Rayleigh, John William Strutt (1842-1919 ; 3rd baron ). The theory of sound . Volume II. 1995.

1/ Les contenus accessibles sur le site Gallica sont pour la plupart des reproductions numriques d'oeuvres tombes dans le domaine public provenant des collections de la
BnF.Leur rutilisation s'inscrit dans le cadre de la loi n78-753 du 17 juillet 1978 :
*La rutilisation non commerciale de ces contenus est libre et gratuite dans le respect de la lgislation en vigueur et notamment du maintien de la mention de source.
*La rutilisation commerciale de ces contenus est payante et fait l'objet d'une licence. Est entendue par rutilisation commerciale la revente de contenus sous forme de produits
labors ou de fourniture de service.
Cliquer ici pour accder aux tarifs et la licence

2/ Les contenus de Gallica sont la proprit de la BnF au sens de l'article L.2112-1 du code gnral de la proprit des personnes publiques.
3/ Quelques contenus sont soumis un rgime de rutilisation particulier. Il s'agit :
*des reproductions de documents protgs par un droit d'auteur appartenant un tiers. Ces documents ne peuvent tre rutiliss, sauf dans le cadre de la copie prive, sans
l'autorisation pralable du titulaire des droits.
*des reproductions de documents conservs dans les bibliothques ou autres institutions partenaires. Ceux-ci sont signals par la mention Source gallica.BnF.fr / Bibliothque
municipale de ... (ou autre partenaire). L'utilisateur est invit s'informer auprs de ces bibliothques de leurs conditions de rutilisation.

4/ Gallica constitue une base de donnes, dont la BnF est le producteur, protge au sens des articles L341-1 et suivants du code de la proprit intellectuelle.
5/ Les prsentes conditions d'utilisation des contenus de Gallica sont rgies par la loi franaise. En cas de rutilisation prvue dans un autre pays, il appartient chaque utilisateur
de vrifier la conformit de son projet avec le droit de ce pays.
6/ L'utilisateur s'engage respecter les prsentes conditions d'utilisation ainsi que la lgislation en vigueur, notamment en matire de proprit intellectuelle. En cas de non
respect de ces dispositions, il est notamment passible d'une amende prvue par la loi du 17 juillet 1978.
7/ Pour obtenir un document de Gallica en haute dfinition, contacter reutilisation@bnf.fr.

THE

THEOU Y OF SOUND.

TRK
.1..1..1

.lnJ

THEORY OF SOUND.

DY
JOHN
WILLIAM
BARON
ST.RUTT,
F.R.S.
RAYLEIGII,
M.A.,
FOttMRKLY
FULLOW
0F
TKIXITV
CAMr!):inGE.
COLLEG):,

VOLUME M.

bonbon;
MACMILLAN AND CO.
1878
[~.Z~/t~MMn'cf~

<E'n)nbWt!t:
A
i'~tffT)!t)nY(').(').AY.t!A
~T')'UKttf[YH)tS)TV)'ttKSS.

CONTENTS.
CHAPTER XI.
23G254.
Equations of
Aerial vibrations.
E.pmlity uf prcswuro in n)l directions.
motion. Equation of contimiity. Spcial form for incompressible ihnd.
Motion in two dimouBioua. Stroam function. Symmotry abont nu axis.
inVolocity-potontial.
Lagrango's thcorom. Stokes' proof. l'itysioU
Equfttion of cnn.
ThonMOti's iuYCfiti~tion. Circuhttinn.
tcrprotntiou.
Expression iu po)nr co-ordiu~teH.
tinuity iu torma of vcL.city-poifmti~.
Motion of meomprossibio Unid in fiimplyeonncctedHpMM is dotcnninod
hy bouudn.)? conditionn. Exteusiou touinttiply connectcd spMcs. Sphcro
of u-rotittionally moviug iluid suddeuly soliditiod wouid havo no rotation.
In-otn.tiotml motiou bas tho Jeast possible euergy. AtmioHy with thoorioff
Gnerai oqu~iou for
of heat nud ctuctricity.
EquatioM of pressure.
sonorous inotiou. Motion m ouo dimension. I'ositi\-u aud uc~tiYc proHargrcssivo wt~vea. Rolatiou bobveou yclocity aud coudeiMatiou.
monie type. Euorgy propat~tod.
Haf tho ouo~y is potoutia), aud
htdf Junotio. Nowtou'H ealcutatiou of yclucity of Hound. Laplaec~ corroction. Expression of volucity iu tutius of rutio of spcifie hcats.
trou)
Experimout of CIcmout and Dosormcs. liaukine's ualeulation
Stokos' invusti~tio)).
Joute')) equivalout.
Possible cilcct of radiation.
of Itcat ha.s
Rapid HtifHng of tho sound. It appears that communiHation
no Beusibk oSect iu practico. Voloeity dopondunt upon toupcraturo.
Exact
Variation of pitchoforgan-pipos.
VclueityofMuudiuwatb)-.
diCcreutial cquatiou for pluuo WttVM. Aiq)iicatioii to wayes of tlicory
of Htcady motion. Ou]y 01 ouo supposition ns to tho law conucctuiH
without tho assi.stpresHuro aud donsity Otn n. wavomaintain ils fo)-!u
type. Puisson's
ttncsof nu improsscd force. ExpiauatiottofchanHoof
Relation botwoou Yotoeity and condensation in a pro~'ssivo
cquation.
Harn.
wavo of tinito ampUtndo. DifKcu]ty of ultiniatu disccntinuit.y.
Limitud initial dist.urbanc<j.
shaw's intL~rats. Rx!Uiann'Niuvestit;atiun.
Expcrimontal deturmixationH of thu -v'dooity of sound.

t'AOU
1

Y)

CONTENTS.

CIIAPTER XII.
r.\<in
-H.f.

2552Rf)
Vibratiousintnbcf). Goocrnifonn for Hinipbharmo~io type. Nodoannd
loops. Condition for (in opcn ond. lu Ktfttionnryvibrations tboro nn)st
ho nodHnt intct')t)s of
Rottcctionof putsCHnt doned ;md opou cu()s.
rrob]f!)ui]icnmpouudvibMttons. VibratiouiufttuboduotucxtoronI
sourcfs. H(~hc))dsf)poi). ]'rf'K''B3i\'o~vod))todiHt))rh(H)ccntn)'t!)t
end. ~fotio)iorif.;i"~ti))Hinthotnboit.sLdf. yot'codTibrfttiuno! pistou.
Kuudt.'ticxpfriment.). Stxnn~u'ycij'fHuits. Vibrations ofthoeoJumu
of air in an o]'~n)]-pi))o.Ru)atiou of lon~tb of \vn.o to lougth of pipo.
Ovcrtoncs. Froquoncy of an organ.pipo d(.'pci)dH upont)to ~s. Comparifiou of YcloeitiHof Houud iu Vfn'iounRttsoH. Exnminntton of
vi)))'nti))Roolumn of nir by motubrnno and mutd. Dy Kiini~s i))nucfi.
Cm'Ttidpipes. Ih'nnehud pipes. CouditiottH to ho Kntisnod at titt)
junctioi].ofco!)noct<id]iipc!H. Yftn~bIotiGct.ic'n. Approxinmtoc~cutatiottofpitch for pipes ofvfn'inMoMctio)]. I])))nc)tooofYaritttio]Jtof
soctiouon prof.ft'e.ssivo
wavoH. Varitttiou of denffity.

CHAPTER XIII.
65

~2<ji7
chmubo'. Cuhicmbox. Pr.sf~tnnt'oof
Acrin) vibration.') in nrectans"
rooma. liectan~ufar t))bo. CouipoHttionof two e<pt)dtrainH of wnvoM.
I!(;f!octio)i)'y)i)'i~;idp!ftnWft)I.Ct'con'HinvestigatiotiofrcOccUonfmd
Mfrnction of pfnno W)t\'Mnt n [duno surface. Law <~f!'inM. Casuuf ttir
nnd watt;)-. Hot]t ixcdin ~scuus. J''rL'fi))<!l's
cxprcsHion. J!('f!cuti(~nat
KU)'fneoofnir)mdhydrof{t'n. Honeftion fj'omwn.nunir.
Ty))d)L)I't)
oxpct'itHcnts. Total t'(!i)uctiou. RnUcctmn from a pln.to of ~luito
ihiHkucss.
U1IAPTER

XtV.

273SD.~
Aj'Liti'dryinitial disturbnneo in an uulimited .at.)uobpLL')'o.J'oiHKon'saolution. Verification. Limitod initiul disturbanco. Cnf<oof two dimeu.
aiotts. Doductiou of Bnhttion for n. disturbauco contiiiiially ronwcd.
Sources of Hound. Hnrtttojuotype. Vorifiottinj~ of soJutiou. Sources
diiitriLutcdovcrn.Rurfaeo.
Shoot ofdottbio
hifi!)itp)n)icw)(l).
Hourcef). Wn.vcsin threo dimnjtax~t' symjuttrieni about n point. Harmonie ty]ic. A coudfnsf'd or mre'ffcdwnvo ctuntot exist niono. Cot)Inititdcirc~jnstn))cc.s. ydoeity-potentinlofa. a
tiuuitytbro))(;hpo)o.
givou h'ourco. CtHeubttion of eno~y Mnittcd. SpcaMnt? trompet.
Theory of conicnl tubes. Position of nodcH. Coinpositiou of vibrations
Interfrence ofsoundufrom
fronitwosiuJpiofiurc'csofliiiopitL'b.
cjcetriodiy jmtiutftinc'd tunin~ fot'hH. l'oints of (iikueo. Existoicu
ottu!) to be infcrrcd from eoutiidcrutious of hy)!)mctry. CuBeof bd).

85

CONTENTS.

vu

Exprimental motliods. Mayor'a oxporimont. Sound shadowa.

Aperturo
:np!)-Gcn.
Huy~htinri'~o~rs. Ccner.T.:f-.).ud~h~(dow!
Obliqua Rcroon. Conditions of approximatoly compictorofloction.
DivorginR Wavos. 'Variation of intcnsity. Foei. RoOoction from
cnrvod snrfacoH. Elliptical nnd parabolic reflectors. Pormat's
principto. Whisporing e~ories. Obsorvationa inStranl'aenthodraL
ProbaMo oxplnnntion. Rosonaneo in buildings,
Atmosphorio rfraction of H0)md. Convoctivotiquitibrium of
tcmporn.turo. Diilorectial
to
of
oqnntion
path
ray. Rofractiou of sound by wind. Stokos'
cxpiMmtion. Lawof rcfra.ctio)]. Total reflection from wind ocrhenc!.
In tho case of rofraotion by wind tito eonrHoof n. Sound
ray ia uot
rovorsiblo. Observations by RoyjMids.
Ty~Il's observations on fog
Law
of
fiRnaIs.
diverROt~oof Mund. SpoakinRtmmpot. Diffrnotion
of sonnd through a small nporturo in nn inCuito screea.
Extension of
Grcon'a thoorom to vcIocity-potoutiKis. Hoimboitz'a thcorom of reci.
procity. AppHcntion to double Bonrcoo. VnnatioB of total onorgy
wititiu ft doscd spaco.

rAa~

CHAPTER
XV.
~C-302
.13,
Sceondurywnvosduc to a VM-intionin tho modium. Botativo importanco of
Mcondury wnvoa dcpondHnpnn tho w~n-Iength. A rgion of a]terod
comprosaibUityacts lilo [t simple sourco, 0.rogion of n.Jtoroddonsity liko
futoubio sourco. Law of inverse fourth powers infcrred
by method of
dimonsions. Exp])umtion of harmonie cchos. Altration of cimmotor
of compound sound.
Scoondfu-ysourcos duo to excessivo nmpHtudo.
Alteration of pitch hy ro]ntivo motion of source M)d
rcipient. Expori.
mental IHnstrationHof Dopp)er'H
principto. Motioti of a simple source.
Vibrations in a rectanguiar chambor dn to intcmal sources.
Simpio
source situatud in au unHmitcd iubo.
Enorgy ornitted. Comparison
with conicnl tubo. Further discussion of tho motion. Calcuia.tion
of
tho raction of tho air on a
vibrating cireular plate, whoBopiano is corn.
ploted by n. fixed Bango. Equation of motion for tho plato. Caso of
coincideucoof uatural and foi-eodperiods.

CHAPTER XVI.
303322.

.1UU

Thcory of rosouatora. RoMnator composod of n piston and air rGscrvoir

rotontml onergy of compression. Poriodiotimo. In a
largo dMB of air
rosonatorHth compression is
sensiblyuniform thronghout tho rservoir,
nnd tho kmetio onergy if.senaibty eonfinedto tho
noiHhbourhoodof tho
air pasMRca. Expression of kinetic
one~y of motion through pasmgos
in tcrrns of cloctrieal
eonductivity. C~)cu)ationof nnturnl pitch. Case
of sovoral eliannols. Snperior and inferior
Jimits to conductivity of
channols. Simple aperturos. Eltiptic
apertnro. Comp.LriHon
witli cir.
cular aportnro of cqual nroa. In
many caseBa catculatiou bascd on arca

CONTENTS.

oniyi.ss.t.ciuut.
S.tponorandmfonor!im:t.stothuudnctivityof
noehs. Correction to
)un~t)t of passage ou acconnt. of opon end. Con~<'t'y~~M~nu..tcdi)yhL.ar:ycyiiih!rIcaItim-faoosofr(..Yo!ution.
Co.npar.son of <-a)cn)atcd and observer pitch.
M..)tiplo resonaneo.
Oatc~at.on of periods foi- doubto rcsonator.
Communication of eno~y
to cxtornai atmospi.crc.
Hato of dissipation.
N.uucrical exam~u.
Lorced vibrat.ons duo tu an cxtornal source.
Hulmitoitz's
th~ory of
opcn pipes. Con-ectioutoJcngth.
Hateof dissipation.
Inf)uoucoof
HanHO. Experitnontal mott.od.s of dctcrmininR tlio
pitch of resoiintors.
DMcuMton of motion oriHiuntioH within au
op<u pipo. Motion duo to
oxtcrual Boureos. Effoct of cn]a~omcut at a closod
end. Absorption of
Sound byresonators.
Qnmcho'. tnbM. Opomtionofaro.souatoreto.so
to a sourco of sound. Rcitiforeomeut of sound
),y re.4oi-intors. Idea)
resonator.
Oporatiou of a rosonator eJoao to a double source. Savart'H
Two or more rcsona.toM. Qu~tiou of
oxporimcnt.
formation of jets
uunug souorona motion.

UHAPTER XVU.
323-~5
Ai)plieatious of Lapiacu'H functiuns to ncuusticat
prohiems. Cunc.ral .sohition
tho
terni
of
tho
mYotv.ug
order. Expre.s.-jioa for mdLd
Di.
vclocity.
vergent wavcs, Ori~iu at aspJ.L.rieal Hurfaco. TL formation of Honoronti
waYCHrcquu-cs in ~nM-al a eortaiu arca of
movinH Hurfaco; othcrwiso tho
mcchauictd couditions uro HatiKti<dly a )ucat tmnsfercucu of air
withont
appMcinh)ocondc))H,ttiou or rarfaction.
StokcH'discussion ofthoeffect
of JatoruI motion.
Lo.siio'saxpc.rimf.nt. C.den)ntiou of numcrica] rosult~.
.Lho tc.rm of zcro ord~.r is i.sua[Iy dutieicut w~n i].a
sound i-iHinatoa in
tho vibration of a 8olid
)jody. licaction of thc snrroundinH air on a
Dg.dv.brnt.nHSphoro.
Incroascofotcctivoiuurtia.
W).cnth~p)tcru
'asma)Imcomparisou~iththowaYo-Ienf;th.t))croiabutlittlL.connuu.
nicntion of euorHy. Vibration of an
eHi]Moid. MuMipIo Hourc~. In
cases of symmotry Laplaco's f.mctions reduec to
L~endrG's functions
Caleulat.ou of tho encr~ M.uttcd from a
vibmtinn sphcrical surface
Caso whou tho disturbaneo ia limited to a f,)naH
part of tho 8phcric.U
Burf~o. Numorical rcsu)tH. Effcct of a sma))
f,p]~.ro Mtuatcd doso to a
Bom'co of sound.
Auatyiical tranMformati.
Caso of coutinuity
throngit polo. Aualyiieal MpressiouH for tl.o
-~loeity.potcntial
Expression in torms of Bes~I'H fnnctious of fractionai ordM-.
Particntar
casos. Vibrations of f~s confinf.d within a
rin:d tipherical envciopo
I:ad)a)vi),rnL<i(.s.
Diauh:tral vibrations.
Vibrations (..xpresscdby
a
Lapiaeo's fonction of tho Hccond ordM'. Ta))]o of wavc.i~~th.s.
li~ativo
pitch of varions toiles. Gnral motion oxpressiblu
by Himpio vibrations
Case ofuniform
initia) ve)ocity. Vibrations ofKasiNchtdodbft.wocu
eoncottricsphurica)
surfaces. Spitcricatsitcctof~as.
Investigation of
tho dtsturbauce prodncod whe)i
].)anc wnvcs of sound inipinge npon a
spiiorical oi)stactc. Expansion of tho vetocity-potentiat of p)ano waves.
Splicro tixcd aud ri{;id. Intensity of seeoudary wavcs.
rriuuu-y wavcs
in
a
sourco
at
a
originatinn
nnito distance. Symmetriod oxprossiou
for socondtu'y wavcs. Case of a f~seons obstacle.
E<iual conipreosi.
bitities.

PAOE
PAULP

CONTENTS.

Ix

C'HAPT.-t XVin.
?'~m
?~3

33G343.
l', Expansion cfYtdoeity-potcntifd in
rroUcmofn.Rpbt'rieatiaycroftdt'.
ExFouricr'Hsorif.'s. Din'orcntm.t quation Bfttisficdbyc.ichtorm.
Solution fur thoc~stiof symmetry.
prcHscdintormsof~n.ndof~.
Conditious to ho Rn.tisdcdwhon tho ples arc uot sources. Rduction
tot'egcndro'Hfnnetions. Conjugntopropcrty. Transition fromnphnricnl to p)n.nn hyor. Densel'Hfunctionofzoroordcr.
Sphorioal
tayor boundcd by pnrallola of Itttitndo. Solution for sphorictd layor
bonuded by smnHcirclo. l'nrticular enscs HoluhJoLy Lcgondro'H funetiouH. Conomi prohicm for unsynnDctricat motion. Transition to
two dimcnfiiona. Comptoto sointinu for ontirc Hphoro in tcrma of
Lftpiftfo'sfunctiins. Expansion of fin nrhitrfn'y function. Fonnuin.
of drivation. Corrospott<lin(;formula in Dcssol'a fnnotions for two
dimensions. Indupcndcnt invcstit;n.tiouof pifmo proLJom. Tranavorsc
vibrations in a cylindricat nvbpo. Cn.8of uniform initia] volocity.
Soctor hounded by mdini w~ts. Application to watc-rw~vcn. Vibntn. circnl~u'eylindur witi) piMio
0
tionn,not ncpcss)u'i)ytraMs\'t~'Hc',witi)in
end~. Cn)np)otoHointion of diff'it'cntial ('(jn~ti~u without restriction
n.Hton.bsoncnnf potarRourco. l''uriunIn.KfdcriYn.tio)). Expression of
vulooity-potentift)by df'scondingffcnn-convorHcntxericH. Cnso of pure').
divorcent wf~vo. Stokcs' n.pp)icationto \i))rntinR Htrinsa. Importn.ueo
of Boundin~-bon.rds.l'rovcntion of latral motion, Volocit-y-potcntitti
rrobloui of
of n lincn.r Houreo. Siguificanef! of l'etardfttiou of
)))M)wavoa impinf.;i))f;upon n. cy)indrica.Iobstacle. Fixcd. nnd ri(;id
nylindor. I\rftthcnmtic[~!yannicH" probicm rolnting to tho trftnavMc
vibrations of nn dMtic soHJ. Application to thoory of light. Tyndfttl's
oxperimonta shewin~ tho sm~IIncHsof tho ohstmction to pound nitorcd
t'y hbricH. whnsoporps f~rr'0))cn.

XIX.
CIIAPTER
280
3443~8
Fluid Friction. Kfituru of viscocity. Cocn'tcientof viscocity. ludcpendont
of tho density of tho gff. ~raxwell'a oxpcrimonts. Cotnpn.rifionof
quations of vincous motion with thoso fipplicnbto to tin cin.stio fiolid.
Assnmption thut ft motiun of ~niform dihtatiou or contrfteHon ianot
opposed hy viacons force. hitoTtcs'
expression for dissipation fonction.
Appheation to theory of ]~ut wnYos. Craduni dceay of harmonie
wavesmaintainctl fit thn ori~i)). To n. first approximation thovolocity
of propagation is nnn.u'c'ctedt'y viscosity. Kumcricn.~cnleuhttion of
coctciontof decity. Tbo ciTeotof viscosity nt rttmosphcric prcHsuroia
scusitilofor vory hiRh notes only. A hiss boooinoa inaudibto nt n,moderato distunec fron) ils h'curcc. lu rnroncd n.ir tho h'ect of viscosity ia
muo!) ineroftscd. Transvorso 'vibrations duo to 'viscosity. Application
to caictunto cffcctHof viscosity on Yibmtious iu nfurow tubea. Holmholtii's nnd Kirehhon''s rcaults. Obson'ntions of Schuoebeli nnd Sccbcck.
Principio of dynamien.1simiifn'ity. Thcory of shipa fmd rnodets. Application of prineipio of Eimi]Mityto dnstic plates.

CONTENTS.
APPENDIX A.
Correction to Opcn
End
Noteto 273
Note on ProgrcssiYc
Wavcs

l'AOP,
9~

~'7

CHARTER XI.
Ai~lUAL VIBRATIONS.
23(!. StNCH t))c atrnosp))erc is thc abnost miivc'rsa] vehicic of
Sound, tbc i)t\'cstig'atio)toft.bc
vibrations of a gascm).s mcdium
bas alwuys boc'n con.sidcrcd tbc pcculifu- problon of
Physic:~
A<'o))st,ics; Lut m ni), (.'xcL'pt :). fcw .sp~-i:dly simple qucstiott.s,
cit)('f!y n'Lt.ing tu Lhc pt-(~):)g:Ltin)tof sumul iu ouc dilnbnstO)), t!t0
]!t:Lt.hc)n:tt,tc.ddinicuitie.s ;u'u such that. pro~rc's.s bas bL'cn
vcrystow.
.f~vu!)when a Utcm'utical rc.sult is oLtuinc~, iL ofto)
])appcns that
]t cannut bc
submit.(('d(.t)t)tctcsbofuxpcrimc))t,indcf:Utkcf
;u;cun).~ nic't.)hj<).sofmL';tMuri)~ thcintcnsityof
vibrations. Iti
woctm dois tu suive thuso
.tncj~rts
oi'thcMubjuctnHUt!~
~'robh'm.s \vh~Memathon~ticu) conditions :n'L!.suf)iui(.-)iLty
sitn)))L-to
:'d))utof solution, :).))([to trust t" thon
a.n(.lt()~cn).r:t.tp)'i)tei))t<js
""t to fcavc u.s (juitci in titc d:n'k wit)t respect to ot)tur
(~ucsiiousm
~Licit wu n~y bu ititurcstcd.
Ja thc prcs~nt c!)aptcr wc shiU!rc'g-:u-d f)ui(!.sns
prfL'ct, H)at is
to say, wc sh~![ aHsntnc th:tt t!)C mutu:U action bct\VL'cti
any two
au
port)un.ss(jpamtu()by
u)<).)surfa!s);o~i~o~Hi'ce.
itcrc:Lftcr wc Hha)t say souicthing' about Unid
frietio)i; but, in
~ncra], acoustica) pl)cn(nn<-))a :u'c not mat<t-ia)]y di~turbct! hy
.('h (]uviation from
pcripct ftoidity as cxists ))i tbc' case of air
a)xtot!)crg:tscs.
Thc cqu;)j:ty of prcsfini'L: in a!) diructtons about a
givcn point
is !), ])'jccMsary
cuns(.'<ptcncc of permet Ouidity, ~hutbcr thcrc bc
rcstor]Moho)t,a.sisprovc<}byconside)'in~t)ic<(pu)ibi-iumcfa
smal) tctrahcdron uudct- tbc
opf.'ration uf tbc fbm) pressures, t])c
J:.Jf.

0
EQUATJOX.S 0F

~r~
P~un.
thcir

'r
w I

FLUJD

L-

"y

O. <)

~au.

'f<L.

~o~)Lht
Oc dUIIOLud
Lcdc~.U~, by l,,

JriUIll is

'h.

r.int.

..in

lil/'('{'s

~<
~nc.nt

,~r,

~=~(.Y<
clmtya
~t'i'

t~,

-1"

'r

'r

~o-f-fux.

,).

,~j,,

r~+~)

"<c.
cl, cly, cl.= iu tlte co-ordill:tlL's
uf tlm ]lUill!, ut whielt l:ltc

;idl'l'illg 1)1{'cl/ilililrillll of a small

cylill(1c-1'",illt i~;tt L'llds, tllU
(ltmsuvf l'U-Ol'Jilla(l:S
lI'c~
cl,c,clJ, cl.r. Tu uhlaill tlm
1'1'l'il'celi\'ldy
l'(I"atillll~ u1' IlIO/iulI\e IaVU,ilt
:,U:<:lIl"dalleu
~'t). D'AJu.nLerL-.s
Priucipf., ,~r,)y
Ac. by Yr.pi~.c
-L~C
t
)
~'crc
~.d~th.~c.-atiuuscni~~idc
.f~uideun~rcd.
Thus

-p~u.,
,,j t(..nn.s of~.
f.
})fll'ticll', ",ltidl(,l~1' it muy 1..Jl"tli:vt aL
tlie t.illlu t is fouud a tlm
puiut rc,,1/,
1l.f~ura HIJlflll illtcl'd uf
t.
~r'

~s,

un,

))u).f ~,t.

.f.bvc)~ity~.U~<.n,
thu otlicr hand
eXIJI'usscsthe

(i,,t ,j,~
chauge in Il tlw \'I,luei ty of tin;
umjimcl

which
is uut 1ixe itl
EH
L~u jM&p~M,buti.iutcs
slrtcc, Lat IIluVOSwit.h
tll() ilnicl.
w.t).t).<;))tn,j.
..ot..U.
~,dt.i.
;hcd.a,

Tu
'Jutf.is
tjtis

237.]

EQUATION0F CONTINUITY.

position in spacc (dut!'min(jdby thc vn-incsof .T,y, z) is rctn-incd

]).
1 11
l f
1. 1
1.] ]
inv.u'i:).biL!,w)t!i<ji)i
itIs:tCL')'t:dnpnrtic!eofthci)ui(lo)iw))Ich
intention i.sfixcd. T)njru)!t.t,io)itbotwucntlK't\okIndsoi'dn't'ui'uttLiatiunwi()t :'(-'s{)ccttf titnu is uxpru.s.su~)
by

;H)())nu.stbt~')e:u'!yc")iCL'i\'cd,t.)")nn']ti~!).I:).)'~cctas.s<)fi)t)porta))t
pru))!c)))swith \\)ti<j)t wcs1tanhuucc'upic(tint.huM(.)tn.'I,t!K!()i.
tiucUun pt'acticaDy di.-iappL':Lt'.s.~Yhcnu\'cr th tmjtiott. i.s vcry
sm:).]!, tlic tut'ni.s ,-(f.~ ~c.d)'Ish
7)
utUmatcly
~=~.

in !'L']..ttiv<.i importance,
n.nd

2~S. Wc havoi'urt,))cr to expruss tt)C condition that therc is

n()Ct'catiunnt-:umihi):).t.i))n<jf]n:)tturittthcintL!riu!'<)ft'!tu<!uI(L
(tt'a sni:dl r~ct:Ui~'))!:u' para!)u!cpipc;d
H'K, /3,'y bcthuud~~
{'.u'idiL'].(.<)th('axt'S(~'c<)-urdinat.(;.s,t)tu<(U:m<.it.yf]n!ttturw))ic)t
p!is.s(.s (~)tuf t,))uInc)udud sp:L<t.'in thne (~ iii exce.s.sof ttmt whidi
f'))t~r'<i'<

thu sn-c:)Hc<)uqu~tion of continuity. WLcn /3 is constiint (wit))

n's])C(;Ltu butti timu and .sp~'t'), t!)< ('((H:)t.io)iansumus titc simple
forui

lu prohicms conncctcd with suund, th vclocitics and the \'fu'iaPutting

tionuf(1cMsityarcusua]!ytr(.LtL!d:)LHsm:t.![qu:mtitic.s.
i: small, and ncg~ctcailcd ttto co~M~'o~,
p =~ (1 +A'), whurc
1
wefind
we
iug 1 1
i!igt)ieproduct!M-y-,Ac.,
zi ~L .Cc.,

.ST!!t.:A~[-r)rx<'Trnv

~s

~ci.)

r~nnc>

~t~u,

<).

fil
~Ln

..s c.t.iy

p~anututhcphmGuf~

~1''
fll'lJitl'rtl'Y. TIlt! 1'nuctiinl
~i:l.r~
't = ('UlIstnIJt,

ti)

11'

"Lid,f.
is call1,t! the

.strcrmi-l'mnctiml, HiIJl'<:
tlll!
cnrves
~1'Imn tlm mnticul is
Htl'ady, tlmt is, Hlways t,lo;
"~<h~
HII1':llnl5~tic~ully,llre Huh,tit.lltioll of ouc f'm~c.tir~n
t

Amutllur ('rUi!!of 111111()l'ttlll('

iv \ll'lI tlmr~: i,~
Hj'llIlnctl')' J'olilld
tllat nf ,(',
li:l'I'ythjll, is tlH~n('xllJ'l)ssiJ,!oil
tlm 1I)tioIl lnkt.'s
il pl:IIIl'S
p!al'
pnsHillg tliruylr tlm uxis
of'SYIIlIIIl'tI'Y. II' llm VI.h:iti"
<" m\
)'!));t.i)d;t))().)(.rp(;)), r)~rf,,f)n.v;.
<'
110
y
Sj'IlIIIIC'tI',)'
i,
~rcun.'i.ity

In ?)h])n.sf,a])
(it~c.
rr

,<)

i-

we s]:dl ll:we fu

in \'iltlll":

'f'))U))mh..n.r
ml,r-1-mlr~_l_.i~,rh III' al. IIIOIW/III"/'
u pel'fect
ditr(~l'eIJtiaI clc~, it, will rurrt:lin
so tur ull
Sldsl.'III/I'Il/'
alld Le tlmn

LAfiHANCE'~
)s THEORE~r.

:23f).]
Hf't in motion

Ly cnn.scrvntive

furcc.s axd

pressures

transnuLt.cd

trL))ntj]n.:c.\t.c't'i<'r,t)K!~)t:Lnt.ittL's

(whiuh Aves)):t]Ldnote, h y

(bpiu-t i'ron zero.

~) c:u) MLiVO'
writc fur

Wt.;ass)nu<jth:Lt.pi.s!).iunctio!t()rF~,andwcH))!)))
hrcvitvy

~.)'h('t.'<[Uatit))tS(')'in"tit'n<)ht:uuedf)'uni(I),(2),~37,tn'e

\nt.h t\vo ot.hcr.sof

tlie

f~rxt r~f.t.tmg to

y:md

.?.

Hy

ttyputhc.si.s,
</Y~r
~y"
s~ th:tt ))y(!i(1'c)'(!)ni~tm~' thf <u's),uft!)C:~)0\'c cqu~tit~ts wit.h
.subtt'acting,
respect tu 7/:UHtL))L'St'r(m() wit,L respect U').nd
i.c)i)nn):)t(! cj !)))'! t!ic hoprcssc'd i')i'ccs,(jb).:Li)UtJ~u'~t:(.tiuus
\\)Hc)tju:).y).)CputI))t.uth('r<))'in

s:uiiu furni
\Ytt)i
1 t\vo
two oLhcrsoi'
ullici;s1thc s,wu;
J
furm givin~
~iviy 1~
-L/C JL/C
Jnthuca~

of.'u) incotnp)'cs.sib!cfh)id,wcmaysuL.st.itut(jfnr
<
</ f/o
i,
,!m(tthn.st)t:)n)
~' + < ](..spqmv!).iL't)t <

which ?u'c thc cquatiott.susft! hyH~hnhoItxtLstitcfunndation

uN)i.stt)C<)rL'msn'sj)t;t't.m~'v'))'tiC(.s.
1)-' thc motinn bc continuons, thc cocfMcicntsof ~,?;, ~in
t))C !~)()vnc())t!'Ltio)).s!u'('!)]!)iuitc.
LctZ'1u!tut(;thuir~)'c:LtL'sL
]nt)nG)'ic:dY!t)uL',f).))ttntLL'HU)uof th nuiUL'ric:d vaincs (~
By i~yputhc.sis,Hisiniti:diyzcro;
tItC<)Uc.sLi(;a is w!)L'thcL'in

LACRAXOE'STHEOREM.
<~ c.
~~y~

,,f time it can b,

is

Tho,

'r'

r~Sf).
c.

