11 



thoir aniplitiulcs to soino extent (•(|ualiziMl hy ['UssiDi;- siihsciiiiciitly tliidiinli ;i 

 mcdinin who«' limits of t-lastirity were so narrow as to Ik- t'xccodcd liy tlu' strains 

 tlio waves caused. Tlie elasticity oCtlie medinni mischt however Ite iicarlv |icrf'eel 

 with res(ieet to tlie verv iiiiiuite vil)rations wliieli accompany or oriüinallv lollow 

 the |)riiici|ial shock', and these woiilil thcrct'nrc lie tvansmitlcd widi ci>ni]iai:itivclv 

 little loss, and (if the ahove enneiusion is eorreet) with a velocity «jrcatei- llinn that 

 i)f the principal movements. Moreover, it apjiears proliaMc tltat a incdinm sn<'h as 

 clay or semi-solid nnid will transmit \ii)rations of short period more ncarlv like 

 a trnc sdiil. and \ilirMtioii- of lonu period nioi-c m-arlv like a viscous li jiiid, and 

 that the former will l)e jirojianatcd with yi'cater velocity than the latter. Jt is a 

 fact of common ohservation that llie •mind wliieh in manv cast's aicmnpanies an 

 earthtpiake is nsnally heard hi'lorc the ariival of ilie >e|i:natel\- perceptiMe nn- 

 dldati'ins of the soil ; and the writ<'r"s oliserxations have shown thaf the records 

 •riven 1)V scisnioMraphs frc.picntlx' exhiliil minntc \il>rations of short ]ieriiid at the 

 Ix'ijinnin«^ of earthipiakcs, eithi'i' <'nlirely precedjne- or sii|Krpospd on the earliest 

 princijial movements, whose ]icriods are longer and which coiitinne to ho traced 

 after the >hort-pcrio<l nio\i nients jiave died onf. 



§ 11. Miivnn()il< iif n piirliclr (hiriiia art E(U'lli']uii/:r. 



The actual constitution !■!' the earfii's crust is so far from Inane- homoe-eniMu-, 

 even on a large .scale, that any tlisturhanee projiaii^ated throui;;h it nui-t he e;i'eatly 

 mf>diticd hy the nnmerons reflections, refractions, and dif1Va<'tions which occur 

 ali'Ujr its rout<'. I'acu when a sinele impidse proceeds from ;i siuelc orii;iuatine; 

 Jioint, it is olivious from what has l)e(ai sai<l al>o\-e that in eeneral nian\ wave.s 

 iiiith of eonipres-^ion and of distortion will reach an ohserviufj; station distant from 

 the siiin-ce, traveliinjr in different lines. Tlie direction of the principal normal 

 im|)uise will not in }i;eiieral he that of a line joiniii!.'' the ohservini;- station with 

 the M)i I rcc. Its line of transit will u^uall\• lie lieiit in a x'eilie.al plane, ln.lh 

 >liarplv hy pas.s:i<jc |hi'oii<;li surfaces which .separate djfli'rcnt strata, and ;^raduallv 

 on account of continuous ehaii}]jes in the elasticity and density of a siiiele stratum, 

 produced, in the u)iper layers es|M'eialIy, hy irravify and hy the presence of mois- 

 ture ill the soil. And in inaiiy cases (liis liciidin<i; will lai^e place in a horizontal 

 plane aUo, with llii- rc-iill that the a/iniuth of the piineipal normal ilispla<'eni<'iil 

 wlii<-h iK-ciirs at a di-lant st.iliou will not coincidi- with llic licarini^ of the origin. 



When, ill addition to (he coiisideialious which lia\c heeii indicated ahove, it 

 is lioine ill liiiiid that the place of nriirin of an c:ii°tlii|Uakc may itsilf lie widiiv 

 exlcniled (as, for cxaiiipic, ill .-lioek> wliiili are caused liv (he formal ioii of a lis-^iirc 

 or " fault " of some |eni;thl, and that cadi jiarl of the initially disiiirlied ri'L'ion 

 will fji'Uer.dlv ori^iiiate no) one liul many sui'ees-ivc impulses, ii i- ejcailhat the 

 niovenienis whii'h niav lie expi-cted to in-ciir at an oli>er\iiiM- vfailon are eoniplex 

 to an alino«^t iiidelinitc dei^rcc. 



To coriiph'tclv <lcteniiiiie the iiKilioii of the ^roiiini at .'inv one point we 

 niii.st ohserve, from the hej^inniii^; to th>' end of the disdirl aucc, the (lisplaeenieiii 



