ON SEISMOLOGICAL INVESTIGATION. Gl 



which it follows that although the largest earthquakes of the year 1899 

 are discussed many earthquakes which are comparatively smaller have 

 been omitted. 



The object of the discussion is to indicate by examples some of the 

 directions in which this extensive system of earthquake observation is 

 increasing our knowledge of dynamical phenomena inherent to the world 

 on which we live. 



The plan of the discussion is as follows : — First, those earthquakes 

 which have been I'ecorded at the greatest number of stations, and which 

 have Icnoivn origins, ha\e been selected from the others and analysed 

 separately. To contirm tlie results towards which these analyses point, 

 references have been made to the more trustworthy records obtained by 

 similar instruments in previous years. The principal objects in view have 

 been as follows. The determination of the velocities with which various 

 types of earth vibrations are propagated and the duration of preliminary 

 tremors at varying distances from origins ; to show that earthquake 

 repetition and echoes are fairly frequent and to point out the existence 

 of phenomena for which satisfactory explanations are as yet wanting. In 

 connection with these investigations references are made to hypotheses 

 relating to the physical condition of the interior of our earth. 



Second, the results obtained by the above analyses are used as a 

 means to determine the foci of disturbances not included in the first 

 section of this paper. These foci, which for the most part are sub-oceanic, 

 in some instances indicate localities where it would be unwise to lay cables, 

 and where we may exjject to find configurations differing from those shown 

 ujjon our physical maps. 



Remembering that very many of the earthquakes discussed represent 

 initial disturbances which were followed by many after-shocks, the map 

 depicting these foci shows the regions on the surface of the earth where 

 in the year 1899 seismic activity was most pronounced. 



2. Velocities of Earthquake Waves. 



The knowledge hitherto at our disposal respecting the velocity of trans- 

 mission of earthquake motion over long paths has been based on records 

 obtained from instruments differing in type and sensibility, all of which 

 were installed in Europe. The result of this has been that, although the 

 registers led to the determination of average velocities along paths of 

 varying lengths, they never gave actual velocity from point to point. It 

 was seen that along paths from 10° to 90° the velocity of transmission of 

 the preliminary tremors increased rapidly with the lengths of these paths, 

 whilst the average velocity for large waves increased but slightly. With 

 regard to the former my own analyses of heterogeneous materials led to 

 the conclusion that, if the preliminary tremors travelled along paths 

 approximating to chords through the earth, then the average velocity of 

 transmission to a distant station was practically dependent on the square 

 root of the average depth of the chord connecting that station and the 

 earthquake centre. This furnished Dr. C. G. Knott with the hypothesis 

 that the square of the velocity of these particular vibrations, which were 

 in all probability compressional, was a linear function of the depth. With 

 this assumption, and with a given initial velocity, the rate of transmission 

 at any point within the earth could be determined and wave fronts 

 dtawn ; and by acceptihg a law respecting the increase of density within 



