AND OTHER WATER WAVES 175 



water does not rise proportionately to the amount 

 driven shoreward, for a head of water accumulates 

 which in time prevents a further rise. The surface 

 drift, indeed, continues, but the undertow is in- 

 creased, so that the amount of water receding from 

 the shore is equal to that approaching it. In the 

 case of a prolonged storm when the wind blows 

 into a bay and there is a steep beach, it seems 

 quite possible that the undertow may be so strong 

 that the bottom current is continuously seaward, 

 although it would, no doubt, be jerky, the seaward 

 motion being checked as each wave -crest passes. 

 These are the conditions under which even shingle 

 may travel seawards from the foot of the breaker. 



If, however, there be no terminating cliff or sea- 

 wall, so much of the shingle as is only reached 

 by the extreme wash of the water discharged from 

 the breaker will, I think, still be driven shoreward, 

 owing to the effect of percolation. 



That off-shore wind produces up-welling of water 

 against a coast -line has been proved on the large 

 scale by observations of temperature and salinity, 

 and it follows that there must then be a sort of 

 shoreward undertow. Its intensity must be much 

 less than that of seaward undertow during on-shore 

 winds, but it doubtless increases the normal shore- 

 ward action of the waves. 



