Doon lies. The lake itself has its head fairly within the amphitheatre, and it stretches northward for six miles with an average breadth of rather less than half a mile. Down this valley there must have been a vast pressure of the ice that filled the wide hollow, and here, consequently, we might expect to find the most marked evidence of glacial erosion. And, in truth, the whole of the glen occupied by the lake bears the impress of the ice. The ledges of rock that project into the water are as smooth and polished as a well-worn pavement, but covered with long ruts and grooves that run parallel with the direction of the valley. Near the head of the loch some of these projections, consisting of hard, dark greywacké, traversed with grey granite veins, have been shorn as smooth as if a committee of geologists had determined to lay open to the best advantage a complete section of the junction of the two rocks. The rocky islets are similarly polished and striated, and show the smooth convex flowing outline so characteristic of roches moutonnées. One of them, indeed, has been called by the people of the district the “whale's back.” Wherever the rock along the side of the valley has been recently bared it retains the same ice-worn surface. Loch Doon is a true rockbasin. At its foot the barrier of rock which keeps back the water is smoothed and striated, the direction of the markings proving that the ice has come

up out of the lake. Not only so. For a hundred feet or more above the level of the water the rocks which rise above the end of the lake are similarly worn. There can be no doubt, I think, that the ice which filled the hollow of Loch Doon went up the slope at its northern end and so passed down into the valley beyond. The deep gorge of Glen Ness, by which the river escapes, seems to be partly the work of the glacial period, but much deepened since then by the wasting action of the roaring torrent which fills the narrow chasm from side to side. Besides the polished and striated rocks, the Loch Doon valley shows abundant moraine rubbish. Among the detritus, granite-boulders brought from the far mountains are especially numerous. They are sometimes thickly clustered in patches along the margin of the loch, or heaped on its islets. Some of the islets, indeed, look like the tops of moraine mounds appearing above the water. In short, there is no locality in the south of Scotland where the existence and effects of ancient glaciers can be more impressively seen, and none where, as it appears to me, the glacial origin of such rock-encircled lake-basins is more clearly evinced.

Since the last remnants of the great snow-fields and glaciers melted away from the uplands of the south of Scotland, there has been a good deal of minor change in the general features of the district. The crags and cliffs where the naked rock comes out

into the light have suffered from long centuries of exposure to the elements, and their ruins are seen below them in streams of loose rubbish and piles of large blocks of stone. Many a runnel has deepened its first channel into a gully that runs as a narrow gash down the smoothed hill-side. The brooks and rivers too have been busy in eating away their banks and lowering their beds. Some of the most picturesque ravines, such as that of the Crichhope Linn in Nithsdale, have been cut by running water since the glacial period. The lower terraces and alluvial haughs that flank the margins of the larger streams, have likewise been made since then. Into the changes due to vegetation-the growth and disappearance of forests, and the formation of peat-moss

-I do not here enter, reserving until a following chapter a brief reference to the nature and proof of such changes over the whole country.

CHAPTER X.

THE MIDLAND VALLEY.

GEOLOGICAL STRUCTURE.

BETWEEN the southern flank of the Highlands and the northern edge of the uplands of the pastoral counties, lies that wide hilly plain or valley which, for want of a better name, I have been accustomed to call the Midland Valley. It is only in the broad sense, as a band of lower ground between two mountainous tracts, that it can be spoken of as a valley; nor can a district so plentifully dotted with hills, and even traversed by long chains of heights, be in strictness termed a plain. Geologically, however, this belt of lowland country is a valley. Either side is bounded by Lower Silurian rocks, rising on the north in the rugged schists and gneisses of the Highland mountains, and on the south in the contorted greywackes and shales of the Southern Uplands. Between these two boundaries the rocks belong to newer formations, which may be broadly looked at as

dipping away from the flanking hills, the oldest strata lying along the borders and the newest along the middle. This arrangement, however, has been sadly disturbed by numberless faults and depressions, so that some of the older rocks are found coming up to the surface far into the centre of the plain, while here and there the later rocks, in place being confined to the middle, stretch across to the margin and even go over into the Silurian district beyond.

Along the northern border of this valley there runs a broad well-defined belt of Lower Old Red Sandstone. It is exposed on the coast between Stonehaven and the Firth of Tay, whence it extends south-westward across the island to the Firth of Clyde. It contains a great abundance of igneous rocks-the product of volcanoes that were active in Scotland during the deposition of the Lower Old Red Sandstone. These volcanic materials form numerous detached eminences, as well as the long hillranges of the Sidlaws and the Ochils. There is no corresponding broad belt along the southern borders of the valley. The Lower Old Red Sandstone does appear there in a tract of moory heights between the Pentland Hills and the Ayrshire Coal-field, and has there been cut through in the well-known ravines and Falls of the Clyde. But it is surrounded by Carboniferous strata which overlie it and stretch away south to the Silurian uplands. The Upper Old Red