ohliiiuoly to rlio taiijj^nt of tlic ciiclr' it drew. Tlio liin's c niul <•' s-.tvi' to sliow 

 the relative jiositions ol' tin- i)ointor.:. The (list visil)i<' (iistmlianrr is at a (on 

 the inner rceonl). The reeonl extends over abont two revolntion-;. This is an 

 exeellent example of the prest-noe of .-liort-]-.eri'i(l vibrations of small aiii|ilitU(li' 

 in eonjunetion with lartrer and slower motions. The luri^-est <Ii.-iila<Tiiieii!, at //, 

 nu';isures 4.2Ö iiim. whin redneed t" ihi' natural scale. 



$ Ö9. Enrthqitake of Sepfemhcr -.'Htli, /.V.V.'. 



A moderate oartlit|iiake cif the ab >ve daf illn.strate.« well tlie arrival of 

 sh(Mt-i'eriod vibnitions in advance of the ]>rinei|>al movements. I'late XV III 

 show.s the reeord, whieh eonsi.sted almo.«t wholly of K-AV motion. It be^an at 

 a and wa'^ traced in the direction of the arrow. From a to h notliino- appears 

 I)iit a rip]ilo of minute .short-period wavi<. The jj^reater motions, whieh b{'o;in 

 at /', have at first short-))eiiod waves mixed np with th<'m ; bnt these disappear 

 after a time, and the eartiupiake dies ont in relatively lonji-period waves onlv. 



In thi.s case the early ] ait is shown with unii>nal clearnos, owinn- t) the 

 fact that previou.s to the .«hock the pointer was held by friction somewhat inside 

 of its position of eijnilibrinm, and c(iiwc(|nently the first slrikins; ean.sed it to 

 creep outwards from the centre. 



The multiplying ratio was (j to 1, and one revolution of the plate t'^k li'.t 

 i-WHind.*. 



§ 60. Enrlhfiuukc of April Mnl, /.V.V.7. 



A .sharp and lonji continued shock (followed by another sli<rlifci- >liakinii- 

 about ten minutes later), which ot^-nrred at lO..'i'» p.m. on A])ril 2.'>rd, ISS.], is 

 recorded on IMate.s XIX and XX. Plate XIX is the reeord niven l)y the 

 horizontal |)endulums of § 24, with a multiplication of 4. One revolution of the 

 plate took 04 second-. The shaking la.sted .«o long as to make it impossible to 

 follow all the motions, but the piincipal part of the disturbance is very clearly 

 .•^hown. Tlu' earliest visible motion is at a on the K-W circle. At l>, nine 

 seconds later, the motions snddeidy become large. The corresponding i)ar( ot' 

 the N-S record is //'. The record covers five revolutions of ihe plate, or li 

 minutes of time. 



I'late XX is a very remarkable record of tiie same ear(ii'|Uake, given by the 

 long |icnduium in.strument of S •>7, with one pointer free to ninve in any a/.imuth, 

 and having a multiplicatii>n of 10. The |>late revolved om-e cvi'ry 72 .«econds. 

 Nothing could show more clearly than tliis docs the extraonlinary irregularity ()f 

 cirthipiaku motion. In fact, it is impos.-ible (o apply liie terms ii.-ed in s])eaking 

 of regular undulations lo the >p;Lsinodic twist.-- and wriggles which llic surface oi" 

 the earth here ('Xi'dUed. 



In I'late XX the direction of motion of the plate inider the pointer was 

 EJUI'N. The principal movements are at the placii marked <■, and these conc^- 

 jxjnd to the recoixl.'* given by the horizontal [lenduluuis at 6 and // on I'late XIX. 

 6 



