was near the Hundred Islands about midway of its length. One old valley headed up in Northwest Bay, and the other probably did not come much south of Sabbath Day Point. Faulting connected them, however, and the damming by the drift of the old outlet to the south at the head of Kattskill Bay and of the normal outlet to the north into the Trout Brook Valley or possibly eastward into Lake Champlain at Blairs bay, produced the present composite lake, with its precipitous mountainsides and wild rugged scenery. The old relations are greatly obscured by the glacial drift. Schroon Lake is another ponded and drowned river valley, with both a sluggish outlet and a sluggish inlet, the former with wonderfully developed terraces on either side, running like railway embankments as they mark old periods of high water. The other notable series of lakes, like the Fulton Chain, have been produced by the drift in the old valleys, which were the great drainage lines before the glacial epoch. With moderate portages they can be navigated long distances.
Some of the smaller lakes are in fault valleys and not infrequently are on the divides so as to be the sources of the streams. The Cascade lakes (Fig. 7) between the Keene Valley and Lake Placid are good illustrations. Although now two, they were once continuous and have evidently been divided by a landslide.
The Ice Invasion of the Glacial Epoch.—Coming after so long an interval during which the Adirondack area was land the Labradorean ice sheet possesses exceptional interest. All the scratches so far observed and recorded point to a source on the northeast. The glacier advanced from this quarter, and, as has been shown in some detail by Dr. I. H. Ogilvie, rode over the highest mountains and apparently filled the valleys with stagnant ice, since, except in the borders, scratches are rare. In fact the Paleozoic strata, in the lower confines, where the Champlain clays soon buried and preserved the scratches, are the most prolific sources of observations. The hard Precambrian rocks have mostly lost them by weathering.
The ice sheet must have found the Adirondacks covered with a heavy mantle of the products of decomposition. The long time during which the mountains had been land could have had no other result than this. It also found them of rugged topography much as now, because if we believe or assume, as is reasonable, that the Cretaceous peneplain was broken up into the flat-topped blocks by the preglacial faulting, the region must have presented a very irregular barrier in the pathway of the ice. The ice has left not a few characteristic topographic forms as the result of its action. Cirques appear on the flanks of several of the higher mountains, as, for example, on the northwestern side of Giant and the eastern side of Whiteface. Projections of the ridge run out in each case at right-angles to the main axis, affording
