ON SEISMOLOGICAL INVESTIGATIONS. 67 



VI. On the Dissipation of Earthquake Motion as measured hy 

 Amplitude and Duration. 



In order to obtain some idea of the manner in which a very large 

 earthquake approaches extinction as it radiates, a comparison has been 

 made of the amplitude and duration of motion of twenty-two large earth- 

 quakes, as recorded by instruments installed at varying distances from 

 megaseismic origins. 



The iirst question which arises relates to the comparability of the 

 records. The instruments which give the records here used are, for the 

 most part, the type adopted by the British Association — viz., Milne hori- 

 zontal pendulums. A single record of a world-shaking earthquake from 

 one of these instruments or from any other instruments would by two 

 observers be interpreted for amplitude in a similar manner. For duration 

 however, this would not necessarily be the case. An examination of the 

 film by one observer might make the duration a few minutes longer or 

 shorter than that determined by another observer. Again, two similar 

 and similarly adjusted instruments beneath the same roof may yield 

 records showing slight differences, particularly with regard to duration. 

 A complexity of factors conspire to render seismograms obtained from 

 what we regard as similar, similarly adjusted, and similarly installed 

 seismographs not strictly comparable. They are, however, to a certain 

 extent comparable, and an estimate of this comparability may be obtained 

 by an examination of a series of teleseismograms obtained from instruments 

 installed over an area like that of Great Britain, each part of which is 

 practically at the same distance from all very distant origins. In Britain 

 we have stations at Shide (Isle of Wight), Kew, Bidston (near Liverpool), 

 and at Edinburgh, each of which is provided with British Association 

 pendulums. The foundations at these places are respectively chalk, 

 alluvium, red sandstone, and volcanic rock. The records for amplitudes 

 for earthquakes with distant origins have been as follows.^ 



Amplitudes. — For seventeen large earthquakes the average amplitudes 

 measured in millimetres were as follows : Shide, 2'1; Bidston, 1*4 • Edin- 

 burgh, 1-4 ; and Kew, 1-5. Three of these are distinctly comparable. 



1. Decrease in Amplitude. 



As illustrative of the decrease in amplitude six large earthquakes have 

 been selected from the catalogue of earthquakes recorded in the Antarctic 

 regions in the years 1902-3. This catalogue is now being published. The 

 dates and origins of the selected disturbances are the tirst six given on 

 page 72. At distances varying between 30° and 165^ from the above 

 origins, records were made of each of these earthquakes at many stations. 

 The actual number of observations are given on page 72. The instruments 

 used were in all cases Milne horizontal pendulums. The amplitudes in 

 millimetres in relation to distances from origins measured in degrees are 

 shown in tig. 2. The small figures on a curve indicate the number of 

 stations at or near given distances from an origin which were used to com- 

 pute an indicated amplitude ; thus for earthquake No. I., at a distance of 

 25° from its origin, one station gave an amplitude of 5-8 mm. At about 

 50° the average amplitude for four stations was 3mm., &c. 



' See Brit. Assoc. Reports for 1901, p. 48, and 1903, p. 82. 