"' ~i~

'TT
..c~
~-~"
"~
tlie sulution of the
c(I1l:1tion

its

~l ,~irrlimv in t1m :u't.n;tl

casc~, n CHIlllot dep;ut fl'lll
Zeru,
,t h.t.),.)
;t!e'U!nse.j.si.j.)
fnrros al'!illg mn e;mlt
lr.lrtinle hro}lurtioll:d tu ils veloeit)', ns
IH,
sel'II }I)'
Illn.)'
sltbstitlltillg' .l-rcn
1'-K i'~ J IC
K 10,
f"r .1~,.).; oh III
ac~, ur
'1' it i;i. J H!rWISU\VIt
ir1(~>j'.
:').
11
ilie,
cxist
'lit
II.
in IJllids,
iiiit
mul vrc clcllcndellt o0 the oclolinc \Iocities uf their
parts.
~H~
th,tt
to
.y
t),t;,r ,)i~rcntMt

~u,c).y;[,ut

~cw
'y
,r,
is
.<)<c.tdw.d.~v<sdMt.,S..
~f.t~mc,t
sllcw that il-, aud
"en t.,
.E,

{'v,u,

is v point tlmt is
.v~
.n
tll;tt
~.r~r'
s
,r; Yalljslll.'s
Y.
ia tllc

w,

lcticm
it

'?
limit, nnt to

to P.~
tlo: first orcler, Lut
~t

~r~
'<)on<t).c
c,
body. ~t'.ocf,tio,,
~"<''~n1>ccn 8CF..s,~)aH t),c
Iml
the ~m"n
ditli:1'vlltial cocflicimts of s witll
l'o.;pect
w'illt if' (li(l
and
HO,
tllcn il, miglt hc infClTcd
that
s cOldeIHOn'1'
1\'gitilllateJy
var)' t'1'01l1zc:ro.
By Htlll!orel!l rIl/(' to 8tll/('S,
tllc II1f~111C11t5
of 1iIOIIIcntuIll:t)mat
.~F:
jllfillitesill1aJ sj:,lmric:tl
nf' Ilui~i
hurtiutl
tu
0(til~ti
i;, ~7, r, lI1uJtidil'd hy tllc
II1t)J}W/ltof
"L~='s.r:
tlie torms11t'pl'IJ']"llt
nf'gl"C:1ing
nn iuertin, \0
olitill'il
14j11:Iti~llH
to
111~1~IiC7ll~IU
tIlL.
11)(-tioli
of L'JL'ell'il'ity
=~
;olldlldol'R.
tlwuuglr111Jiflll'Ill
-(~?;7/7']
un. p.(,7.
~.A.I~r,nir~mie.)7.

HOTATOtTY VELOCITIES.

230.] -1

ns HK' c~mpc'ncut rotatory vclocit.tcs of thf: HuIJ n.t th point to

wijichtheyrufcr.
vanish thrnu~hont, a. spn.cc occupicd ~y moving
If
<)ui() :n~y.s)n:dlsphcrica) portion ofthcttuid if suddcn!ys<jitdiflc(l
A prdf of this
motion of trimstation.
wonM rct:i.in on)y
he givcna, littic iatcr.
pn'p~.sitiuu in r).)K'r:L)iscd iona wi)l
i,:)~r:u)"c;'stt)Cormn thus <'onsi.st.sin th assertion t)):ttp:),rti.c!cs
ut' tinid ut :my ti'nc dt'.stit.utcuf rotation CiUincvur ac~uirc it.
A st)nn;]):tt()it't'(.-rcttt )not)ciof'in\'c;st.ig'!t(ion]):tshuen
:i hi~-)t)y uistructivo vicw
:nL.ptcd hy T!)o)nsu)t, whic)i at'tut'ds
t~'t)nj\v)H)t('s))'!)jt'ct'.
2K).

J!yt)tui\un]:m)Lcut:de'p)atiuus

IntoH'n'tIn~
wiLh

th

Unit),

t,his

c'juidiu'i

~!ong

!ny

fmitc

'I1

!U'c

7\

moving'

wch.tvc

in which sufixcs (~nntc ti'o vah.s of thc bmckctcd function

:)); t1.<;puiats
r~spucti\'c)y. If th arc bu a. complte
~nd
ch'cuit,

CIHC'ULATfO~.

Lr.Q

L")))WO)'(!s,
~7<f

/~)~/t;r,/

,')

nernuins

tinte.

~r~~y;tpt!
con,tunt

;Ll(Ia;ci~

7'UrlIrtG
nll

tlrrurr~hout

is
nppmpl'iatt'Ij' ca.l/l'i! t,llc cinculcction, :ual the prllposill'"
IlIn,)')lu, st.utucl
-morrinr~vcritlr thc ,llrticl ~'e-

'1/1((
II8cumv'tmot.
L.

t).

as~.f~.

'.y~s

").
c.v.<,n.j.j,
.s~ ~hy
.c.
''caco)np)..<G(hn'<rc)!<i;,).
~td.n
c.u

c.
I.h.

Lu
ail)'
'y-t-

.),~
in..j~i.)

~c.r..u~i.n,

u-.)~iU..h~~t,

~o~cn.sc a)) U.~can

..si.
a ,,f,i,

h.

is
tu

vt.uut
<

p..ss.<.

n
n~utu.,fy.cii~
to hc
r~i~jc

..I..n

I.nf.

tr.iU~

uf.st~.r.o.c.pi..)

L,

Ly In.o~tionaHv
i:
;lru s;licl
f.n<h..t

tiunally
~=;
1I1ovillg'flnicl.
'\Vit).in.n ova)sp.cc..s,.d.s<hat
~)
i~)~h.)byanc.))msoi<)
c.rc.n.s
,-cc.n<.it.)c.i
U.f.
if .f
ul' j~
c. bc
,drc..).tir.Li
~vc.n..t.uiu,.diy.U.cre
c](I~(~II
C\\l'I'U (1rawlI wiillill it.
HlIC:1tspa,('cs ~re C:LJIed
simplysurliwu d' :1I1;tnclmr
~r~j~r~
rils" a clmmi <!1Jl'ugoillg rmmci tltc
ring is
reclucille tu a point, and tIlcl'cfol'o
t}I()!'ma)' lie
t'n-cn]:)t).)n n!ot)< t cvoi
.i)t~
v.),c
.dn~
jBut Un- c.ircH).n
.tth~vh.!c
everyc~)
c.rve
round thcrin. ' an.
~).p~
~o
'c.~nccun.s~t~iuu~a])<)~<.).at..Io.
24L When
(,
~+~s~c~ct.h-n<~
in a.y .h..c.ctl.n
is .pr~d
hy th c..n.)i
~cL.~y

.)~h i.sc.)M
cha.f~
L<)
~.vd.ci
y-p~f,

~]

YHLOCrrY-rOTEXTIAL.

tlie
It\S''k')t<)tcn.nyc]<.sc(Ls)))'fnc~,t))-at.co('Howf'utw~.sMros.s
f~S', whcrc is thc r~ of va.
ute.nont <? Is expresse by
~nti~ocu.st-ut'
ti<)n()f~inr'c'it)~<'ntwn.]-(1s:).]'gtLun")-n):Ll.
constant d~)).sity,t))ut.jt:d!nss ut'Omdiutttuc~i.sthos

th<! intcgmtion ran~ing' ovcr thu w)to)c surr;)cc of 6. If thn spf~

~'c cudDi'thctitnc~,
<S'hcfn~ buth :ttthcb'~i"nu'g:L
t)K')~ss]nustv:u)ish;!m(mmH

co-ordmatcfi, thc trans\tthp"UfunnL-.lc.))~ti..)) is ~~crcadi)yf.Lt~)t.~<nrucUyby:q~yh~(i)

tu t~! C(n-rc.s))un<U)~uk')n<;nt.< \'ohm" titan hy tnmsfurtnm~ (2)
m thu
in a~'onhu~-c wib)) t.)'c :u):ttyLie:Urntcs i~)' uUccting ch~u~'os

-whc'nitiH'l~irud

towurk

in(.k'})(;ndL'ntv:u'):th)L's.
thc pt:u)C A'y, so t)~t
Thus, if wu ttd poiar ct)-"n1inat.s ia

r~OPERTy

op IKp.

r :,2.1L

~1.
~!nltlir, m,~tln!1~Ilnll
file

~L~.r
"<y'.sby(~

'c~t(,u..uf~

fur t!I(~IJ1'llld('11I
'1~
il!

cunvcniunt

Ilancl.

-y~p~
flllid witllin
allY ,~iIlJpl.r-c()lIlll'dl'd ('Iusl'd
1'11'(' 8 i, i~wnlilc~ful,S~
vlutmwnimnl l,y
r:
~y
~n'j.,)),j,
~cri,d
"j.~ ~.s.s,f..
in ,y.yt.
~.i.i,n.)h.
')
it (';IIJ
'<
l't'st,
,.t
.)'i.J.L~
~JUH.
al'llllil'e III) IIIO/I'cillal' l'otatioll
1111l1(,)'
tlH: ulu,rutiun uf
)~
.f
the spaeu
value

;u~~
'b~~thcc.

in 7'Imnu~oynul

.-P
.S ~<~<~M~y
~t
~<Mtf)('<jrcni.)rr7~-n

i,
1'1'u1.JJVIJl
.L'~J:f.?, jlossilJe.

tlie iut(-~t;t,~

j;

iL.r(-h.j.~

't.
JiU <tll'O l'lIlIelillns,
"c.
s~~ry;

'~J~<

..h.r.

J'1l.illg OV('I'IIi(} \'111111110,

,juat.

am!

Ju,

'L~r.
L.

S
1

242.]1

MULTIPLY-CONNECTEDSPACES.

11

over thc s'i)-f:).cc of .S'. Undcr thoso ch-cumst:inccs t)~ douDe

p~ntof~'
:)t")y
~) .~t.ItC.s,:md~uiii'r:):.n.
.n~r.dih
</A~ </A~
wonl.s ~
In oth
~hcrcr wonIs
A~
mast ho u(lnal j,o zc:l'l). ln
<~
Z
;<;
<
tnust hc cunstnnt, and thc twom')tio)'s i<k-nticn.L As npnrIr)-ut:).tiun:d Mnd
no motion oft~c
t.icu):n- <<sc, thd-cc:u)bc
withnt D'u votunu! <S',indupc))')unt)yofa motion of thc surface.
T)tc rcstrictiuu t~ shn]))y con))ect.c(! Hp~CL'sis r~-ndc-t-udncccss:uy
ofUn~ns
t))corcn),w)ucl), !t.swas fit-stpuintud
bythci'.u)))~
uttt hy I!c])uh<dtx, is othot-wisc possi)))u.
J.s nu))t))')y-<n)f'ct<(), thc notn.tiona.t
V~~u t)K: s[):x'c
ntoticit is still (h'tummKttc, if h~ido.s thc nunmd vuhteity n.t.
uf thc'. constanL circvcry p<jint uf 'S thcrc bc ~'ivun titu v:du('s
Fur :).
odidinn.s h) id) thc p~.s.sihh' iiTC'co)it'i):ddc cit'outs.
s
comptctu discussion '.fthis ()m.'sti('))tW(3))U)st,t\'rtt)Thomsun's
ori'dx.'d ~u'u~'i)', :ux) coxtrttt, ourscivcs ht'rc whh thf c:L.scof
adu))h)y-conttL'(-t(.'ds})a('c,w))i''h\i)l.suthc(j!furI]lustr~tiu)t.
Lct J/)'C'D
irrot.LtiutKdty.

I"! :)n(!t)d)cs.s tuho within which i)ui<tmovM

Fur this m'~nj~ t)~i-c nmsL cxLst a Ycitjcity
Fin.

put.uht.t:d,w!n)sc<!ift'un'))t[:tl cocrHcic)tts,cxpt-(;ss)ng-, :Lsthcyd'-),

t)n! cump~m'nt vdocitK'.s, :u-c )K!(;c.s-i:u'i)ysin~lc-value'), but
wL~h t)C('-]nf)t itsc)f]K! siu~ic-Y!~))C(). Thc si)np)cst wayof
is to
mnt-igulLyof~,
:tac).U).n' t))C (linicu)typr(-suntc(thytht'
coh~ivn :L ]):u-ncr~i~ta)<rn:K'r<LSs))mn))~,son.-<toc)nscths
is t)h-)). sin)p)y cnnti)mo"s,
p:~s:).n'c. T)~- sp~c ~)/~7~~tA'
an()<!n'cn'.st))(~)!-c)n:)ppii('sto1t.witI)~utjn("1ific:~tion,if!L))owin tho value cf~
;m~!bu tn:uh; f.))-:t])~s.siL)cfn)it'(tit'turt'nco
t'arricr. Ttti.s~in't.'t-cnc~if iLcxIst, is
~nthctwonidcsofU)~

~Y-Cr,xx,,(.T,

,p,

<0

all<l il tilt'
dYllaIll(';/:1111di(,iltillll('XPI'('S,s(:s
h,Jl'o=~
lllu circolulic,;l l'olilld
i11; l'jll~
''ri''j"i'cc.;t,;ui.,u

<h.

sincc
~o.

())),

0\1'

illn

i11'o

1;11'1'

J\IIIV

;i:r:

Jms tlm -lIlIu \'allll' uu

clc~
the
~sU.).
t\o sid"
thct~.sid.,s.

th.sL.u.L
Lc.u.cuh~i.n

.iiO-.n.~
,)

of

cd'

if rc \'<I/lisll,

T).

-J
cl~e

~Lc~

be giV011,
ifep aml cj~-1-V~, IJU twa 1'mnclimns
1!~ur,
~T:
snlisfJ'illgLnplaee's l'Cjuation
alld the snrnc l1uI'lIlal
~i=~
t iUIl :lIId ilm cOllditio)1
.LaplaCl!'S l'(IIIILtIlaL tl)(,l'u
.t~
s];J/ I.JUIII.itl,CI'
('il'cula/ioll
~s~cr
.S' p
\1:
~<cu)..<i..n
~-d.~n~
~J.
~sin.p.

(at;.nvp,v!
't,.

J~

~s

n,,r hy

~rf~,

circ.).t.i.,
-ya.
IlItlll
il'ly-cIIIJIJ('dl.d asn.s \('/1 "as
I~c..
f.,
'y-<~cd
l;illlp/('IIIII}('ct(~d 1;1':1('(' if v
sd in .ion
hy.
tllc wJJilh!11I:ISS
cumes to l'cst so 80UII vs
tltc IJJttioll ut' thu
'y<.ua.s~.
~L.fthL.Lu~LUIIJJH'Lnm'd
'~vI~~ith~Lf.r.).ti~
rcei~ut
t.,bc-!iko surjhof,nr
~s)y
su~, t).
witlliu
taLe comes in l't'st.
~iD.i. t1IU
tJ.c.tubccn.nc..s~
'1.'llis
L.

.H,, ~out.si.~ olltsidu

bec.
i.t,
<
IlIech:lIlil'al
illtc'j>ctati01I,
t lIuLlcl'sLand)
Ilu!'c

~)~.] J

.\XAL()(:Y

W)'rif

H~AT

AX))

EH':(")'):r('!TV.

13

a(tuit) havit'gnocircuhttion,
axdit
(.'))':n']\'w)'!tt ismoanthy
<'xh.;))M)')norSt.uhcs''h<t~)~
tnxx'ifh~~s <t)t
witht'rspt'<
cuhtr n'tation.
i'or, H ait. thc Hmd ())iovh)j.; suhjt.'ct tu n,
v~h"it.y-p~<)'ti:'t) cnLsidu n.spho'io:)) f':(\'ity(~f :t))y r:n)ms bcc<n))L'M)nh!untyso)itt,t]tL! fhud in.sitk' Un.' c:Lvity cnur<jt:nnno
motion. 0' aswu m:ty :so.st:)hj
it,a!)y sp)K'rica[ portion oF
:t)) i)T')t:)tionn)ty n~vrn~ Ouid ht'cumin~' suti(!t'!)ty .sotid wou)d
))()SSt.'s.so)ttyatUt)ti())i(jft)':m.i1:~it)n,<'i'<ru<<~<'o~
A Mi)t)i!!0'proposition w)))app)yt')!L(.'ircul!))'t1iHC,f))'cy1in(1~r
~it)ti)at.<'utls,i!).t.ho(.)..s(jofih)i<t
]uo\'iu~it'rot.:ttio))!(Hyintwo
(iitm'n.sit'jisonty.
Thu )))')(!')!) ofn.n i)tComp)'('i.si))tL'fhnd\)tic)t)n)s~)CL'nf)UCC!
atrL'.stj):))'tak('.sot'<.lK'TL')n:n'kaL)cp)'()p(')'ty(7'))~')""n<)utot)):t.t
ot':))t Hysk')))s'))ich:n'o
set n) motion \it))j))'('st'ri))c<tvcL)('iti('s,
nauu'ty, that th uncr~'y is <))u icast, possib)c. R'!)ny uthcr
nn'l
motion 1)L' jToposml H:(tistyhi~' thc ('qu:t.ti"nofc(mti;)uity
thchotnx.hu'y cumtitioo.s, itst.'Ot'r~y is ncccssat'ity~rc'ittcr t)):)n
thatofth'jmotio))'\vhichwoutdb~.L;'<)'ci';Ltcdn'o)nrust.
Du; f:).ct that thc irrot:ttiu));~ motion of mcomprt-ssihh'!
2tjt!)ud (tcpcnds upon
V(.-]o'-it.y-potcntir).]Sittisfyi))~ Lap)nc(;'s
foum):<.tion of:). F!U'-i'u:L(.)i))~)))!))o~y butwccn
L'(~n:t.t.io)t,isthc
t)~! motion of'.such~
U~tof
thtid.and
uluctricityorituatni
:).))t))tot')n cum)u('tor,v))ic)[ iti.soft.t'not'~rc.'ttso'viceto
hc:t)'
in mind.L Th~ s:un(i )n:).y he s:ud <jt' thc conncf'ticn bc'twccn
wttich dc'pcH(tmat))(j)n!).tical)yo)i
att. thc b[':UiL'))C.sofDiysics
potuntia), i'"r ItoftL')). happuns tl~t thu :i.n:tto~ous thuorcms
:).ru f:n' from cqnidiy uhviou~. For (.'xampic, thc ~na)ytie:d
t)tcui'cni that, if \7~ = 0,

nvcr a do.SL'd surface. is inost r(.u]i]y sn~'cstcd hy t!)C Ouid

intc]']')'ct!).ti"!), Lut one ohtaitK'd nniy bu intc')'pret(.'d i'ur cIc'cLric
or )n:).nctic iorcos.
Ag'am. in thc thui'ry of t)ic con(h)etion of hcat or cicctricit.y,
it is obvious th:it tlK'rc can hc uost.L'ady motion in tho IntL'rIor
()('fS',withoub tnu)H!n)H~i<'uacross sontLi pin't of thc Luundin~Hurf~cc, but Uns, whuu iuturp)'L-t.(jdfor incompres.'jiD.u itnids, ~ivcs :ni
I)npot'tanta)id raLhci'rcctjttditc t:L\v.
T))~)monen r~e

.Vot;), !or.ctY.

EQUATION
2.). t.
!n')t0t)i(.

W)u..n

0F

v..).)city-p.(.nti..d

prus.sm-ct,):!y

1-KESSUHH.
c.xisLs,

Die

f~
Lcquation
`.

!<ut:)'i)L!L.i

t.,

.X.tcr-

},jj.M

'J~-t'\JJ
~~iJ,

r.

Th.samcc.n.iu.s.un~ybc
.-u.n..d~by~
din.ctappH~tinnuf
.nL.d.mca) prn.p).s Lu the circu.n.s~ucc.s
ofinipul.iv~uotlun.
Jf~=
c({u:Ltton(~) fakcH thc' f(.i-)n

If thc n~t.on hc suc), tt.at thu

cn.nponcnt v~)oci<.i.s arc ~)ways ti.u
.s.un.L ~s.
p.int ..f spac., iL i.s c.))<.d
~eo,~
.)
"n.L.nt.f
d~tun,,
Tj.uc.p..tiuu ufprc.s.s.uci.st).cn

uc.
t.o. ~d)
siit~c
fur

~pncahon,
t.
if

of (2). thc vctoeities and

eondcnsa~j;

part

2.N.)

rf.AKE~VA.VEM.

15

T)K!simph;ntkin() uf'\vfi\'c'-)nn<iut)isth!).tinwhK'ht)!C
(.'xcur.sit)n.s()i'u\'uryp!H't)ck':n'c])ar:dtott~:t))X(.)i)m'n't:)rc())c
tinc. Lut tis t)!L-)'L'f~rc!
s:u)tui)t:t!)])):uu'.spt'rpuHt)icu!;n'tu),)t:)t.
(i)s.s))n)i))L;'t)mt7)'=0) sup])usf t)):tt~i.s:(.funcLiuiiut'~(:md<)
ot)]y. Out'L'<~uatiun(!)):i'l'))L'co));.c'H
~)').

)'t'pn:'sunting thc pn)p~).ti")) oi' mdcpcutk'nL W!(.ca in t,])u positive

:)ndnug:tti\'L;(Ih'L'ut.iu)tSwit.)t<.))L:cu!)Utnjnvci()cityn.
\Vit-hmsu(')))unitsns:dtt)Wt.ht;~p~tic:Lti<)nf)f'th(!n}'p)'uxi)i)a<c
(p<))):)Ltinn
(!),~K;\r)ucity()f.sound
i.St'))tiru)yim1(.'}K'ndc)itot'thc
i'urttt uf thc w:n'c, bcm~, i'or c-x:unp)t. ttiu .samc fui' snnptu WtLves

tl.ie w~vc-k'ngt!)
't'ln; condition
tlu,
whatcvcr tbc
vlic~tcver
b~ Tix;
comlil,i~m
Hati.sficd
s~,tisfic~l
vwve-lcn~tl>>na.y
mnybu.
l~ythc
by
pusiLivcWi~ve,:)n<[LhL'rutbrchy th ixitiai disturbimcc ii' n. posiulunu bc gctict'aLud,is
tive Wt).vc

l'L.\XH I~HOaRKS.SSJV~ WA\)-

~t5.

Whatcvf.r tlie
init.iat.t)Ht.urb.'U)cuniayhc: (axd Mand.sanibuth
arhjt.rary), it c-an niway.s bc <))\'i(!,jd into two parts, .s~LiHfyi)~
n:-s).ucUvdy(:{) :u.(-t.), which arcj.mpagat.udundistm-hc(t.
In
')'(-)t);)..nuntw:tvcthc
(H )~ct.i.)ni'pr{);~a<i~nisd)C~)))C:t.s
thatuftjju!n(~)<.nu)'t)tL-c~/i'p:u'Lsui't)~f)ui().
TiK'r;L<(;atw!tic))L-n('r~yi.str:).ns))HLt.(.acru.s.s()niLof:u'c?),()f
!t)'):nt(-p:t)-ati(.t t,)~),c f't-unt(~).prt~)-CH.sivcwavG))~yhurc~u'dud as t))u)ncdtani(;;dut~surcuft)ici))L(.-n.sit.y(,ft))C
radia.t.iun.
.)))t]itjc;t.sooi'a.si]i)j))c\u'c',i(jr~))ich

Jf t).c int~ratiun wit].

rcHpcct to thnc cxtcnd ov< nny Mumhcr of
comjttutu j.criu(]s, or }.r:Lctica))y~hcncvcr its r:in~c
i.s .softicicuUy
thu
tung,
punudic L'rm.s may hu (ntiitted, and \ve mny tnhc

E~EUGY

~5.]

0F

FLANE

WAVES.

17

or by (H),ifj9 dnote thc maximum value of~

Thus the work consumed in gcnerating wavcs of harmonie type
is t)'f! sa.tncf~iwould bc rcquil'cd to givc the maximum vclocity~3
to t))c whole mass of air through which th wavcs extend 1,
In tenns of the maximum excursion 0:hy (7) and (D)

whcro T-(=\<) is thc pcriodic timc. In a ~)e)~ ~e~tM?~thc

H)t'<i:uncalmcasm'e of tlie intensity is pi\)port'iona.lto tho square
oi'tho amplitude direetly, and to the square of the periodic time
Invcrsc)y. Tiie rcadur, howevGr,must be on Lis guard against
supposing that thc mechanical measure of Intensity ofundulations
of din'ut-cntwave lengtfts is a propcr measure of the Joudness of
the con'cspouding souuds, aa pcrceivcd by th ear.
In any p!:uic progressive wave, wLetIier the type he hn-rmonic
or not, tho whole cncrgy is cqually divided between the potential
and Rinctic fom)s. Purhaps the sunpk-st roa.d to this rcsult is
to consider the formation of positive and negative waves from an
initia! disturbancc, whose energy is wholly potcntial~. Th total
nergies of th two derived progressive waves are evidently equal,
and nmke up together the energy of the original disturbancc.
Moreftvcr,In cach progressive wavc the condensation (or rarei'actiojt)i.s one-))alf of tliat which existed at the corrcsponding
point InitiaUy, so that the po<e~(~ energy of cach pro"Tessive
wuve is o!!C-~<M~er
of that of tlie original disturbance. Since, as
we !u).vejust seen, th whole energy is o?te-/<a~'of th same
quantity, it follows that in a progressive wave of any type onohaf of thc energy is potential and one-haf is kinetic.
Th same coTidusion may aise be drawn from the general
expressions for thc potential and kinetic nergies and th relations
betwoen velocity and condensation expressed in (3) and (4).
Th potential energy of the clment of volume c~Fis the work
Thoendic.-it
statoment
of thoprincipio
mbo<Ued
in cqnniion(10)tirnt1htivo
mot withis in )i pftperby Sir W.Thotuson,"On tho possiblodeusityof tho
ImuhnforoHs
valueofa oubicmileof suu-Hght."
mctUnm,
and onthomcchMuctil
7~t'V<tf;.ix. p. 3f!. 18;
Uoxan<[)tut. ~/tt<. ~/<t~. xLv. p. 17!
//. ~fy. (;)
p. 2CO. 187C.

R. ir.

1873.

18

NEWTON'SINVESTIGATION.

[245.

that would bc gained during the expansion of thc con'cspondin~

qua-ndtyofga;: from ita .tctualto its normal volume, thc expansion
bcing opposcd t])ronguout by tlic uonnal pt'cfisurc
At any
stage of thc expansion, whou the condcnsntit)!i is s- the cn'ccdvc
pressure ~) is by 2.J4 f~~s', wl)icl) pressurebas to bc muItipHcd
by thc corresponding incrment of voh)mc fn'c/s'. Thc wholo
work ga.Incd dunng thc expansion from ~~to
is
Jr(l+s)
thcrcforc ~p,~F.
or ~F.
s' Thc gcncra! expressions
1n

for tho potential and Mnctic nergies arc accordi;)f!y

If the p!ano progressive wavoa be of Itarmonic

type, and s
at any moment of time are eircuiar functions of ono of the
space
co-ordinates (a.-),and titorcfore thc mefui value of tlieir
squares
is ono-half of thc maximum value. Hcncc thc total
cuc!y of
tho waves is equal to th kinetic energy of the whole mass of
air concerned, moving with titc maximum vclocity to bc fouud in
th waves, or to tho potcntial cncrgy of tlie same masa of air
wl)cn conclensedto the maximum dcnsity of thc wavcs.
246. Tiie first theoretica~ investigation of thc
vclocity of
sound was made by Newton, who assumed that tho j'dution bctweon pressure and deusity was that fornudatcd in Boyle's law. If
we assume p==A- wc .sec that the vulocity of sound is
oxpressed
in which t!ic dimensions of p
by V/f, or
(= force-area)
arc [37] [Z]-' [2']" and thoscof (= mass voJume)arc
[.Vj [Z]Newton Gxprcssedt))c rcsult in terms of thc 'e~
of ~e/iOM:o~eHeo:~n/M~.f~Aerc,'dcnncdhy the quation
wherc~) and p rcfcr tu thc pt-cssun;a)i(t t)m dc-nslty{tt thc carth's
surface. Thc velocity of sound is thus
or thc vclocity which
would be :tC()nirud)~y body i:d[ing frudy under thu action of
gravlty tin'ougt) ludfthc h~ht uft)t(.' homogcncousatmosphore.

2-]

LAPLACE'S

CORRECTION.

Toobtmnanumurical
)-csu)t "/c r~'uiretcju~w
~pnir<f
simuftancous v.uc5 of f~d
It i.s found by cxpcru.tent ti)at
fit (~ Cent. undor a pressure of 1033
grammes per squ~-c ccntitnctrc, thc dcn.sityof dry air i.s '001293 grammes pci- cubic ntimetr~. If wc takc tho ccntimctrc,
grannnc, f~d second as tho
fmidamcuta! unit.s t)tc (c.O.s. System), tl~sc
dat:imvc

.sothat tho vclocityof sound at()"v'o)dd bc27.')'f)5 mtres

pt.r
short
of
t)ic
rc.su!t
of
scn.), faHing
direct nbsur~tinn Ly abonL
:).hixt)jpart.
Ncwton's
Invc.stig~tit.n c.stab])Mhc.)
tl.at tho vetocit-yof .suund
-shouldbe indcpcndcnt of th amplitudu of tho
vibration, and aiso
of the pitch, but thc
discrc~ney b~wccn InH ca)cu]atcd v~m.
(pubhshcd in 1G87)and tlie exprimenta! v;due w~not
cxplained
until replace pointcd out tbat tbo use of
law involved
Boyic's
thc assumption tliat in tho cotidensatinns ~ud
rarofactions accompanying sound tho temprature renmins constant, in contradiction to th known fact that, wltcn air is
suddcnly comprefiscd
its tc.npcmturc ri.ses. Thc ]aws of
Boyic and Charles supply only
ne relation hctwccn the tliree
quantities, prc.ssurc,volume, and
temprature, ofn.ga.s, viz.
wnerc th temprature Is tnca.su.-edfrom the ~ero
of U.c. ~s
thermotnutcr, and thcrcforc wit!iout some auxiliary
assumptio~it
is nnpussiHe to specify tlie conuecHon
bctwccn and v (or p)
Lapiacc con.sidcrcdthat thc condensations and rarfactions coucerncd in th propagation of sound take
place with such mpidity
that th I~catand cold produced hve not time to
pass away, and
tf.at thcrofore tho relation betwccn volume and
is sensibiy
pressure
th sa.nc as if thc air were confiucd in an
absoh.tcty non-conduct.ng vossch Undcr thse cireumstanecs thc
cl.angc of pressure
currespondmg to a given condensation or rarfaction is
greater
than on th hypothesis of constant
tenipcraturo, and th vcloci.y
oi sound is aceordingty iucrcased.
In quation (2) !ct dnote the volume
the pressure of
and
th unit uf mass, and Ict be
expressed in centigrade dcgrees
n_ n

20

LAPLACE'S
CORRECTION,

[346.

rec~oncdfromth a.bsohttcxcro'. Tho conittionof thc gn.s(if

uniform)is de~nedby any twoof:tt0thrcc quantitiGs~), a.nd
i.h(tUth'dt~tt.ybc -t~)Ld n. td'ins i tiicn!. Thc rc~tiou
varin-tions
of thc tfjrcequantiticsis
betweenth simuku-Dcous

In ordcr to cffect tLo change spccificd by f~) and dv, it is

in general nccess:n'y to communic!ttc hcat to tlie gn.s. C~IHn"'
th nccessM'yqua.utity of hcat ~Q, wc may writc

246.]

EXPERIMENT0F CLEMENTAKD DESORMES.

21

if, as usua!, tlie ratio of thc spcifie Iicats be denoted by 'y.

Lapl~cc's vaiuc of titc velocity of Soundis thcrcfore grcater t)ta.n
1.
Newton's in thc ratio f~y
By Intgration of (8), we obtain for thc relation bctwccn
Maud p, on thc supposition of uo communication of hcat,

Unuer the same

whcrc )),, o,; arc two smuuta.neous values.
circumstaaccs tlic rci~tion 'betwecu pressure and temprature is
Ly(3)

Th magnitude of 'y cannot bc dctermined with accuracy by direct

cxpcrhncnt, but an approximate value may be obtained by a
iiietliod of which tlie followingis tlie principle. Air is compressed
into a resci'voir capable of being put into communication with
thc external atmosphere by opening a wide valve. At first th
tcinperatui'o of tlie compressed air is raised, but after a time
tite superiluous hcat passes away and the whole mass assumes
tlie temprature of thc atmosphre 0. Lct thc pressure (measured
by a manometer) be p. Thc valve is now opened for as short
a time as is sufHcieut to permit thc equilibrium of pressure to
be compictcly estabhsbed, that is, until thc internai pressure
bas become cqual to that of thc atmosphre P. If th experiment
bc properly arrangcd, this opration is so quick tliat tlie air in
thc vcssel lias not sufncient time to reoive heat from tlie sides,
and thercfore cxpands ucarly aceording to the law expressed in
(9). Its temprature at tlie moment th operation is complete
is thcrefore detcrmincd by
V

Th cnclosed air is uext. a.!lowcdto absorb heat until it bM reg:Lincd thc a.tmosphcnc temprature 0, and its pressure (jp') is
thcii obscrved. During tlie last c!~nge thc volume is constant,
:uid thereforo tlie relation bctwecu pressure and temprature
c~vos

It Is horo nssnmod that

eh'on iirat by roisson.

is const.nut. This quation appears to havo beou

~2

RATIO 0F Sl'ECIFIC IIEAT.S.

