84 



REPORT— 1903. 



VII. The Velocity of Proi^agation of Earthquake Vibrations. 



In the ' British Association Report ' for 1902, p. 65, a diagram is given 

 showing the time taken for various phases of earthquake motion to 

 traverse arcs or distances corresponding to arcs of various lengths. 



From this diagram an arcual velocity for the maximum of large wave 

 movement may be derived of 3 km. per second. For the commencement 

 of such movements this would be slightly increased, and would then 

 accord with observations made by Dr. F. Omori, who obtains for this 

 particular phase an arcual velocity of 3'3 km. per second. 



To give actual velocities or average velocities for the preliminary 



tremors, not knowing the paths they follow, is accompanied by un- 

 certainties. What can be done, and is shown in the following table, is 

 from the above-mentioned time curve to calculate velocities on the 

 assumption that the paths have been arcs or have approximated to chords, 

 or we can make similar calculations from a time curve so corrected that 

 11 and 17 minutes are respectively taken to traverse distances corre- 

 sponding to 70° and 150°. The justification of reducing the steepness of 

 the preliminary tremor curve and yet keeping within the results of 

 observation rests upon the analysis given on pp. 5 and 6. 



From the above table it will be seen that if the preliminary tremors 

 follow paths which are arcual, then there is a marked increase in speed 

 of transmission on long paths as compared with the speed upon short 

 paths. If, however, the paths approximate to chords, then velocities 

 which arc approximately constant prevail. The deviation from being 

 actually constant along chordal paths is apparently a slight increase in 

 speed along paths taken nearer and nearer to the centre of the earth. 



The high values of 10 "5 to 12 km. per second suggest a high rigidity 

 for the world, whilst the approximate uniformity of speed within its core 

 indicate approximate uniformity in those properties which determine the 

 rate at which it transmits vibrations. Unless it is assumed that as we 

 descend in the earth electricity and density increase in the same ratio, to 

 which hypothesis there are objections, the inference is that the nucleus of 

 the world has a density more nearly uniform than is generally assumed. 



To satisfy the interpretation given to these seismometrical observations 

 what is required is a globe with an approximately uniform nucleus not 

 less than i j} of the earth's radius covered by a shell which passes rapidly 

 upwards into the materials which constitute the crust of the world. 



' In an article in Nature, April 9, 1 903, p. 638, on ' Seismometry and Geite,' 

 minimum values are given for these quantities. 



^ If these last values are plotted on squared paper a curve for their mean posi- 

 tion gives the following values: 3 to 5, 9-0, 10-4, 10-fi, 10-8, ll'O, 11-3, ll'S and 120 

 km. per second. 



