298 



SCIENCE, 



[Vol. V., No. 114. 



latter flows at an elevation of only 123 metres at 

 Cuyaba near its source), afford uninterrupted naviga- 

 tion for almost their entire course. The tributaries 

 of these two rivers, and the other Brazilian rivers in 

 general, are, however, plateau streams, and have two 

 navigable portions, — one on the upper stream on the 

 plateau; and the other in the depression, or coast 

 border region. The difference of level between these 

 two portions is one or more hundred metres, and the 

 descent has to be accomplished by a series of cata- 

 racts situated at a relatively short distance above 

 the mouth of the river. Of these, the most notable 

 are the Paulo Affonso cataract on the Sao Francisco, 

 and the Sete Quedas ('seven falls') on the Parana. 

 The upper tributaries of the Amazonas, between the 

 foot of the Andes and the Kio Negro on the north, 

 and the Madeira on the south, are exceptions to this 

 rule ; since they descend from the plateau on which 

 they rise in their upper courses, and afford long lines 

 of navigation. They thus reveal the interesting fact 

 that a vast area of the almost unknown upper Ama- 

 zonian region is at a much lower level than the adja- 

 cent plateaus. Of the rivers that flow directly to 

 the Atlantic, those of the province of Maranhao and 

 the Parnahyba, in Piauhy, offer the greatest facilities 

 for navigation ; because they rise at a lower level than 

 the rivers to the southward, and effect their descent 

 to the sea-level by a gradual slope distributed along 

 the whole course, instead of being concentrated in 

 one or more grand series of rapids. 



The Amazonas and Paraguay present peculiar fea- 

 tures in the very extensive alluvial plains that border 

 the maiD river and the lower courses of their tribu- 

 taries, and in the great number of anastomosing lat- 

 eral channels that cut these plains and put the main 

 river in communication with the tributaries, often at 

 long distances above the junction, and these last with 

 each other. These canals are particularly notable in 

 the case of the Amazonas, where they are called para- 

 namirins, or furos ; 1 and it is said that a boat may 

 traverse almost the entire length of the Amazonas 

 valley without entering the main stream. The for- 

 mation of these canals is to be attributed in part to 

 the formation of alluvial islands that are constantly 

 being created by the sediment-loaded waters of the 

 great river. The number and character of many of 

 them, and especially of the faros, seem, however, to 

 indicate a more general cause, and suggest the idea, 

 that, since a relatively slight depression of the sur- 

 face would transform the river-valley into an estuary, 

 it may reasonably be supposed that at some time a 

 correspondingly slight elevation has transformed an 

 estuary into a river-valley. The long distance to 

 which the influence of the tide (it is sensible at Oby- 

 dos five hundred miles above the mouth of the Ama- 

 zonas) is felt, gives an air of probability to this 

 hypothesis. In this case the present tributaries 

 would have been independent rivers, and would nat- 

 urally have had deltas, the canals of which would in 

 part become closed, and in part be transformed into 

 furos, when the estuary was changed into a river. 



1 The former are canals that return to the same river from 

 which they parted ; the latter, those that unite- two distinct rivers. 



The paranamirins would in this hypothesis repre- 

 sent the marine channels of the muddy bottom of 

 the estuary. It is certain that the lower portion 

 of the Amazonas valley still presents so much of 

 the character of an estuary, that a question has 

 arisen among geographers as to whether the Tocan- 

 tins should be considered as a tributary, or an inde- 

 pendent river. The fact that it receives a considerable 

 quantity of water from the Amazonas through vari- 

 ous furos decides the question of its being a tribu- 

 tary, since the Amazonas contributes much more 

 water than the Tocantins to the so-called Para River, 

 which is only the southern branch of the great Ama- 

 zonian estuary. 



Attention has often been called to the curious fact, 

 that, unlike most large sediment-loaded rivers, the 

 Amazonas has no delta. The reason appears to be, 

 that its lower course is still in a transition state be- 

 tween the estuary and fluvial conditions; and the 

 delta is not to be looked for at what is generally con- 

 sidered as the mouth of the river, but higher up at 

 the head of the estuary. In this case the network 

 of canals between the mouth of the Xingu and the 

 western end of the island of Marajo may be consid- 

 ered as the true delta. 



In a certain sense, the La Plata basin is a triple 

 one, since a slight change of level, which should take 

 the head of the estuary to the mouth of the Parana, 

 would effect the separation of the Paraguay, Parana, 

 and Uruguay as three distinct basins. Although 

 smaller than the Parana, the Paraguay should be 

 considered as the main stream on account of its rela- 

 tions with the elevated portions of the continent to 

 the east and west. The Parana, as already stated, 

 is essentially a highland river. Its tributaries flow 

 into it before it enters the depression by the great 

 cataract of Sete Quedas. The only exception is the 

 Iguassti, which has its great cataract of Santa Maria 

 close to the junction. A peculiarity of the Parana is 

 that the eastern margin of its basin is so close to the 

 Atlantic, that one of its tributaries, the Tiete, may 

 be said to rise in sight of the sea. Another peculiar 

 feature is the tendency of its eastern tributaries, 

 especially marked in the case of the Tiete, to flow in 

 a north-westerly direction, as if they were seeking 

 the source, rather than the mouth, of the main river. 

 This indicates a general north-westerly slope in this 

 part of the plateau. 



A few peculiarities in the principal rivers of the 

 direct Atlantic drainage system, which indicate inter- 

 esting points in the topographical structure of the 

 country, may be mentioned here. Such a point is the 

 general parallelism of the Sao Francisco to the coast- 

 line in the greater part of its course, due to the river 

 being confined behind the Serra do Espinhaco, which, 

 terminating to the northward, finally permits the river 

 to escape, and direct its course toward the sea, making 

 a right angle to its former course. The same phenom- 

 enon is presented in a still more interesting manner 

 by the Parahyba, which exhibits a double parallelism, 

 the river making a U-curve in the upper part of its 

 course, and, after a course of about two hundred 

 miles, passing close by its source. This is due to a 



