8 ALEXANDER'S CAMPAIGN IN THE PANJAB [Chap. I
Though erosion had here, too, laid bare deeper strata of debris, pottery of earlier periods was conspicuously absent. The former bed of the Beds was easily recognized in an extensive marsh to the east of Kahnuwdn. The erosive force of the summer floods, which at times still descend it and reach the old bank, was attested by the undercut base of a brick-built Hindu shrine of no great age.
Some five miles farther north, near the village of Nangal, we visited a considerable debris-strewn area adjoining the old river bank and showing similar pottery. Several Muhammadan copper coins brought from this the supplied chronological evidence in agreement with the late ceramic ware. But curiously enough a punch-marked square copper coin, obviously of early Hindu type, was also picked up on the surface. The relief-decorated fragment, Nang. 9 (Pl. I), may also be pre-Muhammadan. Finally I may mention that when we visited a group of mounds which the map marks rising above broken ground near the village of Mandrdjpur, not far from the southward bend of the Beds, one of them was found to bear pottery debris of the same late type as noticed at the riverine sites lower down. The other mounds proved to be natural hillocks; and here our survey ended.
The observations here recorded from a series of sites along the right bank of the Beds are in such close agreement that some conclusions of antiquarian interest may safely be drawn from them. It is certain that the importance of river traffic in former times and the advantage offered by higher ground at convenient crossing places must always have attracted settlements of some size to the western bank of the river. If none of the sites now traceable along it can, by the evidence of their remains, be dated farther back than the medieval Muhammadan period, it is clear that the present western bank and the river course to which it owes its formation cannot be of any great antiquity.
It is equally obvious also that the absence of any earlier remains by the river can be fully accounted for by the bank on which they stood having been eroded by the river as it shifted westwards. We have no means to determine how far back in time this westward move of the Beds bed has asserted itself, nor whether since Alexander's days it has not been checked at times by a diversion in the opposite direction as is common in the deltaic conduct of rivers.3 In any case, however, our survey of the ground along the modern course of the river has sufficed to demonstrate that we cannot expect there, nor farther east, to