392 THE CANADIAN NATURALIST. [Vol. viii. 



River, in a small, but fertile and beautiful valley, the main 

 stream of the Rogue River is crossed by a good bridge. Between 

 this river and the South Umpqua, is a rugged and irregular coun- 

 try, in which steep-sided hills are huddled together, but in which 

 also several narrow but fertile valleys are concealed. The Ump- 

 qua once reached, is followed to Roseburg, whence a railway 

 stretches to Portland, near the junction of the Columbia and 

 Willamette rivers. 



From the Sacramento River to this point all the streams crossed 

 flow westward to the coast, transverse to the proposed Oregon 

 and California Railway, the completion of which will be a very 

 difficult matter. So far no traces of general glaciation, or de- 

 posits like the northern drift, have been encountered. The hills 

 appear to have been subjected to prolonged sub-aerial weathering, 

 the rocks, when bared on their slopes, being generally soft and 

 decomposed at the surface. The soil covers the hills almost 

 uniformly from base to summit, except where the slopes are 

 remarkably steep ; and is probably in most instances a product 

 of waste of rock nearly in place. The bottoms of the valleys, 

 though occasionally flat, and suggesting the existence of former 

 lakes, or that the sea may at one time have flowed into them, 

 are generelly characterized by broad coalescing fan-shaped deltas 

 of the lateral streams. The summits and higher slopes of the 

 hills are generally stony and gravelly, while the valleys have a 

 clayey or loamy soil, which graduates into the former irregularly 

 on the slopes. There is a remarkable absence of any well-marked 

 terraces or benches ; though, besides those already mentioned, a 

 probable terrace was observed about thirteen miles above Rose- 

 burg, on the Umpqua, with an estimated elevation of 540 feet 

 above the sea. The general impression conveyed by the country 

 is, however, that there are no true terraces, which may arise 

 from the fact that the region has never been flooded, or if flooded, 

 that sufficient available material (detritus) for the formation of 

 distinct terraces has not been at hand, or, lastly, on the supposi- 

 tion that the process of obliteration seen actively in progress in 

 the somewhat similarly circumstanced dry southern interior of 

 British Columbia, has here been so long continued as to remove 

 almost entirely the old water marks. The hills are everywhere 

 seamed with gullies which form the terminations of small valleys, 

 all of which are connected, uniting as they descend toward the 

 main stream. The almost complete absence of lakes or ponds, or 



