The direction of the ice-marks upon the solid rocks, and the trend of the "drums," as the parallel ridges of till are termed, show that the ice-sheet of Teviotdale and Tweed gradually turned away to the east and south-east as it swept round the north-eastern spurs of the Cheviots. Now we may well ask why the ice did not go right out into the North Sea, which is apparently the course it ought to have followed. The same curious deflection affected the great ice-stream that occupied the basin of the Forth. When it got past North Berwick, that stream, instead of flowing directly east into the North Sea, turned away to the south-east and overflowed the northern spurs of the Lammermuirs, bringing with it into the valley of the Tweed stones and boulders which had travelled all the way from the Highlands. It is obvious there must have been some impediment to the flow of the Scottish ice into the basin of the North Sea. What could have blocked its passage in that direction? At the very time that Scotland lay concealed beneath its ice-sheet, Norway and Sweden were likewise smothered in ice which attained a thickness of not less than five or six thousand feet. The whole basin of the Baltic was occupied by a vast glacier which flowed south into Northern Germany, and this sheet was continuous with glacier-ice that crossed over Denmark. When we consider how shallow the North Sea is (it does not average more than forty fathoms between Scotland and the Continent), we cannot doubt that the immense masses of ice descending from Norway could not possibly have floated off, but must actually have crept across the bottom of that sea until they abutted upon and coalesced with the Scottish ice, so as to form one vast _mer de glace_. Thus it was that the Scandinavian ice blocked up the path of the Scottish glaciers into the basin of the North Sea, and compelled them to flow south-east into England.[H] Had there been no such obstruction to the passage of the Scottish glaciers, it is impossible to believe that snow and ice could ever have accumulated to such a depth in Scotland. The Scottish ice reached a thickness of some three thousand feet in its deeper parts. It is evident, however, that had there been a free course for the glaciers, they would have moved off before they could have attained this thickness. And we can hardly doubt, therefore, that it was the damming-up of their outlet by the great Scandinavian ice-sheet that enabled them to deepen to such an extent in the valleys and low-grounds of Scotland. [H] In the extreme north of Scotland we find that the Scottish ice was, in like manner, compelled to turn aside and overflow Caithness from south-east to north-west.