C. Davison — British Earth qitahes of 1889. 315 



not quite up to, the surface, the amount of the slip near the surface 

 being of course extremely small. 



Origin of the Sound-phenomena. — The observations on the sounds 

 that accompanied the Edinburgh and Lancashire earthquakes lead 

 to an important result, throwing light on the origin of these vibra- 

 tions and the part of the focus from vrhich they proceed. I will 

 now point out briefly what I believe to be the origin of these sounds, 

 reserving a more complete discussion of the question for a subsequent 

 paper in which other facts bearing on the problem may be con- 

 veniently brought together. 



In both earthquakes, the shock and sound, we have reason to 

 believe, were caused by slipping along well-known faults, the foci 

 of the sounds being nearer the surface than the foci of the corre- 

 sponding shocks. In both, also, the area over which the slip took 

 place must have been very limited in extent : and, while the amount 

 of the slip may have been greatest near the centre of this area, it 

 must certainly have died away towards its upper and lateral margins. 



Now, the seismographio records recently obtained by Prof. Milne 

 and others in Japan show that earthquakes usually begin with a 

 series of tremors very small in amplitude and very rapid in period, 

 from six to eight occurring every second, but becoming slower 

 before the shock takes place. These may last for many seconds or 

 even several minutes. Following, and continuous with, them come 

 the sensible vibrations, of larger amplitude and longer period, about 

 three to five occurring in every second. One or more of these, 

 attaining a still greater amplitude and longer period, of one or two 

 seconds each, constitute what are generally known as the principal 

 shock or shocks. The earthquake closes with vibrations of smaller 

 amplitude, but which have a period so long that no record of them 

 can be obtained. The earliest tremors, on the other hand, are not 

 registered on account of the smallness of their amplitude, and, in 

 all probability, as Prof. Milne -suggests, the " minute movements 

 which have been recorded ai'e the continuation of still smaller and 

 more rapid movements which .... have never yet been rendered 

 visible." It is to these supposed rapid vibrations which form the 

 front portion of an advancing earthquake, that Prof. Milne attributes 

 the origin of the earthquake-sounds.^ We may conclude from these 

 observations that, initially at any rate, the period of the vibrations 

 increases and decreases with their amplitude. 



Now, from different parts of the area over which a fault-slip takes 

 place, there must proceed vibrations differing greatly in amplitude, 

 and therefbi'e also in period. From the central portions of the 

 slip-area will come, as a rule, the vibrations of largest amplitude 

 and longest period ; while, from the margins there will proceed 

 minute vibrations of a period so short that they may be perceptible 

 only as sound. The position of the line separating the marginal and 

 central parts of the slip-area will depend only on the amplitude of 

 the vibrations corresponding to the period of the lowest sound that 



1 Prof. J. Milne, Japan Seismol. Soc. Trans, vol. xii. (1888), pp. 53-62 and 107-8. 



