ON SEISMOLOGICAL INVKSTIGATIONS. .69 



tend to be asymptotic to the ordinate for distance. The distance from 

 an origin at which a curve tends to be asymptotic is apparently dependent 

 upon the intensity of the originating impulse, which may be estimated by 

 the magnitude of the records near to an origin and the distance at which 

 cQects of the impulse have been recorded. 



The rapid decrease in amplitude shown in the curves is probably in 

 large measure due to the spreading out of the waves as they approach 

 their equatorial region, and also to frictional resistance. An estimate of 

 this latter factor has been made by Dr. C. G. Knott. The data on which 

 the estimate is based will be found in the chapter on Seismic Radiations 

 in his forthcoming work on the Physics of Earthquake Phenomena. By 

 a comparison of the maximum amplitudes of the records of an earthquake 

 produced at any one station, a.s the tremors have passed through the 

 minor arc or through the major arc connecting the epicentre with the 

 station, he found that, in virtue of viscosity of the material, the amplitude 

 of the Large Waves is cut down to half its amount at a distance of 41°"2. 



Observations which indicate that certain earthquakes, although not 

 recorded in their quadrantal region, but may be recordable in antipodean 

 regions, suggest that the curves we have given would, if continued, have 

 a slight rise upwards.^ 



2. Changes in Duration. 



An indication of the value which may be placed upon a set of similar 

 instruments as recorders of the duration of earthquakes which have 

 originated at great distances, the following figures are quoted fi'om 

 the British Association Reports in 1901, p. 47 ; 190.3, p. 82. For 

 cwenty-four large earthquakes recorded at Shide, Edinburgh, Bidston, and 

 Kew, the total durations were respectively 1,493, 1,-516, 1,464, and 1,.371 

 minutes. The average durations of these earthquakes were therefore 62, 

 63, 61, and 57 minutes. These figures show that although the instru- 

 ments at these stations wei-e installed upon very difTerent foundations, 

 already described, the records they yielded were in close accordance. 

 They also tend to confirm the idea I have frequently expressed — viz., that 

 with megaseismic movements the crust of the world moves much in the same 

 way as a raft does upon the ocean. All parts of it, whether alluvium or 

 crystalline, respond to forced oscillations. For a small earthquake, by 

 which is meant one which only disturbs a small area, the result is dif- 

 ferent. In this case stations in close proximity to each other may yield 

 records of duration and amplitude which, amongst other things, depend 

 upon the nature of the strata upon which the instruments have been 

 installed. If a list is made of the duration of local shocks, and this is com- 

 bined with a list giving the duration of teleseismic movements, it would 

 appear that the duration of earthquakes increases with the distances of their 

 origins. Such a result, however, is based upon records which are hardly 

 comparable — one is a vibration of the earth's crust, whilst the other, to a 

 great extent at least, appears to be due to a mass movement in the mate- 

 rial beneath the crust. Attention was drawn to this in a British Associa- 

 tion Report, 1892, pp. 225-227. In that report I expressed the view that 

 both types of disturbance as they radiated exhibited a duration which 

 became less and less. Now that we have many records of durations for 

 given large earthquakes, which wei-e obtained at widely separated station.s, 



' Proc. Roy. Soc, Vol. A76, 1805, p. 293 



