for the stones and clay and sand to gather behind it. One may notice the same kind of action upon the bed of a stream, where the sediment tends to collect in the rear of prominent stones and boulders. And we can hardly fail to have observed further that the sediment of a river often arranges itself under the action of the current in long banks, which run parallel to the course of the water. Underneath the ice-sheet the stones, sand, and clay behaved in the same way. Behind projecting rocks in sheltered nooks and hollows, they accumulated, while in places exposed to the full sweep of the ice-stream they were piled up and drawn out into long parallel banks and ridges, the trend of which coincided with that of the ice-flow. The presence of confused and irregular patches and lenticular beds of sand, clay, and gravel in the till is not difficult to understand when we know that there is always more or less water flowing on underneath a glacier. Such streams must assort the débris, and roll angular fragments into rounded stones and pebbles; but the materials thus assorted in layers will ever and anon be crushed up so as to be either partially or wholly obliterated by the slowly moving glacier. As the stones and clay were derived from the underlying rocks, it is no wonder that the colour of the till should vary. In the Silurian tracts it is pale yellowish, or bluish grey, and the stones consist chiefly of fragments of Silurian rocks, all blunted and smoothed, and often beautifully polished and striated. When we get into the Red Sandstone region of the low-grounds the colour of the clay begins by-and-by to change, and fragments of red sandstone become commingled with the Silurian stones, until ere long the colour of the deposit is decidedly red, and sandstone fragments abound. Everywhere the stones show that they have been carried persistently in one direction, and that is _out from the watershed, and down the main valleys_.