AND OTHER WATER WAVES 305 



hollowed piece of pavement, it can be seen to flow 

 gushingly, for the greater regularity of cross - 

 section to some extent co-ordinates the flickerings, 

 and the smaller depth tends to increase the gushi- 

 ness of flow, as has already been explained. 



At the entrance to, and for the first few yards 

 of, the rectangular, flat-bottomed, paved channel 

 these embryo roll -waves pass the eye in too rapid 

 succession to be accurately counted, but there are 

 about 120 per minute. They are here little more 

 than ripples, and the fronts, which face down- 

 stream, are free from white froth. As we walk 

 down the bank of the conduit, we observe that a 

 regularising process is at work, which, in a short 

 distance, produces from this hurried and con- 

 fused crowd of progressive wavelets an ordered 

 series of roll -waves of greater amplitude, with 

 foaming fronts facing down -stream, extending 

 across the channel in a straight line at right angles 

 to its axis. On June 6, 1904, at 465 feet from 

 the entrance to the paved channel, there were 

 notable larger waves ; but there were many minor 

 ones also, and the appearance was still somewhat 

 confused. At a distance of 567 feet from the 

 entrance to the paved channel there was no more 

 confusion, a distinct series of waves, 33 per 

 minute, passing the observer ; whilst after travel- 



