366 DR. CHARLES DAVISON ON THE [Aug. I907, 



impression which it produced was again as if a very heavy body had 

 fallen on a floor, followed by the quivering which such a fall would 

 produce. The duration of the shock was 1| or 2 seconds. 



At Menstrie, the noise began suddenly, and was as loud as 

 thunder overhead. 



(9) 1905, August 3rd, about 6 p.m. 



A very perceptible shock was felt by several observers at Red 

 'Carr. 



(10) 1905, September 21st, 11.33 p.m. (Principal 

 Earthquake.) 



Intensity, 6 ; centre of isoseismal 5, in lat. 56° 11-8' N., long. 3° 45*5' W. 

 Number of records, 139, from 57 places, and 22 negative records from 19 

 places (PL XXVI). 



Of this shock, there are 41 records of the time of occurrence, 6 

 being regarded as accurate to the nearest minute. The average of 

 the latter is 11.33 p.m., and one estimate of this time is given by a 

 signalman on duty. 



On the map of the earthquake (PI. XXVI) are shown three 

 isoseismal lines, corresponding to intensities 6, 5, and 4. The 

 isoseismal 6, only part of which can be drawn with accuracy, is 

 13 miles long, 8 miles wide, and 82 square miles in area. The 

 isoseismal 5, which is determined by a much larger number of 

 observations, is 20 J miles long, 14 miles wide, and contains an area 

 of 227 square miles. Its centre is 3 miles N. 25° E. of Menstrie, 

 and the direction of its longer axis E. 29° N. and W. 29° S. The 

 ■distance between the isoseismals 6 and 5 is about 3| miles on the 

 north, and 2f- miles on the south, side. The isoseismal 4 is 33| 

 miles long, 26| miles wide, and 700 square miles in area ; its 

 longer axis runs from E. 27° N. to W. 27° S. ; its distance from 

 the isoseismal 5 is 7| miles on the north side, and 5 miles on the 

 south. Outside this isoseismal, the shock was felt at Falkirk. 1 

 mile to the south, and at St. Eillans and Balmeanach, respectively 

 3 and 4| miles to the north-west. The total area disturbed must 

 therefore have been about 1000 square miles. 



In its nature, the shock of September 21st closely resembled its 

 predecessors, the only noticeable difference being that the prominent 

 vibrations, which again gave the impression of the heavy thud of a 

 falling body, were preceded, as well as followed by tremulous 

 motion. At some distance from the epicentre, these tremors were 

 imperceptible or nearly so. Thus, at Comrie, which is 15 miles 

 north-west of the epicentre, the earthquake began with a rumble 

 like that of a heavy cart approaching, and as it seemed to pass 

 below the house there was a heavy crash, the rumbling dying away 

 as if the cart were receding in the opposite direction. The mean 

 of 54 estimates of the duration of the shock is 3*4 seconds. 



The sound-area is co-extensive with the disturbed area, though 

 there is a marked decline in audibility in all directions near the 

 boundary. Thus, while the percentage of audibility is 84 for the 



