74 REPORTS ON THE STATE OF SCIENCE. 



VII. On the Direction in which Earthquake Motion is most freely 



2)ropagated. 



We frequently see the isoseists of an earthquake in the form of 

 ellipses, which indicate that motion of a given intensity has been propa- 

 gated farthest in a particular direction. In a discussion of earthquakes 

 observed in the Antarctic regions, see 'Proceedings, Royal Society,' Vol. 

 A76, 1905, p. 293, I showed that certain large earthquakes had travelled 

 round the world in bands, the length of which had their greatest exten- 

 sion from their origins in particular directions. The object of the pi'esent 

 note is to show that certain large earthquakes with approximately known 

 origins in the Northern Hemisphere have been recoi'dable at a greater 

 distance measured towards the west than towai'ds the east. Also the 

 distance to which the motion extended across and to the south of the 

 equator was less than the distance to which it travelled east or west in 

 the Northern Hemisphere. The phase of motion here referred to is that of 

 the large waves which travel to greater distances than their precursors. 

 This particular phase has a practically constant speed, whether its path 

 is across the alluvial plains of Asia, the granites of North America, or 

 beneath an ocean. The crust of tlie world is apparently influenced by 

 movement in a medium which it covers. The number of earthquakes 

 considered has been seventy-four, and they are designated by their Shide 

 number given in tiie registers published by the British Association. Their 

 origins were in one of the following three districts ; — 



Distiiet No. 1 is south of the Caucasus, approximately 40° N. lat. and 

 45° E. long. The earthquakes considered are 588, 595, 598, 704, 705, 

 7136, 948, 1,077, 952, and 1,351. 



District No. 2 is north of Eastern India, approximately 80° E. long, 

 and 30° N. lat. The earthquakes considered are 618, 6136, 663, 626, 

 640, 644, 662, 632, 653, 676, 696, 692, 689, 793c, 720, 684, 679, 832, 

 833, 982, 1,070, 1,052, 1,036, 1,038, 963, 9,836, 1,0446, 1,320, 1,293, 

 1,184, 1,133, 1,240, 1,208, 1,135, 1,264, 1,167, 1,129, 1,153, 1,468, and 

 1,475. 



District No. 3 is east coast of Japan and north of the Philippines, 

 approximately 1 30° E. long, and 30° N. lat. The earthquakes considered 

 are 899, 8966, 935, 1,150, 857, 963, 1,166, 1,145, 1,387, 861, 859, 860, 863, 

 1,111, 1,266, 858, 903, 1,010, 1,274, 1,116, 1,130, 1,163, 1,386, and 

 862. 



In the following three tables which refer to these districts we find in 

 the successive columns — 1st, the names of stations where earthquakes were 

 recorded ; 2nd, the approximate distance of a station from an origin ; 

 3rd, the number of earthquakes which we should expect to have been 

 recorded ; 4th, the number of earthquakes which were actually recorded ; 

 5th, the percentage of possible records. Stations the names of v/hich are 

 printed in italics are in the Southern Hemisphere. 



In Great Britain there are five stations, and therefore the opportunities 

 for obtaining records have been greater than they were at isolated stations. 

 This partly may explain the high percentage of records obtained in 

 Britain. But if we reduce the number of records obtained in Britain to 

 the number obtained at San Fernando in Spain, where there is only one 

 instrument, the general result is but little altered. A similar remark 

 applies to India, where there are three stations. In New Zealand there 

 are two stations. In Manila the instrument is not of the Milne type. 



