traced to the wearing and grinding agency of ice, not in the form of glaciers, but as floating bergs. On close examination, the rock-surfaces are found resolving themselves into rounded and oval bosses, frequently grooved and striated in lines radiating from the interior mountains.

These appearances must be due to the wearing action of floating ice; for, independently of other considerations, it is impossible to suppose that this comparatively low-lying district was overspread by one broad sheet of ice as a glacier. This moulded form of rock surface prevails at least as far south as Kendal and the southern extremity of Windermere. The striations and groovings range towards the south, scarcely ever deviating more than 20° from the meridian; but, as remarked by Professor Ramsay in the case of the striated surfaces of North Wales, it is only when the natural covering of detritus or turf has been recently removed that the finer striations become apparent. In general, the oval or rounded forms of the naked rock-even where it stands, and has stood for ages, prominently out of reach of all protection-is faithfully preserved, while the finer ice-marks have yielded to atmospheric erosion acting along the planes of cleavage (see fig. 1). The hollows are generally filled with drift-gravel in a base of red clay, with boulders of porphyry, which sometimes rest as perched blocks on the slates and grits of the Coniston series. As particular points where the striations may be observed, I may mention the Coniston road, above Hawkshead; the porphyritic rocks, both on the east and west flanks of Skelwith Fell; a remarkably fine example on the Coniston road, half a mile south of Brathay Church; and several positions in the Old Hundred" of Troutbeck, above Ambleside.

Passing now to the consideration of the glacial vestiges of the higher valleys, I shall describe them under the two heads in which they appear naturally to arrange themselves in a glacial point of view.

1st, Those valleys which have been channels both for floating ice and glaciers.

2d, Those which have been occupied by glaciers only.

The first class comprises the lower and larger valleys of the mountainous district, as Windermere, Coniston, Little

and Great Langdale, Ambleside, Grasmere and Troutbeck. They all exemplify glacial phenomena; but, with the exception of their higher reaches, they do not appear to have been occupied by glaciers. For reasons presently to be stated, I attribute these phenomena solely to the action of ice floating down these channels when they were occupied by the glacial sea. In treating of these valleys I shall commence with that of the Rotha, which was the first to come under my observation, and which forms the principal prolongation of the great valley of Windermere.

Valley of the Rotha (Ambleside and Grasmere).—At the head of Windermere, the grooved surfaces-of which I have already given examples-are continued into the valley of the Rotha, and rise on its flanks to a height of about 400 feet above the level of the river, as may be observed along the road to Kirkstone Pass, above Ambleside. All along the alluvial bed of the valley, from Ambleside to a considerable distance beyond Grasmere, the examples of well-formed roches moutonnées are both numerous and striking (fig. 1). On one of these ice-worn bosses, fluted and grooved with striations ranging due south, the new church of Ambleside has been erected; and grouped around are several fine examples. One feature in these bosses deserves special notice. When viewed sideways, they appear as prostrate cones or wedges pointing up the valley, or nearly north; thus proving the movement of the ice to have been from north to south. These bosses are at an elevation of 130 feet above the sea.

On the road to Patterdale, at an elevation of about 600 feet, the striations are exactly parallel to those of the bottom of the valley, though they cross transversely the ridge which divides Stockdale from Scandale, a proof that the direction of motion was here independent of the form of the valley. Higher up, by the sides of Rydal and Grasmere lakes, the striations are parallel to the longitudinal axis of the valley. The flanks are covered with Drift to an elevation of about 400 feet above Grasmere. In some places this formation presents a slightly terraced surface; and its generally smoothed and undulating slopes, similar to those which the Drift assumes in Lancashire and Cheshire, show that it is a truly marine de

posit, and not to be classed with the products of subærial glaciers. It may be traced from Windermere valley into those of Stockdale, Troutbeck, and Scandale, gradually ascending along their sides till it passes into moraine gravel, at an average elevation of 850 to 900 feet.

In composition and want of arrangement this marine Drift resembles moraine matter, of which it is indeed only another form. It is to be recollected, however, that neither the circumstances under which it was deposited, nor the materials themselves, were favourable to the development of planes of bedding. It is composed of gravel and boulders, angular and rounded, polished and striated, imbedded in a matrix of clay of various hues, from dull brown to deep red and purple. This latter colour prevails much in the glacial accumulations of the higher valleys; and I could not but recollect, in connection therewith, the red colour of the boulder-clay of Lancashire. On close examination, it was evident that the clay itself was formed by the trituration and decomposition of the felspathic rocks of the mountains.

