1881.] 



On a New Seismograph. 



445 



at first. The difficulty of getting a truly plane plate and of setting it 

 exactly perpendicular to its axis of rotation, made it needful to allow 

 a small amount of up-and-down motion of the marking points during 

 the revolution of the plate. This has been done by connecting the 



Fig. 4. 



straw pointer d to the other part of the lever c, by a hinge joint, which 

 gives the pointer freedom to move in a vertical plane, but leaves it no 

 freedom to move sideways relatively to the part c. The effect of this 

 would be that half the weight of the straw would be borne by the 

 glass plate, which would give a disadvantageously large amount of 

 pressure at the marking point. To remedy this, a very flexible 

 straight steel spring, g, has been added, which stands out from the top 

 of the stirrup, and bears a portion of the weight of d through the 

 tension of a fine silk fibre, h. This arrangement retains the necessary 

 vertical play of the pointer, without allowing its pressure on the plate 

 to be greater than is just sufficient to rub off the thin coating of lamp- 

 black which receives the record. So long as no earthquake occurs, 

 the pointers continue to trace out, over and over again, the same two 

 circles on the plate. 



Owing to the imperfection of the driving clockwork, it has been 

 found to be desirable to add an appliance for marking time at short 

 intervals on the plate during an earthquake disturbance. This is done 

 by a short pendulum, which remains deflected from its normal position 

 by the attraction of an electro-magnet, througii which a current is 

 permanently passing. A contact breaker, arranged to be broken by 

 the shock, interrupts the current and releases the pendulum, whose 

 beats then mark a scale of time upon the revolving plate during about 

 one revolution, after which they cease to mark. 



Since the seismograph was made it has recorded several earthquakes 

 quite successfully. The first of these (on November 3, 1880) was 

 found to consist of more than 150 successive undulations of irregular 

 amplitudes and periods. The mean period of a complete wave was 

 about - 6 second, and the greatest total amplitude of the earth's 



