268 Journal of the Asiatic Society of Bengal. [N.S., XII, 



be eaten into from its south edge : they determined where 

 the through roads should lie. Two main trade routes from 

 time immemorial have traversed the Terai : and though the 

 passes on the Thibetan frontier have fixed the upper ends 

 of their course in the mountains, their trend towards the plains 

 has in both been towards a great river, not for the sake of 

 river-carriage, but because the rivers' spill afforded the way into 

 the low country. It is most noticeable that the one route 

 coming from the Kuti and Kirong passes inclines westward 

 towards the old course of the Gandak, and that the other 

 coming from the Tang-la crosses the Jalep-la and descends 

 with a westward trend via Daling (now via Kalimpong) towards 

 the Teesta. 



Trade cuts for itself a channel like a river ; and its opportu- 

 nities beget opportunities : the strongest and most enterprising 

 of peoples take possession of the routes ; they cause the chan- 

 nel to be cut deeper, to be better, to be of more use. The 

 trans-terai trade-routes have been no exception to this. The 

 greatest enemy on them was malaria. But with the opening of 

 the ways through the forest the danger from this became less , 

 and the routes gradually more and more excelled any possible 



rivals as the means of passing through the beast-infested, un- 

 healthy belt. 



There are other sand-carrying rivers between the Gandak 

 and the Teesta, such as the Bagmati, the Kumla, the Kan- 

 kai, the Mahananda, and largest of all, the Kosi; and there L 

 reason to believe that man has at times made more headway 

 against the forest on their sand-cones than between them. 

 Thus, though the Bagmati now emerges from the great 

 ' bhaver ' or forest so far in Nepal that I have not been able 



t0 c fSoo™ ® Xact ldea how the forest Umit lies , in Bunnell's Atlas 

 ot 1783 it is recorded that it then made a wide bay, into the 



head of which the river came. The Kumla, whatever the 

 forest line may be like at the present time, enters British terri- 

 tory near the head of a wide bay in the frontier line, which 

 seems to be evidence of a former limit of administration (i.e. 

 reclaimation from forest). On the Kankai the forest now 

 recedes to 28 34 m a deep bay, which has deepened much 

 since 1783. The Mahanada, almost as soon as it leaves the 

 hills, enters cleared land. Lastly, on the Kosi Rennell shows a 

 deep inlet m the forest along its western side. These bays 

 in the bhaver or forest point to a relative success b V man in 

 forcing his colonization over their sand-cones. All the rivers 



s 



named are much smaller than the Gandak and Teesta except 

 the Kosi ; ; in which connection it is most interesting to refer to 

 Captain Hirst s contention (this Journal 1908, p. 468) that 

 the last-named ha Sj only attained its present dimensions in 



w£?WK S °p J r de /™ and far more recenfc *»an that date at 

 which the Gandak and Teesta acquired their importance. 