Several fine examples of roches moutonnées occur above Grasmere lying as cones, or inclined planes, with their apices pointing up the valley; and when we enter Easdale, and examine the surfaces of the rocks on both sides of the valley along the western flank of Helm Crag, it is impossible not to be struck with the ice-moulded forms which they assume up to an elevation of about 1000 feet.

Great Langdale.-The glaciated aspect of the rocks at the entrance to this valley is conspicuous, especially along the flank of Skelwith Fell. About half a mile from Brathay Church, on the Coniston road, the surface of the "Coniston flags" has been bared for several yards, and exhibits a system. of parallel flutings and striations, ranging S.S.E. This, within the limits of a few degrees, is the invariable direction of the striation south of the Brathay, and is independent of the form of the ground. The ice has pursued the same southerly course, whether ascending along the flanks of Skelwith Fell and Oxen Fell, or crossing the more depressed district at the head of Windermere; and it is astonishing how completely the inequalities of the ground seem to have been ignored.

An interesting section in Boulder-clay occurs N.W. of Loughrigg Tarn; it is of a deep red colour, resting on a grooved surface of felspar porphyry. The chloritic slates, which form the flanks of Great Langdale along the northern shore of Elter Water, are remarkably ice-moulded and grooved, to an elevation of 650 feet above the lake. Along the bottom of the valley roches moutonnées protrude, and boulders are abundantly strewn over the surface. The striæ range E. 10° S. in the line of the axis of the valley. At the village of Langdale, ice-moulded, polished, and fluted surfaces are remarkably fresh; and I would here remark, that the fine-grained slates of this part of Langdale exhibit the glacial striations in greater perfection than the harder trap-rocks of other parts, the latter having suffered more by weathering along the planes of cleavage. The oval and generally glaciated surfaces are sufficiently evident in the bedded porphyry which sets in above the village, but the fine groovings and striations are seldom exhibited.

Langdale Moraine.-Although the bottom and sides of Langdale present frequent instances of perched and strewn blocks and accumulations of gravel, there does not appear to be any object which might not have been produced by floating ice, until we arrive within half a mile of its upper extremity. Indeed, the ascent of this valley is so very gradual that it is improbable glacial ice could have moved along it; and the upper limit of the ice-worn surface seems to preserve a perfect level along the sides of the valley, which may be distinctly traced by the eye, from some positions above Elter Water, at 650 to 700 feet above that lake. Combining these circumstances with the fact that we meet with no undoubted instance of a moraine till we approach the head of the vale, it appears to me the more probable supposition, that the glacial evidences of this valley (as already stated) are attributable to the action of icebergs floating down when it was filled by an arm of the sea, to a level of about 1000 feet.

On approaching the ascent to the Stake Pass, at a height of about 500 feet, we meet with a well-defined moraine, formed of large rounded mounds of gravel, and strewn with boulders. This moraine at one time probably crossed the entire valley,

and must have produced a lake, till the mountain torrent, which everywhere has left evidences of its power, hewed for itself the channel which it now occupies. Although at so low an elevation, this moraine to all appearance has never been covered by the sea, and is one of several proofs that the glaciers have occupied the valleys after the sea had retired. (Pl. I., fig. 3.) The head of the valley is surrounded by a steep though not vertical wall of felspar porphyry and slate, stretching from the northern shoulder of Bowfell to the Langdale Pikes. The Stake Pass crosses the lowest portion of the ridge at an elevation of about 1500 feet, and forms a portion of the central watershed. On reaching the level of the pass, I was surprised to find the whole surface covered by mounds of gravel and boulders, extending some distance down the valley towards Borrowdale, and bounded by the crags of the northern shoulder of Bowfell (see Plate I., fig. 2). It is not improbable that this may be the terminal moraine of a glacier which descended along the higher reaches of the Derwent; but its occurrence in such a position appears to me not easily accounted for. The aspect of this moraine resembles a vast collection of tumuli; and fancy might well point to it as the sepulchre of a battlefield.

Plate I., fig. 3, represents the relative positions of the moraines of Great Langdale and of the Stake Pass.

Little Langdale.-Towards its head, Little Langdale parts into three branches. The north branch leads up to Blea Tarn, and contains fine examples of roches moutonnées, perched blocks, boulders, and glacial striations. These last, near the Tarn, point S.E. down the valley!

The central branch contains a well-formed moraine (Plate II. fig. 4), which probably at one time formed an embankment across the valley, in which case it would have produced a lake; but the mountain torrent has scooped a channel between the southern side of the moraine and the mammillated bosses of porphyry which form the ridge.

The southern branch skirts the base of Wetherlam; and, judging by the remarkably smoothed and grooved surfaces of its flanks, has been subjected to an intense degree of glacial action. Around the copper works, the striations are visible