262 ELEVATION OF THE RAUDALES. 
a longitudinal valley between the mountains of Parima and 
the great mass of the mountains of Brazil. 
1 was surprised to find by actual measurement that the 
rapids of the Orinoco, the roar of which is heard at the 
distance of more than a league, and which are so eminently 
picturesque from the varied appearance of the waters, the 
palm-trees and the rocks, have not probably, on their whole 
length, a height of more than twenty-eight feet perpen- 
dicular. In reflecting on this, we find that it is a great 
deal for rapids, while it would be very little for a single 
cataract. The Yellalas of the Bio Congo, in the contracted 
part of the river from Banza Noki as far as Banza Inga, 
furnish, between the upper and lower levels, a much more 
considerable difference; but Mr. Barrow observes, that 
among the great number of these rapids there is one fall, 
which alone is thirty feet high. On the other hand, the 
famous pongos of the river Amazon, so dangerous to go up, 
the falls of Bentema, of Escurrebragas, and of Mayasi, are 
but a few feet in perpendicular height. Those who are 
engaged in hydraulic works know the effect that a bar of 
eighteen or twenty inches’ height produces in a great river. 
The whirling and tumultuous movement of the water does 
not depend solely on the greatness of partial falls; what 
determines the force and impetuosity is the nearness of 
these falls, the steepness of the rocky ledges, the returning 
sheets of water winch strike against and surmount each 
other, the form of the islands and shoals, the direction of 
the counter-currents, and the contraction and sinuosity of 
the channels through which the waters force a passage 
between two adjacent levels. In two rivers equally large, 
that of which the falls have least height may sometimes 
present tho greatest dangers and the most impetuous move- 
ments. 
It is probable that the river Orinoco loses part of 
its waters in the cataracts, not only by increased evapo- 
ration, caused by the dispersion of minute drops in the 
atmosphere, but still more by filtrations into the subter- 
raneous cavities. These losses, however, are not very per* 
ceptible when we compare the mass of waters entering into 
the raudal with that which issues out near the mouth of the 
Bio Anaveni. It was by a similar comparison that the 
