xviii INTRODUCTION. 



j^owever, that the velocity of rivers depends not on the 

 (^hvity of their beds alone, but chiefly on the mass and 

 velocity of the water thrown into their channels at the 

 spring head, and the supplies they receive from tributary 

 branches as they proceed in their course. In the Amazons, 

 the Ganges, the Senegal, the Gambia, and in every river 

 whose course, in its approach to the ocean, lies through a 

 low country, it will be found, that the rise of a few feet in the 

 tide is sufficient to force back, up an inclined plane, by its 

 mass and velocity, the whole current of the river to the 

 distance of several hundred miles, and the farther in pro- 

 portion to the narrowness and depth of the channel beyond 

 its funnel shaped mouth. In estimating the probability, 

 therefore, of the identity of the Zaire and the Niger, as far 

 as the length of their course may be supposed to offer an 

 objection, we should inquire rather into the supply of 

 water than the declivity of the country through which it 

 would have to pass. In this respect, the Niger would be 

 placed under very peculiar circumstances ; its course, 

 lying on both sides of the Equator, and through a consi- 

 derable portion of both tropical regions, would necessarily 

 be placed, in one part or other, under the parallels of 

 perpetual rains, and consequently receive a perpetual sup- 

 ply of water. Now all the representations that have been 

 given of the lower part of the Zaire, describe it as being 

 nearly in a perpetual state of flood, the height in the dry 

 season being within nine feet of the height in the season of 

 heavy rains; whilst the difference in the height of the Nile 

 and the Ganges, at the two periods, exceeds thirty feet. 

 The flooding of the Zaire is therefore periodical, its highest 

 >|state being in March, and lowest about the end of August; 