[2-iG.

s(jU)at,hydhnu)nticnof~:("),

By cxpurnnontsuf

thisnat.)))~

(.')c)))c))t. and Dc'sonn~s dc-

~))tinL:d'y=l~t!bt)t,).JtcnuLhudiso))\'iouHtynots))sc(-ptih)c
of :my grcat iK-cm-i~y. Thc v~hnj uf
'y j'C((uir<jd tu ~ccmcittj
t))L: catcn!atud:n)doL.S(j)-rdvc)uc~iL!.st'.som)d
isl--K)S,oft)tu
su[).st,antt!t.!currcuLuc.ssufw)ticht.)tcruc:m))<jiitt.teduuht.
\Vc!L)-cnot,))owc\-t;'r,(tcpc)i(L'ntoi)th<Jt)hun<)niun:iofi-nnd
fut- unr knowlud~c of thc !nagnit))()c of 'y. Thu ViLtuc (.F
/<
thc spcifie liuat at constant prcs.sorchn.s Lccti dt.;tu)inincd
L'\}n'nnie)tt:d)y by Rc~n:udL; :md :dt))ough 0)1 account uf in.
lurent <)ifiicu)tic.s tho c'xpcrimcntid mu(.]njd tnay i:Lil tu
yic)d
;t Matisfac~ory rosult iur /< tho infurtnatiutt sought fur
may bc
uhtaincd indircetly by niGana of n. rehdioti bctwcoi tl)< two .spcoitc ])cats, Lruu~))t tu Ji~ht Ly t]iu mudum science of Thcrniodynamics.
Iffroiat.hccquatiun.s

Lt as suppose th:Lt (~ = Q,or that thcre in no communication

of))C;it. Itisknowu
that Hic )te:t.td'j\-c)')p(!d(ht)-)ngt)]c compression ofa.napproximatc!y pcrfcct~s,.su<j)iaHair,isatinost
ux~cDy thc thcDnul c'~tivatent of tlie work donc lu comprcs.sin~
It. This nnportant principtc w.s assuined by Mayer iu his
ccicbt-atud liiemoir oa H)L' dynfLnuca~thcory of hea.t, titough
(')t g'ru)))xtswtuch ean hai'ttiy be cousidured mtcqua.tc. Howcvcr
th:Lt nmy bc, t)te priucipic itscif is vcry nariy truc, as bas since
been proved by thc cxpennieuts ufJmdu aud TItonison.
If wo mcasm- Le:i.t in dynamical nnits, Maycr'H principic
o~ tlic understaudiug that there
may bc cxpresscd ~=~~u

2.1 G.]

RANKIN-E'S

C~LCUJjATION.

23

is no communication of hcat. Coniparing this with (15), wc see

tha.tt

By
of that
<)~rcc
on i t.

Rcgnault's cxperniicnts thc spcifie hcn-t of air is -2379

of watcr; and in ordor to raise a gramme of watcr onc
Cent., 423.')0 gratnnie-ccntimeti'ca of work must be donc
Iluncc with thc same units as for ZP,
= -2379x 42350.

Culculating from thosc d~tn, we find ~y=1'410, ~grecing almost

cxacUy with th value dcduccd from th vclocity of sound. This
investigation is duc to Ra.nkluo, who cmploycd it in 1850 to
c:dcula.to tlie spcifie hoat of air, t:i.M;ig Joule's quivalent and
the obscrvcd volocity of sound as data. In this way he anticipatcd tlie result of Rcgna.ult's cxpcriments, which were not
publislicd uutil 1853.
247. Laplacc's thcory bas oftca bccn th subjcct of misapprchcnsion among studcnts, aud a stumblingb~ock to those
rcmn.rkabte persoi~, caDcd by De Morga,n,pM'ildoxcrs.' But therc
Ciui be 110i'L':i.suna.b).c
donbt t)ia,t, antccedcntly to ail calculation~
t!)c Ilypothesis of no communication of hca.t is greatly to bc
prct'urrcd to t)ie cquaHyspcial hypothcsis of constant temporature.
TitCii'cwotdd bc a. reaL di.HlcnItyif tlie velocity of sound were
not dccidcd)y in excess of Nuwton's 'vainc, a.nd t!ie wondcr is
l'~thcr that tho cause of thc cxccs.srcmained so long undiscovcred.
Tlie on)y question which can possibly Le consiclered open,
is wliother n small part of the Iicat and cold dcveLopcdma,ynot
escape by conduction or radiation bcfoi'e producing its full effect,
Kvcrything must dpend on thc rapidity of thc altern~tions.
Hc)ow a certain, limit of slowness, thc hcat in exccss, or dcfeet,
would have time to adjust itself, aud tho tonpcrature woutd
remain scusiUy constant.
In ttlis case thc relation betwcen

24

STOKES' INVESTIOATJON

[247.

presse and density wouid bo th:tt w)uc]t I~~d.sto Ncwton's value

of titc vulocity ofsound. On 'hc othcr !~m], Mhnvca.cci-t:un Hmit
~iqmckm.s, Lhc !,a.-iY.'uufd.~Jt.vc~.jifc~nhcdm.~no)i-L:~ndxcttng vc.ssci,a.s suppo.scd in L~pt~s Uicory. Nuw fdthough
tito circumstiu~cc.sof thc autoa-t pi-obtf!n anj hctt(.;r rcpi-c.scatcd
by titc lattct' t)~n Ly t))c formut' supposition, theru ]n:Lystill
(it m:)y bu sn-id)bo n.sensible d(;vi:tt.io)ifnjni t)ic law of pressure
and dcnsity invulved in Laplace'.s theory, ent:u)i))g so)newli:).t
slower velocity of propag:tt.iunof sound. Thi.s <juustiun h:i.s bccn
carefully discusscd by Stokes In a p~pcr publishcd in 1851',
of wbich tit fullowin~is :ui outlinc.
Th meclianical cqua.tionsfor tho SH;a~motion ofuir arc

Thc tcmpct~turc is supposcd to bc u)ufor)n

cxccpt il, so f:).r
as It is disturbud by tho vibrations
thcmselvc.s,so that if dcuote
t!)e e~ceMof tcmpcratnrc,
ihe cncct of a smati surdon cundcnsfition s Is to producc an
elovation of tcmpcmturc, which m~y be dcnutcd by /3~. Let
be tlie quantit of heat cntcnng ttic dment of volume in
f~
time dt, mcasured by thc risc of
tcmporit.turc thfLt it wou!d
produce, if thcre were no condensation. Thon (t.hc distinction
and
Leing ncglectcd)
betwcen

bcing a function of

and its di6crcnttal cocicuta with

respect to 6pacc, dpendent on the spcial character of thc

dissipation. Two extrcnie cn.sc.smay bc mcntioncd, thc Urst
whcn tho -tcndency to equaHsation of
temprature is due to
conduction, thc second whun tho opei~ting causo is rudiatio])
and the tra.nsparcncy of the nicduun suc.h that radiant licat is
jP/t<y.
Af< (i) i. 805.

247.']

0F EFFECT 0F RADIATION.

not scnsibty absorbcd within

25

distruice of svre wavc-lcngths.

lat
:s tthat
IIC1 1:'>
winch
in 1.1
t.hc
~IJlutter,
;n.d JJI
\lUCI',W
l'lnur Cli:'JC
Li11tthc
l\ l'unnur
c~c <tC
~~x~~
v-a, ;LIll
Ncwton's
(sctcct.edby Stokes for :u)a)ytienlinvestigation,
h~v uf radiation bcing ftssumcd a.s a, suficicnt approximation to
Ihc trutli. We i~ve thoi

lu thc c~sc of pl~no wn.vcs, to which wc shdl cunHnu our

:~tu!ttion, u :Uid Vfuush,wlule , ~),& arc functions of (:uid <)
unty. Eiuuiuatiug~ and Mbctwu (1), (2) a,ud (3), wc HM~

if y be written (in the sa-mcsense as beforc) for 1 + a~.

If the vibrations be IiM-nionic,we may suppose that &varics
as e" aud thc cqutttioii bceomes

Icss tbMi ~7r) if wc wish for thc

but (~ being positive, fmd
wc
expression of t)) wave travelling iu tlie positive direction,
must take tits lowcr sign. Discarding thc ima.giuary pM't, we
fiud as tlie appropriatc solution

TII~ AMPL~'UDE

1S ~fOUE

r~.i~.

Thc ~rst t))in~ tu bc noticed is thf).t t)io Honm! ca.nnot

'bu
i't'upa~tttjd tu dis(.;)ncL'untc.sn Mm Le insensible.
Thu vcioc'iLyafpro~ag.Ltion (F) in

i\ow irom (O)w(j suc t)<at

cannot bc insensible, un!css
is cithcr vcry grcat, ur vcry smai. On tiic first
supposition
frum (11), ut- dircctiy from (7), wc hve
apj)roximatcly, ~=~
(Newton), and on t]tc second, F=~,
(Laphice), ns ought
c-vjdunt)ytu bc t)i(i case, w])0i titc meaning of y in (.-)) is consxicrud. W)):~we now Icaru is t!):it,
if and Mwcrc compamb)e,
thu c<!c:ct.wou)(]be not mcrdy a dcviaiion of
from eititer oi'
tho limiting values, but a rapi.!
stiiling of tbu sound, w!iich wc
J<nu\vdocs uot takc place in nature.
Of tins theorctical rcsult wc
may convince ourscives, as
Stuhcs cxptiuns, witl.ont tbo use of
analysis. Imagmc a m~s
of air to be conimcd witbin {t dosed
cylinder, in which piston
is workcd wit)i a
rc-ciprocatingmotiou. If thc period of tiic
motion be vcry long, tbe temprature of the air rcinains
ncar]y
constant, tlic hcat dcvclopcd by comprcssicn n~ving time to
cscape by conduction or radiation. Uudcr tliesc circumstauccs
t))u prcssm-u is a fnnctiou of vohtnic, and whatcver work
bas
to Lu cxpcnded in
producing a. givcn compression is rcfunded
whcn ihc j)i.stonpasses Dn-ough tite samc
position in thc reverse
du'L'cttun; nowurk is constuned in the Lng run. Next suppose
O'at th motion is so rapid that Hicrc is no time for thc hcat
and cu!d duvchjped hy t!iu condensations and rarfactions
to
c-seapc. 'i~c pressure is stiU a function of volume, and no work
's dissipated. Tiic ou)y din'ercnce Is t)iaL nuw thc
variations
uf pressure arc niorc considcrabfc t!ian hcforc I)i
comparison
~itht))c variations of volume. ~Vesechowitisthathotho)!
Newton s and on Laptac~s hypoU.L'sis,tJK! wavcs travcl without
dissipation, t))ongh with difcrcnt vc'Iocitics.
But in inturmediatc cas~s, whcn thc motion of thc
piston
Js ncithcr so s!ow that ti.c
temprature romains constant nor
.so quick titat t])c ])cat lias no time to
adjust itsu)f, th rcsuJt
is diHurent. Th work
cxpundcd in produciug a, sm~I! condcnsa-

2t7.~ J

INrLUEXCRD

TIFAX TIH VELOCITY.

27

tion is no longer con)p)ct(dy rcfnodod during thu corrcsponding

rart.'t'action on accunnt of thu ditninishcd tutnporaturo, part of
thc I~'at dcvulopud by t))c compression havin~ in Utc tnciUttime
csc~pL'd. ln f:Lcbt))u pa.s.su~'cof ]x.tt by conduction or nuti~tion
front n-wiu'nn't' to a nnitcty co)dur Lody :d\V!tyHinvoh'cs dissipation, a p)'I)K'ip)'j whiutL occupius a futuhunoit.al positi<jn in thc
SL'iunce!of Tt)ct'!)io<)yn:unics. In order thci'(.;for<jto m:Li)~tainthc
tnot.iunot' t))o piston, cncr~'y nm.st bc supptiL'd front -\vithont,
and if t.))(.'robc on)y a liniitL'd store to Le <]r:).wnfron, titu motion
inust. ultitti<<.tofysubside.
Atlothcr point to hc noticcd is that,
p;u':d.)lc, Kwon)d depund upon x, vix. ou
a st:).to of thin~'s whidi frotn cxporirnej)t
suspect. On thc contrary thu cvidcncc
provo that thurc is no such eonncction.

if f/ and
worc cointhc pik-h of t!i<j soum~
wc h.'t.vc no ]'(,'ason to
of observation gocs to

From (10) wc sec that thc faUlng off in thc intcnsity, cstland.
tnatcd pcr wavc-lcn~tl), is a maximum \it)i tan~, or
= ~Y. In this case
is a m~xininm, whcn
by (!))

Calcula.tmg from titesc Jut~ wc fhtd th~t for cach wavclun~h of :).dvnncc,thc innptitudc of the vibration would Le
dn)T.i)iishudinthcrntIu'C172.
To tnkc a.iiumerical cxampic, Ict

la 20 yards thc intcnslty would bc dhnmi.sitcd in th ratio

ofitbout 7 ~'Hions tu (me.
Cun'cspondi.ngtotins,
If* the vahic of y were ~ctnaUy that just written, smtnds of
Wtjthcrethcpttc)tiu<~n-:stiouw(ju!dbcY(jryr!Lpid)ystifiud.
iurcmfL'i'th:tt(/isitiii).ctcithurrnucI)gr<~t(;ror(;Lscmueh!L'ss.
But cvcM so large a Y.duc as 2000 is utt(-!r)yinudmissib)' as
we m~y couvince our~ctvcs by cousidcrin~ th si~ntficancc of
quation (5).

38

EFFECT0F CONDUCTION.

[247'.

Suppose th~t by a rigid envdopc t.i-anspa.t-cntto raJiiUit hut,

thu vohnuc of small ])):),.ss
of ga.s wcru n):u))t:unc<tcon.st!mt,
L!u t.hL-c~);).}-)'~)
to ..L-ni~.iu k..s tht.ri~:u ~idiiiun nt !j
t..neis

-whcre ~1 dnotes thc Init,i:Ucxcc.ssof toupuratnru,

provittg th:tt
tLt'tera time
t)tc cxcus.sof tcmpcr~urc wouht fait to les~tLun
t~tfits 0)-ig:u:),lvalue. To.suppose thatt]ii.scou)d
happening
two thousan<!thof second of time would he In contradicLionto
tho most supcrficial observation.
We arc thcreforc justihcd in assuming th:tt is
<7 voy small
in cotnparison witli
and our cquntions then bucoine ap])rcxiniateiy

Thc effects of a srnall radiation of Iicat arc to Le

for rati.cr ni a damph.g of the vibration than in au sougbt
altcred
vclocity of propagation.
Stukes calculatcs that if Y =1-414,
H~c ratio
F=ll()0,
(A 1) ni which tlie intcnsity is diminis])c<t in
pa.ssiu~ over a
is
<)I.stancc.-r, given by Io~=-0001156~
in foot.scco'ndmcasurc. Altho~h we are not able to
makcprccisc measuremMts
of the iutensity of sound, yet the tact t)mt
audible vibrations
eau bc propagatcd fur many miles cxcludcs
any suc!i value of
as
could
q
apprcciabiy affect t!tc velocity of tmusmissiou.
Ncitbcr is it possible to attributc to thc air sucl a
conduetin<T
pawcr as couid niatcriaUy disturb thc application of
LapJacc'~
In
order
to
trace tho en-cct.sof conduction, wc have
theory.
ouly
to rcp!ace iu (5) by
Assuming as a particu]ar solution

247.]

VELOCITY DEPENDENT

UPON

TEMPERATURE.

29

teaving thc velocity of propagation to tbis order uf approximation

still equal to ~/x'y.
From (18) it appears tliat thc nrst cffect of conduction, as
of radiation, is on tlie amplitude ratbcr than on tlle velocity of
propagation. In truth the conductmg powcr of g~ses is so
fec~c, Md in the case of audible sounds at any rate thc time
durin"' which conduction can take place is so short, that disturbance from this cause is not to bc looked for.
In thc prcccding discussions the waves arc supposed to bc
propagatcd in an opcn spacc. Whcn thc air is confined wlthin
a tube, whosc diamctcr is small in comparison with thc wavcIcngth, thc conditions of th problem are aitcrcd, at least in
th case of conduction. Wliat we have to say on this hea.d
will, however, comc more conveniently in n-noUierplace.
24-8. From the expression \/(~y) ~p, wc sec that in th
same gn.s thc velocity of sound is independent of tlie denslty,
becausc if the temprature be constant, varies as p (~ =~p0).
On th other hand th vclocity of sound is proportional to the
square root of the n-bsolute temprature, so that if f~ be ils
value at 0"Cent.

wherc th temprature is mcasurcd in thc ordinary manncr from

the freezing point of watcr.
Th most conspicuouseffect of th depondence of thc velocity
of sound on temperature is thc variability of tlie pitch of orga.u
pipes. We shn-llsec in thc following chapters that th period
of thc note of a flue organ-pipo is th time occupicd by a. pulse
in runnin~ over a distance which is a dennitc multiple of th
Icngth of th pipe, and therefore varies inversely as the velocity
of propagation. Thc inconvenience arising from this altration

30

YELOCITY0F SOUXI)IN W.ATKH.

[248.

of pitch is nggravatctt ))y thc tact ti)at t))c rccd

pipc.s arc not
snodarly af'fuctud; so that a change of ttjmpcraturc puis fu)
o)'g!)))0))tcftu))H~'ithit.sc!f.
l'rof. Alaycr' ttaspropo.scd tomakuthcconncction
bctwcc'n
tutnperattn-c and wa~c-l~ngt,)~th fonn<)ation ci' a pyromctric
mctitod.bnti amuotawarc
\v))ct,ho-t))c o.xpcrunott h~sc~r
Lccn cm'ricdout.
Tlie con'cetncss of (1) as rcgiu'dsnir nt tlie
tcmpcndurcH ofO"
and 1()U"i~s Ltjcn vurificd L~pcnmcutaXy
by Kundt. Sec 2GO.
In dif'crcnt gascs at givun tc-mporaturu ant]
prc.ssm-Grt is
invers)y proportiona! to thc square rnts nf t)ic dchsitics, a),
IcaHtif Y bc cn)tstant=.
2, For thc non-cptKhjusahtcgases ry doc.s
not scnsibiy vary froin ils vatn< for air.
Thc velocity of sound is not elltirely
indcpcndcnt of thc
do~-ec cf drytic.ssof thc air, sinec at n-givcn prossurc moist air
is somcwhat Ji~))tct-than dry air. It is caicidatcd t]jat at 50
F.,
air satnrated wlt!t moisturc wou!d propagate sound bctwGun
2 and :{fcct pcr second fastcr t)iau if it \verc po'fucDy
dry.
T))c fnnnu)a

may bc applied to c:dcu)atc thc vciocity

~=~
of tionnd in nquids, or, if thut bo I{nown,to
infcr convcrsc]y
tho coufHcIcntof comprcseibiHty. In. the case of watcr It is
found by expcri)ncnt, titat the compression per
atmosphre is
-0000457. Thus, if (//j = 103:; x US],in absoiutc f.f:.s. units,
IIeucc

~=-0000457,

siucc p = 1.

~= 1489 nictrcs pcrsccou(),

which docs notdifur much from thc observcd va)uc (143~).
~4:). In th preceding sections t))C thcory of plane wavcs
bas bccn dcrivcd from t))C gnera! c'() nationsof motion. Wc
nuw proccc'd to an indepoident investigation in which the motion
is cxprcsscd in tenus of the actual position of thc
layers of air
instcad of by mcan.s of thc vulocity potentia!, whose aid is no
fungur ncecssmy inasniuc)i as iu oac dimension thcrc eau b~!
no question of moiccuhu'rotation.
'OnntiAcousttcryronctcT.
Accnrdin~ to thc tiinctic

7'/<xLV.
theory of ~sca,

)).])-<.
t1o voiof.ity

M)!t').y).y,ttndis))roportin)mlt.i,t)R~unmv(.)ocityafthn)))n]<.fu]('f).
]~77.
.)/(5)tn.;)..tU.

1873.
of onun~ is detorminol
)'rc!-t<),

249.]

EXACTDIFFERENTIALEQUATION.

31

If V)
(Ic-nncthc a.ctua! pnsitions n.t timc < of
2/+nt'!Q'bot!n!g i~ym'~of :!)r ~'husou'jUiiibt'iutttpositions :u'c (tt~no)
by .f and a;+~, thc dcnsity of t!)CIttchn'cd s)icois givc'n hy

th expansions a.nd condcnsntinns bci))~ sxpposcd to tn.kc nhicc'

;iccurdit)g to tlie adiabatic titw. 'l'lic tn:s of uniL cf :u-~ of
tlie
J slicc
1.. i8 l and1 thc
1 corruspondin~
1" ]no\'ing rforce Is ~l!' 1
fZj;
~iving for tlie cqnation of motion

Equation (~) is a.n e.wtc~equ!Ltnj)idctmin~ thc actual absci.ss:i

in tci-tns of thc cquitibrium abscis.sn. ruid thc timu. If t))e
motion bo assumcd to bc smal], ~vc mny rc))!acu
oceurs as tho coc~icicnt of th sni:dt qnn.ntlty

wltic))

f~)

by its np-

proxunatc va.!uc unity; and(-t) thcn bccomcs

thc orclinaryapproximatc equation.

If the expansion hc isot))cr)na!, n.s in Newton's titcory, thn
quations con-csponding to (~) !tn<1(5) ;n'c cbt:uu(.'(! hy tnurcty
putting 'y = 1.
Whn.tcvcr
th
(1)

coustitutioh
:m(l(3)

bc

may
of

the
thu

relation
mdium,

betwoon
t!tc

cquatiou

n.nd

o~

(k'pcndi))~
(~f motion

is

Ly

32

\VAVE80F PERMANENTTYTH.

from which p, occurring in

[249.

is to be eHminatcd by mcans nf

.f/ u
it~~1j.
tllu
rc~muiut:
therci:(.:ionbc:)VC(;!)~anL),cxpreHS(.(tirt(l~.
1)c";vw;n
cl:c
2.'i0. lu thc prcccding investigations of acria! wf~es wc
that thc air is at t'est cxcfpt in s') f:n' as it i.s
h:ivc HU}'pf)sc<l
distnrbcd by thc vibrations of sound, but we arc of course at
Itbcrty to attribntc to th who)e mass of fur conccmc'd any
comnion inotiou. If wo suppose that tbc air i.s moving in thc
direction contrary to that of th wavcs and with the s:unc actuat
velocity, thc wavc form, if permanent, is stationary in spacc,
and th motion is .~eftf~ In th prsent section we will considcr th prol)lcm under this aspect, as it is important to ohtain
a)! possibledcarness in our vicws on t!)e mechanics of wave propagation.
If
p~ dnote respect!vcly the velocity, pressure, and
p bc
density of thc nuid in its nndisturbed state, and if
th currcsponding q~antitics at a. point in the wave, wc !iavc
fur th equation of continuity

dctcrmining thc law of pressure ~nder whi~i n.]oncit is possible

for a stutionn.ry wavc to ma.inta.in itscif in Huid moving with
Frum (3)
vc)ocit.y

Smcc thc relation between. th pressure Mi'I thc <1ensltyof

fictuat ga.scsis not t)tat cxpresscd in (5). wc cnnehidc that f).sdf!n:nnt:uning stn.Liunnryao'ial wnvc is un iinpu.s.sihitity,wha.tcver

250.]

WAVE

0F

PERMANENT

TYPE.

33

may bc thcvcJocity?~ ofthc gnera) currcnt, or in othcr wor~s tha.t

a w:),vc Cfmnot hc jn'n])!~at(;d ru)ativc')y to th undi.sturbcd
parts
of thc g~s withriut uttdurguitt~ !in :)ltc)-ation of
ty[)0. Ncvcrtixjtc.s,
wh~n Lho ch:U)~us 01 (jcnsity concenn-tt are sma!], (;')) niny ))G
satisfied :).ppruxi)natc)y; arxt wc sec fron (-t) th:tt. tho vubcity
oi struatu neeussiuy to kou? tho w~vc stit.tiun~t'y is ~cn Ly

which is thc same as thu velocity uf titc v'uvc cst.iinatcd 1'da.tivdy

tut)ic(ini(l.
Tins ))i('t))()d of rc~.i,)'<tin~ tho sn1)jfct sliews, pcrhaps more
thc !)i)turuoft!)< rotation hc<v(!cnve)oci<y
clL':).)-)y(.hana.)]y<)L))('r,
and coudcnHation 2-t.') (~), (')<). In !Lst:U,i<jt):u'yw!).vo-f"r)n:),)(jsa
of vducity iiceompauic.s !m au~tnc'ntt't) dcnsity accor<]in~ to t))c
prmcipic of cnm'gy, and tticrc-f'orc thc fLnd cotopo.sing the condcnscd parts ()fa '\va.vc movcs f<u'\va,)'()more slow]y than tho
)in<]istu)'bc(t purt.iuti.s. RL'Iativciy to t)t Huid thcrcfurc tho
niotiun of t)m c(j)t(h;t)H<!
parts is in tlic sa.mc dh'uctiuil as that
iu winci) thc wavcs arc prcpagatcd.
Whcn thc reh~tion bct,wccn pressure a.)id density is otho' than
that cxprcs.scd in (~), a stationtu'y wavc catt bo m~mtan.icd oniy
by thu aid of an hupru.ssct! force. By (1) and (2) 2S7 wc hve,
ou th supposition that thc moti'.m is stcady,

shcwing that an hnprcsscd force is uccc.ssaryat every place whcrc

Mis variabtc and uncqual to M.
2.')1. Th reason of thc chimie of type v'hich cnsucs when a,
waA'e is Ict't, to it.suJFis nnt <)i~unft to midcr.stimd. ;From t)te
ordinary thcory wc know that an infitutciy Stnali (!ist)t)')):mccis
propagattid with a. certain vdocity
\))ic)t vdoL-ltyi.s !-<-)at:o
to tlie parts of th mcdmni urxti.sturhudhy thc wavt;. ]~t us
considor now tito case of a. wavc so long ti~at titc variations of
R. II.

34

SUPERPOSITION 0F PARTICLE VELOCITY.

[251.

V(')oc!tyand dunsity are insensible fur a. considrante distante

)).s
a)o))~it,:tnd:tt.:),p)acc
w])('rcthcvc!oeity()isnrntf'Jct
i'
ts!)).!t!s~co')'r.t\~
t'h'~)r.
TL'v
1
with w)tich thc sccondary wavc is pn.pnniUcd <hn)))~h t.he
tnediuolis~, !))[).on :'c-('HUt)t<.t't))(;h~:dt)t(,tit)nf,f'th(-]n<j<!iu)n
i.s ~+M,an(I<)u))('))ds upmi
its<')t't))Cwht')t)V(']ucityot'!it]v:).n'-t'
t!!C p!U'tf)t'ti)e]t.H'
W!L\(!!tt, w)))(j)t tftCS))):))) ~:)V(.-isp]:K;).
'V)):).t,)~asbc;t'n~)(](~T,.st'c<))))]!))yw;LV~n))))i)('s:)).-i()tnt))Op!)]'ts
itst.)f')))d<.))usw('H(-c
t))!)t,!tf'(.c)-a.t,imc itD~n
"t't))c)L'))~Yt.'
pt!K.hcr(':t.(-t')'t:un
:u)\)ncc"f
\)oci<ytfist<'b(.
fomnt.isin
i<so)'i~in:J p~.s)<icj)])y;t<)i.st:)nc')n!(),~()<,t()<l)ut.t<)(~-+~)<;
")')s W(']<t;!y(~)))'c.ssi),); i.spn)[)~)tf(twit!t:t.vc)()(;ityn.+?/.
!').sym))n)i(-:dnt)t:)tin)) "=-=.(.+((~-)-)~,wJH.ru/iMa.n!H'bitr:ny
functiun, an Ct~))ati~))(i)'.st<~tai))L'()by Poisson'.
Froni ihc!L)~)U))<'))tj))st(.')~))]oy(-di<.n)i~)(.
f)ppc:u-!)tf))'st
si~)ttt));tt:).))<)-:)ti!~))()t'<y))(.-w;).s!t)K'L't-.ss:))'yit)('i()(;))tint))L'])rog)'~sH
as totho
"t'!).w:)\-c,inth~)('Ht)(.'Ht)yofanyp:u-ticn)iu'supposition
)'c);t<iottL('hvt-cH pressure :)))() dotsity.nndyct,
it \).spro\-('(I m
~)() t)):)t))t t))'(::)St'()i'on(Jp;uLit.'u);n'
!:tw ut'pressure tht'rc
wo))H he no !ttter!)t)o)t ut' type. AVe ),a\'e, ito~'ever,
tacitty
:).ssH)))L'dintitu prsent.se('io)) (htais
<-o))s):)))t,w))ie)~is t:)ntamomt.toa
restrictiuttto
]3oy)e'.s !a\v. Un~'rnuyc'titerifLWcf
isa iuncLiun of/3,a))d theref'ore.~s wu shaH soc

pressure

~)
In th ca.scct't.hchuv expn.ssp<1 in (;')) 2.'i(),t)iC'
presen<)y,of/f.
rotation Letw-'L-u M a.n'tpfora,
progressive \vaveis sxch that
A/(;

j+~

is existant,

asnu~h

:ul\'anee heing !o.stby s)o\vcr

propa~atit)))()nct<)nugniente(t()cn.si<yas
tion ot'thevetucity~.
So far as t))c constitution

ofth<'

is~'ainc'ttby

superposi-

mdium Itscif'isconccrncd

thc;)'t'isnot)nn~topr(;vent,oHras('ribi))~arbit.rnryva))K'stt)bf)th
?/arn)/), hut in a progressive wave a rotation ~et.\yeenH)csc<<)
<jUa))tit.ics]nust)'c'sa.tis)itj().
~Vckuo\v:u)-ea(]y~2-)-))tha.t<his
is thc case whon thc 'Ust.urbancc i~ smaO.and tl)c foUowing
arj.;)nn('ntw')H))oton]y:,1)cw<))at,snch:).r(;)atiu)iistobccxp('ctt'd
in cases whurct))cs(piarcoft))c
motion inust.bcrcta.incd,but
will cvun derinc thu forn) of't))C rutn.tiun.
~rrn)('i)CM)u-)aT)h"'f.]i('~))Sn)).Jo~))~
p.)U. ]sos.

~N~t.

vn.

BETWEEN
251.] RELATION
VELOCITY
AND PEXSITY. 35
W)):)t(;vcr

may

bc t!)c

n 1IIl,: j Il,'itlt!.J;III('
I"II
'n.H.!is~uL.

j~.s.(.vf'p)-~r.ssn-uwLvu
d<)!s;Ltioni.s

Luv

ofpre.ssm-e,

!.s bJ",
1.
.'jy~

tlie

vaiccity

of

:1.) ..<('I'WI. 1
tn.=.
'V~

p]-cpa"!i-

,'PII

thcrc)atinnbctwcc..vc.)oeityan.!cnn-

T<],s

n-I.-d.if.nbc vi.tcd
.t:)y
pnm~it w:vcwi!) ..n~r~,
'r.-u')!.)~int),c.
n~~i\-c <h'r~-tim.. L~usno~pk-hn-cto.
~<c~cof~)~vcpm~n~~vcw~~in~~n~,Uch~H
at
YL.!uc.t.y~n<h)..nsit,y:m.v..ry ~nu)u..t) h,,t bc.-on~.i.nj.ntL~
nc<unn,!atjm,))..t,i).smq~n-w).Ltc.dit,i.,u.smu.stbc.s~!sfi..di.L
"r<t.rt.o],rcvc..t),h<jf.,r,u:,t,n<,f..L,)~t.ivuw~vu.
[ti.sc)(.)-0.at
th<.nn.s~rt..
<c<)U~tiu.h.-t).r,not,:Ln..g..ttiv~wavc~-i)tbc
~-n~.<L.<hL(,anyp.,intwi[),)..j~n.tupu.iM<c.st.~c
of't,!m)~i.,t).c
tho r't,a.hh.tt,p,~
i~m~].nLci~)~uur)n,ud
t.hc.stator
thu~.s.-Lt~<)..sL-L.K.cfr.nni(,n.twm
t!
bu d~cnmn.j<tby
thu cn).L.n~nnpp)ic.)~.
<.us,na))di.st.rb~~s.
ln ap].)yi,rt).i.s
~rcto~.m~h-r
D.c v..)uciti..saHdeund~t,im.s,
nut
c~nuuwc
h, <).pmvai[ms
~)ut~y,l,trd~.ivdy
h, thcnci~.bnuri,~
f.rt..s oi ti.c m..hui~ .s.jtJuLt tliu furni uf
(1) jn.upL- fur t.).c prsent
pm'po.sujtj

wh!d. )sthc

dation

im~rcs.,ve ~vc.
i'.in'n.sfiaw'.

lx.t.woc.1 and~ ncec~aryfura p~.sitivo

E~u~iou (~ .ya.s obtai,,L.d a..aiytic.d)y
hy

1.~ the case of

Bcy!c.s h~v,

is constant, and t).c re!

y~
lion bchv~.u
vciucity ~nd den.sity~ given fir.st, 1 LeHcve,
by
Huhnholt/iij

bc tiic ()c)ts)t.y
co)Tcspon(]ing to = 0.
In this c~c Poissfjn'.s
aHow.s n.s to ~.rm .-L<).fmitc I.)c.-t
intgre
et U.c chang-c oitlie c.-u-Iier st~cs uf t!io
type ~ccompanying
7'/N7.~'nf)).<.l.s~:), p. Hf;.
= J~r~r)~.
(~ y'
,y. p. ioc. 18~2.
n

32

ULTMATE

3G

DISCONTINUITY.

