40 ME. J. A. DOUGLAS OX GEOLOGICAL SECTIONS [April IQI4. 



west of the La Paz valley parallel to the main Cordillera, com- 

 parable with the ridge which Dr. J. W. Evans has shown to form 

 a watershed between Lake Titicaca and the valleys of Illabaya 

 and Sorata. 



Snch a ridge would have divided the eastern Altaplanicie into 

 two parts. On the west probably existed a depression of no great 

 depth, which was levelled up by a widespread alluvial covering of 

 gravel and marl with recent freshwater mollusca, during a long 

 period of gradual recession of the lake- waters, and also on the south- 

 west by a series of gravel-terraces left by the Desaguadero River as 

 it deepened its course in the underlying Cretaceous sandstones. On 

 the east existed a much deeper longitudinal depression, which was 

 probably distinct from that of the lake-system, for it is extremely 

 doubtful whether the waters of Lake Titicaca ever overstepped the 

 intervening barrier. This eastern depression lay parallel to, and at 

 the base of, the Cordillera ; it, also, was completely filled up by a 

 vast accumulation of alluvial detritus. 



Its great depth is well exhibited in the valley of La Paz, which 

 has been excavated b}^ the river to a depth of over 1600 feet with- 

 out reaching bed-rock. The almost perpendicular sides of the 

 valley are composed of horizontally-bedded clays, sands, and gravel 

 containing large boulders of granite and Palieozoic quartzite derived 

 from the adjacent Cordillera ; these deposits are, in places, highly 

 auriferous. Near the summit, as noted by Forbes, occurs a con- 

 spicuous white band of volcanic tuff. 



With regard to the mode of infilling of this deep depression, if it 

 be assumed that a breach through the Cordillera was already in 

 existence, as appears probable, it might be suggested that during 

 the final stages of elevation of the range the water-supply was 

 diminished with the extension of the glaciers ; and the La Paz 

 Biver was unable to excavate its bed with the rapidity necessary to 

 keep the outlet sufficiently open to allow of the removal of the 

 torrential or fluvioglacial deposits, which, it may be supposed, were 

 subsequently swept down in vast quantities from the Cordillera by 

 the melting ice at the close of the Glacial Period. 



Before discussing the origin of Lake Titicaca, it is first necessary 

 to give a brief description of the gravel-terrace deposits of the 

 Desaguadero. This river, flowing southwards from Lake Titicaca 

 to Lake Poopoo, crosses the line of section near the village 

 of Calacoto, where it is joined by its main tributary, the Mauri. 

 Between this point and the village of Callapa three distinct 

 terraces are visible at different levels. The highest, situated some 

 200 feet above the present river, is well developed in the neigh- 

 bourhood of Ulloma, where it is plainly seen lying horizontally 

 on the upturned edges of red Cretaceous sandstones. It is made up 

 of sands, clays, and coarse quartzose gravels, the last-named being 

 often cemented into a compact siliceous conglomerate. It is highly 

 fossiliferous in this locality, and a number of mammalian bones 

 were obtained, all of which also occur in the Pleistocene Pampan 



