0. H. Hershey — Rive?' Terraces of California. 243 



extreme narrowness. At several points from here upstream 

 there are in the small canon, where the rock is excessively 

 soft, traces of a lower terrace whose rock floor is usually 5 to 

 10 feet above the stream. They are indefinite, are due to a 

 shifting of the stream meanders and do not necessarily indicate 

 renewed uplift. From Cecilville downstream they are weakly 

 developed at a number of points, but nowhere are conspicuous 

 and will be ignored in the following discussion. 



In the vicinity of the old town-site of Petersburg, several 

 miles above Cecilville, the river was able to excavate a basin 

 in the area of the Salmon hornblende schist because of its local 

 shearing and partial alteration into chlorite schist. Here the 

 terrace system is typically developed with three main levels, 

 the lower trenched by the river to a depth of 30 to 40 feet. 

 The lower terrace may be traced through the next gorge 

 upstream. At its level there is a slight bench with gravel 

 deposits along both sides of the gorge, and the present canon, 

 with nearly perpendicular rock walls, has a depth below them 

 of 30 to 50 feet. This leads us into the Summerville basin, 

 excavated largely into an area of non-resistant granite. 



The altitude of the river at Summerville is about 3100 feet 

 above the sea. It flows in a rather wide canon (because the 

 rock is unusually soft), and 30 to 40 feet deep. From its 

 edges, before mining operations were begun, a gravel terrace 

 extended back several hundred feet and then the basin floor 

 rose in successive benches to a level of 300 feet above the 

 stream. Indeed, the entire terrace system was developed here 

 in unmistakable form. This is the farthest point up the 

 river at which it reaches a prominent development, as above 

 here the rocks are nearly uniformly hard, the canon narrow, 

 and the stream very high grade. Glacial deposits occur 

 within five miles upstream. Within the canon at many points 

 there are benches, by which the terrace system may be traced 

 into connection with the glacial series. This has been done. 

 However, it deserves treatment as a separate subject, but some 

 of the conclusions will be used in this paper. 



The most important features described in the preceding 

 paragraphs are, that each basin throughout the area discussed 

 contains a broad, flat, gravel floor, and that the streams no 

 longer flow at this level, but in a tiny canon trenched mostly 

 from 30 to 50 feet beneath it into the hard rock below. The 

 contrast between the broad valley floor and the narrow caiion 

 winding about in it is extremely prominent and certainly 

 indicates a change of conditions. The size of the canons 

 relative to that of the streams is everywhere so nearly alike as 

 to point indubitably to a like age for the flat valley floor in 

 each basin. E"ow, the development of the flat valley floor 



