156 THE IMPROVEMENT OF RIVERS. 



has led to the belief that it is best to make the floor strong enough to resist the upward 

 pressure, and to let the seepage find its own escape. 



Too much reliance should not be placed on the ability of sheet-piling to stop 

 seepage. \Yhere it can be driven under the best conditions, as was done at the Wa- 

 chusetts Dam, Mass., where each pile was planed and sunk through soft material with 

 water- jets,* it may prove very effective; but as these conditions cannot be secured 

 on river-work, the chief function of the piling appears to be to obstruct the carrying 

 off or the flow of material. To quote an example: on the upper end of the coffer- 

 dam for a new lock, triple-lap sheet-piling was driven to rock through 2 to 3 feet 

 of sand and mud, the depth of water on the rock being about 13 feet. The piles, 

 which thus required a minimum of driving, were of good lumber, rough but not warped, 

 and were carefully put down. In spite of the favorable conditions, however, when the 

 slope of the river caused a head of about 6 inches between the upper and lower sides 

 of the piling, a copious leakage was visible everywhere through the joints, and this 

 continued until the backing was completed. 



The usual method of constructing an artificial floor in this country is to bolt 

 timbers to the sides of piles, covering them with planking or squared timbers. These 

 should always be calked, unless a layer of concrete is placed over them, as the water 

 will force its way through the smallest opening and gradually enlarge it, and if the 

 foundations are of light material, it may, in time, cause undermining. The calked 

 timbers should be covered with planking spiked down so as to prevent the oakum 

 from being forced up. It is far better, however, to discard timber altogether, for 

 reasons mentioned elsewhere, or else to cover it with a layer of concrete bedded around 

 spikes driven into the wood. 



A type of floor which is in favor in Europe, and which is probably the best of any 

 in use, consists of an inverted arch of cut stone or of concrete, the voussoirs in the 

 former case being cut either for a flat or for a curved arch. The expense of the method 

 has probably prevented its wider adoption, but with concrete the cost would be little 

 if any more than that of a flat floor, and the gain in strength would be very 

 great. In the locks of Dinant and Anseremme, on the Belgian Meuse, where the 

 chambers are 39 feet wide, the floors were formed of inverted arches of 2 feet rise, 

 the voussoirs being 15 inches deep, and laid on a bed of masonry of a least thickness 

 of ij feet. Below this was a bed of concrete 2 feet thick. At the new Suresnes lock 

 in the Seine the rise was made 20 inches in 59 feet, the arch being laid on a solid bed 

 of masonry. 



In one or two instances in America the floor has been placed 10 or 12 feet below 

 the lower miter-sill, and surrounded on all sides with concrete, exactly like a box 

 'without a lid. This box is then filled with gravel or other heavy material, and cov- 

 c'er] with plank spiked to joists, the idea being that the weight on the floor will coun- 



* Proceedings Am. Soc. C. E., April, 1902, Frederic P. Stearns. 



