on Earthquakes in Japan. 363 



interchange of watches, even if several are interchanged simul- 

 taneously, has not proved satisfactory, owing to the variable 

 rate of the watch during its journey. The interchanges made 

 by us were by railway between Tokio and Yokohama, a dis- 

 tance of about 18 miles. 



Comparison of the times at the two places can be made by 

 referring each to a well-regulated chronometer. But for ordi- 

 nary observers the taking of the chronometer-rate is extremely 

 troublesome. 



Clock-stopping. — The ordinary method of taking the time is, 

 as in Palmieri's apparatus, to stop a clock. We have already 

 indicated how an ordinary clock with a balance-wheel may be 

 stopped. A pendulum-clock may be stopped, as Mallet has 

 shown, by dropping a catch on a pin fixed in the pendulum- 

 bob. The best way of relieving this catch, which is a toothed- 

 bar axled at one end, is to hold up the other end by resting it 

 on the extremity of a horizontal wire fixed to the bob of a 

 conical pendulum — for example, the extremity of one of the 

 indices of a conical-pendulum seismograph. The wire ought 

 to be in the same straight line as the bar which holds the bob 

 of the conical pendulum deflected. The whole of this apparatus, 

 which can be made for a few pence, can easily be made 

 small enough to go inside an ordinary clock-case. 



Time-taker. — Rather than the method of stopping a clock 

 at the time of an earthquake in order to obtain the time, we 

 are now using the following method : — A clock with a central 

 seconds' hand is taken. The hour-hand and minute-hand are 

 each produced and bent outwards at right angles to the face. 

 The hour-hand, however, is made slightly longer than the 

 minute-hand, and the minute-hand slightly longer than that 

 indicating seconds. In consequence of the bending upw r ards, 

 the ends of the three hands move in the same plane, instead of 

 being slightly below each other. These ends are furnished with 

 very small heads of wash-leather, which from time to time are 

 soaked with an oily ink. A light flat ring, with divisions on 

 it corresponding to those of the clock-dial, is so arranged that, 

 at the time of a shock, it can be quickly advanced until it 

 touches these heads, and then drawn back. By so doing, 

 the hour, minute, and second of the shock are indicated on 

 the varnished face of the ring. This back ward-and-for ward 

 motion of the ring is effected by means of a crank and 

 connecting-rod, the crank being caused to make half a turn by 

 means of a pulley which is turned by a falling weight. At 

 the time of a shock an electromagnet in connexion with a 

 circuit-closer pulls a catch away from the pulley, and the 

 crank-pin, which is at its lowest position, moves to its highest 

 position, while the weight moves to a small platform. This 



