320 WAVES OF THE SEA 



on turning quickly, there would be a fully-formed 

 roll-wave extending transversely across the whole 

 channel. Once formed, they were never subse- 

 quently dissipated, but, on the contrary, grew 

 quickly as they travelled through the water. The 

 reason for their growth was not evident from mere 

 observation. The outfall into the lake was in fairly 

 regular cadence, the phenomenon being similar to 

 that of the Griinnbach, though not so crisp. This 

 want of crispness was evidently due partly to the 

 sloping sides, which did not keep the wave-front so 

 perfectly transverse as the Griinnbach. The Plate 

 is from photographs taken at intervals of a few 

 seconds, the first photograph showing the outfall of 

 the Guntenbach at a wave-trough, the second at 

 a wave -crest. 



After watching for three months, I came to the 

 conclusion that the mode of origin of the waves 

 in the Guntenbach was entirely different from that 

 in the Griinnbach. In the former there were some 

 slight, long depressions extending the whole width 

 of the pavement which were almost masked by its 

 general roughness. The most marked depression 

 was under the lowest bridge, and somewhat con- 

 cealed thereby. I found that the roll-wave started 

 here more often than at any other point, and I 

 noticed that up-standing edges of the pavement 



