ON TIIK KAUTinillAKK PHENOMKXA OF JAPAN. 



24* 



cmoiits 



ill r-ftr 

 sliovlly 



I. 



,1 I'or (U'- 

 ) (•()ml>iiH! 



tlio inon' 

 Lu><? on u 

 proniim'"' 

 il circuui- 

 iiphs witli 



also open 



ivint? ?^''i^^' 

 ;nino these 

 as follows : 

 loke.d pliiti' 

 ly droppcil 

 lorixontally, 

 ; 13 again in 



fxls ol" ovovy 



m a strip of 



\)late. I'l 



tlie pointov 



iliagvanis, * 



n. 



lakes at three 

 jitely pointed 

 k'ismological 

 jsly followed 

 [\ly produced 



-Near 

 „„ sides oi 

 These are 

 ound. The 

 ■ an eartii- 1 

 icr indices. 

 i°ly setiul 

 , intcrval> I 

 ,uue a smal]l 

 i usually lieli 



deflected, is set swinj^ing by nn automatic aTranpferaent in my honsc. Its 

 lirst swiiij.^ relioveH the catches and sots the phites in motion. Hy means 

 (,f the time ticks it is easy to eomparo tho oeciirroneo ot" any special 

 vibration taken at the varions .stations witliin ono hnndrcdth part of a 

 second. At the corner of my triangle, at Station Nutuhcr I , the ground 

 is moderately liard. Station II. is situated on a small promontory leading 

 out into a marKh and near a shallow pond. Station 111., where tho 

 i:iiiiiii(l is moderately hard, is behind a heavy brick building which stands 

 very near to the almost perpendicular face of a deep inoal. Tlie results 

 which have hithci'to been obtained, are briefly as fol!i)w.s ; — 



1. The diagram extending over the longest period of time and showing 

 tlic largest waves i.s always obtained from Station II. in tho vicitdty of 

 the mar.shy ground,- -the diagrams at tho other two Btations being much 

 smaller. The smallest record is invariably that at Station 111. near tho 

 decj) moat. 



2. At Stations 1. and 11. waves which may bo the s;imo can oc- 

 <'asionaIly bo identified, Itefc the identification of a wave at III., which is 

 cdmnion to 1. and II., is not oidy rare, but it is accompanied by great un- 

 certainty. 



;>. In a given earth(piak(< wo find that tho frequency of waves at tho 

 ♦lifl'erent stations in given intcrvaLs of time is diflbrent. For e.vample, tho 

 number of complete east and west vibrations during tho first twenty 

 seconds of time at the diirerent stations during five earthquakes Wcas as 

 follows : — 



Freiiucncij of Waves. 

 Kiniihir (if Woven in twenty 8Cco)ids. 



From tho above table it is evident that the average period must be 

 (lifTercnt at diifercnt stations. The small number of waves observed at 

 Station III. iu March 31 and April (J is probably due to the smallncss in 

 amplitude of many waves vvhich, because the period of the earthquakes 

 was long, have coalesced iu the diagram to form a straight line. Speaking 

 generally we may say that tho aver.ige period is longest at Station II. 

 near the marsh. 



At any given station, however, the period varies considerably during 

 tlio same disturbance. Thus, in ;&Iarch .j1, tho period of tho north and 

 south motion near the commencement of the disturbance Mas "20 second. 

 A few seconds later it was ••i second. 



A similar result is obtained by the analysis of the diagram taken at 

 Station 11. Selecting the largest waves from the diaurrams of the different 



