But while the upper strata of the ice doubtless followed that particular course, it is obvious that this could not be the case with the under portion of the great sheet, the path of which would be controlled in large measure by the form of the ground over which the ice moved. The upper strata that overflowed the Outer Hebrides, as we have seen, were locally deflected again and again by important obstacles, and it is quite certain that the same would take place with the deeper portions of the ice-flow. It is well known that the sea along the inner margin of the Long Island is very deep. In many places it reaches a depth of 600 feet, and occasionally the sounding-lead plunges down for upwards of 700 feet. It would seem, however, that these great depths did not exist before the advent of the ice-sheet, but that the bottom of the Minch along the eastern borders of the Long Island was then some 250 or 300 feet shallower than now, the floor of the sea having since been excavated in the manner I shall presently describe. It is quite apparent, therefore, that the long ridge of the Outer Hebrides must have offered an insuperable obstacle to the direct passage of the bottom-ice out to the Atlantic. Here was a great wall of rock shooting up from the floor of the Minch, at a high angle, to a height ranging in elevation from 400 feet to upwards of 3000 feet. It is simply impossible that the lower strata of the ice that occupied the bed of the Minch could climb that precipitous barricade. They were necessarily deflected, one portion creeping to north-east and another to south-west, but both hugging the great wall of rock all the way. We see precisely the same result taking place in the bed of every stream. Let us stand upon an almost submerged boulder, and note how the water is deflected to right and left, and we shall observe at the same time that the boulder, by obstructing the current, forces the water downwards upon the bed of the stream, the result being that a hollow is dug out in front. Now, in a similar manner, the ice, squeezed and pressed against the Hebridean ridge by the steady flow of the great current that crossed the Minch, necessarily acted with intense erosive force upon its bed. Hence in the course of time it scooped out a series of broad deep trenches along the whole inner margin of the Long Island, the amount of the excavation reaching from 200 to 300 feet. Similar excavated basins occur in like positions opposite all the precipitous islands of the Inner Hebrides. Wherever, indeed, the ice-sheet met with any great obstruction to its flow, there excessive erosion took place, and a more or less deep hollow was dug out in front of the opposing cliff, or crag, or precipitous mountain. While, therefore, the upper strata of the ice-sheet overflowed the Outer Hebrides from south-east to north-west, the under portions of the same great ice-flow were compelled by the contour of the ground to creep away to north-east and south-west, until