Mounted on a fine horse, lent to him by an Indian, Clarke quickly reached the river, though the mountains to be traversed were rugged and broken in the extreme; but here the difficulties in tracing the course of the Columbia became so great, that it was evidently impossible to do so without spending the whole of the remainder of the year in the neighborhood. So far as we can make out from the various accounts of his expedition, Clarke did not actually reach any of the headwaters of the Columbia, which are situated between N. lat. 54° and 42°, but he made a sufficiently thorough survey of the mountains from which the more northerly of its two branches proceeds, to convince himself of the identity of that branch with the river of Viscaino and Gray. A result of this trial trip of secondary importance was that the leader recognized the hopelessness of attempting to descend the Columbia in canoes, and on his return to the Indian camp, it was agreed that a detour should be made in a northerly direction, so as to strike the river below the mountain pass from which it flows on its way to join its sister stream in N. lat. 46° 5′, W. long. 118° 55′. On the 23d August the expedition was once more en route. The sources of the Missouri had been discovered, the position of those of the Columbia determined, and the country between the two traversed more than once. All, therefore, that now remained to be done was to trace the great river of the West to the Pacific. Accompanied by a number of Shoshone guides, the Americans made their way toward the ocean by a rugged Indian path, halting now at one, now at another Indian village; and after much wandering to and fro, with terrible sufferings from cold and hunger, they entered a district inhabited by a people calling themselves Ootlashoots, who turned out to be members of one of the numerous inland Columbian tribes who inhabit the regions on the Pacific between N. lat. 52° 30′ and 45°.