PROBLEMS ASSOCIATED WITH BHYTHM. 



245 



seems to have affected the lower part of the 

 current more than the upper. 



The trough used with contracted outfall was 

 0.92 foot wide and was contracted, in the 

 manner shown in figure 3 (p. 25) to 0.30 foot. 

 The oblique walls producing the contraction 



I 2 



Velocity 



FIGURE 74. Modification of vertical velocity curve by approach to 

 contracted outfall. 



were 3 feet long. The first velocity station 

 was 1.5 feet from the outfall, at a point where 

 the width was 0.64 foot; the second and third 

 were at 0.75 and 0.25 foot from the outfall, 

 corresponding to widths of 0.47 and 0.35 foot. 

 The Velocity curves, given in figure 74, show 

 the descent of the plane of maximum velocity 

 from a position slightly above middepth to 



FIGURE 75. 1'lan of experiment trough with local contraction. The 

 letters show stations at which velocities were observed, and are re- 

 peated in figures 7G and 77. Scale, 1 inch=2 feet. 



near the bottom. The acceleration, in this 

 case connected with contraction of channel, 

 as well as with the release at outfall from the 

 channel resistance, is a function of depth. 



In the third arrangement a trough 0.92 foot 

 wide, with a slope of 0.58 per cent, was con- 

 tracted at one point in the manner indicated 

 in figure 75. This gave to the stream such a 



FIGURE 76. Profile of waU-r surface in trough shown in figure 75. Scale 

 1 inch=2feet. 



profile as is sketched in figure 76. Velocities 

 were measured at four points, the distances 

 from the point of extreme contraction being 

 2.0 feet, 1.0 foot, 0.5 foot, and 0, and the corre- 

 sponding widths of current 0.92, 0.55, 0.37, 

 and 0.19 foot. The velocity curves are plotted 



in figure 77. The level of maximum velocity 

 is at or near the surface at the point of initial 

 contraction, A, and then drops quickly, being 

 below 0.9 depth in the narrowest strait, D. 

 In this case also the acceleration appears to 

 increase in regular manner with distance from 

 the surface. 



To avoid the peculiarities observed near the 

 outfall, as well as those appropriate to intake 



FIGTOE 77. Modification of vertical velocity curve by local contrac- 

 tion of channel. The letters indicate positions of velocity stations in 

 trough. (See fig. 75.) 



conditions, all other determinations of the 

 curve were made near midlength of the trough, 

 the ordinary distance from the outfall being 

 17 feet and the least distance 13 feet. 



In figure 78 are three curves illustrating 

 the influence of slope of channel. The trough 

 was 1.96 feet wide and its sides and bed 

 were smooth; the discharge was 1.119ft. 3 /sec. 

 Curve A shows velocities 17 feet from the out- 



FiGURE 78. Modification of vertical velocity curve when mean velocity 

 is increased by change of slope. 



fall, with the trough level; B corresponds to a 

 slope of 0.26 per cent and C to 0.56 per cent. 

 With the trough level, maximum velocity 

 occurs at about 0.75 depth, and with both 

 inclinations of trough its position is indicated 

 at from 0.2 depth to 0.3 depth. As curve 

 differs in type from all others obtained, and as 

 it may be made harmonious by rejecting a 

 single observation, it is thought probable that 

 the dotted line better expresses the fact. Its 



