at many points, and range nearly due east, in the direction of the valley. This direction is continuous beyond Little Langdale Tarn; but here the boundary-ridge is interrupted by an opening trending southward, and immediately the striations bend round in that direction. The flanks of Oxen Fell and Skelwith Fell also exhibit the southerly striations; and it is evident, from this and similar instances, that some influence had been constantly acting upon the ice, forcing it to the southward. Where the ice has held an easterly course, it is due entirely to the elevation of the ridges which bound the valleys. If we suppose these influences to have been either prevalent north winds or currents of the sea, it is an additional evidence that the glaciation of these valleys has been. produced by floating ice, up to a certain level.

Stockdale. The flanks of this and the adjoining valleys of Scandale, Rydale, and Troutbeck, are overspread by a heterogeneous accumulation of gravel and boulders in a matrix of red and gray clay, which towards the lower parts of the valleys assumes a somewhat terraced aspect. I have already stated that this is probably not true moraine matter, but the incipient form of marine Drift. Towards the head of Stockdale, however, at a height of 950 or 1000 feet, it passes into a true moraine, which occupies the head of Kirkstone Pass (1200 feet), and extends down on the northern side into Patterdale for about 300 yards. On this side the moraine is more perfectly formed than on the side of Stockdale, and consists of the usual assemblage of rounded mounds of gravel, with strewn and perched blocks, covering polished surfaces of cleaved felspar porphyry. I do not think, however, that the remarkable block from which the pass derives its name is a true glacier-perched block, but has fallen from the crags which bound the pass.

This is the last of the valleys which I shall attempt to describe on the southern side of the watershed. The description of those on the northern side I propose reserving for a future occasion; and shall content myself with remarking, that the great valleys of that part of the district exhibit similar phenomena, the striations and course of the ice having been northward. Thus the flanks of Grisedale, one of the wildest

gorges in the Lake District, having its sources in the heart of Helvellyn, are ice-moulded up to an average elevation of 600 feet or more above the bed of the river. There are also remarkable examples of perched blocks of huge size, and welldefined, if not extensive, lateral moraines. That this valley has been the trough of a glacier, extending almost to its entrance, there is the clearest evidence, in the existence of a large terminal moraine. The elevation above the sea of this moraine is not more than 550 or 600 feet; and its position is within a short distance of the entrance into Patterdale and the head of Ulleswater (380 feet). The moraine is about a quarter of a mile in length, and formerly extended right across the valley; but the impetuous mountain torrent has hewn a channel, and has levelled the ground for a breadth of 100 yards. This glacier, extending from Grisedale Tarn to the terminal moraine, was three miles in length, with an average breadth of 400 yards; and, judging from the height of the polished surfaces along the flanks of the valley, 600 to 800 feet in depth.

If I have hesitated in the cases of the valleys of Langdale, Grasmere, Rydale, Stockdale, and Troutbeck, to admit of the existence of glaciers occupying their entire lengths, I certainly have no such hesitation in the case of Grisedale, where the terminal moraine, placed near the very mouth, leaves no room for scepticism. In the case of the former valleys, the welldefined moraines are situated near their heads; and the glacial phenomena which are exhibited along the bottom and sides of these valleys, and extended beyond into the open country, being such as are known to be producable by the action of floating ice, I am disposed to refer them to this agency. Without entering upon the question whether the valleys of the first class may not, at an earlier stage of the glacial epoch, have been occupied by glaciers, I only here maintain, that the marks of the former presence of ice are such as are capable of being produced by bergs floating down from glaciers which precipitated themselves into the sea, as in Tierra del Fuego, and, at a former period, in Scandinavia, where the rock surfaces along the fiords are grooved and ice-worn to and below the water's edge.

The two great facts to which the glacial evidences of the Lake District (as far as I have observed) appear to point are these: first, that the sea stood at a level sufficient to float ice charged with boulders over ridges and hills, which are now at elevations of 800 to 900 feet; and, secondly, that after the sea had retired, glaciers descended the valleys as low as 500 feet above the present sea-level.

The reader will therefore bear in mind that I have not now entered upon the question of any changes previous to these, the most recent and most apparent.

Tarns. The production of tarns, or small mountain-lakes, by the agency either of moraines forming embankments, or by the scooping action of glacier ice, has been illustrated in the case of the High Alps and North Wales by Professor Ramsay.* Not less satisfactory are the examples of this kind in the "Lake District," where these lonely basins of water abound; and distinguishing between tarns and the larger and less elevated sheets to which the term "lakes” more properly belong, I think we may safely assert that there is scarcely one of the former in whose production glacial ice has not been concerned. Indeed, as a scientific distinction, it might be advisable to restrict the term "tarn" to those lakes which can be traced to the formation of glacial agency.