[25L

it finaUyIcads us to n. diMcultywhich
progressofthcwavc,n.nd
has not a~iyct bccn surmonntcd'. 1. If wc dra.w a curvc to rcprescnt
th distribution of vclocity, taking .'<;for aLscissa a.nd ?' for
ordinatt. wc inay find thc con'cspondin~ curyc after H)C )ap.sc of
timc by the fotiuwh)~ cunstruction. Thron~h nny point on t)'c
Ot-igitia)curvc draw a st.raight ]me in th positive direction para1)''t
to a;, and of k'n~th cqmd to (f7.+ ?~) or, as wc n.)').;co))cc)')u''dwith
thc shapc of thc on'vc on!y, cqual to Il <.
t. Thu ]ocns of tbc ends ut'
thse linc's is th vulocity on'vc aftur a, time <,
Thc
But thi.s Ia.wofdcrivation caxnot ))o)d gond iudcnoitciy.
crcsts of thc velocity cnrvc g'ain conth)U:d)y on thu t)'(m~'hs and
must nt last nvcrtakc! thcm. Aftcr this t))c cm'vc woutd imticatt!
two vaincs of ?/ fcf onc Y.dno of
cca.siog to rcprcst.'nt anythi))~
tha.t eould actuaHy t:)~~ p]acc. Ju fact wc an' not at lihcrty to
push the application of thc intgra! htyorni thc point at which thf
velocity becomc's discontinuons, or th Ych'city cnrvc bas n,vertical
In ordcr to nnd wlicn this happons !ct us ta~c two
tangent.
ncighbonring points on any p:u't of thc Ctu'vf which sh)pcs downwards in thc positive direction, and inouirc attc-r what time this
part of thc curvc beconcs vcrtica]. ]f thc diffrence of ahscissa!
bc ~.r, thc hindcr point will ovcrtake thc forward point in thc
Thus thc tnotion,as dctcrmincd by Poisson's
timc ~(f/~).
e<ptation, bccomcs discoitinuons aftcr n, titnc C(p)at to th ruciprocah takcn positivciy, of tl~c grca.tc.st ngative value of ff~
For cxa-mple, lot u.s suppose tha.t

whcrc !7is thc grcatcst Initia) vdocity. Whcn < = 0, th grcatest

ngative -valueof

Is

so th~t discoutinuity will com-

mence nt thc time <= X 2??'!7.

Wi)cn ~iscnntinnit.y sots in, stritc of things cxists to which
n.)'c inappHcabtc'; fLUt)th suhscthcus)tn.l<!iH'crunti:duqu:).t.i~))S
qucnt prognjHSof thc tnotion ~s nf't hccn duterniincd. It is
probable, as suggcstcd Ly Stores, that s(j)ncsoi'tofrpOcctioawouid
ensue. In regard to this ma.ttcr wc must be carctul to kcep
m th T)icory
of Soumi." .P/< ~ny.Nov.1818.
StokM, Oua dimeulty

251.]
]

EA.RNSIlAW'S

INVESTIGATION.

37

purcly mathcnmtica! questions distinct from physical oncs. In

practtuc wu havu to do with sphcricat waves, witusc divcr~ency
inay ofitsdt'bc sufMcicutto ho)d la chuck tl~c tcndcncy tu d~conIn actuel ga.sc.stuo it is curkun t!t:).t:bufut-c
tmnity.
discontinuity
could enter, th hnv of pru.ssurcwon!d
bc~'in to citante ils funn,
~nd
DtcutOnenccof'vi.sco.sityconidnu lunger ije ncg)cctcd. Hnt
H'csc eon.sidur~tionshve
nothing to <)owit]~ tho mathumatlc:il
pt'obton of (tutunninin~ wh:).t wuuld ))appcn tu wa.vcs of nnitc
a)np]ttu<)uin :t, nicdiun), frce froni viscosity, wlioso pressure is
undur~ cireumstauccs cxactty
proportionn! to its duusity; aud
tlus probJunthus not bucu sulvud.
It is worti~yof rcinark tiiat,
althou~'hwcmay ofcout-se conoive
a w.Lvoof Unitc disturhance to cxist at
aoy inomunt, thcrc is a
i'tmt tu tho dun).ti~n ut' its
prcvious indcpcndcnt cxistc])co. By
dmwing Unes in t!)u ncgativu instcad of in thc pusiLivcdit'ectioa
~'c may trace t))u Jtisturyof ttie
vuiocitycurvc; and wc .sce that
as wc pus)i our lutjulry furth~t-an(t f))rt)nirint.)
past timu t))c furward ~lupusbucunic oasiur a)td t)ic back\v:u-d
slopcs stccpci-. At a
timc.cquai tu thc grcatcst positive value
of~, antceedunt to that
at which thc curve is nrst coiitcmplatcd, th
vclocity would bc
diseuutinuous.
2.~2. Th cotnpicto Intgration ofthc exact cquations
(4) aud
(Ci) 2-t.!)lu t))u ca.suuf a progrc.ssivcwavo was iirst utfcctcd by
Earnshaw'. Findiug ruason fur tifinking that iu a sound wavu t])o
quation

ea)i by jncfuisof tlie arbitrary function bc nnutc tu coiucide

witli a!)y dynanuc:d quation in wliicli t])Cratio of
aiid
t~
i.s cxpresscd in tcrm.s of
Tlie f'ormof thc fnnct.ion F Lc-in~

M~C
7'ruct'~t~u/' </<e
~! ,S'~t\ Jau.C,1850./<t' 2')~. 18CO,
p. 1~;(.

~8

HARNSHAW'S INVESTICATICN.

thus dctormin~ t))e8o)ut,io)) maybccumpicicd

prucc.ssupplicablu Losncb cases'.

[252.

bytLc

usuat

Writingo for Lrevity

J etin pinceof-cl.c ~'e h~.vc

as might aiso ha.vc bccn info'rcd from (.).) 2.')1. Thc cnstant (7
othcnvisc
vanisi)CH,if~(o!),viz.<,V!U)i.s)tw!)(j)iet=I,)-jp=~;
itrc{-)renent.iavu!uuityot'thc)n<;(ih[)u:Lsa\Y)tut(j,!)avin~])()L!)in"'
tu do wit.h ti)c w.t.voas such. Fur a~oA~~e
pro~russivu \vayc t)~
)<vcr si~tts in ttie :uubig(ntics :u-c to Le uscd. Titus in
pta <jf
(!{),wc;ha\'e

fi-uniwhich by (8) wc .sec ti.aty-(a+

) t is an arbitra~ function

1 Loulu'h'
Z~t')i<t<;<
A'</t<ff<<(~f.<,
Ch.xt\

253.]

MEMANN'S EtJUATtONS.

39

of &,or of ll.

Conver.s.-Iy thcreforc M is fu. fu-bitrary funetiuu ut'

,?/(a+;f)<,aiidwc]n!).ywri).u

E<;))ation(9) is Poi.s.son'.s Int~m), con.si.tcrud ht thc

pr~cc()i))g
.section, whurc t)<c sytnbu! x htLs thu sa.)nc mu.tni))~ :Lshure ~t~ch~
tt)~.
2.'i3.
T)'c]))-uh]mn<)ftj!:).))cwtt\'(;.s<)ft)))it(!!un)))it.()(!(.).tt]-!)ctc<!
a)so t))c !~tcnt~)t uf Rictn:mn, wttu.se niutnou- wns
connnutucat~t
tu t)ic Ruyid Sociuty(jt'UuLt.i))~~n on t)te 2Kt.i)
ut'Novuntber, LS.T)'.
Riu)n:mn'sinvnstig:)tio)t
i.siomxk~ontttu~.ncrathydrodytuuniciLt
in
e.t.Ltion.s Invc.st.i.n.-d.cd
~:{7, 2:~8, :Lnd is )mt. rcst.ict.c.t tu :my
p:n'tic)ti:n-):).wot']~.ssurc.
Inon)(;)-,))uwuvur,]H)tn)t(h)iytuHxto~t th di.scu.ssio)~of thi.s }):n-t.uf onr
.subj('(.-t,,a)t-c:KJy po-h:U).s
trc:U.cd n.tgr~Ltcr Icn~t)) t)):).n its physicid
imp..rt:t.nc(i wonM
w;u-t':mt, wn si)a)l hu)'u confnKi uur.sutvu.stu thc e~su of
Boytu'~ hw
of pressure.
Appfying c~u:Uions (i), (2) of 2.~7 and (1) uf 238 to t]ic
cu'cumstimccs ut'thu presunt pruLk-m, wc "'ut

ebencr
'Uub~dicrurtpfJnnxunH
witc.
(iiittixn~n, /)~/nt~/)n<t,t.vm.
i~the/.<r/~t'<<r)'<XY.p.l~

L"fbv(4)nvnne))d]i(..hcrSrLwin~u,)~1~0.
S<:e)d.un.ucxc(;Hnt
nb.itt'tMt.

40

IJMITED

INITIAL

DISTURBANCE.

[253.

Thse quations arc more gnral than Poisson's and Earnshaw'3

ill t)tattheyfu-c
not!imitudtoi]te
ciuseofa single positive, or
ngative, progressive wavc.. From (5) wc Jearn tfiat \v})atevcimay bc Lhe v:dae of7~ corrc.sponding to th punit a; and t]ic timc
t, th s:une vahtc of l' corresponds tu th point .c+(?;.+a)
at
th time <+~;
am) in thcsiunc w~y from (<i) wc sec that
Q rem:Li))su])cI):U)~() \Y)K-n.a.~d ac<)uirc! t)~ n)C[-(j)no)t.s (~-a)f~
n-nd </<rcspuctivL-ty. If 7' and Q Le given at a. ccrtiun instant of
ti)no ns fonctions ci'.r, aud thc i-cpt-esott.fttivc! curvesbc
drawn, we
may duducu Lhc cot'i'c.spondii)~ vatue < ;< hy (-t), and thus, as in
~51,cunstructthecu[-vcsrup)t.i~(j)tti))~t])evatucsufj!and
(~
aftc-rthc sma)! intcrvaloftunu~,
ft-un)w))ie)t thcnuw values
of <t au<] p in their turu bcemac knowu, a)id thc
pruccss can bc
l'cpuatcd.
T))C ctcmcnt of the fluill, to wLich the v:ducs of T~and
(? at
any moment Lctong, is itHc]f nioviug tvith tite vcfucity ;<, so that
th velocitie-s of~and Q rch~tively to tho dment arc
munuricaHy
thc satnc, and (.-quai tu a, th~t of 2' huiog in tlio
positive direction
aud titat of
in thc ncgativc direction.
Wo aro now in a position to trace the
consquences of an
initial disturhancc which is confincd to a finite
portion of tho
jncdimn, e. bctwcon .x= c( jmds-=~, ont.side wldch t)ic mdium
is at rcst and at its norma) (icnsity, so that thc values ofjP
and
arc ci ]o~
]!ch valuc uf P propagatus itsuif in turn to thc dcmcntsof nui.) whicitliu in front fit, and
cad)Yah)of~tot]fosc
that lit: Lcitindit.
Thuimniorlimitof
the rgion In-\vhi<jhPIa
variabfc, viz. thu p)acc witcro
first attains t)tc constant value
ft log~, cotnes into contact rirst \Yit)i thc variahfc va)ucs of
Q, and
a Yariah!c' velocity. At a ddinito timu,
movcsaccurdingiywith
rc()uiri))~ fur ils dL-tL-rmination :t Sf~utioa ofthu din'urcutiat quations, t))c )tin()cr (L'ft ham!) iinut of thc rL-gion throug)) whicii
varies, tuccts thu ttindcr (rig)it iiand) iinut of th rgion throuTh
wtti(di~ varies, afterwhic)) th t\vo rgions scparatet!)em.se!vcs
and indudu hchYCL'ntitem a portion uf ihtid in its
etputihrium
condttion, as appears from th tact t])at th vatucs of~and
are
hot]j (tlugp..
in ttie positive ~avo () lias th constant value
M log~,
C) so that !~= et log~ as ill () 251
r~

in the ngative wave

Att)napnintno f'rrf))' sccms

to haYO
cr~pt into Diem~uu' work,winchiB corrcetLdiu tlie t~tract uf the 7'u/<t/t<' ~< ~t'A-.

253.]

POISSON'S

INTEGRAL.

41

P bas tlie samc constant value, giving as tlic relation botwcen u

and p, M=
Since in cach progressive wavc, wben isolog
Po
latcd, a law prcvaits connccting the quantitics M and p, we sec
that in the positive wavc f~t vanishos with
and in tlie ngative
wavc ~<tvanishcs with <
Tims from (5) wc k-.arn that in a
positive, progressive wave
vanishcs, if thc incroncnts of a? and
t bu sucii as to satisty thu quation <~ ()t+ f~= 0, froni winch
Poissun's intgrt Immodiatuly foltows.
It wou)d Icad us too far to foUoweut titc ana)ytical
dcvclopmcnt of Kiemaun's method, for which the readcr must Le refet-rcd
to tho original mcmoir; but it would bc ijnpropcr to pass over iu
sUcucGan en-or on the suhject of discontinuous motion into which
Rionann and otiier writcrs iiave faUcn. It bas bcen bc!d that a
statc of motion is possible in wbich thc nuid is dividcd into two
parts by a surface of discontinuity propagating itscif with constant
vclucity, ail t)~ nuid on onc sidc of the surface of discontinuity
hcmg in onc unifonn condition as to density and vclocity, and on
the othcr sidu in a second unifonn condition in th sanm
respects.
Now, if this motion were possible, a motion of thc same kind
in which thc surface of discontinuity is at l'est would a)so ho
possible, as we may sec by supposing a vclocity cqua! and
opuosite to that with wl)i(.-hthe surface of diseontinuity at nrst
mo\-c.s,to bc impres.sed upun tite whoie mass ofnuid. In ordcr to
nnd thc relations that must subsist betwcen th
velocity and
density on t]ic onc sidc (~, ~,) and t))e velocity and density on the
othe)-sidc (M,,p.J, wc notice in thc nrst place that by tho
principto
of conservation of matter p,=p,?<
Again, if wc considcr tho
momentum ofa siicc bounded by par:d!el pianosand including the
surface of (U.scontinuity,we sce t))at the momcntum Jeaving th
stice in thu unit of time is for cach unit of area
(/J.=ptM,)~,
wbiie thc momentum cnto-ing it is p,!< T)ie dincrence of'm~
moitum mnst t)Cba)anced by the pressures acting at thc boumhu-ies
of thc slice, so titat

Th motion thus dctct'uuueJ is, howevcr, not possible; it satisfics

42

EXPERIMENTAL

[253.

DETERMINATIONS

tlie
con()it,io))s<~n~ss :md motnottum, but it viutatus thc
cututitiou of cnc~y
~) cxpru.sHudby tiic c~xatioa
i'I.fs fu-~mcnt ha.s b~-ca ah-~dygivo.i In anothcr for.n
in 2.~
whicit wuu)d atonejus) ify us in rL.JMt.n~ t!.ca.ss..mod
rnoti.-n.sinco
it a]))ic:u'.s(.[)~tno stcady motion is
po.ssibfccxcoptundcr theiawof
Fmm ~~u.-Ltiun(.S) of t).t .s~ctiun wu
dcn.sity D.crc dctcrxnnc.t.
can fuxt wliat unp)-('ss~) furcc.s woui.t be
nccc.ssary (.) nia:).t:))i t)jo
jnuLio.. (tcHned ).y (7). It f.p].c:)r.sthat th.; force
.Y, tho.~h c~n~))0() to tllc pkcc ut'
~cuuti.mify, i.s ))):u)u up of two ~rL.s of
app.).sitc si~ns, sincu Ly (7) !t p!t.s.s.
t.),u y:,),)u
~'h..
whoc moving
!u.d tiji.s cxphun.s i.ow
fu)~,vi~p~v;u)i.s].c-.s,
it is t))at tho con<)itio)) n..)~ti))~ to muj.K'ntum i.s
satis~! by (7),
thoug]) thc iu)-co A' bu i~nurL'd :dt.o~ht.-t-.
25~. Tho cx~t cxp~ritncnta] .~crnnxation
nf tlte VL-toci~y
of sound i.s a m:~u.- of~ro:Ltcr
di<)~u!t.y ti.:u. nn~ht i.:Lvc hcL.n
cxpectcd. OL.sc.-v;K.iun.sin thc u]K-.i .nr arc ti~Hc to crror.s f.ou
thc ctt'uct.s of wind, am] frum
nnccrtfunty wi~i rc.spcct to th
exact eon<titio;t of thc ntmo.sphcrcasto
tonpcnt.tureand drync.s.s.
On thc uthci- Jtand wht~i sunnd is
p)-pa~!)tcd throug]i air cc.~
tained in pipes, disturhance ~risos fron fricLimi :md fr.nn
tt-iu).sf.;r
of Iicat; and, aithou~h in) ~t'cat o-rot-s frmu tht-s .sources
arc
tu bo fearcd i)i thu case of Luhu.s of con.sidcrabic
di:u))ut(.'r f,m;h
:ts Hutnc of thosu ctnptoycd Ly R~xatdt, it is dimcuit
to f(;ej
sure that tt~c iduat ptanc wavcs of
tt~-ory aru iicarty cno~h
rcalixcd.
T)~c foDuwin, Table' nontain.s a list of the
principe
meutai dutermiim.tiona \iuc)t )mvo bcen madc hithcrto.
Nnmos of Observera.

cxpcri-

v.
of
<J"Cuut.i)iMutruH.

AcadomcdesScicnccs (1738).
Lcnxcubcrg(lMlJ)

(33~3
f~33'7
s~i
3~o.('
3~')")

(~l()ing-ham(182l)
~ur~u des Longitudes (1822)
MuHandv~nBbdk
'Lusanrjnet,U~.Apri),

1877.

0);' TilK

254.j

VELOCITY 0F

KaincsufObsorvM~.
St.'U))j'J'c)'andA!yt'b;K'k.
;)h-!LV:Lisan<[AhtrL)ns(lS4-).).
W<jrt.hcnn
~tunu (IM71)
Lui~ux.
R~-)t:mJtt

SOUND.

4:3

VctocityofSnuudu.t
U"C(jut.u)Mct.run.
M2~
H~l'C
:4
:j~()'7
3!:3()'7

In Stonc's cxpo'nnGnts' thc course ovcr ~']nch thc Sound

wa.stitnud connnunccdatib
distance of (!-t()fcetfr<)nth<so))rc(sot)tat :niyetTorsarisIt~
fi'oa uxccssi.ve disLurbancu werc to
a ~rcat uxtunt avuidcd.
A mcthod

hasbcoi
proposcd by BusHc)):rf"t- d(jteniu!)in~
thc vctocity ut'sound wit)K~tt t!<c nsu oi'gruat distances. It
car is!).b)c to dcide
dpends ~pu))thc!prccisiotiwiU)w)nc]tt)iu
~')~(jt.))ernh()rttiekHar('sinndt:U)eouH,ornot.
InKonig''s"f()rni<jf
thc cx])cnincnt, two sn)!LHc)ectt'o-]nagnettc countcrs arc controllcd
Ly :).foj-k-intCD-uptcr ( C~), who.sc pcriod is ono-toith of n. scconf~
and givc syncftnmous t.i('.k.sof thc ~atnc pcrtod. \V!)en thc
conntcrs :u'u c)o.sc to~ctitur thu mtdibtc tid(s eoincido, but as otic
c'onxtur is gmdmdiy rcittoved from t])(i on.)',th two scrics of tic)<s
f;d!.!Lsuud<r. AVficnt))u dii!'cruncc of distances is ~bout 3-)<jnctres 1
coincidc-ncc a~ain takcs place, proviog Diat ~-t mctt'es is about
thc distance travcrscd by sound iu a tcntli part of~, second.
'J"/tt;M.187~p.l.
~7'<w/)));.xcfj. 1~.1851.
~tM~.CXVIU.UlU.lHC~.

CHAPTER

XII.

VIBRATIONSIN' TUi~ES.
255. Wn hve an-e~dy( 245) considcrcd tne solution of our
fundamcntal c'pm.t.ion,whcn tiie yciocity-potuntia. in au utilimitcd
Huid, is a.fuuction of onc spaec co-urdmate on]y. Ja thu absunco
uf n'ictioti no ctm.ugcwoutd be ciUt.scdby tlie introduction of a.ny
munber of iixed cylituh'icit.Isurfaces, w)ioso geucrating lincs a.re
parallul to thc eo-ordumt.clu questiou for evcn whmi t)ie Hurfa.cca
arc absent thu ihud ba-sno toidcncy to move across tbeni. Ifons
surfu.CLj.s
of t)ic cyHtKh'ica.1
bc ctosed (in rc'.spcctto its transverse
section), wc ftavc tbu impot-t:uitprubicm of tlic axitd motion of air
wltinn :).cylindricat pipe, whici), wncn once tho tnechanical conditions at thc cnd.s arc givcn, is iudcpcudcnt of anything that may
happcn outsidc thu pipe.
Considerin~ a simple harmonie vibration, wc kuow ( 2-t5)
that, if <~varies as (~

of winch hnniy on;y ti)Crcn.1parts will bc rct.:uncd. Tho first

f~rm will bc must (.-uuvcuiclitwhcu t!ie vibration is tationm'y,or

255.]

HARMONIC

WAVES IN

ONE DIMENSION.

45

nc:u-!y so, and the second whcu thn motion rcduces itself to
pn.stLive,or ncg.~ivc, progressive undulation. Th const:).nts
and m tlic syinLiiciL!sotution
may bc co]np!px, a.nd thus the
<in:d expression in ternis of rcfd
qufmtit.ics wi)I invoive/arbltrary cousta.nts. If wo wish to use re.d ()H!UttHicsthroughout, we
must takc

but thc fmatytic~] work wou]d

gt:.])er!)Hybe longer. W])cn no
:unhi~uity e:).n anse, we sl)a)) sotnetin~s for thc s~hc of
brevity
d)-up,urr<st<n'c, tbo t~torinvolvmg t.)ie tuncwithout oxpt-c.ss
mention. E(~u:ition.ssuch as () are of course
cqu:t))y truc whethcr
thc f~ctorbe undcrstood or uot.
Taking thc Urst form in (3), wc h~vo

If therc bc any point at which citLcr or

is pcrmancnUyzcro,
<,&
thc mtio
7~must bc rc~, ~nd thcn t)ic vibration is
~i'OMury,
t))at i.s,thc samc in p])n.sc~t fdt points
sinmttiLncousfy.
Let us suppose t)tat H)crGis a nodc at th
origin. Thcn when
f/d)
thc condition of which is 23= 0. Titus
~cs,

From thcsc quations we tcarn that

vanis!)Gswhcrever
(!.C
G
sn] ~:=0; th~t is, th~t. hsites thf: ori~in t.hcrc are nodes at tho
points .B=~?~X.,M bcm~any positive or ucgit.tivc:intcgcr. At any
of thcsc piaccs inHnite)y thiu rigid
ptanc b:u'ricrs normid to a;
mt~ht be stretched across the tube without in any way alter-

NODES AND LOOPS.

4G

[255.

Ing thc motion. Midw~y hct.wccn p:~h pnir of conRccutive nfx~'s

thm'u is :). /w/), o)' p)!t.e(.'of no prcssm'c v:)ri;)tion, sincc ~) =~
c'tuttnihiciLtion '\Y)L))tho
ALmy ofthc'sc h~ps~
(()):2't"t-.
cxtcrna) ntmuspiicrc tnight, ht.' npencd, wittmut c:m.sin~ :).))y()ist:)))'hfmcecftLc'tnotKjnJ'romit.it'passing'inoront.
'J~)u)nopsa)'cthc
maximum
ptitccsof maximum V(.')"cit,y,n))(l t)tc nodestitthscof
pressure v:u'i:).tiun. t iutcrv:ds of
uvL'ryUtit)~ is cx:)ct)y rcpcaLud.
If thcrc Le :t. undc n.t.T=~ ns wH as a.tthc nrigin, si))K~=(),
wht.')'c )~ isit, posit.ivc intL'~r. Thc ~')';L\'usttono
nr .=~
~'))ichc!H]bcs<'un(]('d~)y:),ircunt:LinL'di)t!Ld')n))tyc)nm;d
pipe
of [cngLh i.s th<)'~f'o)'(;thut winc)) h:ts :L 't.Vt'u))g't.)t op):~ tu
TJii.sstitt.umott, itwiilhu
observa), )t(')').s~'()J\]):~<vc'rbctJ)c
g:).swlthwhic))t)tcpipnisfi))cd;
h))t<~cin~)))<tcy,o)'th(;
place
)' thL! tune' i)t t)tu nm.sio:).!Hc:()c, dcpL'thts n.).soon thc nature; uf
t))u p:),rti<jutarg:t.s. Th<! pernjdic tiuK! is givun by

Thcothci'ton's
pns.sib)c foradoubiyclonu()pi])c
i~vcporiods
whic)t :u'< sub)))))tti))iL's of t]):Ltof thu gr:LVcst t'~tc, :Utd thc whotc
sy.st,('nifo]'msah:n'tnu)ticsc:t)c.
toinLutttsnnwHUpjto.sc, witLout, htnppm~forthcnn~ncnt.
f[nn't.;howsn(;)t:t. condition ()fthin~sc':U)~t''sc'f'u)'(.t)):)tt))C)'ciH
1. )'u:t<i<n)((i)~ives
:).It)opinstt.K)ot':).])<)()c!tttnej"~mt~=/.
cn.s/< = 0, whc'nco = -)-~ (2/~ + J), whf'rc ?)!is zero <')':), pusit~iv'
t~his c:).se thc ~'t'nvc.st to))(; hit.s w:n'c-tu))~L)t oq'):)l
iutc'gcr..)n
thu nodctu
lof<'n]'ti)ncst))c]t')in't)tr)fthcpipci'u('k')))''(It't'<))<)
t))o Joup, in]ttthcf)t))c['t~nc.sf())'tnwit.hit,:L!):n'n)0)tic.sc:L)(.ft'om
w))ic)), huwuvcr, :d) thc nicnibcr.s ufevoi untcr :u'e ini.s.si))~
25f!.

By incans of a ri~ift han'icr titcre is no (tifUculty in

but. thc condition
scou'i))~ a ]t<"1'tt:uiy<k~i!'u<)po!nt()f:),tub<
for a h'op, i.c. t!)!tt mxtfr uo circumstaucus sha)) tl)c pt'c.ssurc va.ry,
ci\n on!y bG rcit.lixct) n.pproxi)natu)y. In inost cases tbc vnriatiou
oi' press)u'e at n.uy point of n. pipe mny Le nuide stnal) by aHowinn'
a h'ce coinntuuicution with t))G externiLl ah'. T)ms Eu)cr und
L:).grangc nssumod constn.ncy of pressure n,s t!)C condition to be
WeshilUaftcrwn.rdsrcturn
satisficd.rLtthccndofanopcnpipc.
to the pro1))cnt of thc open pipe, aad investi~a.tc hy n. rigorous

25G.] 1

CONDITION FOR AN OFE~

END.

47

proccss thc conditions to bc satisficd a.t thc en(). For our imtnctiiato purpnse it will 1w su~cicut to kuow, w)):)t is imIcL-d
t<)!c)-;i.b)yobvi.jus,t))!).ttlle ope;) cndofft.pijtCjn~y
i~ L'uat.d~
:).f""p,if thc diiunutcruf thc pipe bc n<~)<ctcd iKcotnpitrison
W)th<.))c~tvc-ju~t,]t,p)'uvi(tedthucxtct')]:Ltp)-c.sm'i~t.)t(.!]tc!r'-hLu).)t'))U(t(!<jft))copeitC))<)henotitsu]i'v:n'i:djJ(jf)'<))ns())nc
cause
J!'t)<-p(;)i<)(.-ntofth'))i[)ti~n\vitltinthc!
pipe. Wi~nthcruisfm
m<))t.'n()c)ttH<)U)'cc(~'sunnd,t))c pressurent the c))duf'(.])c pine!
).stho~n)0f).sitwou)d
)'t.'])tt))(;Ha.n)cp)acc',it't))Gp)pewcre
Th
!t\ny.
in)pC()i)))C))tto.secun))~t))nfnHn))t(;!)tot').))(_! conditio)i
i"r a )~<)))~t:),ny(]('sircd point fies in th
inertie ofth(jn):K;])iuc!y
)'r(pti)'n()t<).su.stai))t.h('p)-(.!H.su)'(i.FurthcorL'tif.-idptn'jto.st.'i-iwetnay
('vc-r]<)t~t])isdif!i('.u!ty,!UKU))t.!)~inua. 'nasstc.s.spintonh~kc'dhy
a conp~Hscd sn)-it)~a!so withuuttnaM.s.
T!)C a.s.sumption ofa,
foopatanopcncndofapipc
ji.stant.a)nonntto)ic")cc[.in"'t)tc
mc'rtiaofthcoutHidcnir.
Wcha\-csc'('nt))a<ifa))odccxi.st:)tnnypninLoF

apinc

tL(.-runn)stheasurk.s,r:ni~date(jU!di))t(.'rv!L)s~t.ttniid\vay
L<t~.ju)c:u.:)tpairoff-o))H~c))t.i\'tj; )~)dc.sth('rc)i)t).st])(;ah)op,
a)t(t
t)tatthcwho!cvibratu)a))msth(.sta),i(')ia)-y.
l''i)C.samoc(i)u])).siu;t
f~')].)wsif thercbcatanypuiut
a )<)op; hutit
)nay])L-rf'L'('tiywc]l
thattho-o
arc
n~ititcr Utiles norioop.s, as
h:t,pcn
jt)rt..x:mii))c in
thL! case whcutitu motion i'(j()))~s to a positive or
~cg-ativu prowavc.
gressive
In.st.atiu)):u-yvi))rati(jnthcrci.s)]ntra)).sfc)-(;nccof
('nL!r~y atun~ t])c tube in ei[,)tcr direction, for enurgy canjtot pass
anodt.'ora.loop.
2;'i7. Thc !'r-]a.ti(~)S
bctwfCt) thc ]c)i~'ths nf an nnen or d~cd
pii)G attd t))c Avavc-icn~ths of (,])c itictudctt (-()]u~]i of aurnny ~so
l'c invostigato] hy it.Jtowi])~ t])c inotioa nf n,
~?/7. by ~1,ie]) is
umkTstuod a wavc c<~)(i)tc<!within n;UTow )nnits and
compo.scd
of uniformiy cun<I<;))su<)
nr r!U'cfi<j<!
finie). Jn ]<x)hiu~ at th ])):),t,t~r
i'rom this point of vi<j\vit i.s nccc.ssfu-y to takc intoncunt e:u'ci'nily t]io cil-curn.sta.xccsnndcr Avhic])tftc v:u'iuus t'cftuct.ious takc
])):tcc. Lc't us Ht-st sttppo.se tiiat
cundutt.sct! pu]sc tr.ivcis in tho
positive dirccbion to~u-ds a b;(.n-icr fixct! acro.ss thc tuLc. Sincc
thc CDO-gycouta.mct) in th \va.vc ca)ni"t
cscapc h'om t!tc t)tbc
thcrc must bc n. rcficctcd wave, ami thiT.t this rcftcctcd wave is
a).so n. w:i.vcof condcnsntion nppcar.s from t)tc fact th:).t tl)crc is
no
!oss uf ttuid. Thc same conclusion
may bo arnvcd at in another
way. 'J'hc cfuet of thc hu'ricr may bc inutatcd by thc introduc-

48

REELECTION

AT AN OPEN

END.

