/. Bowman — Physiography of the Central Andes. 401 



lower than in the path of the La Paz. The Mapiri has eaten 

 its way back toward Lake Titicaca until bow a divide but 

 1500 ft. high separates it from the Titicaca basin ; and in a 

 geologic sense the capture of the Titicaca waters is imminent 

 from this direction to a degree but slightly less marked than 

 in the case of the La Paz. Further interest attaches to these 

 future changes in the disposition of the basin waters because 

 of the transition that is now in progress in the Titicaca basin 

 itself. Its waters are gradually receding from its shores 

 (tig. 19), and the discharge of the Desaguadero, its outlet, is 

 likewise decreasing. The climatic change which this expresses 

 is steadily progressing and, if continued, will lead to the com- 

 plete isolation of the Titicaca drainage. The lake will then 

 become salt for a period ; but it will again become fresh 

 when capture from the east or north supplies it again with an 

 outlet. These changes will occur in a short space of time 

 geologically speaking, and will undoubtedly occur if the forces 

 now in operation here are continued in their present direction. 



Further support is given this view by the direction in which 

 the headwater attack is taking place. The La Paz tributaries 

 west of the Cordillera Peal are all working most vigorously in 

 the piedmont deposits that front the range. The courses are 

 arranged in part regardless of the piedmont, in part in strict 

 conformity to it. The slopes of the piedmont are everywhere 

 arranged regardless of the main drainage. On the south side 

 of the La Paz amphitheatre one looks in vain for a slope 

 towards the amphitheatre. The drainage of the interior basin 

 begins at the very lip of the amphitheatre and runs away from 

 it. On the other side the mountain streams descend steeply 

 from the Cordillera Peal and then turn in semi-circular courses 

 (Hg. 14) toward the main axis of the La Paz valley. The 

 structure lines within the detrital material do not dip down the 

 present inclines of the valley head, but in a contrary direction. 

 "We must conclude from this evidence that the piedmont was 

 formed long before the La Paz river headed west of the main 

 range and that the present course of the river is due to head- 

 ward gnawing in very recent geologic time, a process still is 

 active operation. 



It is important to note that the rainfall and other conditions 

 are here very special indeed, and that a similar explanation 

 cannot be assigned to the other rivers of the plateau whose 

 courses are out of sympathy with the structure. For example, 

 the streams that cut across the old mountain axes between 

 Cochabamba and Oruro are well within the eastern divides 

 that bar the rain-bearing winds. They occur in a dry region 

 and, furthermore, one in which the warping of the peneplain 

 has been distinctly less marked than in the region east of the 



