BRITISH ASSOCIATION. 



245 



38 miles from Garrichrow ; westward to Dalmally, about the 

 same distance ; as far north as Glenlion, 30 miles ; and 

 southward to Alloa and Stirling, 20 or 30 miles. All the 

 shattered chimneys noticed near Duniva, were on walls, &c. 

 running N. and S.; those on E. and W. walls being un- 

 touched. The injured buildings stood on a gravelly soil ; 

 but the distance from rocks below was unknown. There 

 was nothing in the weather previous to the earthquake, to 

 give any notice of its approach ; indeed, after a course of 

 some years' observation, no exact rule in this respect has 

 been obtained ; even a period of wet weather, which was 

 formerly thought the constant forerunner of frequent and 

 violent shocks, is not always succeeded by them ; and, on 

 the other hand, earthquakes have occurred when the sky 

 was clear and open. The spot from which the earthquake 

 shocks in Perthshire appear to originate, being situated 

 about a mile to the north of Duniva, it is not difficult to un- 

 derstand why walls running N. and S. were affected ; and 

 those from E. to W. untouched. 



3. On the 9th September, 1841, another pretty severe 

 shock was felt at Comrie, about 1 0' before midnight. The 

 following morning the Association's instrument in the steeple 

 was inclined f of an inch to the south ; that in the Comrie 

 House i an inch to the north. This disagreement in the 

 indications may perhaps be accounted for by the occurrence 

 of two other shocks in the course of the night, and pre- 

 viously to the examination of the instruments : the weather 

 during the two preceding days was remarkably wet and 

 close. 



4. On the 8th June, of the present year, two shocks were 

 felt at Comrie, between 1 and 2, a.m. The horizontal pen- 

 dulum recently sent to Mr. Macfarlane's house, indicated 

 an upheave of the ground to the extent of a quarter of an 

 inch. From the review of all the details, it seems probable 



s2 