[257.

tton of a.sinn)ar and cquidistant wavc of condcnsat!on moving in

thc ngative direction. ~nc th two wavos are bot)) condenscd
ai'td J.C):Op;!g~i.~(.iUt<<{t'!i.y <!i!'<'C!.io~!<<'(.'h~'it.i~sut tho
ibtid composing tlicm arc C(jna)and opposite,and iherutbrc neutralise one another wltoi tlic wavcs arc snpcrposcd.
If thc progrcss of t])C ngative rcf)cctcd wavc bc intcn'nptcd
t)y a second barricr, a.shnilar rcitectiou t:).kcsp)acc, and th ~'avu,
still ronainin~ (;ondui)S(;<1,
regains i(.spositive chiu'af'tur. Whcn a
(tistancc bas bncu tt'avc))('dC()ual to twic<jt)ic ]L'))gtbf'f tbc pipe,
tbo ori~)n:d statc of thii~s i.s cotnp)ctt;)y rcstorcd, ?u)(!thc snmo
that
cycle ofcvcuts rf'pcats itsc)fin(h'finitdy. Wc Ica)'~ t]n!r('<'f)'c
th pcriud within a <)')<th)y
doscd pi])<jis t))c tinic! cccxpiud by a
pu]sc in travdHng twiec t)tc Icngtl) ofthc pipe.
Thc case of an opcn end is Homc'whatdiffrent,. Thc suppicJncntary Jtf'~ativciwav(;nuccssary to imitatc thc cHuct of thc <~)Cti
end umst cvid(;nt]y hc a wavuof rarcfaction c:tpahtcuf ncntral~in~
th positive pressure of tbc eondcnscd primary wavc, and t)n)s in
the act (jf rcf)(jctiona wavc chao~c.sit.sc!):trac).('rfro)n eonthjuscd
to rarc'ticd,or fnou rarcfx.'d to c<n)d(.')).scd.Anotbur way of con8id(.-i-ingthc)n:ttt.('r is to observe that in a positive condeuscd
pu)nc thc momcntum ofthu motion is forw:u'<)s,am) in thc
absence of th neR~ssary forces cannot bc changL'dby th rcrtcction. But forward motion in ti)c rc-uccl~t ngative wavc is
indis.so!ub)yconnected with ttic rarcficd couditioti.
bar'kWhcn both ends of a tube arc open, a puise tr.T.vc'Ding
wards and forwardswithin it is compiL't'dyrcstor<d to its original
sta-tc aftcr travo-sing twicc tbc ]cngt)i of tbc tube, .sufft'rixgin th
proecss two rcucctions, and t))us tbc rctation bct\vcen Icngth and
period is thc sarnc as in th case of a tubu, whosc bnds aro both
closed but whcn onc end of a tube is opt'n and thc othcr cloncd,
a. double passage is not Huniciunt to c)osc thu cyctc of chaxgc's.
Th origif~a)con<)e))S('d
or rarencd ch!tract<;reatmot bc rccovcrcd
until aftor two )'(.'H(j(;tionH
from thc opcn cm!, and aoco)'')i))g)yin
thc case contomp)at(jdt))c p~'riodis t)n; tinic n'fptircd by thu puise
to travcl overyu; titnc.s thc k'ngth oft)'c pipe.
258. Afiur t1)c fid) discussion of thc rorrcsponding prnb]cms
In thc chaptcr on Strings, it wiii not ht) m'ccssary to say mucit on
thc cornpoundvibrations ofco]umnsof air. Asasimp)ccxat)tp)o
we may take tl)c case of a pipe opeu at ono end aud ctosedat thc

258.]

PRODLEM.

49

othe)-,whichisM)(Mcn]yhrougLttorcstn,tthctimc
=0,-<('t(jr
Lcing for sonc tune I]) motion vi).h a uniform velocitypand)~! te
its ieu~t.)). Th Initia! statc of t)te cont:uncd air is tiK-ti onc 0!'
unif'ornivducity p:u-nHu]to .-K,
{mdoffrocdom from compression
:ui(! r:u'<jfactic)i. If wc suppose t)):Ltthc
ori~iu i.s at tl~e clu.sett
un<),Du!gf;ne]-it.I
sofutiuuis by (7) 25;'j,

con.st:).nt,.s.

thccocfnS!ncc~!st<)LcKcroirnbi;).])yfor:tnv;L)u(-Sfjf~
cio.t.s
nnt.st vanish; thccocfHcient.s arc to Le (IcLcTinl)icd
t'y thc condition that for~H vidu~ of.~ 'bctwccn 0 :m<]
~'ho-c Lhc.sum)n:tLioucxtcnds
vaincs of?' frotu
tc~fintc~-al
to . Thc ~cte)'mi))!itinn of thc cocfficiotts jd fron
(2) is
(-fbctc<) int).c usu:dway.
Muitipfying hy sin~d
intgra t)ng fron 0 tn we gct

Jn U)c
cascofatubestnp])(.<]att.)tcorigi)tan(]opcnat
.7;= l, lut ~= cos ?i<
Lh tlie value of t),c potcnt.ia):tt thc
open en.)1
(h~ to !Ut cxt.<nfL) .scmrofso.nxL
Dcf.onnit.i.~ r and 6' ili
<)!ahon(7)25~, \vff!)t.)

Itappcar.sthatt)~ vibration within th tube is a

minimum
whencos~=~l,
thati.swhe.~i.samu]tip]oof.inwhk.hca.sc
Ho
4

50

FORCEDVIBRATION.

[259.

thcrc Is a nodc at a:=

Whcn Is an odt} multiph of ~X, cos/<~
vanishes, and thcn aceording to (1) the motion wcuid bccomo
n.f.
lu f.)ns (w;)' the sujtpstt,icn thut Lhc pressure at thc
openend is in(tcpendcntofw!)!).t h.q)pcuswithin th tube bruaks
down; and we c:m oniy mf'cr that the vibration is vo-y large;,in
eouscqueucu of thc isoc!n-o))i.sm.Sincc thcre is a. ]ode at .r=0,
thurc must be a loop w!icn is an odd
and wo
!nultip]c of
concludc that in thc ca.sc of isochrouismt)ic variaLiotiof
pressure
at the opcn end of thc tube duc to the externe! cause is
cxact)y
ncutr~Ised by thc variation of pressure du<;to th motion within
the tube itsdf. Jf thcrc were rca))y itt thc
opon end a variation
of pressure on thc whoc, th motion Must increasc without limit
tn thc absence of dissipativc forces.
If wo .suppose that the ori~in Is a loop instead of a
node, H)c
solution is

whc)'c<~)=eos~ is thc givcn vainc of at tlie open end .~=~.

Iti this case thc expression bccomes ItiHt-nte,\v)tc]i
/c~=~t7r,or
= ?~X..
Wo will ncxt considcr tho case of a tube, whosoends arc both
opcn and cxposcd to (!i.stut-))ancesoft.hc saxic penod,n)nkin"-6
cqual to 7/c'"<,7~ rc.speetivdy. Un)css -t.hcdist.urbanccsat t))C
ends arc in thc s:une phase, one at Icast of thc coc-mcients7f, 7~
must be complex.
Taking the nrst form in (3) 255, wc Iiave as th gcnera!
expression for
If wc take thc o'Igin in thc midd]e ofthc 'tube, and assume that,
t.hc vatucs /e"
7~c'"<con-ospond rcspcci.ivcly to ~=~, a;=- l,
wc ~nt to dtermine ~1 a.nd 7~,

259.]

DOTH ENDS

OPEN.

51

This resutt n..ght .Iso bo dcduccd from

(2), if we con.sidert!h.t
thc rc.tun.Gd motion ariscs fro,nt).o
superposition of thc motion
winch 13<h,cto t)~ disturbance
/ca!cul.tcd
on th hypnthosis
t'~Lt tl.co~or end ~=-~ is a
!oop, on thc motion, which is
duc to A c'"<on ttic Ilypothesis that th end .e=
~is ]oop.
Th vibration cxprc.s.scd
by (.t) c~nnot bc ~y,
un!c.s.sthc
raho
7. be r~t, th.~t is unicss tho di.stu.-b~necsat
t)~ ends bc
in s.nnhu, or ni opposite,
pba-scs. Hcncc, cxccpt in t).e cases
rcscrved, therc is no loop anywftcrc, ~nd thcrcforc no
pf.~c nt
a bmnch t.ibc ca.i bc cunncctcd
along which sou.Klwill not
~h.ch
L'opropi~ated'.1.
At th Tniddicofthc tube, for which ?=
0,

th variation of pressure
(proportion~ to vanishcs
~inr,that
yy +A =(), tj~t j~ ;(di.stm.b:tncMn,t thc ohi.s bc cqu~l un.!
if
"i
phases. U.i)css this condition bc .s~Is~d, t)~
~o.s~c
expression bccomcs infini(,o, wiicu
2/ = (2~ + 1)
At a point disant
from tlie middte of tho tube t!)o
express) on for is

v~nisnng wh,n 7/=7f, that is, wlicn th di~turbanceaat the end.s

cq.ud ,,nd ni thc ~7~ phase. In
gcncnj Lccomcsinfiuitc,
w'icn s.n~=(), oi-2/=~.
If

~t

P~arc
w.H ~e

onc
duc

end

of an

to an

cxternal

i.s expre~d

Ly

un]i,uitcd

tube

thcre

Le a variation

of

a train
of progrc.s.sivc
source
wavu.s
inward.s
pror.~atc<)
from that
end.
Thns, if th Jcn~th
t)ic tube Mcasnrcd
froiu t!,e
t).e velocityopun end bo

Potential

~=co.s(~).
~"J'

corrcsponding

tu

of

(Phil.
co.np.nn~ the ~cn.itic.
of .ourcc.s of .onnd of the ,~nu
of tho
iH
th .~u,.ce. to bc
v
the
nntil
-'1"
"ibr.t.0..
't~
"?'
~
'
"tric
c.su)c
a(.c 'of
~y'"R
point of J""cti.n tho dist.n.h.
this al,

tho
test it nppcars tL.t
""M
nsbmnptton is 110tthooretic)t])y cornet.
I.itcb
hitcli.

E~

43

l-'ORCHD VIBRATION

52

0F

PISTON.

[25~.

~)=cos7~at~=0;sut)):)t,ii'thcc!tusRoftLL'di.stm1j:))]cc\YitJtiH
thc tubu bc t))c p:ts.s:(~(jof :t.train of' progressive WiLVus
aen'.ss(.hu
opcitcnd, thci)i.tt'nsity\YiUnnt]iL'tu1juwinb(;thcs:uneii.s]nt))~
space outsidc. It nmst nut bc for~'Lt.cuthfLt.th di~tneto' of thu
tnL(iIs.s))ppos(;(t to bI)tfihitc)ysm~niucu)ttu:t.riHunwit))t)tL:
k'))gt))of:Lwavc.
Lct us ncxt suppose thnt thc sonrec of t))Cmotion is within thc
tnl)uitsc)f,'h)u ~(.-xinnpteto thc inexorabletnotiunof;). piston
~t the origin'. Tho constants in (~) ~5.'):).)'cto hc detenninctl
thc COI1<
couditions
ttIons tt!ia.tw!)uu
lat Wlen ~,c=
cos ?lt (s~y),
1.)ytlie
by
,t;=0,0, ~'=cos?)<
th:Lt,
.111(l and th:lt,
.IJ
il:lJ
T)ms ~=-t:m~,
-hcn ~=~, ~=0.
~7~=1, :nid thc ext))'<rjt)f<)t'f~i';

T))c

motion

i.s

a. )nini]num,whcncosA~=i!,t1)at

is,

whcnttie

!t.))<'ft))CtnbHis.'(.!nu)tipIcuf~

Wi'cn

is

ah odd mu)tij))c uf
t]ic place ocenpicd hy thc
pi.st()nw()n!()buano<.k',ift)Ki<)pcncu(twcrci-ea))yaIo()p,hnti)t
ttti.sc~sctiics~ntiun
i'ai)s. T)tc cscapenfoncr~y iront t))c tuhc
p)-C!VO)tst]iccncr~y frum accumulatmg beyond a ccrt~m puint-;
but no account [-:ui Le tiL~oi of <)ns so !on~ as UtC opoi end is
trca.tud ri~orousiyasfLioop.
Wcshidt
rc'sutne t))u question n)'
rsonance at'Lct-wc hve conHido-cd in i~)\d<;r (tt.'tail thu thoury ot'
t))(- opcn end, whcn wc shatt bu ab)~ U~duid with it )nor(j satisfactoj-ijy.
j!i )i)<(-mnnnc')- if th point ~'== bc
thc cxprcHsion for is

a nodc, instcad of :),h.-op.

~t)"'st.hc)))ut)<)nI.s:t.mi)ii)nu)n\v!tt;n~isnn(K]d)n))!t.i[))c;of~,
inwhichca.set.iicori~in
i.siUoop. Whcun.snn
cvcmnuttipicut'
.t))(;o)-i~i)i.st)C)u)dbc:).
notk', wh~h isrurbid.tcubyUtucotif))Int)tis e:Lsc~ecor()it)gto(~)thc
tiotisot't.hc'.jucstiun.
motion
Luc()H)C.sin()nit.c,w!ncIt)t)c:)n.stt)atinthc!d)S(;'ucoofdis.sipativc
furcus thu vibmtiuti wou)d Incrc:~o witituut limit.
'TitMcpruUc)tLsarucotMiJct'cdbyl'ui.ssu)),.VJw.<)).<t.n.t'.M)j.

2GO.J

KUNDT'S
EXPERIMENTS.

5:3

2(!0. Tbc cxpcrintcnta.! Investigation of acria) wavcs within

ptpus ))as bucn e~'cctot with nsi(iorab!c .success by M. Kundt'.
To gcncratc \av(;s is o:L.sycnougb; but it is not so
ca.sy to iuveut a
tnctbud by wbich thc.y can bu cf~ctnatiy examina.
J\[. Knndt
(H.scttv~rcdt));Ltt)tc uu(!u.s of.st:Lti(juary \v:L\'c'se:ui b<j tnadc cvidcnt
by ttx.st. A iitt!u iinc s:).nj r iycoptxtimn .sL'(j(t,.shakot over t)jc
ititcriur oi' ;). ~ss tube cotitaiumg- a
v!b)-iLt.i)t~cobnnn of :).ir,
(H.s~u.su.sitscti' in t-ccun'ij)~-p;Ltt.urns, by means fi' wbich it is e~sy
to dut~nninu t)i(; po.sit.i<)t).s
of tbe Dottcs amt to measm'c th
it)t(;)-L)s ))ctw<x-n t))0)). h) Kun(]t'.s cxp~'itncnts t)te
ori'nn of
(bu sumid w!~ iu thu iut~ItmUmtt vibnLtion of
~giass tttbc~dbjd
tbu sounding-tubt. :u)d tbu (tust-fi~u)-cs were furfncd
i]La.scco!id
and far~r tube, c.d)cd t)tc w~vc-tubu, t))u fattL-r
bctng provutcd
\(!) :t !U()VC:).b)u
.stoppur fur tbn purposc uf adju.sting Itsicnc.'t)).
Tbu otbo- end of tLu wave-tubc was fittcd witit a cork t])ruu")t
wbicb t)tu Muun()i))~-tabc pa~s~d h:df w.y.
By snitab!o fncti~i
t)~ .soondm~-tubc wa.s can.scd to vibratc in its
~ra.vc.st mode, so
that tbo ntral jn.'ittt \vas nod:J, andit.s ititeriur
extremity (doscd
Avithn. cork) cxcitud act'h'd vibratiffn.s in titc Avave-tubc.
Bymcans
of tbe stojtpfjr Un- k'n~tb of tbc coftunn ofair c-oubi bu
adjustcd so
as to ui~ku t)t(i \'i))r)Ltio))sas vi~jroos as
po.s.sib)c-,whicit bappcns
-~bcn tijc intorvat bctwcan tbc stopper and t)te end of
tbc
sounding-tubc is a. iuu]tiptc of hati' t)ic wavc-!c)]gt!i of th
.Sound.

Witht)tis ~ppamtus Knndt wa.s abio to

compare thc w~veJungtb.s uf t))c same sound in various gasc.s, ft-om wbicb t!)e rc-h).tivu vcbcitius cf propagation are at once dcducibic, but tbo
rcstdts
wct-o not cntirc-Iy satisfactory. It was found that tbe Intcrvfds
of i-ccurrencc of tlie dnst-patteDi; were !iot
strictly cqua], and,
what wa,s worse, that titu pitch of t.be sound Dot
constant
ft'om onc cxpL'rinieut to another. Thc.se dcfeets wcrc traced
to a
communication of motion to th waye-tube
tbrougb the cork, by
w))ich thc dn.st-f]gures were di.sturbcd, and t!]C
pitch mad in-c~uur
)n consquence uf unavoidab!c variations in tbc mountinn- of
tbc
apparatu.s. To cbviatc tllem, Kundt replaccd tbc cork, wlnc)i
funned too stiff~ conncction betwecn the
tubes, by layers ofsbcet
iodiarubber tied runnf) with sitk,
in this way a ncxible
obtai)~
and pcrfectfy air-tight joint; and in ordcr to avoid
any risk of tbc
coniparison of wave-Jengtbs bcing vitiatcd by an altcration
ofpitcb,
7'f~. ~otf. t. cxx. p. 3:)7. 1H(!8.

54

KUNDT'S EXPERLMENTS.

FSGO.

tho a.pp!U-atuswas modined so as to makc it possible to excite

thn t.w.'sysh~ns of dus'-f~tros :i::u)'n!u;!y
:t)! u) r~p~~c'. tu
thc .jiu'tc souod. A coitatcraJ adviuttage of th ncw metl~od Cnsisted ui thc limination of temporature-corrections.
lu thc improvcd "Double App{t)-iit~s"th
sounding-tube was
cau.scd tu vibratc
its seco?;(/~;o~ by friction appticd nGar
th )nid()ie; n-ndthu.s tl~ nodcs wun- fonnc<tat thc points di.st:uit
from th cn(is by ottc-fom-t.hof th Jength of thc tube. At cac)t
of thcsc points conncetion wns madc wit]t an
imh'pcndcnt wa.vclubc, providcd wit.h iin ndj)tstab)c stoppa', and witi) brancb tubes
aK! stop-cocks snit~bic for admittin~ t)tc varions
gasc.s to bc
cxpLTimuntcdupo)). Jt is (-vident )hat dust-tigures fonnct! in thc
two tubus corru.spondri~oron.s)y to t)te salnc
pitd), mid th~t t)~.rcforc a. compansou (~' thu intorvais of rccurrcncu iund.sto a. correct
dutcrtninatiou of th veiocitius of propi~-atiol, undur t])(!circmnHt:Luccsof tbc uxpurintcut, fur thc two ~ascs -\vit))~-)ncht!)Gtubes
arc tiHed.
'J'bc rcsnits at which K~ndt arrivcd wcrc as fo))o\ys:_
() Tho vdocity of .sound m a tubu din)i!))Hhcswith tbe
di:nnct<jr. Abovc a ertain ditunutcr, Ituwuvur,thu
citangu is uot
pL:recptibIc.
(/') Thc dntlinution of vclocity iacrcascs wit!i th \va.vcofthc totic cmp!oycd.
I(j])g-t.h
(c) Powdur, sciittcrcd in a. tube, duninisbcs the velocity of
sound in narrow tubes, but in widc oncs is without cH'cct.
(~ lu n:u-row tubes th effuct of powder iticrcases, when
it is very rniety dividud, aud is strong)yagitated in
conscfjuoice.
(c) Rou~bc-ni)~ tho intct-Ior of a uarrow tube, or
inerGasiu~

its surface, di)uini;jhcst!)u velucity.

(/) Iti widc tuhcs thse changes of vclocity arc of no importance, so that thc mcthod may be uscd in spite of thcni fur
uxact dctenninations.
(~) Tiie inHuoicc of tho intcnsity of .sound on th velocity
eannot bo prov'ud.
(/<) With thc exception of thc h-.st, thc wavodcngths of a
to!)c as shcwn by dust arc not an'uctcd thc modeof
hy
excitation.
(i) Jn wido tubes thc vdudty is indt'pcndott of pressure
Lut ia sn)a)t tubc-s thc vcluclty iucrca.suswith tlie
pressure.

2GOJ

KUNDT'SEXPERIMENTS.

55

(D AU the obscrved eimngcs hi t.hc veloc~ty worc due to

fncUon, .uK' cttpecj.Jfyto exchangc of bat betwceu th air and
thc sidcsofUto tube.
(/.) Th velocity of sound at 100' agres cxactiy witli that
given by thcory'.
We stiaU rcturn to thc question ofthc
propagation ofsound in
narrow tubes a.s ai'cctcd by thc causes mcntioncd aboy
(~'),and
s))a!) ti)cu investigate tlic formula givcn
by Hchnholtx and
Kirchlioir.
2(il. In th Gxpcrimcntsdcscnbed in thc
prcceding section the
act'ial vibrations arc j~ee(/, t))e
pitch bcing dctcrmiucd by tbc
cxt~rnal source, an() not (in any apprciable
dcgree) by tbc Icngtb
of tbe column of air. Indcud,
strictly spcaking, aU .snstain'cd
vibrattons are forccd, as it is nut in tl~c
puwer of frue vibrations
to inaintain thum.selves, cxcept in t])e idcal case whcn tbcrc is
absutntL-]yno fi-ictiun. Ncvcrthc)css tborc is an important pmct)cal distinction bctween th vibrations of a colunui of air as
excitcd by a lougitndiniLHyvibrating rod or
by a tuuing-fork, and
sucit vibrations as thosc of th
organ-pipc or cttcn)Ical))annonicon.
I)i thu latter cases t)to pitch of'tho sound
dpends prineipa)!y on
thc Icngttt of th acriat cotumn, thc function of tbc wind or
of tbc
Hanio' buing mcrc)y to rcstore t)~
cncrgy lost by friction,and uy
cummunication to thu cxtcrnal air. Tbc air in an
orgau-pipu is to
bc considcrcd as a column
swinging almost frecly, thc Jowcr end,
ncroHM
wbie]i t!.c wind swccps,bclug trcatcd
roughiy as opcn, a)i(t
thc upper end as closcd, or
open, as th case may hc. Tbus the
v-avc-Iengtb of tl.e principal tonc of a stoppcd pipe is four times
t))c tungth of thc pipe, and,
cxcept at th cxtrcmities, tbcrc is
ncjthcr nodc nor Juop. Th ovcrtoncs of th
pipe are thc ofM
bai-tnontcs,twctftb, bigl.er tbird, &c., corresponding to t)ic varionss
subdivisions of th column of air. lu thc case of tbe
twcifth, for
exan)plc, tbere is a node at tbc point of trisection Hearcst to thc
Fron.Mn~o
in tho momoiraireadyeited,fromwLicht].ouotiec
expressions
to L.Yocuntcu.phttodcon.nthotoxtis,-ri..cip,d]ydcriYc.d,
M.Juu.dtappears
.uuat.ou of Lis iu~ti~ious, butlamu~bio
to ~indM.yl~crpublication
ou
thosuLj(jet.
-'Tho

subjcct of ~nsitivc fimes witb aud without

pipos i.s tre~ted in condut.,1 by Pr.,f. Tyn.hdt in )~ ~-r-rk ou
Suund; but tho nu.c).ani<f
..d..i~lc
t as ch~ of ph.,no.non,~ JH still
vcry impcrfL.cUy uudc.r.stoud. Wo .)u~ retnm tu
ttmnsubseqtK'ntchai'ter.

5G

EXPERIMENTS

0F

SAVART AND

KUNm.

"pcn end. and a, ]o.,p at the othcr point of trisection

[2G1.

midwvy

'tw'thcnrst.a)'dth~s(o'~rd(')-d~fth~p:p.
ru thc case of th <.pcn
or~an-pipo both end.s are )nops, and
thL-rc .nnst bc at luast cnc interne n~.dc. Tj.c
wave-fcr~~H. of tho
th icngthof the pipe, ~uch is dividcd
r'nc,pattonc~twice
into two sniular
parts by a noduI)i thu iniddic. Fnj!n tt.i.s wc sec
<~ foundatiun ut' thu
ordi.~uy r.ttc t)uit Lhu pitch of an open pipe
is
For rasons
to 'csatnc~th~ufa.pp~[pip.d')Laifit.s]o,~th.
bc ~norcf.,))y
in a sub.s.[Uunt
e.xpi.i.d
co.n.ectc.d
ch~pt.
with onr prc.sL.,it i.npc.rfL.cLfreinent
uf t),c opcn cm), thc ru)c is
.y
appruxinmtdy eon-.ct. Th opc.n pipe, ditrurin~ in th:.s ruirmn U,c stuppud pipe, i.s
~t
cnp.)c of soundinq thu whu]c .sc.rics
ut ton~h.nnin~
thc )~r.jiunicsc~fuuuded
upun its
~~c. In t)~ case of t)~ octave t)K.rc is :t
)nop at t!.c ccnLro of U.u
t'.pc :u)d nodc.s at t).c pun.ts tnidway b~twcun thc cuntrc and the
(.trc]mt)cs.
Since t).u frequcncy of thc vibration in a
pipe Is proportionrd
to tlte vdoc.ty of
propagation of .sound in thc ga.s with which tl.c
f")"~ is r.iicd, thc compari.son of thc pitd.cs of thc nutes
ohtaincd
thosarno pipe in
dinbruutgasc.s i.s auobviou.sn.ethodof
'n
dGtenn.nn., thu vclucity r~fpn.pa~ati.n, in c~scs whcrc th.
impos~L'lLy of ..hta.uiDg a .snHicicntiy long co)nrnn of thc
~as prccludcs
tho
<,ithc dn-cct moD.od. In this
appiicntion C'fdadnl ~it], his
"sna ~.sc
H.o way. T).c suhjcct ~.s rc.snnK.d at
iatcr
i~
d.~c by.s~city
D~on~~ and by Werthein~, w)~ obtainud
fair)y satisfactoryrcsnit.s.
2<'2. Thc condition of tlle air in the
inicrior of an or~n-pinc
,,Sa..rt~ho
I~dIntJ
-t~~
~i
.strctd.cd m.n.bmncon
~Uch a htHo sand .vas
1"~
~ttcrud.
In ~o nc-ighbou.i.uud u)anode thesandrc.maincd
und.sturb.d, but, a.s a fo.p wa.sapproac!.cd, It danccd .vi~
~~biy
.c
and muru v,go..r. But
by far th. n..st striki,~ funn of tin.
:.s that
invcntcd hy K.ini~. In Dus nK-th~ thc v:bra~
~r~.n.nt
"c.tc.d
),ya.sn.dt ,a.s fianK.fud t).r.gh
a tnbc .hich
n. cu,nnn.,ucati.n ~It). a
cavity caHcd ~nanon~tric
capsule.
i..L~c~
1.xr,c.p, 11a.
t,
'-f'~<.<~(~tw.ii.'l.]~,t.xx)l.p..t;(f.
'O.f/fC/ftM., t. xxtY.),);
~j

~C7~

2G2.]

CURVED riPE.

This cavity is boundcd

on onc sidc by n. mouhra.ne on which

t)~ih"I:)g:nr

:)oN.

A:ifhc~i~nbr;inc~ih)'itt.i-c:idcri))gthc

57

capacity of thc capsule Y!u-iah]L!,thc suppiy of gas bccon~'s unThc pcriodisof

sit-iKiyand t))c Hamcir~o'jnit.tcnt.
course too
~)na!t <or<!)cint<n))ith.-)ttotun))i~stit.sdfn.ssuchwhcn
tho
i)!unc'ish)('l-:t.'(1nt.st(;a<H!y. Bys)t!t~in~'t)m])t':u),()r~-itttt)m!ii(!
c)t':L]tin\'caL)ctnir)'(n',ti)crusu[ut[on
int.onturcoricssd~t.achct~
ini:tg'('.sn)aybccf'i'L'cte;d; but<jvc!twithoutrcs()h)tiunt!iecdtcred
<h:D-:)ctcrof' thc f):unc i.s cvi()L')ttirom its gcncml
ftppc~r~ncc. In
thu app)ic!Ltiou to orgiUl-pipes, one or niorc cap.s<))csarc tnounted
on a pipe in su<Ii !i. umuncr that U)e tuonhrancs nrn in contact
\it)t t))e vil)rati!)g culunui ufair; nud' thc diffurcnRc in th Hamc
is vury markcd, iLCcordmg-as the associated ca,psu)c is sttua.tcd at
n,)iodeorat:).!o')p.
2G3. H!t.h(;t.() weItn.vc .snpposcd thc
p!pc to hc stmight, but
itwi)) rcaditybc anticipt~) t].t, whctt thoernssHecLion i.s.smaU
n))d does xot vin-y in a.rc:L,stnughtnc.ss is tiot n nmttcr of
importan. C'ouccivc a curvc([ axis of running n.)"))g thc jniddfe of
<)K-pipe, and ici tlle constant scctjOtt
pc']-pct)dicuiar to this axis
bc ~S'.Wtien t])c grc~tc.st diatn~tcr of <S'is
voy smaU in comparison
with th wnvc-iungth of tho sonnd, thu
YclucLty-pctcnti!),!
huconn's nea)-!y inviu'iahJc ovcr t))u section;
applying' Grccn's
thcorL'n) to thc sp~cc houndcd hy th iutcnor of thu
pipe and hy
two cross sections, we gct

shcwing that
dpends upon .r in tlie samc wnyas if t))0 pipe
wo'c strai~ht.. By means of uquation (1) t.))e vibrations of itir hi

~S
8

DRANCIIEDPIPES.

f3G3.

curved pipes ofun.form section jnaybc

c~I)yinvc!,t!~tcd,a.nd the
rc.s.dt.s!U-et).c ri~orousconsc-quencesofuurftnKL-uncnt~)
oquations
~Y!t-L{,m,<j,ji~
.C'~pi.sjjsmaH.
In
t!K'
case of t!nn tubes such as wuu!d ~0
th~nitcfy
use.! in e.xpcnmcnt,t).cy s~i at
any ratc to givu !L vurygood
reprcsentft.tiunof \viiat actu:J)y Iu)ppuns.
26- Wc now p~ss on to tlie c.jnsidcmtmn of ccriinn cases
of
connc.otcd tu!s.
In t].e ~cco.npanyi.~ f~uru J~
rpp.o.t.s a
t)'t p)pu, wh:di divi.]c.s!(.tJ9 iuto two b~nchc.s
At 7~
Z)/~ 7~
t-hcbranc!s rcunitcnnd fonn a. si.~)u t.ubu7~
Thc sciions
thc .si.~)c tubes and uf thu L.-aucfiu.s~rc .Lssumcdtu bu
unifurm
as ~'c)t as vury stn.d).

In thc first instance ]ct ns

suppose that a positive wavc o<
arhitnn-y typu is advancing in ~1. On its arriva) at thc fork D, it
witt~ivurisc to positive wavcs in .Z~md C, and, unjcss a ccrtam
condittun hc .satisfic.),tu a n~ativu rcf!uctc(]wavc lu
Lut tlie
putcntlal ofthc positive wavc.shc denotcd by~
/bcing in
cach case afunctiou uf ~<; i and let tlic !~HcctL-d
wavc'bo
~'(~+<-<<). Thcn th conditions to bc satisfiud at D aru nr.st that
thc pressures .shaHbe t! same for thc threc
pipes, and second)y
that tlie wlh.icvclocity of thc Huid In s)udl bc
equal to th sum
of t))e who!c vdocitics of t)ic Huid in 7~ and C'.
Thus, usinrr
C'
to
dcnotc
th arcas of t]ie sections, wc hve, 2i4,
-J, Z?,
whcnco

~of'n'h.\nsa).pn~tn~.tonninot).o
refluer
nn~~frncteJwaYcs
thc junc~-n of ~o tKL.s uf MC'Juns 7.
+ r, and
rc.spc.ctivcfy, arc b'iv~ Ly

2G4.]
It appcars tha.t/, ~nl/
tion. if

DRANCHEDriPES.

5!)

are always ttie Siunc. Thcre is no rduc-

iho wavc thcn auvanccs in 7) and C cxact!y as it wouh! ])ave

donc iu ~1, ha([ titere bccn no break. If thu lengtb.s of tho
branches bc'twecn and be cquai, and th sectionof bc uqual
to t)tat of/), th waveson arrivai at JE'combine into a wavc prop:Ltu(t :L)u))~ !Uida~:un thurc is no ruftccdou. Thu division
uf L))ctulx.:bas thus Lcun ab.sohttclywitbout ufect; ~nd smcc t!ic
s:nno wontd bo truc fur a ngative w:t.vc p:sing froin J~to ~1,
wc nt!t.ycondudc gettcndty ti~at a tube may be (tividcd into two,
or more, branches, :dl of thc M:unchngth, without in auy
w~y
infiuoicin~ tl)u law of acri:d ~'ibnUio)),providcd ~I)~tth whole
socHotirc!ii:u)ic(jnst:).nt. If th Icn~.hs of titc bra-nchcsfrom 2)
to 7~'bc unuquid,t!iu rcsult is difTercut. Bcsn]usth positive wavo
in titcrc will bu in gnerai ngative rcHectudwavesin 7j'and C.
Ti)e tno.stintcrcstittg casu is whun tho wave is of harmonie type
a.ndonc of t)iu br:u)chcsis longer thftn thc other by :). multipleof
If t)tc diHcrcnccbc an CMH.
thc rcsnit will bc
multiple of
t]ic samo as if t!tc branches wci-cof cqual
Jcngth, and no rcncction
wIHcnsue. But suppose that, wltile and (7 arc cqual in section,
onc of them is longer than tbe othcr by an o~ multiple
of
Sincc t)to waves arrive at J~ Ln opposite phases, it foHows from
synunctry that tlie positive wavc In ~must vanish, and that th
whieh is neccssa.rily tbe same for a!l thc tubes,
pressure at
must bo constant. Th waves In 7~ aud 6~are thus rencctcd as
frum au open end. T))at thc conditions of th question arc tbus
satisned may aiso be sccn hy supposing a barriur takcn across tlic
tube 7~in th neighhour)iood uf jE'in suc!i a way that tbe tubes
J~and C'communicato~'ithout a change of section. Thc wavc in
cach tube will thcn pass on into th otbci- without Interruption,
and thc prossurc-viu-iationat j~ heiug th rsultant of equal and
opposite componeuts,will vanish. Tins bcing so, t!ie barrier may
bc rcmoved without altcring tt~e conditions, and no wave will be
propagatcd along F, wbatcvci- its section may be. Th arrangcJt/eM. r~)f!ft'<H~,
t. i!. p. 305. Th rcrutorwillnot forgotthat both
J'nisso)],
dittmetcra
mustbesmnU
in comptu'isou
witlithoWt).vo-lci)~th.

