Vol. 60.] CAERNARVON EARTHQUAKE OF 1903. 241 



A'. Origin of the Earthquake.^. 



From the seismic evidence, we obtain the following elements for 

 determining the position of the originating fault: — (1) the mean 

 direction of the fault must be parallel, or nearly so, to the longer 

 axis of the isoseismal 7, that is, it must be from X. 4<» ; E. to 

 S. 40° W. ; (2) the hade of the fault must be towards the side on 

 which the isoseismals are farthest apart, or towards the north- 

 west ; (3) the fault-line must pass a short distance, a few miles at 

 the most, on the south-east side of the centre of the isoseismal 7 : 

 so that, in the epicentral district, its course may be submarine, or 

 it may pass through or near Clynnog or even a mile or two farther 

 to the south-east : and (4) the fault must be of some magnitude, 

 extending about 8 miles both to the north-east and south-west of 

 Clvnnog. 



On the map of the epicentral district (tig. 2, p. 239), are shown 

 two faults reduced from the Geological-Survey map (sheets 75 & 78) : 

 one traced for a distance of 14 miles from Aber to Dinlle on 

 the coast of Caernarvon Bay, the other for S or 9 miles from 

 Bettws Garmon to Clynnog. Of the two, the former satisfies the 

 seismic conditions more closely. Its average direction is X. 52° E. 

 and S. b2 z W., it hades to the north-west, and, according to Eamsay. 

 the downthrow of the Silurian beds on that side is between 400U 

 and 5000 feet at Pentir (3 miles south of Bangor), and between 

 2009 and 3000 feet at Dinas (4 miles farther to the south-west). 

 If the fault, after leaving Dinlle, is continued under the sea as far 

 as Xevin, trending rather more to the south, it would occupy 

 approximately the position assigned to the originating fault. As 

 no other large fault is known to exist in the epicentral district, it 

 seems probable that the Caernarvon earthquake was caused by a 

 slip along the Aber-Dinlle Fault. 



The region of the fault-surface occupied by the seismic focus was 

 about 16 miles in length, extending from near X'evin to near 

 Caernarvon -. and the amount of displacement was almost uniform 

 throughout, dying away somewhat rapidly towards both ends. 

 Though two maxima of intensity were observed at some places, and 

 were indicated on the seismographic record at Birmingham, there is 

 no evidence that the focus was discontinuous. The displacement 

 appears to have been of that simple type to which the great 

 majority of slight earthquakes owe their origin, and to have been 

 distinguished onlv bv its great length. 



c j 



The accessory shocks fall naturally into two classes. The first 

 includes those, six in number, that were strong enough to attract 

 the attention of several or many persons ; the second includes six 

 tremors (three of them accompanied by sound) and two earth- 

 sounds, but all so weak that their occurrence in each case rests on 

 the evidence of only one observer. 



The fault-slips corresponding to the former class were confined to 

 the north-eastern margin of the priucipal focus, or to its immediate 



