HORIZONTAL PENDULUMS. 133 



periods of oscillation of the EW and NS component horizontal 

 pendulums being respectively 18.8 and 35.5 seconds. It will be 

 observed that, in each component, pulsatory oscillations of period 

 4.0 seconds are superposed on the proper vibrations of the 

 pendulum, the motion assumed to be horizontal (see § 16) being 

 multiplied 37 times. 



Fig. 14 is a similar diagram, with the same multiplication 

 ratio, obtained on tlie 28th January, 1898, showing one of the 

 most remarkable storms of pulsatory oscillations, which I have so 

 far observed. The average period of vibration of the oscillations 

 was 6.5 seconds, while the maximum range of motion in each 

 horizontal component was 0.15 mm. 



16. On the nature of the pulsatory oscillations. 



The small slow movements, termed here pulsatory oscillations, 

 are not, in general, identical with the microseismic disturbances 

 indicated by ordinary short-period simple pendulum tromometers ; 

 or those indicated by a Paschwitz Horizontal Pendulum, when the 

 rate of motion of the photographic paper is small, say a few 

 centimetres per hour. The movements in these cases are simply 

 the proper oscillations of the pendulum, and they have apparently 

 no connection, at least no simple and general connection, with 

 the pulsatory oscillations. Fig. 15 represents an instance, in 

 which the proper motion of the Paschwitz Horizontal Pendulums 

 was great, yet unaccompanied with any pulsatory oscillations. 



It is a very important matter to determine whether the 

 pulsatory oscillations are purely horizontal movements or com- 

 ponents of tiltings due to undulations of the ground like waves 

 of water. If horizontal movements only, their recorded amplitude 

 would not depend on the proper period of vibration of the 

 horizontal pendulum, but simply upon the ratio of the multi- 



