486 FROM PIKE LAKE TO FORT UNION. 



from thence to Wall ah- Wallah is a level plain of barren sand. Below the fort, high basaltic 

 bluffs again rise on either side, affording some of the most picturesque scenery upon the river. 

 They continue for some twenty miles, below which they are interrupted, the plains coming down 

 to the river for a considerable distance. Towards John Day s river (the Mah-hagh) the bluff s 

 recommence, rising two or three steps and surmounted with hills. Thence approaching the 

 Dalles, the country becomes more and more broken. It is noticeable that in the basalt upon the 

 Columbia, particularly in the neighborhood of the Des Chutes river, (Wan-waw-wi) the basaltic 

 columns often converge at the top without the horizontality of the beds being interrupted. At 

 Willow creek, (the Hockspe,) the artemisia disappears, that river forming its western boundary 

 to the south of the Columbia. Below the Des Chutes the hills are freckled over with mounds, 

 not only on the summits, but the sides also; and it would seem that the forest once extended to 

 this limit, and has probably been driven back by fire. These mounds differ entirely in character 

 from those mentioned as occurring on the plains, and from those crossing the gravelly prairies 

 near the head of Puget sound, the origin of which is clearly due to water. 



At the Dalles the basalt is overlaid with coarse conglomerate, chiefly formed of masses of por- 

 phyritic rock imbedded in sandstone. This occasionally stands up in needles, which have been 

 protected from the weather by boulder caps. The hill-sides would seem to indicate that the 

 barrier at the entrance of the river into the mountains was suddenly broken, and the superficial 

 earth and strata swept away. Traces of terraces remain on the north side of this entrance at 

 a considerable elevation. The basalt, associated occasionally with conglomerate, lines the river 

 to within twenty-five or thirty miles of Fort Vancouver. The origin of the cascades has already 

 been referred to. There seems to be no question that a slide from the mountains on the south 

 side has actually taken place, and that the water has thus been backed up above it. No evidence 

 exists of any subsidence of the low alluvial shores, by which what is called the sunken forest 

 has been submerged, but that its position is simply due to this rise of water. 

 I have the honor to be, sir, your obedient servant, 



GEORGE GIBBS. 

 Capt. GEORGE B. MCCLELLAN, 



Commanding Western Division N. P. Railroad Exploration, c. 



FINAL REPORT OF LIEUTENANT C. GROVER, OF HIS EXAMINATIONS FROM PIKE LAKE TO FORT 



UNION. 



SIR: Pursuant to your instructions, on the 25th of June I left the Red river trail at Pike lake, 

 in order to pursue a more southern course to Fort Union, on the Missouri, than that marked out 

 for the main party of the expedition under your command. 



The principal topographical features of this immediate locality are those of a broad rolling 

 prairie, falling off gently to the south, and partially drained by small streams tributary to the 

 south fork of the Chippeway river. Many small lakes, a general peculiarity of the prairie 

 country previously passed over, are also found here ; they are mostly the result of imperfect 

 drainage, being fed by falls of rain, and are much less numerous in the dry season. Those that 

 appear to be permanent have sandy bottoms, and are generally thinly skirted with scrub-oak, the 

 principal growth of this section of the prairie. The formation is a loose drift, having more or less 

 gravel in its composition, according to its elevation. Granite boulders, of from a ton s weight 

 down, are sometimes seen, but they are not numerous. 



The south fork of the Chippeway river, which I crossed nine miles from Pike lake, winds 

 through rather a depression in the prairie than a valley, as the difference of level between its bed 

 and the adjacent country is but small. It has a general width of about thirty yards, and a depth 

 of about six feet. Its bed, as well as the bed of its small tributaries, is sandy. The velocity of 

 its current is about four miles an hour. Its banks are not wooded. 



