94 



conjunction with those from Christchurch, Wellington, and Perth, have confirmed this observation, and we 

 have now a number of instances where the movement from an epifocal area has travelled round and 

 through the world, to re-appear as a recordable quantity at its antipodes. 



It is not affirmed that in the region between an epicentral district and its pole seismic movement did 

 not reach the surface of the earth, but only that even with instruments very much more sensitive than the 

 Milne type motion has not been detected. The phenomena under consideration might also be described as 

 antipodean resurgences, convergences, focal effects or contrecoups, each of which, however, might be objected 

 to as implying an explanation for this antipolar relationship. 



In the preceding registers we find the following 19 illustrations of possible re-appearances, viz. : 

 Numbers 1, 32, 34, 51, 53, 59, 83, 89, 91, 93, 95, 96, 108, 111, 115, 117, 120, 129, and 130. 



Out of these it seems that with earthquakes Nos. 1, 34, 83, 89, 95, 96, 117, 120, and 129, the surviving 

 phase has been P 3 . At Hamburg, Strassburg, and other stations where there are pendulums with a shorter 

 period and a higher multiplication than those of the Milne type, PI has occasionally been recorded, e.g., 

 this is the case with Nos. 1, 93, 111, and 130. In other instances the polar responses have been nearly 

 simultaneous, a conclusion, however, which for many reasons may be more apparent than real. 



The interpolar transit of a wave of the P 3 type may be compared with that of a deep-sea wave down a 

 rapidly widening and then up a similar but rapidly narrowing estuary. The dimensions of these estuaries 

 are assumed to be large. When half-way on its journey the height of the wave and its energy per unit 

 area would be less than at its commencement or its terminus. It might, therefore, traverse the central 

 area and not be noticed, but because of subsequent convergence it might become recognisable at points 

 still farther from its origin. 



With very large earthquakes the movements were recorded all over the globe, and from experiments 

 now in progress at Pribram, in Bohemia, the seismograms obtained at a depth of 1150 metres, although 

 they show a diminished amplitude, differ but little from those relating to the same disturbances 

 recorded on the surface. The earthquakes we have to consider are of this type, but less in magnitude. 

 Let us imagine one of these smaller efforts to start over an epifocal cap subtending 10 at the centre, and 

 that this expands as a ring 5 in width until it reaches the quadrantal region. The area of the cap or ring 

 in the two positions will be approximately as 1 to 11, and if we neglect loss due to friction and assume 

 constant energy, the intensity will be diminished in like ratio. With such conditions it seems conceivable 

 that a disturbance might be missed in the quadrantal region and recorded at its antipodes. The distance 

 to which motion would invade the superficial region between the focus and the quadrantal region would 

 depend upon the intensity of the disturbance at its origin. 



The reappearance of PI, which is probably a condensational wave, may be accounted for by assuming 

 that reflections are focussed in an antipodean region. 



8. Seismograms, Pulsations, Magnetograms, and the Value of g. 



It is now well known that at certain observatories magnetic needles are frequently disturbed by unfelt 

 earthquake motion. To throw light upon the consequent irregularities which from time to time are 

 shown in the magnetograms at particular stations, horizontal pendulums have been established. The 

 records given by the latter instruments are due to mechanical movements, but whether the corresponding 

 perturbations shown in the magnetograms are due to a similar cause is by no means certain. At one 

 station teleseismic movement may disturb surrounding and subjacent magnetic materials, with the result 

 that needles at that station may respond to magnetic effects, which would not be the case at stations 

 where the neighbouring materials which had been equally disturbed were non-magnetic. 



At Ross Island the basalts are distinctly magnetic, while Mount Erebus and other recent cones indicate 

 that physical and chemical characters, and also the arrangement of magnetic materials, have suffered 

 change. 



The varying activity of Erebus suggests that these hypogenic processes have not yet ceased, and with 

 seismic disturbances it seems probable that large bodies of magnetic magmas and rocks are, at least 

 temporarily, disturbed and altered. We might, therefore, anticipate that the larger seismograms obtained 



