Dr. C. Davison on Earthquake- Sounds. 49 



most important fact brought out by this analysis is that the 

 highest percentage is only 70. The continual approach 

 towards a percentage of 100, which is characteristic of Italian 

 earthquakes, is here absent. That in eight years as many as 

 65 earthquakes should disturb areas of more than 10,000 

 square miles and yet be unaccompanied by recorded sound, 

 while the audibility increases as a rule with the disturbed 

 area, is indicative, in all probability, of the defective 

 powers for hearing deep sounds possessed by the Japanese 

 people *. 



Other Countries. — The evidence for other countries is much 

 less complete than in the preceding, and I will therefore 

 merely give a few examples. Of 521 Greek earthquakes in 

 1896 recorded by Dr. S. A. Papavasiliou, 160, or 30*7 per 

 cent., were accompanied by sound. Of these, 309 were felt 

 in the island of Zante, and 101, or 32'7 per cent., of these 

 were attended by sound f. Of 39 earthquakes recorded in 

 New Zealand in 1891, sound was heard with 18, or 46 per 

 cent.f. In the Charleston earthquake of 1886 the audibility- 

 percentage is low in nearly every State, being 19*5 in the 

 central State of South Carolina, and, among the neighbouring 

 States, 21-2 in Virginia, 30*8 in North Carolina, 17*9 in 

 Tennessee, 20*5 in Georgia, 12*5 in Alabama, and 23*0 per 

 cent, in Florida ; but it is possible that these low figures are 

 due to the incompleteness of the records §. The most reliable 

 records for any non-British earthquake are probably those 

 which are obtained from Prof. F. Suess's study of the Laibach 

 earthquake of April 11, 1805. In this case the audibility- 

 percentage is 60*8 for the whole disturbed area, and 96' 7 for 

 the central province of Carinthia||. 



Relation between Sound- Audibility and Geological Structure. 



In earthquakes which are strong enough to be recorded by 

 seismographs, both the period and amplitude of the vibrations 

 depend on the nature of the surface-rocks. It is therefore 

 reasonable to suppose that some relation must exist between 

 the audibility of the short-period vibrations and the nature 



* The type of comparison is only mentioned in 13 cases in Prof. Milne's 

 catalogue, but it is worthy of notice that 10 comparisons are made to 

 wind, 1 to thunder, and 2 to explosions. 



t " Observatoire National d'Athenes," Bull. Mens. Seismol. 1 annee, 

 1896. 



X G. Hogben, Austral. Assoc. Rep. 1892, pp. 1-33. 



§ C. E. Dutton, Amer. Geol. Surv., Ninth Rep. pp. 410-528. 



I! Wien. Geol. Jalirb. 189(3, p. 523. 



Phil. Mag. S. 5. Vol. 49, No. 296. Jan. 1900. E 



