ox TEKRESTKIAL SUBFACE WAVES. 313 



o£ an f avly autumn I found that there was practically no bore, although 

 in springtime I had seen a good one with a smaller tide. The iBshermen 

 opined that the sands between Awre and Frampton, a few miles down 

 the river, had shifted in .'•ome way, which would be found to account for 

 the failure of the bore at Newnham. Accordingly at the next succeeding 

 tide I took up a post of observation at Hock Crib, where I could com- 

 mand a view over the extensive sands in the broad straight stretch 

 towards Severn Bridge, which constitutes the commencement of the 

 proper estuary, as well as of the last bend of the winding channel of the 

 river proper looking towards Newnham. The main stream of the river 

 had established itself in a channel somewhat near the right bank, and up 

 this the 'first of the flood' came with a good 'head' to it, the bore 

 appearing as a crested breaking wave stretching quite across the low- 

 tide channel. It thus continued until it reached a point close to Hock 

 Cliff, at the head of another channel which skirts the concave left bank, 

 passing Frampton. From the position now reached by the incoming 

 flood, water now poured hack into this channel, the further advance of 

 the tide up stream being stopped. Presently the back-flowing stream 

 met another coming up the Frampton Channel, and the turbulent waters 

 then began to spread over the large expanse of sand intervening between 

 the two channels. Not until these were nearly covered did any percep- 

 tible rise of tide make its way up the river channel towards Newnham. 

 The total rise of tide is unaffected by this course of events, but the 'head' 

 which the flood had gathered is lost and the bore is spoilt. It was 

 pointed out in a paper on sand waves in tidal estuaries that the flood 

 tide tends to follow the chords of the arcs made by the sinuosities of an 

 ebbing current. According to the information collected by the late 

 Mr. Frank Buckland, the Severn in a wet season tends to collect in the 

 Frampton Channel, and in a dry season in the Awre Channel. 



In the estuaries of the Mawddach, Dovey, Ac, dealt with in ' Geogr. 

 Journ.' August 1901, the last of the ebb and the first of the flood respec- 

 tively had settled into separate channels and there was no bore, but a 

 circulation of the tidal waters. 



That the continued action of the flood tide upon the sands minimises 

 the bore whicli it at first produces is further shown by the observation 

 of Mr. D. J. Wintle, of Nevvraham-on- Severn, who reports that 'the 

 best heads are two tides previous to the highest tide of the moon — say 

 for four tides before the highest of the moon. The very next tide after 

 the highest tide may run within a few inches of the same flood-mark, but 

 will have a comparatively poor head and lack crispness.' 



On the Size of Waves as related to the Bate of Admncc of a Cyclone. 



The greatest waves will be developed in that part of the cyclone in 

 M'hich the direction of the wind coincides with the direction of advance 

 of the cyclone, and I wish to call attention to the fact that, along this 

 line of action, of all the waves which the velocity of the wind is capable 

 of increasing, that length will enjoy superior opportunities for growth 

 whose group velocity is equal to the rate of advance of ^ the cyclone, the 

 storm either outrunning or lagging behind the transmission of energy in 

 waves of any other length. The velocity of the group in deep water is 

 half the velocity of the individual waves. 



It was pointed out in the last report that the period of the longest 