~0

BRA~CHEDPIPES.

L
r-?(;

~ntnuwundcr
considration v.asinvcutcdhy
Rc~dK~andh~s
~'uun cmp)..y(.d Ly<~)in(-kcaud
otLur.sforcxpt.ritm.nta) purposcs
:tj,p)!~<.tiu'tt!,i..)~s)):)rh.r~;).(),i.
i-, .].
H)ep~o.on)ono)Y.tsc]t
.s<,ft.r~.n~d
'.oa.s:u)..x;).np)c(,ri,,tcr~ru.c.t..)~)t.d.t)HT..c-hufH)uLj..(.tiu.I,u).<).(..san)cc.n,(,t
!'<-.snit) w))(.). O.~ru~Io-is
]~)tu.su).pu.
t).;)t.L)n-}H,sit.ivcwavt..s
n~utmii.sccachoth~-iti
7'nh!t).:).tL)n.n.t!,m);)tt..rc.nds
Jt,inust
.K~.Tbcf.~<~hjn))..tdhTuisnoh,ss<,fc.n~inint.~fcruncu
b))t.)iya.)itr(.rc))<.)i.-<tnbu<iun;
diverti
~i..)K.r.;yi.s
fro.,)'
""cp!itn.)pj[,car.si)tanut)KT.
I"t!h.})n,st.t(.asuthcp.).sit!vu
~vcin~
n't.).)-is
c.m~ys(..H~ywithi<.
nu wavea)u..rr~'
O~rc. ~hvo],s.s.),)u
alternative.
-it!r.ya(.cnnm~t.s
'n Dbranches, or dscit passes
in't.),c f-onant-aa
La<ka)<t
In.~urtos~
n~at.vewav.
what r<.uiy]~pj~.s,
lut u.s trace
t))e pro~ress of thu wavt'.s rcfkcttt back at 7'
Thse wav(.s arc-~qua] !na~tittxic
a)u) .start fn.tn 7;:n
oppnsitc p).a.s~;i.)<
pa.s.sa~ihjin7;tu~<)~~
~rav..]
.)i.sta.. ti.anthuoth~-Ly
!atLT
au (,<)
and
,uu)tip)uof'
thc.rdurc ou arriva! at
bo iuc.anp!u(u ac.-ur.i'ancp
t).y~ij)
whirh
Ut.(h-rth(.sc<.irc)..n.stanc~t)~.ycun.b.ncInt~.si.)L;Icw:u'(.
tr.~(.su~ati~.)y
a]o~.t,an.itl.c.rc
i.snn rcf)ccti.,n. W).~t),c
ne~atn-c ~avc ruac)K..str,c c.nd of thc tu),G J, or i.s .,t.hurw..sc
<!ist.trbc.d ni its euur.se, t).c ~hoie or a
part may !~crdJcctud, and then
tl.c procs .s rupoatcd. But h.,wcvcr <t.c~
this ~ay haj<pcn thcrc
will bc i)o wavu
a)u,~ F, un!c..s.sl,y ac.uunda.tam in oon.sc.qucnceof
.'L
ur pcri.~s, thc. vibration ia thc
hranehc.s bccon.G so
c.an~)cn
g-rL-t that a .sma)I fraction ofit eau no tonner bc
u~k-ctcd.
Fig.CG.

Or wc n.un

U.ns. Suppose t).c tube.

FeuLuH'Lya a

2G4.J

BRANCHED
PIPES.

Gl
L

bu.n'icr as bcfurc. Thc motion in th ring bcing duc to furcc.s

actixgat D is nL-ccss:t.niyMytnmot.ricfdwii)t t'unjK'ct to 7~, and
thu point \Yh:ch(!ivid'jt!7)/~<)i)tt.oc~))!t.l
parts. J~ncc'j~'Ls:).
vibration i.s .st~ti(jna.)y. Tinsbein~, t))ocasc,ata,
n<'(t(.audt)ic
])<)i))tA'tnh!t:utt.fro))iJ'/<))tclL)K.')-si(]c,th(jrunm.stbc!T.]o(m;
an~tt'thcb;u-r)ui-Lorctn()Y(-!dih(.'rcwi)Isti![bc
notcmiL'ncyto
producc vibration in 7' H'tttepf.'rimct~roft~cring-bca.inultiplu
ot' thcrc may bc vibration wit))i)~ it oi' tho ncriod ia question,
i)u)cpu)n)<j))tiyuf:t.ny)at(;rat(ip(.'ttings.
Anyc'nnijination nfcfjnncct~d tuhc-s maybo trcit-tcd in .T,si)ni!:u~t:mm.'r. Tho gnera), principfc is t)tat at ai)y jmictiun a
spaue
takoi
can bL;
lar~u enougtt tu inctude n)I t)ic rgion t)u-ou<')i wiucit
Fij;.

57.

thc \vant of uniformity afTcctsth huvof th wa-vcs,and yct so smaU

that its iongcst dimension ma.ybc ncg~cetedin comparison with
Undt-r t.hcsccirctunstancc'st)te nuid within thc space in (tucstion
jnny be trcatcd ns if t)tc: wiLYu-Icn~thwcrc infinit, or thc nuid
itsc-if incutnprGs.sibJc,in wiuch case its velocity-potcntial wou)()
satisfy ~<~ = 0~fu!!owing t,hc 8:une I~vs as ctectricity.
265. Whcn thc scetiou of a pipe is variabtc, thc probtcni of thc
vibrations of air within it canuot gcncraHy be solved. T)ic case
of conic~ pipes will be tre~tcd on a,future page. At prsent wc
will invosti~te an approximate expression for thc pitch
ofancarty
cylindrical pipe, takin~-first th case whci'e both ends are closcd.
Thc metiiod tha,twill be cmp!oycd is sixiilar to that nsed fur a
string
whose dcnsity is not (nutc constant, ~ !)I, J-tn, depending on thc
pnnciple that thc pcriod of a. iroc vibration fu)n!s th stationary
condition, and may tLerefore be calcuiated froni th potcutial and
kinetic cncrgicsof any hypothctica) motion not dcparting far from
tbe actua! type. In accordancewit]i this p]:(.nwe shall assunte that
tbe velocity no)-)nalto any section ~S'is constant over thc scctioti,
as most be vcry ncarly t)tc case wlien tho variation of is slow.
L<'t ~V reprcsent thc tota) tmnsfcr of ftnid at tune across the

VARIABLE

SECTION.

f2G5.

<~
section ntat x, ~?.
rcckoned f.-om thc
G()ui)ibnum condition thcn
reprosents the total vclocity of tlle currcnt, .nd .Yrcprc.scnts
actu~}.ty
.f.
thu kiucuc
~i,~t!~
of
t)ic
motion
Avithinthc tube is
enorgy
cxprcs.sc-dhv
Rr~ntir\ti

fhc rcsult mayhc cxprossc<!

convenicntly in tcrms of A~ thc corrcct.on t .at rnust bc ,n.dc. to in ordcr ti~.t
U.e pitch ,nay Le
-a!cu)ated
thc o..hnary fur.uul~ as if ~e,.c
con.st..at. y.r
~.n
)."c value of
A~ wc I)avc

2G5.]

VARIABLE

SECTION.

63

Tho cn'cct of a \'ai-iation of section is

greatest near a nodc or near
a loop. An cnhirgRmcnt of sucMonin t)~ Ursbcase lowers
tt~c
p!tc)t, fu)d in thc second OLscmises iL At th points
bctwecn th nodc.sand loops fi .slight variation of section is midway
without
cf~ct. Thc pitch is thus dccido(Hy:dtcrcd
by an enlargement 01contr~tion uc~ t]tc middtc of tl.e tube, but tho iaftucnce of
a
.s!i~))tcunic:dity woufd bc inuch less.
Thc expression for A~ in (8) is
applicable as it stands to the
gmvcst tono on)y, but wc n)ay apply it to thc
tone of thc harmonie scalc, if wc modify it by th substitution of cos
2~
for cos
6
In ttte case of a tuhc o~eHat both ends
(~) is rcplaccd by

instead of (8). T),c piteh of thc sound is now raiscd

by an
cn)argcnic.ntat thc end.s,or hy a contraction at th middie, of the
tubu aud, as bcfure, it is unaifected by a slight gcnct-al
couicality
2GG. Thc case of processive w~vcs
movh~ in a tube of variab)c scctio)i i.s a!.so
intcru.stmg. 1~ its ancrt form th probicm
wontdbc onc of grcat dimc.dty; but wherc tho
change of section
is vo-y graduai, so that no considrable altration occm-s
wit!un a
distance of a grcat niany
wavc-Icngths, t)ic princip)c of cnergy
will guide us tu an approxin~tc .solutioi). It is not difncult to
see
that in thc case supposed t!tcrc will bc no sensib)e rcncction
oftho
wavcat any part ofits course, and that thcrcforc t)ic
of tlie
cnergy
motion must romain uncha.ngcd'.
1. Now wc know, 24;-),that for
a givpn area of wavc-front, tiic cnergy of a train of simple wavcs
is as the square of thc amphtudc, from which it foltows
tliat as
thc wavcs advanc tlic amptitudeuf vibration varies
as
thc square root of t!)c sectiun of thc tube. In nJt othcr Invcrscly
respects t!)c
typc of vibration remains fdjsdutcdy undtangcd. From thse resuiLswc mayget a gnral idca of tlic action of an
ear-trumpct.
P/ ~y. (5)i. p. 2('.i.

G-~

VARIABLE

DENSITY.

F 2 GG.

It appears that according to th

ordinary approximate cquatious,
t!)erc is uo limit to t)tc concentration of sound
produeibie in a
t!'b<o!n!ua[!yt!r.uiiHh)]tg.jf:t].~j.
TJic saioe mctimd i.s app)ic;ib)c, w~cn titc
dcnsity of the
)ncdiuni varies .siowiyfroni point to
point. Fur cx:u)))))< thc
amplitude ofa sound-wuve movins- upwu-dniM thcatmo.sphc
]nay bo detcnniucd hy t)tc condition that titu cn(;)~y ronaina
unc)i:u)ged. From 2.t.-i it ~ppcars th~t th a)np)i~de is inas t)ic square root of thc Jcnsity'.
vct-i-iuly
A de]i(.nto

nrisos ns to tho u]ti)!))itc hto of fio)]nrot)H wavcs

qu~ti~n
))rnpaRat(.d
It sL.mId bu ro.tarhcd
upwardH.
thut tu rnrc fur thc (teadoui~
infiucneo of viscomty
iHtimehmcrcnseJ.

CHAPTER

XIII.

SPECIAL PROBLEMS. REELECTION AND REFRACTION 0F

PLANE WAVES.
2G7. BEFdRH undcrtaking tho discussion of the gcncra.1 quations for acria) vibrations wc may eunvcnicntty turn our attention.
to a fcw spuci:d problumS) rc):Lting princip~Uy to motion in two
<)in)L')).sions,which arc susccpt,ib!u of rigorous fu~d yct cotnpan).tivuiy Hi)np!c solution. In tins way ttic l'eadcr, tu whotn thc
suLject is ncw, will aequirc soae famiU.u'ity wit)i thu I()cas aud
tnuthuds cniptoycd bcfui'c attacking more fot'miJtt,ble dificultics.
In thc prcvions cha.ptcr ( 255) wc hivestigatcd thc vibra.ttons iu
one dimension, which may tnkc p!:tce piu-nnul to th axis ofa. tube,
of \vbi(.;hboth ends arc c)oscd. \Ve wiH now i)j<[nirc wbat vibrations
!t)-L'possible wiL!)in :t closed rect:Lng)dnr box, dispcnsing witb th
restriction tb:).t th motion is to he in one dimension on]y. For
ctu'h simple vibt':Ltion, ut' whicit t!)<jsystcm is cap:).b)u, ~) varies as
a circufat' i'unction of thu time, .say cos/<:~ whcrc /c is somo
aud therci'ore by tbu gcncral difrureacoostant huncc =
tia) quation (9) 2-M

Equation (1) must ho Sfitisficd throughuut t!hj w))u!e of thc

"'cindud Ytunm. Thc surface coudit.iun to bc KLtisfiud ovur thc
si\!sldc.s<jfthuboxissnupty

~'hure

t'bprcsott.s au cjoncnt. of th normn.! to t,!ic sut'fn.cc. It

uuly fur spcial v~ducs of A:t.)iat. it is possible to s;.).tisiy(1) :).nd
(~)si)nu]t:LtK!un.s)y.
!.U.
5

AEMAL

VIBRATIONS
Lf2G7.

Taking three edgcs which meet as axes of rcctungular co-ordi~ates, .~d supposing th~b tho lengths of the
edges are respeetivelv
wo
a, p, 7,
kuow ( 255) t)):~

y, r are Intcgers, arc particul~r solutions of thc

~hcre~,
prob!cm
~y any of thc.se forms quation (2) is satisned, and
providcd'
that
be equal to
or r
thc case may be, (1) is also
It is cquaHy vident that th
satisf~d.
Is
satjsnud over aH tho surface Ly tlie furm boundary cquatiou (2)

whcre a..sbcforc

7'are intcgcrs.

T!~
ail particule~ncml soh.tinn, cl.tained by compounding

solutions incinded undcr (~), ia

arc arbitrary constants, and tlie

summation is
i~ch.land~to ail
L-xtcnded
mtcgrai values uf~, ?..
This solution is
sufncientiy gnrt to covcr the case of any
nut.al stato of thln.s ,vit)un th
box, not involving rn.]ccutar
rotation, ihc initia! distribution of vdocitics
dpends upon th
initial value of
or
and by Founei-s
/+~~+~~),
theorem can bc ropresented
by (5), .suitab!e vahics bcii~ ascrih.d
to th eoc-mciunts~1. In likc
nianncr an arbitrary initit distribntion of
~nduisation (ur rarfaction), <)c.pendingon the initial
"1"
~scribing suitabic vahies to th
~e~nt
coefflcicnts
r
13.
Th investigation
might be prcscnted somewhat dirfurent]y
by connnencii.g wit!i a.ss~ing in accordance witb
Fo~-icr'.

267.]

INA RECTANGULAR
CIIAMBER.

tleorem that tlie 0gnera! value

of~at
titofurm

67

time t c.-ui bc cxprcssed il,

m

.a wh~ch the cocfic.cnts C

n..y dpend upon <, but not upon
i).c c.prc.s.sion.s
iur y and
y.
woaM thcu bc funned an<t
-shown to .nvuive onJy t).c .squ~-es of t!~
coemcicut.s 6', ~,d f.-<n
tLc.se expressions ,vou!d foHow thc nrn.na)
c.qu..Hons of n~iou
counectmg each uormd co-urdin:~c (7 wit)t t])c tune.
Th gravcstn~deof vibration is th~t i.~
~.hieh tl.o entirc
n.ot.on ..s
.nd thcre
P~))dtothc)~c.st<Ii,ne,,sinnofthuLnx
..sno ,nt.crnal
nodo Thu.s,if !,c thc
g.-catest of tlic ti.rec .sidcs
a, 'y, WC:U-Ctu t:).kc = 1, =(~ = ()
In thc e~c of a cubie~
bo.x~=~=~,
(t) WeL~VC

and t!.cn i..stea<)of

As in thc case of thc membrane

( H)7), w).cn two or more
pmn.t.vc modus hve t)~ .s~mc pcriod of
vi),n..ti~, otiior n.odcs
of like pcnod mn.y bc derivcd
by composition.
Thc trcDy incite .Grics of
po.ssihfe .si,np]e componont vibrations is
notnc.ccssartiycomp)ete]y rcprcscntcd in pa.-ticularc~es of
compnund v.b~dons.
if, for c~.npju, wc suppose thu contents of
the box in its .niti..d condition to bc nc.it!.cr
condcnscd nor ra.-cficd
to hve
~yp~-t~nd
unifor.n velocity, whose
componcnts
j)araf)ci to t!)c axes of co-ordinates fu-e
rcspcctivcfy
no .s.mptc vibrations arc
for
whicii more than' o~
gcncrated
t'K. thrcc
r is finite. In tact each
uu.nhcrs
component initia
vdocny rnay hc con.sidcrud sopamtdy, a.nd thc pr.~Jo.n is sin.if.uto t])at solvoi in 258.
5-3

NOTES OF NARROW

PASSAGES.

[~f.
f2G7.

In future chaptcrs we shaH meet wlt)i other

examples of ttic
vibri)tionsof tir within cotnpIuLdyc!oscdvusscis.
Some of thc natuml no~-s oi- thc air ccntiiincd
witbin a rootn
may gencra))y Le detected on singing th sc~Ic.
Prob~y it is
somewiiat in this way that blind
pcople are able to estime tho
size of rooms'.1,
In long and narrow
p~sa.gcs thc vibrations pamUcI to thc
)cngt)i are too slow to affect tho car, but notes duc to transverae
vibrations m:Lyoften bc hcard. Ti~ relative
proportious of thc
varions overtones dpend upon the
place at winch tlie disturbaiice
is crcatcd".
Insome ca.scsof this kind th pitch of the
vibrations, whosc
<hrectio!iis principally transvur.su, is infiuuiicod thc
occurrence
by
iongitudinal motion. Suppose, for cxamp!c, in (3) and
(4), that
=
<y=1, )- U, and thM a is much groater than /3. For the
principal
transverse vibration p = 0, and = 7r
Bat besidos this thcro
arc other modes of vibration in which thc motion i.s
principaHy
trausvcrsc, obtaind hyascribing to small Intgral vaines.
Thus
whcn))=l.

shewing tha.t th pitcb is ncariy the samc as bc'forc'.

2fi8. If wc supposery to bucome
inHnitcJy grc: t)ic box of
the p~ccding suctiun is tr.-uisformcd into an
Infiintc rc.ctan.~fa.H.bo,wbosc sides ~-0a and /3. Wfh-itevcr
nmy be th motion uf
t)~ an- w.tbm tbis tube, its
vclocity-potcntial may be cxprcs~d
by Fuunur's t!tcorem in the scrics

n-markaUo iustnuco is qnotod in

Young'n A'~Nr~ J'/n'~o~y<y. u. p 272,
rom Darwiu-. ~M.<
d87.
Tho !atc bHnd Jn.tico Fiold,
walhed for t]~
f.rst tnnu into .ny r.oM, whcn ho on.u
vi.itcd me, an<l aftor spc.a)<inK a few won).
This r..<.a is about M leet
.<n.
long, 1~ ~ijc,
12 high
ail which lie gucs.c.c
''y tho efu- wiHt ~rent nccurncy.
Oppe), J~ /<~M~
~<
ya,-~~e )r)7u)~ <.rr~<e)t J~./f<<f))t~ut;
7''<))'~r/;r/< ~.r /~)/.<;A', xx. p. l;i0.
"H'
in n,y h.M in whieh it is
t~ss,
p~iMo.
hy
~nf; t).. r~ht note, to c..cite fr.o vihr.tiun.s of
,<cun~'
~,d
,n.uy
it
Jurati~
t. L~ h.pj.~H that (Lu n~n.nt
uuto in aficctcd ~it). di.tinct bcnts. TI~. bn.~Hh
uf thu j~s~~r is ,t)joHt
f~t, and tho ),<.i[;ht nt.uut
fc.ct.

268.]

RECTA.NGULA.R TUBE.

G9

whcre th coenkicnts arc independent of .r and y.

By tho use
f.f this fonn wc sccui'e the fu)f1))ncntof th~
cn~fHt~jt
h.')unf!)!.ry
t)~t LhcrGis to he no velocity across tho sidcs of the tube; the
niitm-cof
as a, fnnction of 2! and < dpends
upon tlic other
cunditionsoftho problem.
Let us coisidcr thc case in which the motion nt
every point is
!t:u')nunic,and due to a normal motion nnposed npon a ba.n-ier
ntrct(;)tingacross t]tc tube at = 0. Assumin~ to be proportiun:U
to e"~ :Ltali pohtts, wc hve thc usn:d dltrcruuti.d
quation

winch by tlie conjugatc pt-opcrty of tho functions must be s.itis~d

scparatcly by cach term of (1). TIms to dotermiuo
M a
iuncttu)i of z, \vc get

Thcsolution of this quation diffcrs in form

according to th si~n
of thc coefHeicntof J~. Whcn
and y arc both zero, the eocfHeicnt is ncccssariiy positive, but as and incrcasc th coefficient
q
c!)angcs sign. If th coefficient bG positive and bc called
tlie gnerai value of
may be written
whcre, M thc fa.ctor e"~ is expressed,
are ~solutc
constants. However, the first terni in (4)
expresses a motion
prop:~g:ttedin the ncgn.tive direction, which is excludcd by the
comHtionsof tho probicm, and thus we are to take
simply M tlie
term corresponding to
y,

In this expression C~ may bc

complex p~ssing to rca.1qua.jititics
and t~king two ncw rca.1arbitrary constants, we obta.Ia

Wc hve now to considcr th form of th solution in casca

whcre the cocfHcicnt
of~ in (3) is ngative. If wc caH it
tlic solution con'esponding to (')-) is

70

MMCTANUL'LAR TUBE.

)-~U.
FSCS.

of witich th first term is to be

rqjceted as bceoming In~initcwith z.
We tbua obtuin cun-uspoidn)~ to f5)

Thc solution obtaincd by

combining ~} the part:cu!ar sohitions
givcn by (5) and (7) is the gnera! solution of t!.e prob~m, ~d
atlowsof value of
over th section ~=0, arbitrary at
every
point lu both amplitude and piiaso.
At a grt distance frmn thc source the tcrms
given in (7)
become insensibje, and thc motion is
repre.scnted by thc tcrms of
(.~ done. Thc cnect of thu tcrms iuvolviog high values of~
am) y
is t).us connuud to thc neighbourhood of thc
source, and at
n.odcratc (ti.staucc.sany suddcn van~tiuns or discontmuitics
in tilc
n.ot.ou at ~=0 areg.-adua))y cased oif a..d
oblitcrated.
Jf wc nx our attention on
any particular simple mode of vibration (for which and do not bot,],
vanis).), and conoive tho
iruqucncyof vibration to increasc from ~i-o upwards, we see that
t)te eHeet, at first connned tu t!.o
ncighbourhood of thc source,
graduaiiy cxtcnds furthcr and furthL.r,nnd after a certain value
is passc-d, propngatcs it.seif to an inimit
distance, thu criticat
h-c.p.cncybcing that of tho two di.nensional free vibrations ofthe
corrcsponding modo. Below th critical point no work is requircd
to 7~.~
th .notion abovc it as much work must be doue
at
= 0 as ..scarncd otf to
innnity in tiie samc time
2of). We will now examine the rcsalt of th
composition of
two tra.n.s of p)anc wavc.sofi~rmonic
typc, wiiosc ampUtudcs ~nd
wave-lun~ths are equa!, but whosc directions of proj~gation ~c
iuch)~) to onc anothcr at an
ang]e 2~. T)ie probtcm is one of
two dnncnsK.ns o.dy, In~much as
everything is th same in
ptanes pcrpcndicular to thc H.tcsof i.iturscetiou uf tlie two sets of
\avu-ft'onts.
At any moment of time the
positions of ttto p]anes of maximum
condensation for cac), train of wavcs
rn.~ybo rcprc.scntcd by par.dte) hncs drawn at equat intervais on thc
plane of tlie papcr,
and t)~so fines inust Le
supposeJ to move wit), a vdocity hi a
<hrcL-ti..n
pc~L.ndic.,)ar to D.cir length. If- buth sets of lines Le
drawn, t)ic p.~pcr will be divi~cd into a System of
equa! parailuio-

2 G9.]

TWO EQUALTRAINS OT WAVES,

Tl

grams, which advance in tlie direction of onc set of diagonals. At

cacli corner of a paraUeIogra,mth eonduns~tionis doubled by the
superposition of tlie two trains of waves, and in th centre of each
paraHuIogrnmthc rarefaction is a, maximum for tlie same rcason.
On cach diagonal there is therefore a series of maxima.and minima
condensations,a.dvaneing without change of relative pusitioti and
with vclucity ft cos a. Bctweeu eacli adjacent pair of lines of
tnaxima and minima thcre is a parallel Iine of zero condensa.tion,
ou which thc two trains of waves neutralize one another. It is
uspceia.Hyremarkable titat, if tlie wave-pattern were visible (like
tfie corrcsponding water wave-pattern to which thc whole of tlie
prcceding argument Is- appticabhj),it would appear to move forw:u'ds without citange of type in a direction dincrent from that of
cithci- component train, aud with a velocity ditcreut from that
with wliieh bth coniponent trains move.
In ordcr to express the result analyticfdty, let us supposethat
thc two directionsof propagation are C(p)allyincHncd at an anf)Gc<
to tlie axis uf x. Tlie condensationsthemsetvcs may be dcnoted by

It appears from (1) that thc distribution of on the plane a~

advauces pa.ml)elto th axis of unchanged in type, and with a
uniform vclocity a cosa. Considered
dcpcnding on is a
maximum, wbcn sin a is equal to 0, 2~ 3\ &c., while for tlie
iutermediatc values, viz.
A.,&c.,s vamshcs.
If a = 7r,so that thc two trains of waves meet one another
dircctiy, tlie velocity of propagation paraUel to x becomes iuHuite,
aud (1) assumes th form

72

REELECTIONFROM FIXED WALL.

[2G9.

Thc problem that wc hve just hccn considcring Is in rcality

thc same as titat of th runcction of a train of plane wavcs hy an
infinitc p):mc walh Sincc th expression on thc right-hjt.! sidc
of quation (1) is fin evcn fnnction of y, s I:i symmetrical '\vit)i
respect to t)te axis of ?, n.nd consequcntty there is no motion
a.crusst)tn,bnxis. Undef thse ch'cttmst:u)ccsit is vident th:tt thc
motion cou)d in no way bc aKcrcd by thc intr(j(]uct,ion ~lon~ th
~xis of a; of a.n.absotutdy immov~b!c w:U). If a bc th angtc
betwcn tho Murfttccund thc dircetioti of propng'n.tionof thc Incident wttvcs, tbc vciocity with winc)~thc pices of !n~xinn)!'icondensation (con'CHpondingto the g)'c;itc.stdvotion of w:T.tcr-W!L\'cs)
movc a.)ongthc w~His ft cosa. It may ho noticcd t!):t.tthc n.cn:d
prcs.sm'cs])a.veno tcttdcncy to move thc wa.)tns n. who)c, cxccpt in
titC case of nbsuhttely pcrpcndiculur incidence, since thcy m'o at
any moment us mnchncg:).tivc as positive.
270. So ion~ as th mdiumwhich is t!ie vchiclc of soun(tcontinues of unbrn~cn unifonnity, phmc wavcs !na.ybcpropagatcd in
any directiou with constant vclocity and with type unchangud Lut
a disturhancc!cnsucs wlten tho wa.vcs ruach any part whcrc thc
tucchanical prcpcrtics of thc mdium undergo a. change. Th
gcttcral proDeni of thc vibrationsof a vai'iabic modimn is probabty
of
quite bcyond thc grasp of our prsent mathcmatics, but n~a~~y
th points of physical intcrcst arc misod in thc case of phmc
wavGS. Let us suppose that tttc rncdium is uniform abovo and
bctow a certain innnitc plane (~=0), but that in crossing that
p!ano t!)erc is an abrupt variation in th mccbanical propcrtics on
which tho propagation of sound dpendsnam~y thc cf~t~'eN.s'tOn thc nppcr sidc of thc plane (which for
!)t7~ynnd ttte ~C!
distinctness of conception we may suppose horizontal) a train of
plane wavcs advanccs so as to meet it more or less cbUquety thc
nrobtnni is to dtermine tho (rcfractcd) wave which is propagatcd
onwards within th second mdium, and aiso that tlu'owu back
into tbe nrst nicdium, or reftected. Wu bave In th nrst p]ace
to form t!ie cquatioua of motion aad to express tite boundary
conditions.
In t)ic uppcr mdium, if p bc thc natural Jensity and s thc
condensation~
density = p (1 + A'),
and
pressure = J*(1 +~1&'),

270.]
]

REFRACTFO~
OP PLANEWAVES.

7:3

whcre./) )H cncfificicnt (Icpcnding 01 thc

compres3:M)ity,f~n(~P
is L).cuadistm-bud pressure. 1~ ii)~ !n:u)ncr ni tlic Jowc)-jnudium

thcnndisturhcd

pressure b~ng th s~mc on both si(]cf! of ~=0.

T;~i)~thc:~is
ut'~p:u-!Ut(jlt.)t))o]incof
intm-sueti..n.,t't,hc
p):t))<t~f U.c Av:wcswitli t!.c surfhcc uf .scparatiuu.x=0, wo hve
i'<.))'Lhcuppcrnicdiu)u(~44),

Thcsc chtions m~st bc s~tisHcd

a!l points of thc fini.). FnrUicr
thc b<)un~!HycondiLion.s
aH points oi'tho
rc~)u-e(~) th~t
Murf~ccof HCp:u'atio)i thu vulocitics
pet-pcndicu~r to tlic suiTacc
must bc t)ie s~mu for thc two ituuls, or

lu onicr to rcprcsent a. tmm of waves of harmonie

type, wc
m:t.y nsHumo and <~ to bc proportional to e'<~+~+') whcrc
+ = cmt.st. ~:vcs thc direction of t)ic ptane of thc wavcs. If
wc :iti.su)ncfor t))u incident wavc,

GREEN'S
1

INVESTIGATION

'1

Fs~Q.V.

tlie rcficcted and refractcd waves

may be repr~nted
by

respect:vc!y

Th
coemc.cnt of < is
ncccs.sariy tlic same in ait thrcc waves
on ~ccountof tlie
periodicity, and ti~ecoemcicut of y nu.st be ti.c
samc, .~ncct).c tracus of a)l th waves on thc
p!.nc ,f section
must n.ovc togctl~
With regard to ti.c coefHcicntof if
appc~ by substitution in thc diHcrentud
that It.s si~n
quations
in p~Ing i-ro~ thc lucidcnt to
th rcncctcd wavc'. In
~h~ged
fact
Now
&- V(..+ ~) ,s th sine of tlie
angle Included between the
axis of x and tlie norn~! tu thc
plane of t]~ w~vcs-in optic.1
t)~ sine of the
an~u,
a~]c of incidence,~d & ~(. "+ is in
T7~
of'
1~
angles
be c.I!cd
(~ asserts th.t sin~: sin~ is
cqual to the constant rat.o
.cU-J.ncwn law of sincs. TI.c )~ of re= ~-the ~!cw
f.act~n .nd raction
simply from tlie fact that the vc]oof
city propag~n normal to tlie wave-fronts is
constant in cacli
n~dunn that
to say, indcpcndcnt of the
~c~
of thc wavet.ken
in
front,
conucetiou with thc equ.! velocities of tlie
traces of
aH thc waves on the
phtnc of sparation (
sin = F sin )
It renoms to satisfy th
boundary conditions (7) and (8).'
Thse mvo

This ccmp!ctes the syn,bo)Ica!

solution. If
sue that. if the incident wave be

(auj

bc rca!, wc
'v~

270. jJ

OF BEFLECTION

AND REFRACTION.

75

is ho-c obtaincd on thc supposition t!i~t t)ic w~vesarcof harmonie

t.ypu; but sincc itdocs not. involve and t))<j)-uiHno change of
phase, it may bc cxtenjcd by Fuuncr's thcorcm to waves oFauy
typu whatuver.
It' thL-rcbu no rcficctcd w~vc, cot
cot =
from which
aud (1 + cot' ~) (1 + eot' ~) =
wc dcduce

which shcws that.providcd thc refractivc index

F Fbc inter)UL~)iatcin value bctwccu unity aud p
tiicrc is aiways au
:u)~c of incidoicc at which t)ic wavc is cun~jlutclyintrujuitted
but otho'wiso titcrc i.sno such tUT'c.
Smco (18) is not altercd (cxccpt as to sign) by an
i))tcrch{).ngc
wc
infer
that
a
&c.,
wavo incident m tlie second
of<9,
jncdimn at :).nangle is refiectcd in t)io same
prupurtioa as a
wave incident in thc first mdium at an a.n'du
As a numerical cxampic Jet us suppose that
thcuppcr medimu
is air at atmospho-ic pressure, and thc luwcii-mdium watcr.
Substitnti)]g fbr eut its value in tcnns of and thc rufractivo
Index, wc gct

FRESNEL'SEXPRESSIONS.

