[klotz] earthquake WAVES 51 



43 -4 



Now that the first harmonic analysis of earthquake waves has 

 been made, one may properly ask what does it teach us or what have 

 we learned? A completely satisfactory answer can not at present be 

 given. Some expected results have been obtained, but also some 

 unexpected ones. 



Considering the different kind of waves in their order, we note 

 that for the P waves, which the analysis limits in the lower limit to a 

 period of 1 second, shorter periods if recorded are not analysed. Yet 

 we know that often the P waves are of the latter order. When we 

 approach longer periods, periods from about 5 to 6 • 5 seconds, we enter 

 into the region of microseisms and of the period of the Bosch pendu- 

 lums; the effect especially of the latter it is difficult to evaluate or 

 eliminate. One trouble about the period of the pendulum is that it 

 is not absolutely constant and one is not disposed to stop the function- 

 ing of the seismograph frequently to make a redetermination of that 

 period. The range of course would be less than a second. The N-S 

 component, as already noted, has air damping, while the E-W has 

 magnetic damping, making the pendulum practically aperiodic and 

 hence not subject, or at least subject to a far less degree, to the motion 

 of the pendulum than the N-S is likely to be. 



It will be observed that for the P waves the periods of approxi- 

 mately 4, 6 and 7 • 5 seconds stand out prominently on both compon- 

 ents, and may be taken as the predominant waves of the earth particles. 

 The long period waves amongst the P waves, especially the one of the 

 order of 30 seconds, I draw attention to without being able to correlate 

 them with any previous data available. 



For the S waves we have hitherto recognized as a common wave 

 length a period of 8 seconds or thereabout, and this the analysis shows 

 too. The large amplitude for 5? 7 period on the N-S component — air 

 damped — looks suspiciously like pendulum effect and resonance. 

 The freely moving pendulum when disturbed will make but few 

 oscillations before coming to rest, so that if disturbed again during the 

 interval of the period (compound) under consideration would most 



