704 DARBYSHIRE [CHAP. 20 



17.3 cm; (c) shows the effect of the addition of a reflected wave and how a 

 mixture of two pressure variations of two different frequencies can be ob- 

 served ; (d) shows the effect at 37.0 cm when the effect of the progressive wave 

 has completely disappeared and all the pressure variation is of the double 

 frequency. No measured pressure variation was more than 9% different from 

 the theoretical value. 



It is not possible to assume, however, that the water is incompressible if 

 the time for a disturbance to be propagated to the bottom and back is com- 

 parable to the sea-wave period. This is the case in deep oceans and, when the 

 compressibility is taken into account, there is no uniform unattenuated pressure 

 fluctuation at large depths but a compression wave whose planes of equal 

 phase are horizontal. This compression wave is generated by the gravity 

 wave motion in the higher layer down to a depth of |A below the surface and 

 is twice its frequency. Resonance effects will occur when the depth is about 

 {^71 + 1) times the length of the compression wave. 



In nature, trains of waves of equal period can meet head on when waves get 

 reflected off the coast and meet the incident waves. Also, in a fast moving 

 depression where the winds can occur in opposite directions over different 

 parts of the ocean, trains of waves moving in opposite directions can be gen- 

 erated. A similar situation will occur when one depression follows very closely 

 behind another. Using values of wave heights and spectra usually found in the 

 middle of storms and (for the coastal case) values of reflection coefficients 

 usually found off cliffs and beaches, Longuet-Higgins found quantitative 

 agreement between his theory and observations. 



Work was carried out in 1950 which confirmed this theory by considering 

 the spectra of corresponding sets of waves and microseisms. Three cases were 

 considered where only one isolated storm in the Atlantic Ocean was producing 

 waves and microseisms. It was clear from all three examples that there was an 

 increase in microseism activity and that eventually, when the swell arrived at 

 the coast, there was a close relationship between the swell and microseism 

 spectra. Fig. 5 shows one example. The wave spectra show a new band of 

 swell of over 20 sec period arriving on 14 March, 1945, at 1500 h and from this 

 time onwards there is a very close correspondence between the two sets of 

 spectra with the microseisms having half the wave period. There is a new band 

 of activity on the microseism spectra, however, on 13 March, 1945, at 2300 h, 

 16 h before the appearance of the new band on the wave spectra. Similar 

 results were obtained for the other two occasions, the microseisms appearing 

 in one case 28 h before the waves. In all three cases, the new band of microseism 

 activity appeared when the conditions were such that, over a large sea area, the 

 winds had veered round sharply. 



This work showed, as Banerji had found, that the increase in microseism 

 activity preceded the swell and that microseisms could be used as storm 

 warnings. Also, because the microseism period is half the sea-wave period in 

 the storm area, the wind speed in the storm area can be determined. 



The two -to -one ratio of sea- wave period to microseism period has since been 



