Vol. 60.] THE CAERNARVON EARTHQUAKE OF 1903. 237 



of the ground is magnified 13-7 times by the pendulum, it follows 

 that the enlarged diagram represents the actual motion multiplied by 

 134. The whole movement is divisible, as usual, into three parts — 

 the preliminary tremors, the principal vibrations, and the concluding 

 undulations. The preliminary tremors are first perceptible at 

 10 b o^ 56* a.m. (Greenwich mean time), and lasted for 13 seconds. 

 The enlarged diagram shows hardly any trace of them ; but when 

 the original record is examined under the microscope, they appear 

 as minute notches, 51 in number, on the trace. The average period 

 of the tremors was. therefore, a quarter of a second. The principal 

 vibrations began at 10 h 6"' 9 V , and lasted 26 seconds. The total 

 number of vibrations is 40 ; but the first 19 are, as a rule, of much 

 greater amplitude than the rest. They have an average period of 

 •63 second, while that of the remaining 21 vibrations is '67 second. 

 In the 2nd and 19th vibrations, which are the largest of the 

 series, the range (or double amplitude) was *023 millimetre in the 

 east-and-west direction, or "024 millimetre (that is, about one- 

 thousandth of an inch) in the direction of the epicentre. Taking 

 the period of these vibrations as '63 second, the maximum accelera- 

 tion would be 1*3 mm. per sec. per. sec, or about one-tenth of that 

 of the Derby earthquake of 1903 at Birmingham. The concluding 

 undulations began at 10 h 6 m 35 s . On the enlarged diagram (fig. 1, 

 p. 236), twenty-seven may be seen, with an average period of 

 1 second; but, with the aid of the microscope, they can be detected 

 until 10 h 7 m 40% though so obscurely in some parts of the trace 

 that their exact number cannot be ascertained. The total duration 

 of the disturbance was thus 1" 44V 



Sound-Phenomena. 



The boundary of the sound-area is indicated by the dotted line 

 in PI. XX. It is 147 miles long from north-east to south-west, 

 136 miles wide, and contains about 15,700 square miles, or, say, 

 three-fifths of the disturbed area. In the whole of the latter area, 

 88 per cent, of the observers heard the earthquake-sound. In the 

 central district, the sound was unnoticed by very few persons, 

 the percentage of audibility being 100 within the isoseismal 7 ; 

 99 between the isoseismals 7 and 6: 98 between the isoseismals 6 

 and 5 ; and falling to 48 in the surrounding zone. The rapid 

 decline in audibility near the boundary of the sound-area is thus as 

 marked as it was in the case of the Derby earthquake of 1903. 



The number of observers who describe the sound is 291. Of 

 these, 45 per cent, compare it to passing traction-engines, motor- 

 cars, etc. ; 29 per cent, to thunder ; 7 to wind ; 8 to loads of stones 

 falling ; 1 to the fall of heavy bodies ; 7 to explosions : and 3 per cent. 



1 There is no trace of the second impulse registered at Bidston at 10 h 7 m 30 s . 

 At the beginning ot the diagram in fig. 1, there is a slight disturbance, which 

 was, I believe, caused by some particle of dust or roughness of the paper. It 

 will be noticed that the second half of the more prominent vibrations are super- 

 posed on a larger curve, which is due to a slight swinging of the pendulum. 



