^8 Mr. reknell’s Account of the 
when a branch of the Ganges is fallen fo low as to occupy 
only a part of its bed, it no longer continues in the line of its 
old courfe ; but works itfelf a new channel, which winds from 
fide to fide acrofs the former one. I have obferved too, that in 
two ftreams, of equal fize, that which has the floweft cur- 
rent has alfo the fmalleft windings : for as thefe (in the prefent 
cafe) are folely owing to the depredations made on the banks 
by the force of the current; fo the extent of thefe depreda- 
tions, or, in other words, the dimenfions of the windings, will 
be determined by the degree of force aCting on the banks. 
The windings of the Ganges in the plains, are, doubtlefs, 
owing to the loofenefs of the foil : and (I think) the proof of 
it is, that they are perpetually changing ; which thofe, origi- 
nally induced by an inequality of furface, can feldom, or never 
do 
I can eafily fuppofe, that if the Ganges was turned into a 
ftraight canal, cut through the ground it now traverfes in the 
moft winding parts of its courfe, its ftraightnefs would be 
of fhort duration. Some yielding part of the bank, or that 
which happened to be the moft ftrongly aCted on, would firft 
be corroded or diffolved : thus a bay or cavity would be formed 
in the fide of the bank. This begets an inflection of the cur- 
rent, which, falling obliquely on the fide of the bay, corrodes 
it inceflantly. When the current has paffed the innermoft 
pdrt of the bay, it receives a new direction, and is thrown 
* It has been remarked, that the courfes of rivers become more winding as 
they approach the fea. This, I believe, will only hold good in fuch as take the 
latter part of their courfe through a fandy foil. In the Ganges, and other rivers 
fubjeft to confiderable variations in the bulk of their ftreams, the belt marks of 
the vicinity of the fea are, the lownefs of the river banks, and the increafing 
anuddinefs of the (hallows in its bed, 
obliquely 
