We see now why it is that the hill-slopes opposite the Minch should, as a rule, be so much more sterile than those which slope down to the Atlantic. The full force of the ice was exerted upon the south-east front, in the rear of which there would necessarily be comparatively "quiet" ice. For the same reason we should expect to find much of the rock débris which the ice swept off the south-east front sheltering on the opposite side. Neither clay nor sand nor stones would gather under the ice upon the steep rocks that face the Minch. The movement there was too severe to permit of any such accumulation. But stones and clay and sand were carried over and swept round the hills, and gradually accumulated in the rear of the ice-worn rocks, just in the same way as gravel and sand are heaped up behind projecting stones and boulders in the bed of a stream. Hence it is that the western margin of Harris is so much less bleak than the opposite side. Considerable taluses of "till," as the sub-glacial débris is called, gather behind the steeper crags, and ragged sheets of the same material extend over the low-grounds. All the low-grounds of Lewis are in like manner sprinkled with till. Over that region the ice met with but few obstacles to its course, and consequently the débris it forced along underneath was spread out somewhat equally. But wherever hills and peaks and hummocks of rock broke the regularity of the surface, there great abrasion took place and no till was accumulated. Thus the position and distribution of this sub-glacial débris or bottom-moraine tell the same tale as the abraded rocks and glacial striæ, and clearly indicate an ice-flow from the south-east. This is still further proved by the manner in which the upturned ends of the strata are frequently bent over underneath the till in a north-westerly direction, while the fragments dislodged from them and enclosed in the sub-glacial débris stream away as it were to the same point of the compass. Not only so, but in the west of Lewis, where no red sandstone occurs, we find boulders of red sandstone enclosed in the till, which could not have been derived from any place nearer than Stornoway. In other words, these boulders have travelled across the island from the shores of the Minch to the Atlantic sea-board. Having said so much about the glaciation of Lewis and Harris, I need not do more than indicate very briefly some of the more interesting features of the islands further south.