In designing a set of sluice openings or regulators, it is sometimes the custom to make the total area of waterway the same as that of the stream in its unobstructed condition. There is no particular reason why it should be the same. In a description of the Assiut Barrage (_Min. Proc. Inst. C.E._, vol. clviii., p. 30), it is mentioned that one of the reasons for placing the floor lower than the river bed was that the width of the waterway of the barrage was less than that of the river. The bed has to be heavily protected in any case, and the proper principle is to fix a velocity which is considered to be safe and, the maximum discharge being known, to determine the area of the waterway accordingly. In the case of a very wide river like the Nile, with a well-defined channel, it is inconvenient to make the distance between the abutments of a work much less than the width of the channel, but so far as velocity is concerned, the floor need not usually be lower than the bed. The protection given to the channel on the upstream side of the barrage (fig. 41) seems to be rather greater than necessary. The thickness of the floor (9 feet 10 inches) seems excessive. The thickness originally proposed was much less. Of the many kinds of apparatus described in this chapter each possesses some advantages and disadvantages. Gates require a bridge with powerful lifting apparatus, and are suitable for large bodies of water and great depths. Comparing needles with planks, the former can be worked by one man and admit of rapid removal, and require far fewer piers. Planks require two men, and are sometimes liable to jam, but obstruct floating rubbish less than needles, and in shallow water give rise to less leakage. Whether needles or planks are used, masonry piers are most suitable where sand or gravel are liable to accumulate on the floor, or where there is much floating rubbish. The hinged frames are suitable in other cases. Falling shutters of the Chanoine type admit of very rapid lowering, and can be used without a foot-bridge. The drum weir is perfect in action, but its cost is high. At any system of sluices the regulation should be so arranged as to minimise the chances of damage to the bed and banks where this is at all likely to occur. If the gates are opened only near one side of the structure, there will be a rush of water on that side, and serious damage may occur. The opening should be done symmetrically and, as far as possible, distributed along the whole length. Until experience has shown it to be unnecessary, soundings should be taken at regular periods of time downstream of every important work where scour can occur. When scour is found to have occurred at any particular part of the work, the rush of water at such places should, as far as possible, be prevented, and a chance given for silting to occur.