﻿20 <?N TPIE COURSE* OF THE GAKGE3 



change which, in lefs than' two years, had taken place; 

 a confiderable portion of the main channel, which, at 

 the period abovementioned, had contained nearly the 

 whole ftream of the Ganges, being, at the time I lall 

 faw it, ib comJ3letely filled with fands, that I hardly 

 knew myfelf to be in the fame part of the river. The 

 fands, in fome parts, rofe feveral feet above the level 

 of the ftream ; and the people had already begun to 

 cultivate j'j'.r/bo and rice, in the very fpots where the 

 deepeft water had formerly been. Two illands, of 

 confiderable extent, appeared to be quite new ; and 

 the channel, in fome places, had been reduced, from 

 the breadth of an Englifli mile, to a furlong or lefs. 

 The main ftream, having forced its way in a new di- 

 reclion, did not at this time pafs nearer to the inlet of 

 the Howleah than 3|- miles, nor nearer than 2 to Hor- 

 rifonker. 



This remarkable change, I was informed, took 

 place during the extraordinary inundation of 1796; 

 at which time the floods had rifen to an unufual height 

 in almoft every part of the country ; but it muft have 

 been chiefly while the waters were draining off, that 

 fuch an immenfe body of fand could have been depo- 

 fited. The inlet of the Hotohah had been, in ccTnfe- 

 quence, rendered fomewhat difficult of accefs, and I 

 was obliged to make a circuit round the new iftands, 

 of feveral miles, to get into it; but, notwithftanding 

 the diverfion of the flream, I found a confiderable 

 depth of water remaining under the main bank, al- 

 though fcarcely any current was vifible from Horri- 

 fonker to Dewangunge.* On my approach to this 

 place, I was informed that the'paftage was no longer 



pra61;icable 



^* Water is found under the banks of the deferted channels of the Gan- 

 ges for a confiderable time after they ceafe to be navigable in the middle of 

 their beds, the fpace immediately under the bank being generally the laft 

 that is filled up. 1 have likewiie obferved that, diiriiig the growth of the 

 Iflands, the fand ufually gathers round the upper point of the ifland, and 

 rifes to a confiderable height, before the fpace which it leaves between ilfelf 

 and the bank of the illand is completely hlled. 



