that may be employed in Earthquake Measurements. 47 



results may be obtained by using weak springs and surround- 

 ing the ball with some more viscous liquid, such as treacle or a 

 mixture of tar and pitch, if only the alteration of viscosity by 

 change of temperature can be readily compensated for, and if 

 it be possible to easily employ a good recording apparatus 

 when using much friction. On the whole our calculations 

 point to the employment of a small amount of friction and 

 strong springs ; but for the following reasons we feel that our 

 calculations should be chiefly employed only for directing and 

 making use of experimental knowledge : — first, because we 

 assumed the friction to be proportional to the mass of M and 

 to its velocity, whereas in reality the friction is probably com- 

 posed partly of a constant term depending on the recording 

 apparatus, and partly of a term proportional to the square of 

 the velocity of M, and to a certain extent proportional- to its 

 surface ; secondly, because extremely little is at present 

 known regarding the nature of the earthquake shock. Even 

 the period of an earthquake vibration does not seem to have 

 been measured with any approach to accuracy ; the informa- 

 tion obtained in some cases from the stopping of clock-pendu- 

 lums is quite unsatisfactory, since the limits between which 

 we can place the period of earthquake vibrations so as to stop 

 an ordinary pendulum-clock are wide apart. 



Section G. 



Fig. 5 (PL IV.) shows roughly our first idea of the construc- 

 tion of a seismometer in accordance with the principles we have 

 enunciated. A leaden ball of some 400 lbs. mass is supported by 

 five strong spiral springs inside a strong iron case, rigidly fixed 

 to the rocky crust of the earth; four of the springs are horizon- 

 tal, and one vertical, and all have the same period; so that if 

 there were no friction the centre of mass of M would describe 

 an ellipse when M is freely vibrating. In order to get a record 

 of north-south, east-west, and up-down motion of M, three arms, 

 two of which (A B, C D) are shown in the figure, carry pen- 

 cils pressed by means of small spiral springs on a band of 

 paper moved regularly by clockwork in a horizontal direction 

 at right angles to B I) — the clockwork, as in Professor Pal- 

 mieri's and other instruments, being set in motion at the com- 

 mencement of the earthquake. The arm AB is rigidly fixed to a 

 small piece AE at right angles to it (see figure of ball enlarged); 

 this again by means of a pivot at E is fixed to E F, which is 

 rigidly attached to the ball. A pin at A, supported by the frame- 

 work of the instrument, allows A B to move round it, and so to 

 record vertical motions of the ball ; and the pin A, having a 

 certain amount of lateral motion in the slot, combined with the 



