June 12, 1885.] 



SCIENCE. 



483 



A RECENT JAPANESE EARTHQUAKE.^ 



Ax unusually great earthquake was felt in and 

 about Tokio on Oct. 15, 1884. The annexed auto- 

 graphic record of it comes, with the following par- 

 ticulars, from my former assistant, Mr. K. Sekiya, who 

 is now in charge of the seismological observatory of 

 the Unive; sity of Tokio. It was given by a horizontal 

 pendulum seismograph of the kind recently described 

 in Science (iv. 516), and it 

 has many features in com- 

 mon with the examples of 

 records shown on p. 517 of 

 the same volume. But in 

 the present case the ampli- 

 tude of the earth's horizon- 

 tal movement far exceeds 

 any thing that has been re- 

 corded since observations of 

 this kind were instituted, in 

 1880. 



The figure shows the rec- 

 ord reduced to about one- 

 third its actual size. The 

 undulations on the inner 

 circle have been traced by a 

 pointer which registered the 

 north to south component 

 of motion, and those on the 

 other circle by another 

 pointer, which registered 

 east to west motion. The 

 pointers are prolongations 

 of horizontal pendulums, 

 and trace their records on a 

 revolving sheet of smoked 

 glass, which in this example 

 was started into motion by 

 the earthquake itself, through 

 the agency of a delicate elec- 

 tric contact-maker. The 

 plate is driven by a clock- 

 work train, which, after 

 starting, quickly reaches a 

 steady rate under the control 

 of a fluid friction governor. 

 The speed of rotation was 

 one revolution in eighty- 

 two seconds. The short radial lines mark seconds 

 during the first part of the disturbance. The record 

 on the outer, or east to west, circle, has been turned 

 round so as to bring it into synchronism with the 

 inner, or north to south, record ; and the earliest mo- 

 tions are distinguished in the cut by the use of a 

 somewhat heavy line. The records begin at a and a, 

 and are traced in the direction of the arrow, which 

 is opposite to the direction of motion of the glass 

 plate. At b the east to west record comes to an abrupt 

 stop, owing to the displacement there having been so 

 great as to carry that pointer off the plate altogether. 

 The inner record extends over nearly four complete 

 revolutions, showing that visible motions of the 



1 From Nature, April 23. 



ground lasted for about five minutes. During the 

 first half-dozen seconds, while both components were 

 being registered, there is a tolerably close agreement 

 of phase between the two, showing that the displace- 

 ments were then not very far from rectilinear. The 

 greatest motion in this part of the disturbance took 

 place five seconds from the start. At that point the 

 actual motion of the ground was 3.7 centimetres 

 from east to west, and 2.2 centimetres from south to 



north. The displacement of the ground is mul- 

 tiplied four times in the original record, or about one 

 and a third times in the reduced copy given here. 

 The two components taken together represent a move- 

 ment of the ground, from one side to the other, of 

 no less than 4.3 centimetres, — a quantity which is 

 in striking contrast to the ' 5 or even 7 millimetres ' 

 which, after three years' experience, I named as the 

 amplitude to which in a Tedo earthquake the dis- 

 placement from the mean position 'occasionally 

 rises.' 



So far as can be judged from the north to south 

 component alone, the most violent motions were 

 over in about ten seconds; but for some minutes 

 afterwards, the oscillations, though very much re- 



