182 DYNAMICAL GEOLOGY. 



ridges) are combined into an amphitheater or circ. In Fig. 163, ArsB repre- 

 sents tlie course of the stream, as in Fig. 162; and Ae/B the eroded ridge, 

 which has lost at e much of its height. 



The ascent of tlie mountain by following the valleys is in such a case 

 wholly impossible ; it can be accomplished only by finding the ridge that has 

 held on to its summit connection with the peak. On Tahiti the ridge by 

 which the author made his ascent to h, the peak called Aorai, about 6000 feet 

 in height, narrowed to two or three feet, and for a short distance to a single 

 foot, putting risks into the excursion, since the slope either side fell off for 

 1000 to 2000 feet at an angle of 60° to 70°. Between h and a (the highest 

 peak, Orohena) the "divide" was reduced in height more than 1000 feet, and 

 the summit at h was but six feet broad. All the outlines of the original 

 crater had disappeared. The lavas usually lie in beds dipping seaward, but 

 those of the central precipices were without bedding. 



From the steps in the work of erosion over such isolated volcanic moun- 

 tains it becomes evident that further progress would result in narrower, 

 thinner, and if possible steeper ridges ; and, even when nearing the end, in 

 sharp crests and ridges, which finally would be likely to disappear through 

 weathering agencies. A flattening of the mountain would come at the very 

 end, and not be a step in the progress toward it. 



These explanations show that a river rising in high mountains has (1) its 

 torrent-portion, and (2) its river-portion, along which it is bordered by flood- 

 grounds. 



The river-portion consists (1) of an upper section of rapid waters, along 

 which erosion at bottom is continued, and the amount removed exceeds that 

 of deposition ; (2) a section of feebler descent and slower flow, where the 

 removal by erosion in floods does not exceed that of subsequent deposition, 

 so that the stream has ceased efficient work. It has reached base-level — as 

 the condition has been termed by J. W. Powell. This base-level section 

 may end below in a decrepit portion, over which deposition along the bed 

 exceeds the amount removed in floods, so that thus a silting up of the chan- 

 nel, and also a corresponding rise of the flood-grounds, go on. 



In the small Pacific islands these sections of the river-portion of a stream 

 are short and not always present. But on the western side of Maui there are 

 remarkable examples of a decrepit ending ; for, while the valleys in the wet 

 and cool mountains are wide and profound, as the map shows, the stream over 

 the leeward (and hence nearly rainless) plain at the western foot is reduced 

 to a narrow trench, which part of the time is dry. 



3. River valleys of the continents. — Over a continent where declivities are 

 long, and the gently sloping plains have large extent, — often hundreds of 

 miles in width, — each of the divisions of the river-portion of a stream, that of 

 rapid-working waters and that of base-level, is often of great length. More- 

 over, along many streams there are often several base-level portions, made by 

 obstructions ; but where this is the case, as Powell remarks, it is evidence of 

 the relatively recent origin of the stream ; for the wear of ages tends to 



