ON SEISMOLOGICAL INVESTIGATION. 183 



Pacific there are at least ten sub-littoral districts where earthquake 

 frequency may be about half that of Japan. If this is accepted as 

 probable, the sub-littoral seismic activity of the Pacific is represented 

 by 2,500 shocks per year, some of which have been accompanied by 

 submarine landslips and consequent changes in the configuration of the 

 ocean bed. When these latter are gi'eat, it is assumed that ocean-waves 

 are created. If we consider the seismic activity round the coasts of the 

 other oceans and seas which cover our globe as being, when taken 

 together, equal to that of the Pacific, then for the world, out of a possible 

 10,000 shocks per year, 5,000 of them have their origin on the sub-oceanic 

 continental slopes. 



To get information about the second group, or earthquakes which 

 have originated far from land, we have to turn to the voluminous 

 catalogues of Perrey, Mallet, Kluge, di Ballore, Fuchs, and other statis- 

 ticians. Such extracts have been made by Dr. Emil Rudolph in his 

 papers, ' Ueber Submarine Erdbeben und Eruptionen,' ' who gives us an 

 account of 333 sub-oceanic earthquakes and eruptions. Because the 

 greater number of these shocks are of volcanic origin, they will be more 

 specifically referred to in the next section. The distribution of these is 

 various, but here and there they herd together, indicating localities where 

 changes are comparatively rapid. One favourite locality for submarine 

 disturbances is in the Equatorial Atlantic, about 20° W. long., and again 

 at 30° W. long., near to St. Paul's. For each of these regions Dr. 

 Rudolph gives about thirty-seven shocks, in depths of water exceeding 

 1,000 and 2,000 fathoms. 



The chief source of information for our last group is, however, derived 

 from the records of horizontal pendulums. Taking a list of them published 

 in the ' Transactions of the Seismological Society,' vol. xx., by the late 

 Dr. E. von Rebeur-Paschwitz, out of 301 records obtained in twenty- 

 seven months, there are only 25 which can with certainty be traced to 

 their origin. Out of the 176 which remain, 105 were almost simul- 

 taneously recorded at places so widely separated as Potsdam, Wilhelms- 

 haven, Strassburg, Nicolaiew, and Tokio, and therefore cannot be disposed 

 of as being due to some accidental disturbance of an instrument or to 

 small shocks of local origin. Each of them was a disturbance affecting a 

 very large area, and indicates an initial impulse of great magnitude. 

 What is true for the observations in Europe has also been true for my 

 own observations in Japan, and also in the Isle of Wight, the only 

 difference being that in Europe the stations were from 300 to 600 iniles 

 apart, whilst in Japan and the Isle of Wight the stations were usually 

 near to each other, and never more than 30 miles apart. In some 

 instances, however, earthquakes of unknown origins were recorded in 

 Japan and Europe, and it is fair to assume that in these instances the 

 whole world had been shaken. 



One disturbance noted by the author in Japan on June 3, 1893, had a 

 duration of five and a half hours. It was also recorded in Birmingham, 

 Strassburg, and Nicolaiew, at which latter place the duration of motion 

 extended over eleven hours. Amongst unfelt earthquakes, both for magni- 

 tude and duration, it exceeded all that have yet been recorded. 



Because the character of the unfelt movements, the origin of which 

 cannot be traced, is identical with the character of those which have been 



' Beitrage zur Geo2>hysiJi, Band I. and II. 



