ing former physiographic features and producing others anew. This phase of glacial study may be best presented under three headings.
(a) Glacial Erosion. The eroding action of pure ice upon firm rock, varying in hardness from that of limestone to quartzite, is apparently slight and limited to a smoothing and polishing effect. When the glacier is shod with rock fragments, as is frequently the case, and has considerable thickness, the erosive effect may be great if the action is prolonged. Hard rocks are gouged, scratched and planed and the fragments reduced to pebbles, sand and clay. The glacier's rock tools by which this action is accomplished are bruised, battered, planed and scratched and the edges and corners are more or less rounded in a manner entirely characteristic of glaciers. When a glacier of considerable thickness moves over a jointed, stratified rock, especially if the dip of the strata is in the direction of the movement, masses of rock may be detached bodily, giving rise to what is termed plucking. By this action a glacier may leave its bed rougher than it found it, and furnish the sites for lakelets, such as the exquisite lakes Agnes and Louise. An unusually fine example of this type of glacial erosion may be seen near the head of Paradise Valley, where blocks of quartzite as large as small houses have been disrupted from the parent bed and shifted but a short distance. Standing upon the undisturbed portion of the beautifully glaciated bed and looking down the valley it is difficult to escape the conviction that many feet of strata have been similarly removed. Many valleys in the Rockies and Selkirks appear to have been deepened and given their characteristic U-shape by alpine streams during the maximum period of glaciation. Their side walls, up to a certain height, have been smoothed and mountain spurs uniformly truncated, as well shown upon the Lake Louise side of Mt. Fairview. Glaciers exert this erosive power to their
