TFIE SOURCES OF THE GANGES. ' 415 



finally to the rivers which they compose, one or other of these dire6lions. 

 It is probably true^ that the sources of the Sampoo or Brahmeputra and 

 its tributary streams are separated only by a^narrow range of snow-clad 

 peaks from the sources of the rivers which constitute the Ganges, or which 

 serve to swell its stream: and the whole province. of Laddk\ elevated and 

 rugged as it is, most likely; declines from its southern limit to both the^ 

 north and west. 



This notion is supported by the information received from traders 

 who traffick between Hindustan ^nd Tibet j. as Lieut. Webs, has re- 

 marked; and it is countenanced by routes from C<25/zwzr to Ladak, with; 

 which Major, WiLFORD furnished me, and which were colledled by himi 

 from merchants accustomed to tr?ivel between these countries j. 



In short it can scarcely be doubted, that the snowy mountains^ seen 

 from Hindustan and especially from Rohilkhdnd^ are the highest ground 

 between the level plains of India and the elevated regions of southern 

 Tartary, Whether the- altitude of the highest peaks of Himdlaya h& 

 (quite so great' as Lieut. Webb infers from observation, 1 1 will, not 

 venture to aflirm« The possible error from the uncertainty- respedling 

 the quantity of the refra(51:ion is considerable ; and, owing, to disappoint- 

 ment in the supply of iiistrutnents,. no barometrical observation could 

 be raade to confirm or check^the conclusions of a trigonometrical calcu- 

 lation. Withoutbpwever , supposing the Hzm^^^^ to exceed the Andes j 

 there is still room to argup, that an' extensive. range of mountains, 

 which rears, high above the line of perpetual snow, in an almost tro- 

 pical latitude, an uninterrupted cliain of lofty peaks, is neither surpassed 

 nor rivalled. by anv other chain of mountains but the Cordilleras of the 

 Andes. 



6 



