314 



DISCOVERY 



record movements from all points of the compass, it 

 is necessary to mount two pendulums at right angles 

 as shown in the figure. 



In practice, one of the points of support is fitted 

 with screw adjustments to regulate the degree of 

 verticality. The more vertical these tw-o points are 

 set, the more horizontal becomes the plane in which 

 the pendulum swings. This has the effect of lengthen- 

 ing the time for each swing, and the corresponding 

 sensitivity increases as the square of this time of 

 swing. Periods up to 30 seconds per swdng are easily 

 obtained, when the pendulum becomes from 200 to 

 300 times more sensitive than the usual seconds 

 pendulum in a clock. But even so, when it becomes 

 a question of being operated by the tiny wavelets that 

 survive after radiating through solid material for 



thousands of miles, the recorded movements are much 

 too small to be of service, and this is where the 

 designer or experimenter meets his great difficulty, 

 namely, how to add a further great magnification 

 when the restoring force of the pendulum has been 

 reduced to its lowest ebb. 



There are two alternatives : [a) increasing the weight 

 of the pendulum ; (b) reducing the friction of the 

 magnifying devices to the utmost limit. 



The former has been carried to one extreme, both 

 by the Italian and German designers. Pendulums 

 weighing as much as two tons are in use. In this case 

 it is usual to use mechanical levers and pointers re- 

 cording on smoked paper ; and in some instances 

 siphon recorders using ink on white paper. 



Fig. 3 illustrates in diagrammatic form a seismograph 

 recording on smoked paper. 



The rod of the pendulum is 5 ft, long, carrying a 

 weight of 300 lb. The lever magnifies the pendulum 

 motion thirty times. The outer end carries a pivoted 

 glass needle which is almost in equilibrium, but has a 

 very feeble bias and so falls lightly upon the smoked 

 surface of the recording drum. The recording pointer 

 is raised from the smoked surface once per minute by 

 means of an electromagnet not shown in the figure. 

 This breaks the continuity of the line and so imposes 

 the time upon the trace. 



The drum is constantly revolved by a clockwork 

 motor, and at the same time made to travel endways 

 along the recording box. The trace drawn by the 

 pointer forms a heli.x from one end of the drum to 

 the other as a result of the two combined motions. 



When a pendulum is disturbed, it requires a con- 

 siderable time to come to rest. To obtain a true 

 record of the ground movement, it is necessary to 

 eliminate this natural tendency of the pendulum. 

 For this purpose damping devices are used : such as 

 vanes fitted to the pendulum and arranged to float 

 in an air chamber or oil bath. A later and superior 

 method is the utilisation of a copper vane floating 

 in a strong magnetic field. 



In the case of the heavy type of pendulum it is 

 possible to fit an electric alarm. A bracket arm fitted 

 to the outer end of the pendulum carries a platinum 

 horseshoe, one leg of which rests in a globule of mer- 

 cury ; the other leg stands adjacent to a second 

 globule whose distance is adjustable. These globules 

 are the terminals in an electric-beU circuit. 



The second globule is adjusted so as almost to touch 

 the platinum contact. In this position the smallest 

 oscillation of the pendulum rmgs the alarm. It was 

 this device, the only one in this country, which roused 

 the writer from bed a few minutes after the recent 

 earthquake in Japan had occurred. 



In the later and more delicate forms of seismograph 

 the pendulums are sometimes as light as i lb. ; the 

 magnification is then achieved optically by moving 

 beams of light recording on photographic film. 



In another tj'pe the magnification is obtained elec- 

 trically. The pendulum is fitted with a series of coils 

 of copper wire which float in a magnetic field. 



When the pendulum moves an electric current is 

 induced in the coils and passes on to a recording mirror 

 galvanometer. 



Causes of Earthquakes 



The cause of an earthquake has always been a 

 subject for speculation. Neither Religion nor Politics 

 have excited more controversy than this vexed ques- 

 tion. The multiplicity of theories is a direct result 

 of the paucity of real knowledge. It is generally 



