Springs, Rivers, and the Sea. 129 



the declivity of those formed by nature is 

 \ r arious and uncertain. 



The velocity of the water of a river 

 Ottght to increase in proportion as it re- 

 cedes from its source : but the numerous 

 causes of retardation, which occur in riv- 

 ers, are productive of very great irregu- 

 larities ; and it is impossible to form any 

 general rules for determining such 

 irregularities. 



The unequal quantities of water (ari- 

 sing from rains, from the melting of snow 

 &c.) which are conveyed by rivers at dif- 

 ferent seasons, enlarge or contract their 

 widths, render them more or less rapid, 

 and change more or less the form of their 

 beds. But independent of this, the size 

 and form of a river is liable to be continu- 

 ally altered by the usual flowing of its 

 waters, and by local peculiarities. The 

 water constantly corrodes its bed where- 

 ver it runs with considerable velocity, 

 and rubs off the sand, or other not very 

 coherent parts. The corrosion is most 

 remarkable in that part of the bottom, 

 which is under the thread 'of the river, or 

 where the water descends suddenly from 

 an eminence, as in a cascade or water -fall* 

 L2 



