176 Dr. C. Davison — British Earthquakes, 1893-99. 



usually employed in describing it. Taking the four eartli- shakes 

 together, we find that 6 per cent, of the sound comparisons are ta 

 thunder, 12 to a load of stones falling, 35 to the fall of a heavy hodij, 

 and 47 to explosions. Now, in the slightest earthquakes, such as 

 those described in this paper, the corresponding percentages are 31 

 to passing waggons, etc., 35 to thunder, 3 to wind, 8 to a load of 

 stones falling, 5 to the fall of a heav}' body, 14 to explosions, and 4 

 to miscellaneous sounds. But it is only in isolated earthquakes that 

 this is the case. The earth-sounds following a great earthquake or 

 the sounds which accompany the weakest after-shocks resemble 

 those caused by explosions or the fall of a heavy weight.' 



It follows from this that the centime of disturbance must have been 

 very small in the case of every one of the earth-shakes ; and this 

 conclusion is confirmed by the small size and the approximate 

 circularitj' of the disturbed areas. 



These two conclusions regarding the small depth and size of the 

 centre of disturbance are strongly in favour of a local origin of the 

 shocks. In the Rhondda valleys they are generally, I believe, 

 attributed to rock-falls in old mine-workings; aud this is no doubt 

 the origin of some of the shocks, for such falls from the roof are 

 known to occur and to lead to a subsidence of the ground above. The 

 chief difficulty in accepting this explanation for the recent dis- 

 turbances arises from the absence, so far as known, of any fallen 

 masses.- The Kilsyth earth-shake was referred without hesitation 

 by one of my correspondents to a blast in one of the whinstone 

 quarries in the town ; but, as another points out, the blasts have 

 occurred daily for twelve years without any suggestion of an 

 earthquake. Others, including several miners, consider that the 

 shock was caused by a fall of rock in one of the old pit- workings 

 by which the ground below is honeycombed. The shock at 

 Pendleton is apparently, though not distinctly, ascribed by 

 Mr. Stirrup to a slip of the Irwell Valley fault, and the proximity 

 of the fault to the centre of disturbance is a point in favour of this 

 theory. 



But this, of course, is not a peculiarity of the Pendleton coal-mines. 

 The beds about Kilsyth are cut up by a series of faults, and in 

 the Ehondda Valley faults pass close to the centres of the disturbed 

 areas of the earth-shakes. Perhaps a more significant fact is that 

 the coal-seams are generally worked right up to the fault. This is 

 certainly the case, as Mr. Stirrup informs me, at Pendleton, and it is 

 probably so at Kilsyth and in the Ehondda valleys. 



Now, by the withdrawal of the coal, the rock above is deprived to 

 a great extent of its support, and tends gradually to sink down and 

 close up the worked-out seam. Nowhere can this tendency be 

 greater than where the rock is severed by a fault from that which 

 adjoins it. Here the sinking Avould take place by a series of fault- 

 slips, each of which might give rise to a rather strong shock on the 

 surface of the ground above. But, as the slip would only afi"ect 



1 Phil. Mag., vol. xlix, 1900, p. 58. 

 = Natur.', vol. li, 1899, p. UO. 



