DoBSON. — On the Reclamation of Biver Beds. 101 



flood-waters were the source of the fertihty of lower Egypt, river works 

 were undertaken with a view to obtain the advantage of the overflow on 

 the greatest possible area, and the flooded lands were raised annually by 

 this process, as well as the river bed itself. 



This cannot happen in the case of torrents carrying large quantities of 

 shingle and sand, which are brought down in times of flood and left in the 

 channels as the velocity of the stream decreases, and although much may 

 be done towards raising the adjoining lands by inducing an overflow of flood 

 water upon them, the floods will always bring down from the upper levels 

 masses of material that it has no power to move along the lower course. 



This is the problem the Italian Engineers have had to deal with in the 

 management of the Po and its tributaries. 



The general system in Europe for conservation seems to be to wall in 

 the river between embankments with the endeavour to make it carry its silt 

 as far forward as possible, making traps for catching shingle ; where possible 

 raising adjoining lands by ponding up flood- water upon them ; and many 

 other expedients for keeping the rivers from changing then- courses and 

 wandering about thek fans into new courses. The great drawback to this 

 treatment is, that however much (within certain limits) the river may be 

 embanked, it continually raises its bed, and the banks must be raised in 

 proportion. 



To such an extent has this been done that some of the Italian rivers are 

 now far above the level of the adjoining lands, so that the surface drainage 

 of the country has to be carried in other channels, involving all kinds of 

 complications in the drainage ; and should the embankments fail, immense 

 volumes of water may be dammed up on the low-lying land by the banks 

 which were built for their protection, entailing great loss of life and pro- 

 perty. Such a disaster occurred not long since in Hungary, at Szegedin, 

 on the Eiver Theis. 



When we consider the enormous cost that would be incurred in 

 attempting to control our great torrents by any system of solid embank- 

 ments, it becomes apparent that we must either invent some plan which 

 can be carried out on a much cheaper scale, or else let matters take their 

 course. 



Doubtless weak places may be defended by groins and spurs ; and the 

 more chance will there be of success so long as only one side of a river is 

 being worked upon. But so soon as works are undertaken on both sides of 

 a river, they will have to be more or less continuous embankments, which 

 must be made of permanent materials, involving immense expense. 



On the Canterbury Plains the conditions are most favourable for the 

 foruiation of broad shingle beds, the loose shingle banks offering no resistance 



