TllL KaKTll. 147 



this case, the water in the midst of the current loses a part of 

 its weight, from the velocity of its motion j while that at the 

 sides, for the contrary reason, sinks lower. It sometimes how- 

 ever happens, that this appearance is reversed ; for when tides 

 are found to flow up with violence ag^ainst the natural current 

 of the water, the greatest rapidity is then found at the sides of 

 the river, as the water there least resists the influx from the 

 sea. On those occasions, therefore, the river presents a con- 

 cave rather than a convex surface j and as in the former case, the 

 middle waters rose in a ridge, in this case they sink in a furrow. 



The stream of all rivers is more rapid in proportion as its 

 channel is diminished. For instance, it will be much swifter where 

 it is ten yards broad, than where it is twenty ; for the force 

 behii'd still pushing the water forward, when it comes to the 

 narrow part, it must make up by velocity what it wants in room. 



It often happens that the stream of a river is opposed by one 

 of its jutting banks, by an island in the midst, the arches of a 

 bridge, or some such obstacle. This produces not unfrequently 

 a back current; and the water having passed the arch with 

 great velocity, pushes the water on each side of its direct cur- 

 rent. This produces a side current, tending to the bank ; and 

 not unfrequently a whirlpool ; in which a large body of waters 

 are circulated in a kind of cavity, sinking dov.'n in the middle. 

 The central point of the whirlpool is always lowest, because it 

 has the least motion -. the other parts are supported, in some 

 measure, by the violence of theirs, and consequently rise higher 

 as their motion is greater ; so that towards the extremity of the 

 whirlpool, must be higher than towards the centre. 



If the stream of a river be stopped at the surface, and yet ba 

 free below ; for instance, if it be laid over by a bridge of boats, 

 there will then be a double current ; the water at the surface 

 will flow back, while that at the bottom will proceed with in- 

 creased velocity. It often happens that the current at the bot- 

 tom is swifter than at the top, when, upon violent land-floods, 

 the weight of waters towards the source presses the waters at 

 the bottom, before it has had time co communicate its motion 

 to the surface. However, in all other cases, the surface of the 

 stream is swifter than the bottom, as it is not retarded by rubbiiig 

 over the bed of the river. 



It might be supposed that bridges, dams, and other obstacl. * 



N 2 



