440 



Prof. J. A. Ewing. 



[Feb. 10, 



III. u 0n a New Seismograph." By J. A. Ewing, B.Sc, 

 F.R.S.E., Professor of Mechanical Engineering in the Uni- 

 versity of Tokio, Japan. Communicated by Sir William 

 Thomson, F.R.S. Received January 17, 1881. 



The difficulty in earthquake measurements is to find a point which 

 does not move during the disturbance. This condition is, in certain 

 cases, fulfilled approximately (as regards horizontal motion) by the 

 bob of a pendulum,* and, for this reason, pendulums have been fre- 

 quently used as seismometers. A long pendulum, whose period greatly 

 exceeds the period of the earthquake waves, suspended from an ex- 

 ceedingly rigid frame, makes a fairly good seismograph, the earth's 

 motion relatively to the bob being shown by a pair of indicating levers 

 at right angles to each other, with their short ends in contact with 

 the bob near its centre of gravity, their fulcrum s fixed to the earth,, 

 and their long ends lightly touching a plate of smoked glass which is 

 kept revolving uniformly by clockwork. If the bob were to remain 

 quite stationary during an earthquake, this would give two curves, 

 showing on a magnified scale two rectangular components of the 

 horizontal displacement of a point on the earth's surface in conjunc- 

 tion with the time, from which the amount and direction of the actual 

 horizontal movement, and its velocity and rate of acceleration at any 

 time during the shock, could be deduced. There are, however, two 

 independent reasons why the bob of a pendulum seismograph does not 

 remain stationary during an earthquake shock. In the first place, it 

 is disturbed by the friction and inertia of the recording levers. [It 

 may be observed here that a considerable multiplication of the 

 motion is essential, at least in measurements of the small earthquakes 

 which are common in Japan, for the actual motion of the earth's sur- 

 face is usually only a fraction of a millimetre.] It is not difficult, 

 however, to reduce this source of disturbance to an almost insensible 

 quantity by a careful construction and arrangement of the joints, &c, 

 and by making the mass of the bob excessively great relatively to that 

 of the levers. The other and by far the more important cause of dis- 

 turbance is the motion of the point of suspension of the pendulum. 

 The motion communicated to the bob by this means would reach 

 a maximum, and altogether eclipse the true earthquake motion if the 

 period of oscillation of the supporting point agreed exactly with the 

 pendulum's free period. By making the pendulum long enough, and 

 the supporting frame rigid, this extreme case can easily be avoided, 

 but the effect is never entirely absent. The writer has observed that 



* And as regards vertical motion also if the bob be suspended by a spiral spring. 



