plateau upon which snow gathers. Down some of these mountain- glens flow glaciers from five to ten miles long. These flow down the hollows, and end in muddy rivers, which flow on in the hollows through drift till they get to the sea in the fjords. Marks made by these, and drift upon and under them, grooves upon rocks, lateral and median moraines, and banks of drift, boulder-clay, and sands and gravels, arranged by the river and by the ice down to the delta, are all ranged parallel to the -sides of the rock-groove, to the ebb and flow of the tide, to the run of the main river, and to the motion of the glacier. Everywhere are marks to prove that the Norwegian glaciers are but remnants of glaciers, enormously greater, which have dwindled away.
In Gweebarra Pass I found marks near the sea parallel to the course of the river, which flows S.W. into a miniature fjord called Gweebarra Bay. Having seen Bergen glaciers and these two sets of marks, there is no difficulty about the meaning of this record. 1st. A glacier flowed S.W. seawards from the watershed down 750 feet some five or six miles to the fjord, and thence went off into the Atlantic. 2nd. Afterwards, when that glacier dwindled and shrank and melted, a smaller mountain-glacier still crossed the track of it, at right angles, from N.W. to S.E., descending from the top of Snow Mountain by a very steep incline, more than 2000 feet in a couple of miles or thereabouts. 3rd. That glacier dwindled and disappeared. But water melting from winter snows and rain-water follows both tracks. The glaciated rocks in Gweebarra Pass are wet by streams which run from the top of Snow Mountain down to the lake from N.W. to S.E., and then run from the lake S.W. along the main groove into the sea.
3. Glacier -forks. — In Norway, and in all countries where glaciers exist in any notable proportions, two commonly join and flow on together. At the point of junction they press alternately upon rocks, which they mark alternately, producing cross grooves upon a flat surface, or grooves in different directions on opposite sides of a rock between the streams. In front of Derreen House, at Killmakillogne Harbour, in the Kenmare river, in Kerry, at the junction of two deep glens, are marks of this kind of which I have copies. Cross striae are common elsewhere ; but here the cause is apparent.
4. Local systems. — In Iceland is a scarped hill called Erik's Jokull *. The sides are cliffs with talus heaps ; the top is a dome of ice whose base is a plateau of horizontal beds of igneous rock. As the snow-dome rises, the weight spreads the plastic base. All round this local system are stones crushed off the broken edges of flat beds of igneous rock ; and these are ranged in curved mounds and heaps about the base of the dome. These terminal moraines belong to the hill, and they were pushed outwards towards the circumference. They were formed under the ice ; for nothing but the sky is above this local system. At one point this snow-dome has extended its base down a hollow, and there is a small river-glacier of the usual
- See ' Frost and Fire,' vol. i. p. 428.
