AND OTHER WATER WAVES 317 



There is one portion of the conduit paved more 

 smoothly, in larger and better trimmed blocks, than 

 the rest. Here, when there was little water, the 

 wave development proceeded best ; afterwards, on 

 entering the rougher part, the wave was somewhat 

 broken up. But when the depth of water was 

 greater, there was but little growth of the wave in 

 passing over the smoother pavement, the growth 

 becoming more rapid when the wave began to ride 

 over the transverse joints of the rougher pavement. 

 Thus, for the best development of the wave, there 

 should exist a certain proportionality between the 

 amplitude of the transverse corrugation and the 

 depth of the water. 



The longitudinal joints of the pavement tended 

 somewhat to destroy the transverse wave, especi- 

 ally when any irregularity of channel caused 

 the wave to pass somewhat obliquely across 

 them. 



There was a good deal of lateral wave motion, 

 which sent a considerable wash against the sides 

 of the channel ; but these, being nearly vertical, 

 reflected the wave back again, and thus tended 

 to keep the front transversal. About half-way 

 down the conduit the slope is somewhat abruptly 

 diminished, remaining thereafter less than in the 

 upper half. It was always a matter of surprise to 



