THE SOURCES OF THE GANGES. 443 



became a shallow and almost stagnate pool, and the latter a small 

 stream; and both having, in addition to springs and rivulets, a consi- 

 derable visible supply from the thawing snow, it is therefore concluded, 

 by analogy, that the sources of these rivers could be little, if at all, 

 removed from the stations at which these remarks were coUe^led. 



2d, ' The channel of a great river is usually a line to which the con- 

 tiguous country gradually slopes ; and perhaps on this account, in the 

 mountainous country, (as information and experience have taught me,) the 

 sides of a river always furnish the most pra61:icable road in the dire(5lio!i 

 of its course.* Now, if the Bhdgirat'hi and Alacanandd rivers had a 

 passage through the Himalaya^ it should follow, that the channel of its 

 stream would form the Ghdtti by which the snowy range became passa- 

 ble. But, since this principle holds good in pra6lice, and since it is 

 utterly impossible to cross the snowy range in a dire6lion which the 

 channels of these rivers might be supposed to assume, I consider that at 

 least all former reports are determined fi6litious. 



sd. ' I HAVE conversed with two or three intelligent natives, whose in- 

 formation I have found correct in other instances, and who have, in pil- 

 grimages and on business^ traversed the northern skirt of the Himalaya ; 

 and 1 have their assurances, that no river, except one, exists westward of 

 the Mdnasarovara lake ; that this stream is called the Saturuz (Satalaj ) 

 fiwer ; and that it turns southerly, west of Jamoutri. 



' The extreme height of the Himdlaya is yet, a desideratum ; but by a 

 mean of numerous altitudes of a conspicuous peak, taken at different hours 



* The only exception to this maxim is perhaps in the case of a cataract, such as the falls 

 of Niagara, where a river dcijcends precipitously from au elevated ledge of rock. Uut no 

 &u<;li cascade of the Gan^s has been tbuud. H. C. 



