GEOLOGICAL SURVEY OF THE TERRITOEIES. Z6d 



becomes a question why the river has thus deserted its own proper valley 

 and has turned off so far to the east to find another outlet. Four old ter- 

 races are strongly marked along- the river, and the third, above its present 

 level, forms broad plains, on both sides of the river, below the mouth of 

 Buffalo. Moreover, the old river-gravel, consisting mostly of quartzites, 

 runs to the top of the island-like knobs in the center of the valley as 

 well as of some of the hills which border it. Crossing westward over 

 these broad plains, which are mostly covered with sage-brush, except 

 about the isolated ponds, which seem to give evidence that much of this 

 third terrace has been worked over by the river and cut up by bayous, 

 we cross a narrow belt of spruces covering the surface of the fourth ter- 

 race, and then descend to the old deserted channel of the river, which is 

 nearly on a level with the third terrace. Crossing this valley directly 

 toward the mountain, we come at once to a series of high, steep, narrow 

 ridges, covered with immense masses of granite and heavily timbered. 

 Within thalast of these concentric ridges, we come to a small lake, lying 

 at the mouth of a deep caGon which runs far back into the mountain ; and 

 here at last we have the clew to the mystery. This canon, like all the 

 other large ones of this range, was occupied by a glacier, whose termi- 

 nal moraines now hem in this lake; and, though the old river- valley 

 seems to have been not fully blocked up by the flow of silt which passed 

 beyond the moraines, so that its general features are still plainly marked, 

 yet it is evident that the deposition which caused its bottom to slope 

 from west to east so much as we now find it to, was sufficient to con- 

 siderably delay the wearing down of this channel. The canon nest north 

 of this was also occupied by a glacier — a moraine-dammed pond will 

 probably be found in the mouth of it — and doubtless its debris aided in 

 checking the erosion. At that time, the flow of water through all this 

 region was much greater than at present, and the duplicating of chan- 

 nels in the broad bottom was far from being uncommon. If a channel 

 had then joined the eastern streams, though only as a bayou in time of 

 flood, there was reason for its being ultimately cut down, so as to drain 

 the western channel, which is the true valley of the Snake. The sum- 

 mit of the highest and outermost moraine is partly covered with river- 

 gravel, as if the stream had at least once reached this height after the 

 glacier began to recede. This crest is 122 feet above the bottom of the 

 adjoining part of the old river-channel, 222 feet above the present 

 level of the included lake, and 366 feet above Jackson's Lake. These 

 and other levels across the valley were taken at my request by Mr. Ber- 

 ing, who has furnished the data for the accompanying section from the 

 lake, across the old channel and the present one of Snake Eiver, to the 

 eastern hills. The length of the section is so great that it became necessary 

 to distort it by increasing the elevations considerably beyond their pro- 

 portional size. (Fig. 51.) We have called this lake, which is about two 

 miles long by a half mile wide, Leigh's Lake, after our guide, Kichard 

 Leigh, (Beaver Dick.) It appears to be mostly shallow, and has a small 

 island near its center. Can it be possible that the glacier, which was 

 formed by the flow from two canons whose junction is only a very short 

 distance above the lake, split again after it emerged, so as to leave this 

 island uneroded? JSTo special depression of the moraines opposite the 

 island was noticed; but examinations just at that point were not care- 

 fully made. ■ On the other hand, if we pass to the north end of the lake, 

 we find a flat divide, not 5 feet above its level, and less than a hundred 

 yards across, which separates it from a large beaver-pond, whose waters 

 escape through two or three other ponds and marshes to the head of 

 Jackson's Lake. A very trifling cut would thus give the lake a second 