Of the tarns belonging to the Southern Watershed, perhaps the most interesting examples, as connected with glacial vestiges, are Stickle Tarn, Easdale Tarn, and Blea Tarn, which I shall describe in this order :

Stickle Tarn.-This basin is fed by a stream which descends a gorge scooped out along the north-eastern flank of the Langdale Pikes. The sloping ledges of porphyry, which form a conspicuous feature at the upper sources of the brook, are striated E. and W. in the direction of the valley; and at an elevation of about 1800 feet, lateral moraine shingle, formed of deep red gravelly clay, with perched blocks, is strewn along the flanks. The tarn, which reposes at the lower end of this valley, is bounded through half its circumference by a wall of precipitous cliffs, whose extremities are united by

*Quar. Jour. Geol. Soc., vol. viii., &c., " Peaks, Passes, and Glaciers," 3d edit. NEW SERIES.-VOL. XI. NO. I.-JAN. 1860.

D

a well-defined moraine-thus forming and completing the basin. The interior edge of the moraine is but slightly curved, and forms rather the chord of the arc. It consists, as usual, of large mounds of gravel and clay, on which boulders lie scattered in all positions; and over the south-eastern extremity of the moraine, the brook to which the tarn gives birth leaps forth, and is precipitated in a succession of cascades into Langdale. From the ice-worn character of the rock surfaces along the sides of the tarn, it may easily be inferred that the glacier which deposited the moraine, and thus became the primary agent in the formation of the tarn, at one time cascaded down the declivities.

But in this remarkable little tarn, there is an object almost unique in its kind, and which the tourist frequently notices without being aware of its true meaning. At about the centre of the tarn, a smooth, oval boss of rock, about twelve feet in length, rises about two feet above the water at its centre. On this is perched a natural monument of the transporting agency of ice in the form of a boulder, nearly round, and (judging at a distance of 100 yards) about four feet in diameter (Plate II. fig. 5). It has a most singular effect, thus placed in solitude, and isolated, by a circular sheet of still water, from the blocks, which strew the banks of the tarn in profusion. By its permanence in so critical a position, it shows how completely the lake has been protected from storms, and from changes in the level, as a combination of these circumstances would probably have long since swept the block into the depths of the lake.

Easdale Tarn.-Crossing to the east, over a tract of rocky ground of an elevation of about 2000 feet, we arrive at the edge of the deep basin which incloses Easdale Tarn. That the valley of Easdale was occupied by a glacier, there is undoubted evidence on all hands. A moraine of large dimensions stretches out from the base of the crags above the tarn, and threatens to block the valley, which perhaps it once did. The flanks of the valley are strewn with boulders and perched blocks of huge dimensions; and the dark mammillated masses of trap are conspicuous, especially south of the lower end of the tarn, where they resemble the upturned sides of a ship,

grooved with glacial striæ ranging east, visible even at a distance when the beams of the western sun glance athwart their flanks. This smoothed and ice-moulded character of the rocks is apparent to the junction of the valley with that of Grasmere; and is also exhibited along the flanks of Helm Crag, which forms the boundary of the valley on the northeast side. It is probable that the lower end of the glacier entered the head of Grasmere valley by the channel at the foot of Helm Crag, while a portion of the ice lined the summits of the cliffs over which the waters of Sour Milk Ghyll are precipitated, and thus assumed on a smaller scale the appearance of the lower extremity of the "Glacier des Bois."

I have already referred to the moraine at the head of Easdale Tarn. A second and less elevated moraine, partly supported by a ridge of ice-worn trap, strewn with boulders, forms an embankment to the lower end of the lake itself. The brook has worn a channel to a depth of 100 feet into the moraine without reaching the solid rock, so that it is probable that the lake rose higher before the channel was worn to its present depth; and as the process proceeds, it may sink lower.

The perched blocks on the mountain sides along the southern flanks of Easdale are remarkable. They may be seen at elevations of nearly 2000 feet (estimated); and one example, conspicuous at some distance on descending from the tarn towards Sour Milk Ghyll, occupies so critical a position on a shelving ledge, that had the district ever experienced an earthquake, it would infallibly have been precipitated into the valley below.

Blea Tarn. This little sheet of water occupies a hollow near the pass between the two Langdale valleys. On the east and west the basin is bounded by lofty crags, whose surfaces are glaciated to a height of 250 feet above the tarn. On descending from the tarn towards the main valley, the examples of roches moutonnées and perched blocks are numerous and well pronounced. The striations range S. 20° E. A ridge of these ice-moulded bosses of porphyry stretches across the lower or southern side of Blea Tarn, and at first appears to be the agent in banking up the lake; but on following the ridge to the spot where the brook escapes, it will be found that