TRANSACTIONS OF SECTION E. 925 



Extent of the Plata Bas{?i in Ancient Times. — I shall try to show that the 

 Plata drainage area was, in a recent {geological period, mucli more extensive than 

 it is to-day ; that its most northern limit was in 10° 44' south latitude, and that 

 nearly the entire waters which now unite to form the Madeira River, the main 

 affluent of the Amazon, once flowed southward into a Pampean sea, which pene- 

 trated north, over the plains of the present Argentine Itepublic, to about 19° 

 south latitude. 



To elucidate this proposition, I must call attention to the topography of that 

 great Bolivian basin across which the whole northern and eastern slope of the 

 Bolivian Andes and the western slope of Brazilian Matto Grosso send their 

 abundant drainage to the falls of the Madeira. The present elevation of the upper 

 fall has been found by instrumental survey to be 547 feet above sea-level. The 

 distinguished engineer, Julio Pinkas, who made the survey, estimated the drainage 

 area of the Bolivian basin at 400,000 square miles. This is somewhat in excess of, 

 but perhaps more accurate than, my own estimate, made when I descended the 

 Mamor6 River and the Amazon, in 1871. Elsewhere, I have shown that the 

 Bolivian rivers lie upon a great plateau, high above the river Purus, as well as 

 above the lower Madeira. The Andes form the south-western and western rim of 

 this plateau, and, between the Mayn-tata and Purus, push low hills north-east 

 towards the falls of the Madeira. On the eastern side are the Matto Gro.sso high- 

 lands, and, south-east, the low Chiquitos sierras of San Juan, the Sunsas, San 

 Lorenzo, Ipias, Chochis and Santiago, some of tliem overlooking the Argentine 

 Gran Chaco, and having a southern drainage into the river Paraguay. The grand 

 rim of the Bolivian basin has two breaks; one leads to the Amazon Valley, by the 

 falls of the Madeira, and the other, in longitude 62° west from Greenwich, and 

 latitude 18° south, opens into what is now known as the Plata Basin. Further on, 

 I will attempt to show how a dam was gradually thrown across this southern gap, 

 until its elevation, once much inferior, became superior to that of the ancient lip 

 of the falls over which the Bolivian rivers now plunge. 



The mean flow of the Madeira River, at the moment of receiving its united 

 Bolivian tributaries, is, according to Keller, 8,654 cubic metres per second, equal 

 to 305,616 cubic feet. Pinkas makes it 6,874 cubic metres. These must be rough 

 calculations ; for the mean flow of a river of such variable conditions could only 

 be accurately measured by a series of observations extending over several years. 

 I believe it is equal to that of the Parana and Uruguay at their junction. 



Outline of the Basin. — The Matto Grosso highlands, overlooking the Bolivian 

 basin, are composed almost wholly of red sandstone, overlying argillaceous schists ; 

 but, near the head waters of the Guapore river, which .skirts their western base, 

 are found rose-coloured gneiss, talc schists, and sandstones, on which frequently 

 rest large areas of a recent alluvium, locally called canrja. 



The Chiquitos sierras rise, on an average, from 1,400 to 2,000 feet above the 

 sea. (Minchin mentions a peak 3,600 feet high.) I agree with Dr. J. W. Evans 

 that, as described by D'Oi'bigny, they present evidences of belonging to the earlier 

 Brazilian highlands, rather than the younger elevation of the Andes. The western 

 section is a wall of friable, ferruginous sandstone, sometimes flat-topped, while the 

 extension, towards the Paraguay River, is a compact sandstone overlying talc 

 schist, which, towards the north-east, rests on gneiss in decomposition. The San 

 Juan and Sunsas sierras are the most northern of the Chiquitos group. The former 

 is mostly of gneissic formation, but the latter is of sandstone resting on talc schist. 



The Bolivian Andes, which face the Amazon and the Gran Chaco, are almost 

 entirely composed of feldspathic sandstones, micaceous and blue slates, clay and 

 sandy shales, at times showing a thickness of from 10,000 to 15,000 feet. In 

 riding from Cochabamba to Santa Cruz de la Sierra, I have especially noticed them 

 exposed upon a gigantic scale. 



The upper fall of the Madeira, called Guajara-mirim, runs over the ferruginous, 

 conglomerate rock called canrjn. This rests on argillaceous sandstone, whicb 

 crumbles easily by the action of running water. The canga gradually becomes 

 undermined, and, breaking in pieces, is moved by the currents into deep water : 

 thus the elevation of the fall is gradually lowered. Keller gives a notable example' 



