SECT. 5] MICROSEISMS 715 



which makes contact both with a series of studs attached to counters and 

 with a series of strips which only cover half the range of movement of the 

 needle (see Fig. 9). From the number of counts in a given time for each stud, an 

 estimate of the r.m.s. value for each component is found. The period is found 

 by the movement of the needle over the half strip so that a circuit is made or 

 broken at every zero crossing. The correlations are worked out by using two of 

 these movements so arranged that a circuit is made when the needles are on 

 the same side on both and broken when they are on different sides. The long 

 strip which is split at the centre is used for this purpose. The correlation co- 

 efficients can then be worked out by the formula given by Tomoda in Japan 

 (1956): 



r = sin [7r(n+ — 7i-)/2(w+ + ?!-)], 



where w+ is the number of counts when the needles are on the same side (i.e. 

 both positive or both negative) and n- is the number when the needles are on 

 opposite sides. It follows that, from the proportion of time the current is 

 switched on to the total time w+/(w+ + ?i_), r can be calculated. 



These instruments and the seismographs are portable and it is proposed to 

 set up mobile stations and combine the tripartite principle with the correlation 

 technique, as with this method there is no limit to the length of the side of the 

 triangle and distances of 100 miles should be possible. 



7. Other Work 



Some work with mobile stations has already been done by Bernard (1952) in 

 France in investigating the variation of microseisms from one place to another. 

 Work has also been carried out in other countries. The late W. M. Jones (1947, 

 1949) related microseism amplitudes and periods with movements of storms 

 past New Zealand. Work has also been done by 0. A. Jones and his collaborators 

 at Queensland, Australia (1947). In Australasia, storms are usually isolated and 

 so it is easier to study their effect. Thus Upton (1956, 1956a) has been able to 

 deduce a formula for the attenuation of microseisms with distance. He found it 

 to be of the form aocl/J?'-, a form suggested by Longuet-Higgins (1953); a 

 similar formula was found by the writer (1957) for sea waves. A great deal of 

 work has been done in recording microseisms in Antarctica. Imbert (1954) 

 studied those recorded in Adelie Land. There, because of pack ice, they could 

 not be generated near the shore and those recorded had to originate in the deep 

 sea. Microseism activity could be associated with the onset of barometric lows 

 and a relation was found between the microseism amplitude and the speed of 

 the lows and the rate of rotation of the winds. A detailed investigation of the 

 wave heights and periods to be expected in the storm area supported the wave 

 interference theory. Similar conclusions on Antarctic studies during the I.G.Y. 

 have been reported recently by Eppley (1959). 



