Devil's lake in glacial times.—While the ice edge was stationary in its position of maximum advance, its position on the north side of the main quartzite range was just north of Devil's lake (Plate [XXXVII]). The high ridge of drift a few rods north of the shore is a well defined moraine, and is here more clearly marked than farther east or west, because it stands between lower lands on either side, instead of being banked against the quartzite ridge. North of the lake it rises about 75 feet above the water. When the ice edge lay in this position on the north side of the range, its front between the East bluff and the Devil's nose lay a half mile or so from the south end of the lake. In this position also there is a well defined moraine.

While the ice was at its maximum stand, it rose above these moraine ridges at either end of the lake. Between the ice at these two points there was then a notable basin, comparable to that of the present lake except that the barriers to the north and southeast were higher than now. The melting of the ice supplied abundant water, and the lake rose above its present level. The height which it attained is not known, but it is known to have risen at least 90 feet above its present level. This is indicated by the presence of a few drift bowlders on the West bluff of the lake at this height. They represent the work of a berg or bergs which at some stage floated out into the lake with bowlders attached. Bowlders dropped by bergs might be dropped at any level lower than the highest stand of the lake.

Other lakes.—Another glacial lake on the East quartzite bluff has already (p. [120]) been referred to. Like the Devil's lake in glacial time, its basin was an enclosure between the ice on the one hand, and the quartzite ridge on the other. The location of this lake is shown on Plate [XXXVII] (s). Here the edge of the ice, as shown by the position of the moraine, was affected by a re-entrant curve, the two ends of which rested against the quartzite ridge. Between the ice on the one hand and the quartzite ridge on the other, a small lake was formed. Its position is marked by a notable flat.

With the exception of the north side, and a narrow opening at the northwest corner, the flat is surrounded by high lands. When the ice occupied the region, its edge held the position shown by the line marking the limit of its advance, and constituted an ice barrier to the north. [9] The area of the flat was, therefore, almost shut in, the only outlet being a narrow one at t, Plate [XXXVII]. If the filling of stratified drift which underlies the flat were removed, the bottom of the area would be much lower than at present, and much lower than the outlet at t. It is therefore evident that when the ice had taken its position along the north side of the flat, an enclosed basin must have existed, properly situated for receiving and holding water. Since this lake had but a short life and became extinct before the ice retreated, its history is here given.

At first the lake had no outlet and the water rose to the level of the lowest point (t) in the rim of the basin, and thence overflowed to the west. Meanwhile the sediments borne in by the glacial drainage were being deposited in the lake in the form of a subaqueous overwash plain, the coarser parts being left near the shore, while the finer were carried further out. Continued drainage from the ice continued to bring sediment into the lake, and the subaqueous overwash plain extended its delta-like front farther and farther into the lake, until its basin was completely filled. With the filling of the basin the lake became extinct. The later drainage from the ice followed the line of the outlet, the level of which corresponds with the level of the filled lake basin. This little extinct lake is of interest as an example of a glacial lake which became extinct by having its basin filled during glacial times, by sediments washed out from the ice.

Near the northwest corner of this flat, an exposure in the sediments of the old lake bed shows the curiously contorted layers of sand, silt, and clay represented in Plate [XXXVIII]. The layers shown in the figure are but a few feet below the level of the flat which marks the site of the lake. It will be seen that the contorted layers are between two series of horizontal ones. The material throughout the section is made up of fine-grained sands and clays, well assorted. That these particular layers should have been so much disturbed, while those below and above remained horizontal, is strange enough. The grounding of an iceberg on the surface before the overlying layers were deposited, the action of lake ice, or the effect of expansion and contraction due to freezing and thawing, may have been responsible for the singular phenomenon. Contorted laminæ are rather characteristic of the deposits of stratified drift.

After the Ice Had Disappeared.

As has already been indicated (p. [101]), the irregular deposition of glacial drift gave rise to many depressions without outlets in which surface waters collected after the ice had disappeared, forming ponds or lakes. So abundant are lakes and ponds and marshes in recently glaciated regions and so rare elsewhere, that they constitute one of the more easily recognized characteristics of a glaciated region.

After the ice had melted, the mantle of drift which it left was sometimes so disposed as to completely obliterate preglacial valleys. More commonly it filled preglacial valleys at certain points only. In still other cases a valley was not filled completely at any point, though partially at many. In this last case, the partial fillings at various points constituted dams above which drainage was ponded, making lakes. If the dams were not high enough to throw the drainage out of the valley, the lakes would have their outlets over them. The drift dam being unconsolidated would be quickly cut down by the out-flowing water, and the lake level lowered. When the dam was removed or cut to its base, the lake disappeared and drainage followed its preglacial course.

In case the valley was completely filled, or completely filled at points, the case was very different. The drainage on the drift surface was established with reference to the topography which obtained when the ice departed, and not with reference to the preglacial valleys. Wherever the preglacial valleys were completely filled, the postglacial drainage followed lines which were altogether independent of them. When preglacial valleys were filled by the drift in spots only, the postglacial streams followed them where they were not filled, only to leave them where the blocking occurred. In the former case the present drainage is through valleys which are preglacial in some places, and postglacial in others.