15 Armstrong and Tipper (1948) have summarized the evidence that the regional ice movement followed a large arc trending southeast, east, and northeast through north-central British Columbia. The amount of erosion accomplished by the Cordilleran Ice-sheet in northern British Columbia is believed to be slight (Holland, 1940, p. 20; Watson and Mathews, 1944, p. 37; Armstrong and Tipper, 1948, p. 306). Davis and Mathews (1944) have pointed out that the activity of the ice- cover in the Cordillera was probably at all times influenced to some extent by the underlying topography and that conditions in the mountains probably never developed from a mountain ice-sheet phase into a true continental ice-sheet phase. During the mountain ice-sheet phase, erosion was limited, and was restricted to smoothing and rounding of peaks and ridges, without effectively reducing the relief. The land surface so produced has been almost completely destroyed by later valley and mountain glacia- tion in the western part of Aiken Lake map-area, and is preserved in only a few places, rarely more than 50 acres in extent, on the crests of ridges (Plates II B, III A). Elsewhere in the map-area, bold, rounded summits and smooth ridges have, apparently, been but little modified since the disappearance of the ice-sheet. VALLEY AND Mountain GLACIATION The characteristic land forms in Aiken Lake map-area are erosional forms produced by valley and mountain glaciers. The most striking of these forms are the heavily glaciated trunk valleys of the major streams, and through valleys (Davis and Mathews, 1944, p. 408). The east and southeast trending parts of all the large stream valleys are broad, relatively straight or smoothly curving, with low gradients and with a wide floor and concave slopes giving a U-shaped cross-profile (Plate I). These valleys are connected by broad, low passes and straight, through valleys, also with typical concave cross-profile, that divide the terrain into more or less rectangular units. One of the best-developed through valleys extends from the valley of Discovery Creek, south of the map-area, northward across the valleys of Osilinka River and Thane Creek, is followed for a short distance by Vega Creek, and continues across Tenakihi Creek Valley to the major trough of Tutizika and Mesilinka River Valleys. Almost in a direct line with this valley, a through valley extends north from Blackpine Lake to Swannell River Valley, down the north-trending part of Swannell Valley, and across to Ingenika River Valley in line with Pelly Creek Valley. A score of other through valleys may be recognized in the area and although all must have been initiated on older weathered or stream-worn gaps, their development is essentially the work of ice erosion, and the resulting reticulate valley pattern is never developed to an equal extent by stream erosion alone. The through valleys and glaciated trunk valleys would appear to indicate that the area was subjected for a relatively long time to an ice-cover that completely filled the valleys, but whose motion was controlled by the valley walls. This stage of ice occupation was passed through at least twice, and four times if the area was freed of ice during an interglacial period. There is some evidence that a large part of the erosion was done by valley glaciers that preceded the last advance of the ice-sheet.