4 
636 PROCEEDINGS OF SECTION H. 
course, to cause more trouble from the erosion of the banks. 
I have known cases where towns and villages situated below 
great bends in rivers have strongly objected to the bends 
being straightened for fear that they would be injuriously _ 
affected by the increased height of the floods. It may be 
shown, however, that although the bends may delay the 
approach of the flood, they cannot diminish the quantity, so. 
that straightening them would have no injurious effect 
below. Straightening bends in rivers is, therefore, alto 
gether beneficial, provided the expense to be incurred to- 
protect the banks is kept in view. 
If a river is so small that the work of training it through 
its length can be done, then it is often possible to greatly 
change in the character of the river, and with it the effects 
of fioods. By protecting the banks, forming proper curves, 
and training the channel to a calculated expanding width. 
downwards, the formation of shoals and banks may be 
prevented, and the river so assisted to increase its depth 
that the current may become. nearly uniform throughout, 
and, if it has sufficient velocity to scour away its deposits, 
the channel may permanently maintain its depth, and so 
greatly diminish the bad effect of floods. 
This training of rivers is, however, generally done on tidal 
rivers, like the Seine between Havre and Rouen; but, if 
done intelligently, almost any part of a river may be advan- 
tageously trained; the uncertainty in the results, if any, 
lies in the uncertainty as to the material contained in its 
deposits, because the training that may be effectual in fine 
sand may not succeed in coarse sand, still less in gravel. 
As the deposits in rivers are one of the chief causes of the 
damage by floods, it is essential to ascertain if possible the 
causes of them, and whether they can be removed, 
partially or wholly. Of course, the beds of rivers are 
extremely irregular, varying with the hardness of the soils 
or rocks over which they pass; but such irregularities are not 
of the nature of deposits, which is the subject under dis- 
cussion. The deposits in a river consist of the material 
worn away from the watershed, deposited in the rivers 
on its journey down to the sea, which it reaches sooner or 
later, except such part as is used up to make new land. 
The length of time taken for the journey to the sea varies 
according ‘to the material; fine silt and mud may reach 
the sea in a single freshet, sand takes longer, gravel longer 
still; while boulders may take centuries, and in some rivers 
may never reach the sea in that form; also, in many rivers, 
especially the large ones, gravel and coarse sand may not 