[270.

whieh shows that th ratio of

cotangcnts dhnini.shcsto xcro, M
h.crc~os from xcro to ~out 13",afLM-whic!.it htconc.s
im:~i'),:u.y,
tutal
m<)ie:).t,ij)g
rcf!cct:on, tmwo shafi sec prcscntty. It n~st bu
ronc.tnbcrcf) thut in ~Jying optiez! tcrms tu
acoustics, it is thu
w~er t!jftt nutst Le concuivcdto be th 'rare'
medimn. Thc ratio
oi'duusitic.sis abuut 77U 1; so that

Evcn at pcrpcndicular Incidence th rfection is

scnsihiypcrfcct.
If both mc-di~ bc gascott.s, = If th
temprature Le c.mfitant; fu)d evcu if thc dcve]op)ncut of ]tc.tt Ly
compression be
takca into account, thcrc will bc no sensible difcrcttcu
bctwucn
and
in t])o case of th si)np]c g~sus. Now, if
=~
=siu~
p, /)
and thc fonnult~fui- thc intcnsity of tlio
siu'
renceted wa.vcbcconics

comcidmg with that givcn by Fre.snc! for light polarized pcrpcn'hcufarly to t].e plane of incidence. In nccordimcowitli Brcw.stcr's
huv tlie rdicetion vanishes at tlie
angle of incidcuce, wl~osc
tangent is F'
But, if on thc othc!- hand'/),=p, tho cause of disturbanco
buing thc change ofcomprcssihiJity, we I)~ve

agrccing~h Frcsncl's fonnu)a for !ig).t pohmzcd in t).c

p~nc
of !nc.(h;ncc. In tUs c~sc t!to rL.ficctcdwave
docs not vauisfi at
any angle of incideucc.
In geuGi~J,wlicn

= 0,

370.]

REFLECTION DUE TO TEMPERATURE

so that thcrc is no rcfMon, if

= p~ p, and theu
~ascs F'

AND MOISTURE.

77

~=

Suppose, for cxamp]c, that aftor p(.rppndicu)ar incidence rc~'cdott takcs place at a surface
scp~-atin~ air au() ]iydro~ou. Wu
iiavu

Thc mtio of intcnsitics, which Is as the

squft,-Gof tho ~p]itudcs,
is ~-t-02 1, so tliat about onu-t)[ird
part is ruficctcd.
If thc di<L.rcnccbetwccn the two mcd~ bc
sni~)),and wo
very
writc ~=F+~,
(24.) bccumcs

If the f~rstmdium bc air nt 0" Cent., and thc second

jnedium bo
air ~t C'eut., r+ ~F=
r~H--003CC<; so th~t

Tho ratio of th intcn.sitics of th reHccted and incident sounds

is
t))~rcforu-83x]0~x~:l.
As annU~r examptc of tlie sfunc ]dnd wc
may ta~c Miecase m
which t))c <h-.stmudium is dry air and t))c second is air of
thc
s.unu tonpcmture satm-atcd wit)~ moisture. At ]UCot.
:ur
.s:tt)tmtcdwith moi.stui-0i.s li~htcr t)~n
dry air Ly abuutono p!trb
v
iu 2~0, so t])nt
J)c:u-)y. Huncc wc conclude from (25)
~~=~~
tlmt th(! ruHcctcd sound is on!y about onc 77~,000"'
part of thu
incidunt suund.
Frotn thse calculations wc sec that rcf!cetiunsfroni warm or
moist ail' jnust gcncrany be very smalt,
tinjugh of course thu entbct
jnay accomulatc by rptition. It mn.st aiso Le rumonbcrcd that
ill practicu t!)L:transMiun frum one state uf
thin~s to the ot)tcr
would bo gradl1:tI,anll tint abrupt, as thc
prcHuntthcnry supposes.
Jt' t)K-sjta occxpicd byt))Ctr!U)sitiun amfmntto a considrable

TYNDALL'S

EXPERIMENTS.

r'270.

fraction of' thc wavc-Iength, t))c raction v-ouh)

be matpria))y
iesscued. On tins account wc rnight
expect grave sound.sto travel
a
through heterogeneous mdium less fn-<y t)t;,n ..)ctjt.cya~n.h,.
Thc rcnection of sound from sur~ccs
scpamti,~ portions of
ot
d~crcrt
~s
dcns.tics hn.scngaged the attention
of Prof Tynd-d!
who h~ dcvi.cd suvcm)
striking c.pcri.nonts in iHu.stmtionof thc
.sul.jcct Bor cxa.np)., sound fro.n a high-pitchcd rccd was
conducted t)mn,gh a tin tube tcwant.s a sensitive
f!,une, w).ic]~servit
ns an n.dicator. By thc
of a c.al-ga.s ~a.no issui,~
intuition
from au ord.nary bat's-~iny humer Lutwecn
thc tuhc and th~
~.ns~ve f!a.nc, t)~ grcatcr part of th~ cHect
couJd bc eut oif
Not dy so, but by holdillg thc uamc at a
suitaUo an.de thc
sound cou!d he rui!cetcd thmugh anuther tube in
.sumdcnt nua'ntity
to excite a second sensitive ~une, which but fur
the interposition
of the rcnoctuig Hamowuuh! havu rcinainsd
undi.sturhcd.
Th prGccdiug expressions
(JG), (17), (18) hold good in cvery
case of rc.ffcct.oufrom a 'dc~er'
medinm; but if th vclocity of
sound bc grcator in the lo~vc-rmdium, and ti.c
angle of Incidence
cxcccd the critical ang)c, becornes
imaginary, and t).e formuh,.
require modification. In thc latter ca.~ it is
i.npossib)c that a
rcfracted~avc should exist, sincc, cvcn if th.
aug)e of rfraction
~-c UO its trace on th.
p~o ofscparation rnu.st neccssardy
outrun the trace of the incident wavc.
bc written in p)a of
If
T~e~e?~ t~ite

th symboilcal quations are

from whieh by
discarr!ing th in~ginary parts, wc ohtain
'yo~/)(f,3rddition,p.282.

270.]

TOTALREELECTION.

79

~ese formu~ indice total reficction. Thc

disturbance in t).o
is uot a ~vc at a)t in t).o
o.-dina.y scn.sc,~n<[at
asec~d
.sl.ortmdium
d~anco from thc surface of
sopu-aticn (.. ncgative) becomes msc~bJe. C~IcuI~ting from
(12) and cxprc.sing it iu
tcrms of and wc Aud

shcwing t!mt ti~c distm-bance does not penetrate into the second
mcdtum more thfui a. few
wa.ve-Ienn'ths.
Thc difFercnccof phase bctweeli the
rcHc-ctcdand th inddcnt
waves is 2e, wlicrc

Since thcro is no loss of energy in reflection and

th
rfraction,
work transirutted in any time across
any aroa of the front of thc
incident wave must be cqmd to th work
transmitted in the same
tn.ie across co.-rcsponding areas of the rcrieetcd
and refractcd
waves. TItesc con-cspondingareas are
plainly in tlie ratio

L~W

0F ENERGY

VERIFIED.

f~O.

-~n
u.e .n.rgy cun~~ou, ~d
agres with the rcsult of nu.itipfy.ng togethcr tbu two bonudary quations
(13),
WJien tho vcloeity of
propagation is grever ia th lowort!~u
th uppcr mdium, aud the
angle of incidence excecd. thc
critical ~g!c no enorgy i.s tr~mitteJ
into the second inediu.n.
othor words thc reficctiou is total.
Tlie method of tho present
invcsti~tiou is substantia))y ~c
a.ne as th.t~pjoyed
by Grccn :n p.per on the ReHectJand
Icract.onof Sound
T). case ofpcrpcndicu]. incidence
~.t u.vc.st.g.ted byPoi.s.~ who
cbtained
(3) and (2.t). ~).eh I~d i.wcvcr bc.n fonnui~corrc.sp.n,);n.
airc.dy givcu tj Y.u
tho rcf)cct.cn of
Ligi.t. lu a sub~~cnt ~oi/Poi~
c.n.s.dcrcdt!.e gnera! c.sc
ofobtiquc incident H.nitinghimscif
.owcvc, to g.cous n..di. for ~ich
Boyie-s law hoids~od,
d
a
.y
.cryccmpi.c.tc. ana]ysis an-ived at a rcsult
cqui~icnt to
Hc a!so vor.hod th~t t),c
-').
nergies of the rcftected aud ro~-acted wavcs make up that of Htc
hicidunt wavo.
271. If
cxtcnded do~i'to]y
v~d. w,th comp],.to
.nlfon.ity in its ~ch.
do
parties
tr.ns.n, ted wave is prop.g~d onw.rd.
'B~jf
at
eonti,I)y.
c'gc i~ thc co.sihi)ity,
o both p.r of th w.vo wHIbc
density
throwll back, .nd ou .riv~t
tlie b~
~.=0; will hc divid.) iuto t.o parts, o e
11~ ~t u..di..n, .ud eue
r.ficct.d b~/to b. ag.~
d.k.d at ..=
,d .su.n. Hy
f.JIo~i~. thc pr~re.s.s of thcse
c ,.t.u .f the pr.hi..
may be .bt.i.L;t,
~ctodand
traus.uttcd ~.s bcing
c.,npoundcd of an incite
'c.
~"th~
is
tlie .~hod
/r''i"
In
u.s..diy ad.pt.d
()p,ics for thc c~poud
'i~
;?
.s~
-s
huu~tiy c.p)anK.d but it ducs not
appc.r to hve any ad~nt..m
.a
~cr. straightf~rd
auaJy.si.s. r. t.f.c f.Uo~. ri
.Ld
cu~nc ou~dvc. te thu a.hcre
th J
~~n
."cdu.,n is .u.njar ill its
thu
p,
~<
= ~rM;.
rlr l'In,tilrrt,
<~
~r'~i"

7'fu<.~))).~
]~jg
t. Jt. p. iJOg
~7,
1. X'l',

;i17,

1,-j;Jl,

IQjf)

271.]

PLATE

OF FINIT

TIIICKNESS.

g).

Jn or(]cr to pass to rea! quantities, t)tcse

expressions must ))c
put into t).c form TPe" 7/'c~ &e )-e< we fmd corresponding tu
thu incident \va.vc
R. If.

(;

S2

REELECTIONFROM A PLATE

s).cwingt)~L cxccptf.thc
.hcr~i.n r,f p].e,
!ncd)U)nm~ht as wut! ijavc Leen unifonn.

r-27l

tho ,vho)cof t).c

If bc small, wc h~-c
apprcxHnatcfy for L),crcf)ccto.]

wavc

~-mu]aapp)yu.g~hcnthop!at.cist)uuin
eom~n.son ~ith
tlie w~c-!cngt).
cos~. it appears t)~t for a givc.
Sinco
=~
ang!c uf incidence th a)np!itud~ varies
or as
im-c.-sclyas
Jn any case t!.c rcHection
vanishe.s,if cot~</ =
t)~t is, if
~bc.n~n

intc.g.r. Thewnve i.sthcnwhu))yt,-au.smittc.).

At i.crpe..<)icL,larincidcn, thc
I~cusity or t].c rcftc.etion is
express) by

Let

us

nc.v

suppt

t!.it

th

.ecuud

mdium

i.s

ii.eu~prcssibic,

.so

OF FIXITE

371.]
t}):Lt

oui'

cxprcssKtM

TlifCKNESS.

83

bceomc.s

shcwing ))ow thc amonnt of roHcctiou dupcuds upon th )'uhttiv.

xut.s.sc.s
ot'HucItt~)!U)titiesof Ute mdiaas itavc vu)utncs mthc r~tiu
of
It is obvinus tt):).t tho {icud mdium huhave.sfiku n.
rigid body !md act.s on]yin virtuc of its ino'tiu.. If thi.s bc sufticient, thc ruf)uetio)iinay bcMmc!scn.sibtytot~).
Wc hi~vo)it)Wto cousidcr thc ca.suiu whicti (~ is i)nagin:uy.
I)i thc symbo)ic:dexpressions (5) aud (U)co.sft~amt t si~n~ :u'c
arc pure hnagiuanus. Thus, if we suppose that r/=~
0=~ :u)(tintroducc tho not~tt'm of thc hypcrbuhc sine and cosine (~ 170), w~ gut
rca], wltitua, a+-,

a-

0j
81

~0 LOSS 0F ENEUGY.

H'S.X
transl11ittcd(Lecotilit

f~

encrgies of
flll"tllc wholo
eller~y of tlle incident

front
aro c(lual fur ail threc
=~F"
it
~h" (12), (B),
tiulls

cil'

waiL:-

is ully )lOCrssnryto uclcl tlic

cquati0118(7), ol7 in e(ll1a-

272. 'l']esc calcnlalions of

rcilcction fln(1 refr<~etionumler
he carricr1fllrtlJer, Lut
~Ltllcir intcrcst
lic
rat/1er oytical tll:llI
licotisti(-,11.It is important to beal'
iiiii](1
tk
110
l'tlLlyy

1~ destroyed

I)y ail)'

IlIlJnber

of

i,oflectioll,4

S~
tiull ulwa~~s
iu HllOthcr,
l'l.:al'l)(!al'ing
011 aCC01llltof tlm
;~rc.,tt dit}'ercl1ccuf' (1C11SlLICS
l'efluction is
liclnitll11attl!r, ~oululs 1)1'o(IticL-(l
iu ;lir arc )lot
cvsily coml11l1lliI:WUI)(ls,
il>1Iudel'water,
IYIIUSI:
U1'1bj11
wit.c.c<. );
(liflictilt,yiu air. ~L~"iofwooL),ora.
mctaJiic
distances with very littlo loss.

g
CHAPTER XIV.
GENERALEQUATIONS.
273. 1~ conncetion with 1)~0 gnrt probicm of aurial
vibrations in thrcc dImcusiujiHone of thc first
questions, whic]~
natu!U)y of~rs itsdf, is tit dctern)i;iaLio)iof thc motion in an
uniimitud atmosphre consquent upon
arbitt-ary initia.1 disturbances. It will be assumcd t))at thc disturbancc is
small, so
that thc ordinat-y~pproxiniatuquations arc
applicable, aud furt'hcr
that the initial vu!oclticsare snch as cnn bc dcrived from a
vclocity])otcntial, or ( 240) that tllere is no CM-c/
If th Jatter condition bc violntell,the probtcMiis onc ofvortex
motion, on whieh
wc do not enter. \Ve s]iaH idso
suppose in the m-st place that no
cxtcrna! f..rccs act upon the uuid, so t)iat tlie
motion to bc
i.s
duc
invcstigatcd
soldy to a disturbanco actuaDy cxistin~ at
a titnc (<=0), prcviou.s to which wc do jiot
push our inrp~-ius
Thc mct.hod that wcs!.a!l c.npjoy is not
very dinTo-entfromthat
of Poisson hy whom thc proHon was first
succcssfn!!yattackcd.
If M.,
bc tlie initial velocities at the point a-,
z, and 80

thc initial condensation, wc hayc (

2-),

by which
ditrereutlal

t!.G ituLi~l
coefficient

values
with

of thc

volocity-potcntial
to tunG
respect

and
arc

of

its

Jetermined.

1 Sur
l'int~~tin..
quc.]qncs <juation~ lindairos aux di~rcnccs
pM-ti~M
et p,u'hcul~remcnt
do l'quation ~n.rato
du luouvcmeut de. fluides
6iaBtiq~
~~i. <!t; <Y)~ft<x~t. m. p. 121. 1820.

8G

ARBITRARY

INITfAL

DISTURBANCE.
.jiujmjM~tjjt,.

f273.
)Z/,),

TLc pmUc,n L~furcus is to

dctcrminc at tm.e <!from thc ab.~
rnt.iat v.du.j.s,and tho ~ncnd
c.juaLiott~pp)ica).!<!.t !) t!m~ .-m<!
pinces.

Whcn is ]<nown,
it.s dcrivativcs~ivc
ZD t)tc ccrnponcnt;vetocitics at.
:nyponit.
Th~ symhuHca)sohnion of

(:3),,my hc ~-ritten

~hcrc~and

~fu-c twofu-hit.nu-y fun<;t.ionsof.~ ?/ nn<1 ~=

/-T)'
the
init;)
Tuc.,nnc.c~an.Lh
vah~of~u.t~~hi~
.shaJi < c.tc ~d
7''rc.spuetive)y, it is oniyncec~rytuub.scrvc.'
th.~twhen<=0,(-l.)~i\'cs

which
thc question of the
Into~rctation of od.) po~.s
in
equ~on
Hlly"
syllibulic
\lol]y evcl.
In the
was a faction of .r
c~y, .vo s.w ( 245)
c.~hc.-c
t!.t its ~uc
fur .nyp.int..t
on thc nit
tin.c~.pe,.dcd
valllcs .f
at
thc p.i.ts ..h~
and
cc-n~
we~
at
-I.
+~, and .s w.Iiy i.
~J
a)I uthcr points, In thc
pru.sentca.s. Lho.si.np)~ supposition
is
point 0 clepenclson
t..n
acs of nu.t at points .situatcd on f)~ ,f,~
of t~
spf.crc .-I.o.scc.nLrc i.s Oand radius~
,n< as tLcrc eau Le no
!)~
r'ccovc.ra~her,wc.rc
's
Jcd to ~nvcst.
t),, cxpr..ssioa for ti.c ~can vainc
cf
.c.onovcr
a sphuri.a) surface In tcr~ of
~csuccc-.iv r.
tud coc~cicnts of the funetion at thc
ccutrc.
By the syn.boJieat f~ of Mac!ri.'s tj.ccrcm
the value of
.t~~
~:7
point P on
tlic
l'
be
writtell
lay

273.'J

1AR13ITRAP.Y INITIAL

DISTURBANCE.

87

tho centre of thc .spftcrc 0 bcing thc ori~ui of co-ordmatcs. In

tlie mt.c~t'ft.tion
over tlic sm'fuce of tlie sphre
lm
~o
I)chavnas c<)t)nt;tnts;wc may dnote them tcmpora.rityby 7~, ?~
so t]):it ~=~+~+?r.
fui dment of
Thus,)- buhig the mdius of thc sphre, and
its surface, sitjcc', bytt)u syntinctry of tlie sphre, wo may repla-ce
/.<;+/)/+/Mn. 1 tlie
riuncttou
<-Z 1
le
s:unc
sa1!lC
11l1CtIOll
of
0
functtonof
tlllctlOn
.? Wlt
wnho~t
lOUt
..)y
by
nny
nllY

+ +

attcring- thc rcsult of tlic intc'gra.tiou,

Th mea.nvalue of~ovcr thc surface ofthc sphre of radius 7' is

tttus expt'cssedby tlie l'csutt.uf the opcratioa un 2'' of thc symbol
\
space,

with respect to Mgu!ar

0
of, if ~~o- dctiotc intgration
or,

or in words,~) at any point at timc < is thc mcan of thc initial

vaincs of 6 over tlic surface of' th spho'c described round t)ic
point ni question with mdius r~, thc wlioc niultiplied by <.
By Stokcs' rdc ( 95), or by simple Inspection of (5), we sce
thfit t)ic part of dupcmling on t)ic initia values of <~ mn,y bc
(tcrivcd from t))n,tjust. writtcn by diffcrcntia,tmg with respect to <
and changing th arhitnu'y fmieti'jn. T)iCcomptetc value of at
thuc is thcreforu

VERIFICATION

0F'SOLUTION.

~3

whichi.sPoissoii'sresuIt'.

On aecouutofthe importance of the

prsent p,.oL!cn. it ,n.v
~t p
~r it s.t..sHc.s
t~t
7?
th gcncr~ dIHfcrenti~
c~ua.ticn (3). T,~
f~
~.rin, i..iud
t o'L[.u I
t~p~ntth~.tt~culy,~d
Mytnbohccquatton

Ncw~~
satisHcd.

i. t~
~J~

Sincc th second part

of I, cLtaincd from th Rr.t hy dl~rcn
t~un, ,t aiso n~ustsatisfy thc fundamcntal
quation.
Wkl~ respect to thc iuttial
ecuditions we sec th.t ~hen is ma.tr
to
zro
iu (8),
o~ual

~S-~
mutiacAc l'h~~ik, 1, 517,

in
1876.

~rchh~

273.]

LIMITED

INITIAL

DISTURBANCE.

89

of which the first term boeomcsin the limit 7~(0). Whcn < = 0,

sinec the oppositcly situatcd c!u)nent.scancct In th Hmit, \vl)cn

th radius of th spherical surface is indefiuitely ditninistted. Thc
expression m (8) thcrefbrc satisfics th prcscribcd initial conditions as wcl] as th gnral din'ercntial quation.
27~. If t))Cinitial di.sturbancc be couuned to a spacc ?~ thc
Intgrais in (8) 27~ arc zro, nnless somc p:t.rtof thc .surface!ot'
tfiu sphre ?'=<~ bc includcd within 7'. Lct ~bc a puint cxtcl'n.'d
to 7', ?'t a.nd ?'~th ]':u1ii(~'titc h'n.stand grcn.tcstsphcrcs dcso-ibcd
about C)A\'hic)teut it. Thoi so lon~ as ft<<)\,
rcmn.inscqual
to xuru. Whcn H<lies butwccn ?\ und)'
may bD nnitc, but fur
v.'duesgrcatcr tban
~)is ag'ain zro. Thc di.sturbanceis thus at
aoy monKjnt coinncd to those parts of'.sp~ccfor which r(<is lutcrncdiatc butwccn and ?' T])CHmit ot'thc wn.vcis th oivclopc
of sphres with radius at, whosc centres arc situatcd on tho surface
of T. Whcn < is smat), ttus System oi' sphcrcs will hve an
extcrior cnvelopc of two shccts, ti)c outer of thcse shcets bcin'r
exturior, and thc inaer intcrior to tho shcH formcd by thc asSL-mh!agcoft])Csphres. Thc outer shect funns th outer limit
tu t])c portion of thc mdium in which th dUatation is diffurcnt
from zro. As < Incrcascs,t!ic inncr stjcet contracts, and at Jast its
opposite sides cross, and it changes its character from bcin"- pxterior, with rfrence to the sphres, to interior. It then cxpands,
an<tforms t)ic inner Loundary of thc shull in whieli t!)e wav of
condensation is compriscd'
Th successive positions of the
boundaric's of th wavc arc thus a scrius of piU'aUGi
surfaces, and
each boundary is propagatcd normaDywith a vcloeity cqual to
If at tho time <= 0 thorc bc no motion, so that thc initial
disturbancc consists merety in a variation of dcnsity, the subsquent condition of things is expressedl)y th first terni of (8) 273.
Lct us suppose that thc original disturbance, still hmited to a
imite rgion y, consists of condensation onty, without rarfaction.
It might be thought that thc samc pccuharitywould attach to thc
C<<w&.
'y'f)i:.
ix.p. lu.
Stokes,"DyHUtUtcul
Thcoryof Difraet.iou,"

CASE 0F PLANE WAVES.

Lresulting

wavc

Dn-oughout

thc

who)c

of

f'2~.

Its

subs~uent
course
but
as Prof.Stores bas rcniarked, such
a conclusion wouhtbe erron'eous'
For vah.es of th tune )ess than
r, -a t].c poientia! at is ~roit then becoiacs ngative
(~ being positive), aud continues nc.~at.ve unt.I ttv.uu.shesagain wheu <=?-,
after wl.ie)i it ajw~ys
ren~n.s equ.'d to xero. W)u!c is
t!.c mdian at
is in a .statoof cmKk.nsatIou,but di.nini.shiug,
a~ incroases a~u.i to .en, th~
statc. of t]ic med.um at is onc of nu-cfactioa. Thc
wavc prom
gatcd .-utwants cnnsi.st.sthurcforc of two parts at )cast, of ~-hich
thc first is cundc-nscdand thc. !ast nu-efi~.
~~),atcvcr ,nay hc t).c
charactcr of thc. ~mai di.sturhanec.wit)un
t),c <i,d y;
of6
tlt anyextc.ja) poillt (J is t,),c.sa.nu .LSt)h.
Initit Ya]uc,an<[therciurc, sutcu a~=t)ic n.c.aucon~n.sati.)).
thu pas.sa.reof
dun)~thc wavc.,d~cndiu~ on U.u
is ,ero. Undcr"'thc
int~r~
hcad of sp)~.nca! wavcs wc shali h~vu occasion tu
rcturu to this
suhjcet ( 27!)).
Thc gnral solution cmbodied in
(8) 273 must of course
embracc the part.cuiar c~su of p)anc ~vc.s, but a few
words on
this application may not bc supurfiuous, for it
nn~ht appear at
first si~ht that the cUcct at a ~iveu point ofn
di.sturbancc i)iiti:U)y
connnL.dto a sficc of thc tncdiunt oyclosed between two
paraHet
planes woutd not pa.s.soit- in any tinitc titnc, as wc know it ourrj.t
to do. Let us suppose for
is zero throug-In~t
shnpHcity t!.at
and U.at wit)mi thc. slice in
<p~atiou thu initial value -A is
constant. Fron. the thcory of p)anu waves we ]n.ow that
at any
~rbitrary point the di.sturb.-inccwi)) fioaDycca.se aftcr thc
iapse of
a time .such that <!<i.s c<{u:dtu thc distance
(~ of thc point
undcr considration froni th furtller
boundary of the initially
disturbcd rgion ~diiie on thc oti.er hand, sincc thc
sphcrc of
radius ~< continues to eut the rgion, it woutd
appear D-omtbc
formuk
t))at thc di.sturbancc continues. It is truc indced
gnera)
remains tillite, but this i.s nut incon.sistentwith rest.
that
It
will in fact appear on cxatnination that t]~c me:m value
of
t)~
mdius
of
t!.c
multipiicd by
sphre is th .same whatever may
bc t)te position and sixe of tl.e spbcre,
provided on]y tjiat it
eut con.plcte)y through th
n.gion of original disturbance. If
a0f/, cpis thus constant v-ith respect both to
space and timo,
and accordingjy tbe jnedium is at rc.st.

275. J)] two dimc.n.stons,when i~

indepcndent nf~ it might
bc supposcd t)):)t the
corrusponding f..r!nu)awou)d hc obtainedby

275.]

TWO DIMENSIONS.

91

thc circlc of cqnd

smiplysubstitutingfor th sp)ici-cof radius
r:uHus. Tilis, howcvcr,is not ttie C!isc. It mity bo provcd t)iftt
thc mcan value of a function 7~(~ ~) uver thc circu!)ifei')juCtj
of a
circlc ofnidius ?' is ~(~7) 7~, wh<jrc z==~T,

difcnng t'romwhat is rcquired to satisfy thc fun(t~mc~t:Uquation.

Thc correct rcsntt applicn.btnto t\vo dimensions nifi.ybc obtn.incd
i'rom thc gcncnd frtnula. Thc ctoment of sp)tct-ic:Usutf.LCocM
ibc
) i ) )'f~'<~ Ayhc'rc
1/1
nmy
rcpfaced by COS
t', 0 arc plane poln.t-co-onh'1'
is thc angle bct\vccn the tangent phmc and that in
na~'s, an~
which th utotiut) takcs place. Thus

where thc intgration extcnds ovcr thc arcn, of t!ic circle )'==<
T))e other tcrm might bc obtalued by Stokes' ruie.
This solution is ~p))lic:iblc to thc motion of Jaycr
of gns
bctwccn two piu'a.lld -phuics, or to th:it of fm unUoutcd stretchc~
mutnbmnc, wtuctt dpends upon thc s:unc fundftmcuta.t quation.
270. From thc sohttion in terms of Initial con<Htions wc May,
as usu:d ( (!G), t]cdnce t)io eH'nct of' a eont!nn:dly rcnewcd disturhancc. Let us suppose tliat throughout the spa.cc
(w)tich
will uttirnatc~y be !i)ade to vanish), a, uniform disturba.ucc
T!)0 rcsultiu~vatuc
cqua) tu (~')~, is communicn.tcd at tmic
of 6 at timc t is

whcrc ~S'dcuutcstlie part of thc surface of the sphcrc )' =(<-<')

SOURCES0F SOUND.

j~~Jt'J.
f-S/C.

nttc.rccpLcdwitinn 7~,a qnantlty which vanishes,unicsa a

be
(t
bctween
tho
compresscd
nan-ow Iimits r, and )~.
Ultimatcly
~y Le rcplaccd by 7.
and ~(<') by
and th rc~);
sult of tlie Integ~tion with
is found by
respect to
writinc
Ifcnce
(tlicvo)umc)fur/,9<

tha thc di.tnrbancc

o~inating at ~ny point .sprcadsitsc!f
.shown~
.yun..utn~I)y ni ali d.recMui.switli votocity and
with an~plitudc
thc
<)i.s~ncc.
mvcr.sdy
Sincc ~ny nutnbei- of particuJar
v~y.n~
sduL.01~n~y bu superposcd, tho ancrai .sotutiuu
oi-t!ic c.juatiou

?-dcnoting thc distance of thc cicmcnt ~F.situatcd

at
y~from
0 (at wftich <~i.s cstimatcd), and <P
thc value of for thc
1')
at th ti.nc t
point
CompJc.ncutary terms, satisfying
t].oui,r), ait sj~cu th cqu~tion
=
u~y of course occur indepcm)c))t)y.
In our prcvious notation (
2't-J.)

.Lndt ,s assumcd t]~at

A'~+r~+~i,
complote di~rcntiaJ
l'orc~ undcr ~'].o.scaction t).c ~cdiu.u coutd not
a.)ju.st itsclf to
arc
c~n .bruun,
c.xcludcd;as for in.stanee, force uniform in
n~rr.
ultudc and direction witllin a
and
space Y~
outsidc that
vani.shing
spnce H.c nature of thc disturl.ancc d~K.ted
by is pe,.],~
sn by con.sidcring t).c extrcn.c ea.sc
~c.n
,i,
c,
a
.s.naH
through
voininc, ~-).ichis suppo.scdto <!i,nini.sh~it),out
ht.nt, wh~ thc magnitude of incrca.sc.sin such a. manncr that
tite
whoc ctrect retins finite. If thon we
integrate quation (~

2 7 G.]

IIARMONIC

TYPE.

through fLsma)t space Including thc point at which <Pis ultinm.tL'iyconcuutrat.cd,-c Und in th limit

shewing that thc effect of <t*may bc rcprcscntcd by a proportiona)

introduction or abstracti(j))of nuid at t))Cp)ac:uin (~uestio)i. Thc
simptcst source of sound is thus an:i.io~ousto a, fucuslu th thc'n'y
of conduction of tioat, or to au lectrode in thc theoryof cicctricity.
277. Ttiu prcceding oxpressions n,rc~cncmt in respect of thc
relatiou to timc of tite functions conccrncd, but in :d)nost idi t))C
applications th.'ttwc s!)a!l Iiave to ma)<c,it will bc convcniL'ittto
analyse thc tnotion by FouricT'.sthcorcm and trc~t scparatciy t))
sunp)e harmonie tintions of varionspcriods,ai'tcrwards,if ncccssat'y,
if si)np)u Inu'conpounding tii(ircsult.s. Thc value of < and.
!nonic at cvcry ))oint of spacc, may bc cxpresscd in tite form
.7)!coH(?~+e), 7t' and ebcing indcpcudott of timc, Lut variable
from point to point. But as in such cases it of'tunconduccs to
simplieity to add thc term ~sin(M<+), malung :dtogcthcr
or jf~c".e' wc wiMassume si)np)y that aU thc functions
T~e~
winch hntur into a problon are proportiottat tu c" tite c"ef~icients being in gnral cotnpiex. After our oprations are coniptcted, tit re:d and imaginary parts of thc expressions can hc
separa.tcd, cither of theni by it.setf constituting a solution of t)te
question.
and thcdirfercntia!
Since~) is proportiona! to e' ~=?r<
cquation becomcs

To adapt (3) of th prcccdmg section to thc prsent case, It is

?*
that
tlie
substitution
torcmark
oniyncccssfn'y
of~fbr<Is
1.
or e'
thus
(-'iTeetc'd
byiutroducing tl)e ftetor c

VERIFICATION0F SOLUTION,

f'277

and tbc-solution of(I) Is

to winch may bc addcd

any suiutiun of
Ifthc
r~.o.)

di.stu.-bing forces bc ~iin

thmnghwtnchthcy~

t!.c
s~n,

~c-]L.n~e-n,aybcro.nov~fn,,n~dcr
and
at n. sufhcicnt.
distance

+ ~=0.

with

buvuty.s)na])i,.co.np!U-i.son
ihuint~d
wc

may

<),u

U.c.samcph.and

ta~c

In ordcr to vcriiy that (3) ~tisfie.s the

diHcrcutiate.n.atioit CI)
wc m~y procccd a.sin thc
thcory of th common potcl.ti~I. Coneue cicluent of t).e
iut~rat ut a time, wo Ii~c Hrst to
Mdcrn~
shcw that

-satires

~+~=~
~j,
course is to cxprc.s.s
in p.]ar co-ordi.i~es
.s.m~st
thc c!cmcnt itsuffas
pulc, whcn it appels tha.t

referrcd to

t),at (3) .s.tisncs

= (), ,t a!) points for
+
cr
w!.K.). va.n.shc.s. 1,, t).c case of a
puint at w].ich do~ not
van.si. ~e may put out of account
thu elc.nen~ situatc.) at a
~H<cd).~ncu(ascont)-ib)[tij)g on)y icnns .satis~ing \7~+~=Q)
and io.- t)jc etcmcut at an innuitesin.a!
dlst!mccrun)~c c-bv
umty. Thns on th w)tu)e

cxactfy as in roi.sun'.s t!ieorcni for the cnrnmon potenthd'.

