116 Geology of South America. 



the products of human industry discovered by M. D'Orbigny 

 at the Rio-Securi. 



It is, doubtless, sufficiently difficult to trace with certainty 

 the line of demarcation between the ancient raised beaches, 

 and those beaches which are from time to time elevated by 

 earthquakes on the coasts of Chili, as well as between the 

 alluvia of the present day and the vast alluvial deposits of the 

 great plains of the interior of America. Nevertheless, the fine 

 sand, sometimes containing shells, which covers the Pampas ; 

 the medanos, or ancient downs of the same plains ; the sands 

 which form elongated hills in the east of the province of Co- 

 rientes ; the gravels and sands of the great Bolivian plateau ; 

 the immense alluvia of the environs of Santa-Cruz-de-la- 

 Sierra, of the plains of Moxos, and of the province of Chiqui- 

 tos ; all these deposits, more modern than the Pampean for- 

 mations, cover them too generally, and too uniformly, not to 

 induce us to suppose that they are traces of a general pheno- 

 menon. The same is the case with the deep denudations, so 

 different from those produced by ordinary running waters, 

 which have furnished their materials. 



Ancient Beds of Torrents. — This is the natural place for 

 noticing one of the most curious observations made by the 

 author. M. D'Orbigny has pointed out at Cobija, at Arica, 

 and over the whole coast of the Pacific Ocean, ancient beds 

 of torrents, which, subsequently to the last movements of the 

 surface of South America, furrowed the whole slopes of the 

 Cordillera, from the summits to the shore. He is convinced 

 that these ancient beds of torrents, occurring in a region 

 where it has not rained in historical times, have not been 

 derived from local rains, but must be attributed to masses of 

 water which descended from the Cordilleras alone. At the 

 present day an aqueous cloud is never seen on the mountains 

 of the western side ; a patch of snow is never visible on this 

 slope of the Cordilleras. In order, then, to explain these tor- 

 rents whose traces are observed over a great space, it is ne- 

 cessary to suppose that the Cordilleras for a time received 

 rains or snow which they no longer receive ; an aqueous phe- 

 nomenon must thus have occurred on the mountains analogous 



