AND BOTHER WATER WAVES 327 



cadence in the outflow. The depth was, therefore, 

 apparently too great for the effect of frictional re- 

 sistance to produce an intermittent flow, and the 

 pools of the pavement were now too insignifi- 

 cant to act as they did with a smaller depth of 

 water . 



I have never seen a stream with a uniform 

 depth of more than 4 inches adopt spontaneously 

 the intermittent flow in a series of roll-waves. The 

 conduit of the Chauderon River at Montreux, with 

 a depth of 6 inches, did not flow thus ; neither did 

 the much larger Veveyse River, in its conduit at 

 Vevey, flowing swiftly at the time of snow- 

 melting, with a depth of 48 inches. 



Nevertheless, I found (after my eyes had been 

 trained by watching well-marked roll -waves for 

 some months) that waves progressing down-stream 

 through the water co -exist with stationary waves 

 in shallow streams, and may even be detected in 

 deep ones, although they do not (except if caused 

 by floods) there develop or grow so as to become 

 conspicuous or important relatively to the other 

 phenomena of flow. Thus (commencing with the 

 Guntenbach ), on looking closely at a standing wave 

 (in which the depth of water was only about 

 2 inches) I noticed that the water here flowed 

 gushily i.e., the excrescence of the bottom, which 



