NARRATIVE AND ITINERARY — CASCADES OF THE COLUMBIA 89 



peak so steep that to ascend it from the river side would probably be impossible. Near its base 

 is the mouth of a small stream called Wind river, which appeared to be formed by the junc- 

 tion of two branches behind the mountain. On the opposite bank the ridges, rising very abruptly 

 from the river, seemed to be formed of loose stones, which would render excavation for a road 

 very difficult and dangerous. 



It is in this vicinity that the celebrated submerged forests are mostly found. They consist of 

 numerous dead trees, stripped of their smaller branches, but still standing upright in thedeep water 

 near the river banks, and presenting every appearance of having grown there. As these trees 

 could never have grown under water, their present position has given rise to much speculation. 

 It has been suggested that they may have been transplanted from the neighboring mountains 

 by vast avalanches. It is possible that this may be true in a few places, but not in all, as they 

 are sometimes found where the position of the mountains precludes the idea. Another theory, 

 which I think much more plausible, is, that formerly a great slide occurred at the Cascades, 

 about fifteen miles below Wind mountain, and formed a dam; which, by raising the water above 

 it, submerged and killed the forests growing on the banks. The appearance of the Cascades 

 tends to confirm this idea. For four miles and a half the river rushes through a gorge, bordered 

 by high and very precipitous ridges, and only about nine hundred feet wide in the narrowest part. 

 Above, it expands into a kind of lake, about a mile and a half in width, containing several islands. 

 We reached the landing above the Cascades early in the afternoon, and were much pleased to 

 find that it was not raining there, as it very often does in September. I took a careful reading 

 of the barometer at the water's edge near the landing, and another at the foot of the principal 

 rapid, about a quarter of a mile below; subsequent calculations give a difference of level of 34 

 feet between these stations. I then walked with Captain Wells to the lower landing, about four 

 miles and a half from where we first went on shore, and took a third reading ; from which the 

 total descent in that distance was afterwards found by computation to be 61 feet. During high 

 water the portage is much shortened, as the boats can ascend nearly to the foot of the principal 

 rapid. 



The wild grandeur of this place, for which Rapids would be a more appropriate name than 

 Cascades, surpasses description. The river rushes furiously over a narrow bed filled with 

 boulders and bordered by mountains, which echo back the roar of the waters. The path, winding 

 through a thick forest on the northern bank, suddenly crosses an Indian burial place, where 

 whitened bones strew the ground on every side, and fill-one with amazement at the vast numbers 

 of the dead. Petrifactions are abundant ; and stumps closely resembling those of trees recently 

 cut, are often found to be solid rock, with bark and woody fibre perfectly preserved. Salmon 

 pass up the river in great numbers ; and the Cascades, at certain seasons of the year, are a 

 favorite fishing resort with the Indians, who build slight stagings over the water's edge, and 

 spear the fish, or catch them in rude dip nets, as they slowly force their way up against the cur- 

 rent. I passed through a rancheria of these savages, on my return in the evening from the 

 lower landing. Their hideous faces, strongly painted on the darkness behind them by the 

 flickering light of the fire around which they were crouching, and their mournful howls blended 

 with the baying of their dogs at my intrusion, harmonized well with the ceaseless roar of the 

 river, and the gloom of the forest cemetery, among the dry bones of which I turned to grope 

 my way. 



September 18. — This morning I returned by steamboat to Fort Dalles, after measuring 

 the width of the river, by triangulation, in two places near the principal rapid. It proved to 

 12 X 



