifestations of volcanic energy in the West of Scotland, extensive changes in the topography of the land have been effected by the operation of the subærial agents of degradation.