here abrupt and precipitous, consisting of cut banks of stratified clay; in other places more receding, but by a gradual slope rising, beneath dense foliage, to an equal elevation.
At this season of the year the water being high and the current swift, we made good time, covering a distance of sixty miles for the first full day's travel. About noon on the 2nd, having reached a narrow part of the river, very remarkable massive walls of ice were found upon either bank, some distance above the water's edge. These walls were of irregular thickness, and from eight to ten feet in height; but the most striking feature about them was that they presented smooth vertical faces to the river, although built of blocks of every shape and shade from clear crystal to opaque mud. They extended thus more or less continuously for miles down the river, and had the appearance of great masonry dykes. The explanation of their existence is doubtless as follows: Earlier in the season the narrowness of the channel had caused the river ice to jam and greatly raised the water level. After a time, when the water had reached a certain height and much ice had been crowded up on the shores, the jam had given way and caused the water to rapidly lower to a considerable extent, leaving the ice grounded above a certain line. Thus the material for the wall was deposited, and the work of constructing and finishing the smooth vertical face was doubtless performed by the subsequent grinding of the passing jam, which continued to flow in the deeper channel. After the passing of the first freshet, and the formation of these great ice walls, the water had gradually lowered to the level at which we found it.
