230 On the Sites of Nikaia and Boukephalon. [No. 3. 



more level, a depth of soil, which cannot be fathomed, it spreads out 

 into a wide sheet of water forming islauds, indeed, but islands which 

 almost as soon as they are formed begin to melt away in the set of the 

 yearly inundation, which, having no rocks nor channels of shingle to 

 determine its current, takes a different course every year, shifting* 

 from side to side of the extensive basin. The action of the wind 

 upon so wide a surface of fine sand, aids this caprice of the current. 

 The waters find their channel of last year obstructed by sand, and put 

 forth their strength in a new direction washing away the islands of 

 last year and depositing sand-banks, which every year rise by the 

 deposit of silt until they become islands : but which are always subject 

 to overflow or even dissolution in heavy floods. Sometimes indeed 

 when the river comes down with unwonted power and finds the old 

 channels obstructed or grown very devious, it sweeps onward over the 

 country and receives an entirely new channel in a directer line, isolating 

 a portion of the country so large that it continues to be an island for 

 centuries and is inhabited and cultivated. Such islands however are 

 rare in the Hydaspes below Jelum. I know of only two or three. 

 They may, when very extensive, be mistaken for the opposite bank of 

 the river by a person who cannot command a bird's eye view of the 

 stream. But the other kind never can be thus mistaken. 



Let us once more return to the river channels on the escape of the 

 Hydaspes from the mountains. These every year sink in depth, until 

 they have cut through the strata of finer shingle and penetrated to the 

 pavement of massive and firmly cemented boulders which no ordinary 

 torrent can move. There the furrowing action of the current is 

 arrested, and the figure of the channels is preserved, by the solidity of 

 the scarps, and the only change that can ordinarily happen to them in 

 the lapse of ages is the gradual and yearly wear of the banks at 

 the salient curvatures, and the consequent decrease in the depth of 

 the stream. If indeed such a terrible inundation as that which 

 occurred to the Indus about twelve years ago should happen to the 

 Hydaspes, the soil of the higher islands would of course be swept 

 away and they would become like the secondary islands, shoals of 

 shingle, strewed with sand, and remain so for ever ; there being at 

 * In the river Sardeh I have known these aberrations amount to eight miles or 



