Page:A Treatise on Geology, volume 1.djvu/314

 298 in the channels of rough streams, or in ordinary gravel beds.

The former boundaries of the glacial sea may be in some degree conjectured by the distribution of the northern drift, for it must have extended beyond the area of that drift. The depth to which the mountain regions from which the most abundant erratics have been distributed were sunk, may be in some degree conjectured; the Grampian, Cumbrian, and Cambrian Highlands, may, perhaps, have been 1500 feet, and the Alps, perhaps, 3000 feet lower. Under these conditions all the lower grounds of northern Europe would be submerged; the masses of mountainous land would appear above the waves, covered with snow, and surrounded by icy floods.

In this broad ocean, cold sea currents flowed most abundantly, though not exclusively, from the north and north-west, and drifted the rocks of Finland and Scandinavia to the plains of Russia and North Germany; the syenites of Criffell to the slopes of Skiddaw, the granites of Shap and Ravenglass to Yorkshire and Staffordshire. Agitations of water, anterior to and coincident with the elevation of the land toward its present height, have disturbed and modified the materials left by the ice-rafts, and mixed them with materials derived from wasting coasts, river floods, and the wearing of the sea-bed. These we shall now consider.

While some remarkable cases of dispersed boulders have engaged the attention of geologists following in the track of Saussure and De Luc, thousands of examples offered themselves of accumulations similarly at variance with the existing agencies of water; but they were never accurately studied till they acquired a new interest from the discussions of De Luc, and the splendid researches of Cuvier into the bones of quadrupeds which