't-!cTi)o)nso))nnJTmt.'tiA'<.7'/<f/i;.JtU.

278.]

SURFACE

DISTRIBUTIONS.

95

278. Tho ci'uctofa. force <I~(ti.strihtttcdovcr a.surface <S'ma.y

})Cobtitiaed ~s a.iituitin~ case frotn (!)) 277. (iTis rcpI.LCcdby
~~(~S', (lunoting thc t))icknc.ssof thc l<t.yer;{uid m the Ihuit w
TiLus
may writc b =

la
i.sthcsn-niou
ihctwos[<]cscf~butthcrc
Thcvatocff~
(ji.scoutimuty I)t its dcriva.Mvcs. If<~ Le dra.wu outwards n'ont ~S'
nortn!tl)y, (4) 276 ~ivus

OC

INFINITE

PLANE

WALL.

L-,
[278.

Thu siUti mcLitmI is applic:tb)c to thc ~oncra.! case whcn thc

tnotiou is uot rcstrictcd to bu sinipic harmonie. Wc hve

is dcnoted t!te nornuU velocity at the plane

whcreby F~j
for thc c)emcnt J~ at the time t (?'- ft), that is to say, at a time
)'ft antcdent to that atwhidt~iscst-imated.
In orderto complte th solution of th probton for t!)e
unntnitedmassoffhndJyingononcsidcofaniniiniteptane,we
hilve toadd th most gnerai value of<j&,consistent witit F=().
Tins part oftite (juestiun i.sIdentical v'it!i tlic ~eucra) probtem ui'
rcileetion from an inHuit.cri~id plane'.
It is vident that th cft'cctof tl)c constraint will be reprc.scntcd
by th ititroduetion on th other sidc of t))C phuic of fictitious
initial displaccment.sand forces, formin~ in conjunction witit thse
actually cxl.sting on t!ie nrst side a systcm ped'ccdy symmetrieal
with respect to the plane. Whatcver the initiai values of~ and
<~may be belon~in~ to any point on th first side.thf .samemust
bc ascribed to Its ;'M~7~c,
and in )ikc mantier whatever function of
th thne inay be at thc fu'st point, it nmsLbe conceivedtu t)e th!
satno function ofthc titneat theother. Underthesu circunistances
it is c!eM' that fur aH future dnic ~) will be synimetrical wiLh
respect to the plane, and thcrefore th normal vefocityzro. Su
far then as thc motion on th hr~t side is concernet),thcre wili he
no ch:)ngc if t))e plane be removed, and t))e Huid continucd
iudcnnit.L'iyin ail directions, provided thc circumstnnces on tht;
second side arc t)ie exact reftection of t))ose on th first. This
being nnder.st.ood, thc gnral solution uf th problem for a
nuid boundcd by an Inrinite p)ane is containod m thc formuiaj
(8) 273, (3) 277, and (8) of th prsent section. They give thc
resuit of arbitrary Initial conditions (~ and <~),arbitrary nppticd
forces (<P),and arbitrary motion ufthe plane (J'').
Measurcd bythc resn)ting potential, a source ofgivcn magnitude, i.c. a source at which a given introduction and withdrawai
ofnuid takes place, is thus t\vice as enectivc when close to a rigid
plane, as if itwcrc sitnatcd in th opcn, and thc t-esult is ulti1ROR.
'P('if.n,J~;fnMi'(;)'crn~j)f)~<f'cyut)'~)f<t.vn.

DOUBLE SHEETS.
_u-.
"E,

m. 1

'1..

~7

~t~~T
source
concentrated in a point
close h~u
a corresponding
e norm~ ~n
'J.tJ3o~ui.Hccof th plane itself,
Th operation of the
plane is to doubte th effective pressures
which oppose th
expansion and contraction at th
.so~.ce
'ed~
double
total energy
and since this energy
is diffused r'~
only
space,
~y~
~~(~
~1~
amplitude,
or potential (~
2.I~),
Wc will now
.upposc that instead
the prcscribcd
of~=0,
condition at tho infinite
pL.ne is th.t ~=0. In this case the
fictitious distribution
on the second side of
of
the plane
rnust s If fT"" of
on
first side, so
the sum of
values at two
corresponding points is always zero. Tins .ocurcs
that on th plane
ofsymmctry itself shall vanish throughont.
Lot us next suppose th.t there
arc two parallel
surfaces
~h~r
small ~~1
and ~'at th
o~n~
second
is equal
opposite to the value
of
on the first. In
thcre is by (2) finite
crossing
change
in the value of
to thc amount of
but in
d7&
'e th
crossing
same finite change occurs in the reverse
direction. When
is
reduced without Iimit, and
replaced by
will
th
dit
same on the two sides of the double
shcet, but there will be
diseontinulty in th value of < to thc amount
At tho
of
same time (1) becomes

po~or~
""T"' sign
Positive on the one side and
native
R.II.

~"rfacc-potentiaL
on th other, due to the
7

98

SPIIERICALWAVES.

[278.

must be underaction of the forces at <S. Th direction of

stood to bo ~o~e~'f~the side at which <~ is to be estimated.
279. The probleni of sphefical waves diverging from a p&!nt
has aiready been forced upon us and in some degree considered,
but on account of its importance it demands a more deta.iled
treatment, If tlie centre of symmctry be taken as ple the veloreduces to
city-potential is a funetion of )- only, and ( 241)
Th equation of freo motion (3) 273
or to
~+~
?'
~r.
ar
</7' r a?'

As in the case of one dimension, thc first term represents a wavo

advancing in th direction of r increasing, that is to say, a divergent wave,and tlie second tcrm represents a.wave convcrgiugupon
tlie pole. The latter does not in itself possess much interest. If
we confine our attention to tlie divergent wavc, wc h:t.vc

the same relation as obta.insin th case of a plane wave, as might

hve been expected.
If the type bo harmonie,

279.]

CONTINUITY

THROUGH POLE.

99

If a divergent distance
bc conflucd to a
sphencal she
w.t~r.:
~iLhcur. ~L
t!
udthur condens~ion nor
veioctt~ tho chamctcr of the w~e Mlinutcd by remarkable relation, first poiuted out hy Stokcs'.l, From
quations (4) wc have
shewing th~, th v.~ue of/(~)
I, the s.-unc, viz. zero, both
in.s.deand outside tbc she]] to which t])e w~ve is
IImited. Henec
if
M
bo radii less aud grcater tliau tlie
by (~
and
cxtremo
radn f tho shel!,

wnich is thc expression of th relation referred to.

As In 274
we sec that a conden.scdor a mrencd wave
cannot exist alone.'
When th radius beeomes grt in
comparison with th thickness
tho variation of m thc
intgral may be negleetcd, and (8) thon
expresses tliat th ~eu~ condensation is zero.
In applying th general solution
(2) to dcduce th motion
rcsuttmg from arbitrary initial circumstances,we must remember
that in its prsent form it is too
gnera! for the purpose, since it
covers the case in which the ple is itself a
source, or place where
fhud is ~ntroduecd or wit)idrawn in violation of
th quation of
contimuty. The total cnrrent across the surface of a
sphre of
radius r is 47n- or by (2) and (3)

an quation which must hold good for ail

positive values of th
argument".
By th known initial ch-cumstancestho values of Mand s arc
determined for th timo ~=0, and for a!l
(positive) values of r.
J'/t)7.jt/<!f/.
xxxiv.p. 52. 1819.
Tho Mtutiou for sphcricat vihrntions
mny bo nbt.nucd without tho usa of (1)
by superposition of tmins of piano wavGs. reiated
similarly to th polo, Md tra.
veUtnK ont.warda iu nJl diroctions cymmetrioUly.
73

100

I~ITEAL CIRCUMSTANCES.

If thse initla.1 values be rcprcscntcd by M~,and

f~.m(3)M-!d(~)

[279.

wo obttun

by which thc function yis dctcrmincd for at! ngative arguments,

and th function for a. positive arguments. T)ic for)Tiof for
positive arguments follows by moins of (0), and then th whoc
subsquent motion is dctcrmincd by (2). Th form of F for
nogative arguments is not rcquired.
Th initial distnrh~nccdivides itself into two parts, Irn-veHin~
in opposite directions, in cach of which ?'~ is propa.gnted with
co))sta,ntvelocity (?,and tlie inwn.rdstr:ivc)!ing wa.vcis cot~tinua.Uy
reflected a.t thc poJc. Since the condition to be thcrc satisfied is
?'<~=0, thc case is somewhat simHfu-to tha,t of a pa.ra.Heltube
tcrminated by an <T/je~end, and wo may thus pcrha.ps botter
understand wloy thc condcnsed w;ivc, arising from tlie Hbo'n.tion.
of a mass of eondensed air round the polo, is ibHowed immcdiatcly
by a wave of rarefaction.
280. Returning now to th case of a train of harmonie waves
travelling outwards continually from tho polo as source, let us
invcstigato th conncction between the vctocity-potentia! aud thc
quantity of nuid*which rnust bc supposed to be introduced and
withdrawn altcrnately. If th velocity-potcntial be

wc !)a.vc,as in thc preccding section, for the total cun-cnt crossing

a sphere of radius ?',

when ?-is smal) cnough. If th maximum rate of introduction of

<))!)(!bc dcnot.cd by J, titc correspondingpotential is givoi
by(l).
It will be obscrvcd that. when tite source, as mca-suredby
is
finite, thc potential and thc pressurc-va)-i:tt!on(proportiona! to <&)
are inDnttc at thc po]o. But tbis dncs not, as might for a moment
be supposcd, i'npty an inDnitn mission ofcncrgy. Jfthc pressure

280 J

ENERGY

EMITTED

PROM GIVEN SOURCE.

101

bc divided into two parts, one of which lias th same

p!tase a.s
t).c vclocity, aiid tlie other tlie .same
phase as th acceleration, it
will bc found that tl)e former
part, ou which th work dpends,
is ~nite. The mRnite
part of tlie pressure does no work on th
who)e, but mcrcly kecps up tlie vibratiuu of tt~e air
immediately
round t!ie source, whose effective inertia is
indenhitely grt.
We will now investigate th
energy emitted from a simple
source of givea magnitude,
supposing for the sak of greater
generaHty tfiat tlie source is situated at tlie vertex of a rigid cone
of aotid angle M. If the rate of introduction of iluid at th
sourco
be A cos
we have

Of thc DgLt-hn.ndmcmbcr thc first tcrm is

a.nd
entirely pGriOfJIc,
ni the sccoud thc mean vatuc of sm":(<)-)
Thus in ttto
is
long l'un

It will bu remarked that when the sourco is

given, the amplitu<Jcvaries iuver.sc)yas M, and thei-oforc the
intensity Inversety
as M'. Fur an acute cone tite
intensity is greater, not on!y on
accouut of the dunimition in tlie soHd angtc
through which the
MathcmaUcttI
Cmubfid~o
187G.
TripusExMuinatitiU,

~02

SPEAKING TRUMPET.

[280.

sound is distributed, but also because the total

energy emitted
from th sourceis ~t~eJfincreaa.d.
When th source is in tho open, we hve only to
put M= 4.77-,
and when it is close to a rigid plane, <u= 27r,
Thc results of this article nnd un
iutercsting application ill th
thcory of the speaking trumpet, or (by tl)e law of reciprocity
10}),294) Itc~nng ti-uinpot. If th diameter of the large open
end be sumil iu comparison with tl)e
wa.ve-Iength,th waves on
arrivai 8uHcr copiuus reflection, and the ultimatu rcsuft, which
must dpend h).rgclyon thc prcise relative
Jcngths of th tube
and of the wavc, rcquircs to he determiucd
by a diifurent process.
But by suHicicntIy protouging thc cunc, this rdtcction
inay bo
duninisbct!, and it will tend to ceasc witen tho diameter of th
open end inchtdcs a large junnbcr of wavc-tongttts. Apart from
friction it would thei-cforebc possible
by diminishing c.)to obtain
from a given source any desired amount of
energy, and at tho
same thue by InHgthening t!ie cne to sccure th
unimpeded
transfureucc of this encrgy from th tube to the
surrounding air.
From the thcory of diffraction it
appcars that tl)c sound will
not fal! of' to any grt extcnt in a latral
direction, unless t]ie
diameter at th largo end exceed hn.If a
wave-lcngth. Th
ordinary explanation of the eftect of a commontrumpet, dcpending
on a supposod concentration of
7-~8 in the axial direction, is thus
untenablo.
281. By means of Eulcr's equation,

wc may casily establisli a

thcory for conical pipes with open ends,
Midogous to thM of Bernoulli for pam]Iel tubes,subject to the same
Innit~tion as to th stnaDncssof th diameter of tlie tubes in comp~isoti witli t]te W!ive-Icn~thof the sound. Assuming tliat tho
vibration is stationary, so th~t
p-~ is cverywherc proportioual to
cos /M~wc gct from (1)

281.]

THEORY Or CONICAL TUBES.

103

Tho condition to be satisfied a.t an open end, viz., that there 13

to be no condensation or ra.reiactiou,gives
= 0, so tha.t, if th
extrme radii of tho tube bo )\ and
we have
whencc by elimjnation of ~4 :7?,sin < (~ ?-J = 0,
=
or
where ??tis an integer. In fact since the form of the
general
solutton (3) and th condition for an open end are th same as for
a paraDc)tubo, the rcsult that the length of th tube is a
multiple
of th halfwave-length is necessarily also the same.
A cone, which is complete as far as tho vertex,
may bc treated
a.s if tlie vertex were an open end, sincc, as we sa.w
in 279, the
condition ?'<~= 0 is there satisfied,
The rescmblancc to th case of parallel tubes does not extend
to the position of th nodes. In th case of th gravest vibration of
a parallel tube open at both ends, thc nodc occupies a central
position, and th two halves vibrato synchrououslyas tubes open at one
end and stoppod at the other, But if a conical tube were divided
by a partition at its centre, th two parts would have different
pcrioda, as is ovidcnt, becausc th one part differs from a parallel
tube by being contracted at its open end where the effect of a
contraction is to dcpress th pitch, wItHe the other
part is contracted at its stopped end, \vhercthc effect is to raise th
pitch. In
order that the two periods may be th sa.me,th
partition must
approach nearer to the narrower end of thc tube. Its actual
position may ue determined analytica!!y from (3) by equating to
zero th value of
o'
When both ends of a conical pipe are closed, tho
correspondin~
notes are determined by eHniinating jl
between the cquatious,

~04

TWO SOURCES0F LIKE riTCn.

[281.

if y, and ?'j, bc very great, tan'' A-?',and

tan'' /<r, :t.re botti odd
multiples uf ~7r,so th:t.t ~-r, Is a. niu!tip!c of .,as thc thcory
"f par:~)~ tnh~ t.'tp'ires.
282. If there 1)Gtwo distitict sources of sound of the same
the vclocity-potenti:d at a poiut
p'tctt, situftted at 01 tuid
7'whofie distances f:'oiu
arc r, and ?' may bc expresscd

whcre A and
arc coc~eicnts rcprcsenting the magnitudes of
thc sources, (which without luss of genorality may bc supposcd to
hve th same sign), and N rcpreseuts tlie retardation. (considered
as a distance) of the second source reiatively to th nrst. The two
trams of spherical waves are in agreemcnt at any point P, if
whcrc 7n is an intcger, that is, if P lie on any
~+ 'x ~'t=
one of a systeni of hypcrboloids of rvolution ha.ving foci at
and 0,. At points )yh)g on the intermediate hyperboloids,
represented by ?a+ af p-~= + (2/~ + 1) tlie two sets of waves
are opposed in phase, and nentraHze one another as far as thcir
nctual magnitudes permit. Th neutratization is
complete, if
7\ ?',= ~1 ~C,and then th density a.t7~continues pGt-manentIy
unch~nged. Th intersections of this sphre with thc system of
hypcrboloids will thus mark out in most cases sevcml circlcs of
absoiate silence. If the distance C\Oj,between th sources be grt
m conparison with thc Jengtil ofa wave,and th sourcestljcmselves
bc not very unequal in powcr, it will bc possible to
dpart from
t)te sphre ?'j :?'~=~1 Z?for a distance of several
wave-Icnn'ths
without appreciably disturbing thecquatityof intcnsities, and thus
to ohtain over finite surfaces several a!ternations of sound and of
almost complte silence.
There is sone diniculty in aetun.I]yrea.!isiuga satisfactoryInterfrence of two indcpendent sounds. Unicss the unison 'be extraordman]y perfuct, tlie silences are only momcntary and arc
co!)sequcnt)ydinicult to appreciatc. It is thcrefore bcst to employ
sources whicli are mechanicaHyconnected in such a
way tijat th
relative phases of thc sounds issuin~ from them cannot
vary. The
situp!cst plan is to rcpcat thc first sound by renection from a Hat
W!t!I( 2G9, 278), but th cxperiment tbun Joses
somcthin'r in
dircctncss owing to the fictitious charactcr of the second source.
Pcrhaps tlie most satisfactory furm of t]jc experimeut is that

282.]

POINTS0F SILENCE.

105

deseribcd in the Philosophical Magazine for June 1877 by

myscif.
"An intermittent olGctric'un'cnt obtaincd fronia. fork
interrupter
128
vibrations
making
pur second, cxcitcd by mcans of etectronagtiets two othcr forks, whose frequeney was 25G, ( G3,64.).
Tficse latter forks were placed at a distance of about ton
yards
and
were
apart,
provided with suitably tuncd resonators, by which
their sounds were reinfurccd. Thc pitch of the forks was
ncccss!n-I[yideutica!, since th vib~tions werc furcc(t by electroin~guetic forces uf absulutely thc s.unG period. With one car
closed it was found possible to define thc
ptuccs of silence with
considrable a.ccm-iicy,a motion of about :m incb
bcnig sufficient
to pro<)ucea markcd rcvival of somid. At a
point of silence, from
which the line joiniug tlie forks subtended an angle of about
GO",
the apparent strikiug up of onc fork, wltcn the other was
stopped,
had a very peculiar eH'uct."
Another method is to duplicate a sound coming
along a tube
by means of branch tubes, wiiose open ends act as sources. But
the experimcnt in this form is uot a very
casy one.
It often happens that considrations of symmctry are sufEcient
to indicate tlie existence of places of sitence. For
exampte, it is
vident that therc can bc no variation of density in the coutinua.tion of th plane of a vibrating plate, nor in the
equatorial plane
of a symmctrical sulid of rvolution vlhrating in the direction of
its axis. More gcner:d!y, any plane is a plane of silence, with
respect to which the sources are symtnetrictd in such a manner
tLat at any point and at its image in thc piane there are sources
of cqual intcnsities and of opposite phases, or, as it is oftcn moro
convcnientlyexpressed, of the sa)ue phase and of opposite amplitudes.
If any numbcr of sources in th same phase, whose
amplitudes
are on tbe whole as mucii ngative as positive, bc
placcd on th
circumfcrcnceof a circlc, thcy will give rise to no disturhance of
pressure at. points on the straight linc which passes tbrou~h thc
contre of the circle and is directed at rig)it aug)cs to its
plane.
This is th case of the symmctrical LcU( 232), which emits uo
sound in tlie direction of its axis*.
Th aceurate exprimental Investigation of acrla.1vibrations is
bosct with cousiderable diniculLles,wiuch have been
only partiaily
'J"/t~(5),m.p..lCO.

1877.

EXPERIMENTALMETIIODS.

f282.

Eurmounted hitherto, In order to avold unwished for

reflections
it is generally necessary to work in tbc
open aIr.wberRd~ate
sue)i as a sensitive namc, is dimcult of
~pp-u-atus,
management.
Another
impeduncnt arises from the prsence of tho experimenter
h.msdf, w)mse person is large enough to disturb
materially the
state of tlungs w))ich he wl.shcsto examine.
indicators of
sonnd may be mentioncd membranes stretchedAmong
over cups, the agitation being made apparat
by sand, or by small pendulums re~Ing
hghtiy aga.nst thcm. If a membrane be simp)y stretched across a
hoop, both its faces arc actcd upon by nearly the same forces, and
consequentlythe motion is much diminished, uniess the membrane
be hu-ge euough to cast a sensible
sbadow, in which Its hinder faco
may be protected. rrobabiy the best mcthod of
t!ic
examining
intensity of sound at any point in the air is to divert a
portion of
it by mcans of a tube
ending in a small cono or resonator th
sound so diverted being !ed to the
car, or to a manometric
capsuic. In this way it is not difncult to dtermine
of
places
silence witli considerabic prcision.
By mcan.s of the same k:nd of apparatns it is possible to
examine cvcn the phase of th vibration at
any point in air, and to
trace out the surfaces on which thc
phase ducs not vary'. If th
nterior of a resonator be connected
by flexible tubing with a
manomctric capsule,Nyliicli influencesa small
gas name, thc motion
of thc namc is rclated in an invariable mauner
(dependincron the
apparatus Itself) to the variation of pressure at th mouth of tho
resonator and in particular thc interval between the Jowest
drop
of thc name and thc lowcst
at
thc
pressure
resonator is Indepcndent
of the ahsolute timc at which thse effects
occur. In Mayer'a
expcrimeut two fimes were empioycd, placcd close
togetber in one
vertical Iine, and were examined witb a
rcvotving mirror So long
th assocmted resonators were
undisturbed, the serrations ofthe
two Hames occnp.ud a Hxed relative
position, and this relative
position was also maintained when onc resonator was moved about
su as to trace out a surface of invariable
phase. For further
dutails thc readcr must bc referred to the
original paper,
283. Whcn wavcs of sound
inlpinge upon an obstacle, a
of
the
motion i.s thrown back as an cho, and under covcr
rort.on
of tbc obstacle therc is formed a sort of sound
shadow. In order
hovever, to produce shadows in anything like
optical perfection,
1 Mlyor,P;,t/.
~), sLiv.p. 321. 1672.

283.]

souND snADows.

107

tho dimensions of th intervening body must be considerable.

Th standard of comparison proper to tho subject is tho w&velength of the vibration it requires almost as extreme conditions
to producorays in th case of sound, as it requires in optics to
avoid producing thcm. Still, sound shadows tlirowu by hills, or
buildings, are often tolerabiy compiute, and must be within th
exprience ofaU.
For closerexamination lot us takc first the case of plane waves
of harmonie type impinging upon an imtnuvable plane screen, of
infmitesimai thichness, in which thre is an aperture of any form,
tbc plane of th scrccn (.v= U) buing paraDel to tlie fronts of the
waves. The velocity-potential of tlie undisturbed train of waves
may be takcn,
If the value of

over the apcrturo be known, formula (6)

and (7) 278 aHowus to catcuh).te th value of at any point on
the further sidc. In tlie orduiat'y tlicory of (Ufra.ction,n.8givon
in works on optics, it is assumod that thc disturbance iu th plane
et' thc apertm'e is t!ie sfunc as if the Bcrt) were away. This
hypothesis, though it eau acvcr be rigorously exact, will sufHce
when tlie aperture is very large in comparison with the wavelungth, as is usually th case in opties.
For the undisturbed wa.vewe have

th integration cxtcnding over tlie area of tlie aperture. SInco

we sec by comparison with (1) th:bt iu supposing a.
~=2-nprimary wave brokeu up, with th vicw of applying Huyghens'
priuciptc, ~<Smust be divided by \?', and tlie phase must ba
a.cceleratcidby a qun.rtcr of{),period.
Whcn ?' is large in comparison with the dimensions of tho
n.pcrture, thc composition of tlie Intgral is best studied by the a.Id
of Huyghens' zones. With thc point 0, for which is to be
cstunated, as centre deRcribca series of sphres of radii increasing

108

nUYGHENS'

ZONES.

[283.

hy thc constant ditFerence~, thc first spticrc of th series being

of such radius (c) as to touch th pjanc of t!ic scr~-p']. On t')is
plane are tttus markcd ont a. sries o; circ)e.s,whose radii p are
givcn by ~+c'=(c+~\)',
or~=~c~,
vcry llcarly so that
the rings into witich the plane is divided,
heing of approximately
equal area, make contributions to cp which are approximately
equal in nnmerical magnitude and a)tcrrmtc)y opposite in Hign.
If 0 lie decidcdiy within thc projection of the area, tho first tcrm
of thc scrics rcprcscnting titc Intgral is finite, and the tenns
-\vlnc!tfollow are atternately opposite in
sign and of numerical
jnagnitudc at first nearly constata, but !d'terw!U-d8
diminishing
to
as
th
gradnally zro,
parts of thc rings intercupted within thc
aperture become less and less. Tiic case of an aperture, wliose
boundary is cquidistant from is cxceptcd.
In a sries of this description any tcrm after thc first is
neutralizod almost cxact!y hy haf tlie sum of tliose whicii iinmcdiatety prccedc and follow it, so that th sum of the who!eseries
is rcprcsentcd approximately by hatf the nrst tcrm, which stands
over uneompcnsated. We sec that, provided a sumcient uumbcr
of zones be ineluded within th aperture, the value
of at tho
point 0 is independent of th nature of the aperture, and is therefore th same as if there had been no scrcen at ait. Or we
may
calculate directly thc effectof th circle with which th
system of
zones bcgins; a course wliieh will have th
advantagGof bringing
out more clearly the significanceof thc change of
phase which we
found it necessary to introduec when th
primarywave was broken
up. Thus, let us conoivethc cirele in question divided into innnitcsimal rings of equal ai-ca. Th parts of due to each of
<~
thse rings are eqoal in amplitude and of
phase ranging uniformly
ovur haf a complte period. The phase of the resultant is thercforc midway betwcen tftose of thc extrme lments, that is to
say, a quarter of a period bchind that due to th lment at
thc centre of thc circ)e. Th amplitude of the resultant will bc
less than if aU its componcnts had Leen in th same
phase, in
th ratio ~sin.<;
Tr, or 2 -n-; aud thereforc since th area
of the circle is TrXr,ha]f the encet of tlie first zone is

thc samc as if tlie primary wave were to pass on undisturbed.

283.]

HUYGHENS'

ZONES.

109

When the point 0 is well away from th projection of the

n.pc!n-c, tho rp~dt is fjui~ .~ffacut. Th scr'CHrq'rc'scnLingt))o
intgra,! then converges at botti ends, and by the samc rcasoning
as before its sum is sccn to bc approximately zro. We coneludo
that if th projection of 0 on the plane a:==0 fait within th
aperture, and be nearcr to 0 hy a grcat many wavc-Iengths than
the nearest point of tho boundary of t!)e aperture, thon th
disturbance at 0 is ncarly thc samcas if thcrc were no obstacle at
a)! but, if the projection of 0 fa)l outside the aperture and be
nearer to 0 by a grcat )nany wavc-Icngthsthn.)]th nearcst point of
thc boundary, t!)oi the disttu'bancc at 0 practicaHy vanisjics.
Dus is the thcory ofHonnd rays in its simplest form.
Thc argument is not very di~ront if the screen he oblique to
thc phme oftho waves. As hcfo-e,tlie motion on tho further side
of tlle screen may bc rcgarded as due to th normal motion of the
particles iu the plane of tbe aperture, but this normal motion now
varies in phase from point to point. If the primary waves procccd
froma source at Q, IIuyghens' zonesfur a point 7~arc th sries of
where
citipses represented by + = P~ +
and ?-are
the distances of any point on thc screen frnm Q and 7~
rcspectively,
and Mis an integer. On acconnt of tho assumed smallness of in
comparison with ?', and ?'~ tho zonesare at first of equal area and
make cqual and opposite contributions to thc value
of <~ and
thus by t))Gsamc rcasoning as heforewe may conclude that at
any
jx'int decidecllyoutsidc the gornetrica! projection of the aperture
the disturbance vani.shcs,while at any point decidcdly within tlie
geometrical projection the disturbance is thc samc as if thc
primary wavc had passcd thc screen unimpedcd. It may be
rernarkGdthut the incrcasc of area of th Huyghens' zones due to
obliquity is compensated in t!)e calcuiationof the intgral by the
correspondingly dintinisbed value of th normal veloeity of the
uuid. Tho cnfccblementof thc primary wave between the screen
and th point .P duc to divcrgency is representcd by a diminution
in th area of tbe Huygbcns' zones below that correspondinrrto
plane incident waves in the ratio ?',+ ?'~ ?'.
TIicre is a simple relation between the transmission of sound
tljrough an aperture lu a screen and its reflection from a plane
leficctor of thc same form as the aperture, of which advantage may
sometimoa bc taken in pxpcriment. Let us imagine a source
sirnitar to (~ !Uidin the samc pliase to be placed at (~ th t~e of

110

CONDITIONS 0F COMPLETE REFLEXION.

L.
[283.

Q in th plane of th screen, and Jet us suppose th~t th screen

is
removed and repfaccd byap!.tc whoseform and
position isexactiy
that of tlie aperture; then we hnow that tt~e
effect at of the two
sources is nnu~ucnccd by th presence of thc
plate, so that th
vibration from Q renoctcd from the
plate and th ~hration from
Otmnsm.ttcd round tho p]ate togethcr make
the same vibraup
tion as would be rcccived from (? if thcrc werc
no obstacle at a!!
Now aecordmg to th as.su.nption winch we
madc at thc b~inof
this section, the unimpcded vibration from
niug
Q may bo
as
regardod composa ofthe vibration that nnds its
way round th
and
of
ti.at
which ~ou!d pass an
plate
aporturc of Die sa.ne form
in an infinite screen, and thus thc vibration
from Q as tr~mittcd
throu~h the aperture i.9equal to the vibration from as
<2' reriected
from th piato.
In order to obtain a
nearly complte reflection it is not nocesthat
the reneeting p]atc inctude more tl,an a
sary
small numhor of
zones.
In th case of direct reflection the
Huyghens
radius p of
the first zoue is dctenniued by tlie
quation

wherc c aud c, arc th distances from the

rejeter of thc source
and of the point of observation. When thc
distances conocrned
are grcat, the zones becumc so
large that ordin.~ry wdfs are
insufficieut to give a con.ptcte
rc.Heetion. but at more moderato
distances cho. arc
cft.u nea.-Iy perfuet. Tf~ area neccssaryfor
con~cte reflection depcuds also upon th
~avc-!cngth and thus
it happens that a hoard or plate, which wou!d
be quitc inadquate
to reflect a
musical note, may rcficct
a hi~ or
very
fairly
brave
sound
tlie
of a high ,vhi,t!e. In
experiments on reHection by
scrccns of moderate size, the
principal cliflictilty i. to gct ricl
uf
th
d.rcct sound. Thc ~p~t
suSc~ent y
plan is to reflect
the sound from an eicctric beit, or other
fairiy steady source,round
.un
the corner of a large buildin~
28~ In th preceding section we have
apphed Huy~hens'
to
th
case
where the primary wave is
principle
supposed to be
broken up at t!.c surface of an
in~ginary phu.e. If wc reaiiv
know what thc normal motion at thc
pkne is, we can calculate
1 ~~)!.3/<t~. in.
(5) p..1C8. 1S77.

S84.]

DtVERGINGWAVES.

th distnrbance at any point on th further side

by a ri~oro~
process. For surface other ti.an tlie p!ane the
problem ].as no<
beciisolved genemlly; nevertheless, it is not
difBeu!ttoseethat
when tlie radii of curvature of thc surface are
very grcat in comparison witli thc w~e-Icngth, the e~et of a normal motion of an
Gtement of the surface must be
very nearly th same as if tho
surface were plane. On this
understanding we may employ tho
same mtegral as before to calculate tlie
aggrogate rcsutt As Il
matter of convenience it is
usually bcst to suppose thc wavc to be
broken up at what is calied in opties a
that is, a
~e-~r/ac<
surface at every point of which
the~Me of th disturbance is 'tho
same.
Lot us considcr tho application of
Huyghcns' principle to
calcu!ate the progress of a given
divergent wave. With any poiut
at which th disturbance is
required, as centre, describe a sries
of spheres ofradii coiltinuaHy
increasing by the constant dinfercnee
tlie first of tlie sries being of such radius
(c) as to touch tho
given wave-surface at C. If 2i; be the radius of curvature of tho
surface m any plane through 1' and C, the
corresponding radius p
of the outer bouudary of the
zone is given bv the ennnHnn

If the surface be one of revolution round

th arca ofthc first
n zones is
and since p2is proportional to M,it fullows
that the
zones are of equal area. If tlie surface be net of
rvolution th
arca of the rirst zones is reprcsentcd
~p'f~, where is th
azimuth of the plane in which p is
measured, but it still i-eniains
true tliat tlie zones M-c of
equal arca. Since by hypothesis thc
normal motion docs not vary
rapidiy over ttie wave-surface th
dtsturbances at P duc to th various zones are
nearly equal in
magnitude and alteriiatoly opposite in sign, and we conclude
that,
as in th case of plane wavcs, th
aggregate effect is the haf of
that due to th first zone. Ti~e
phase at Is according!yretarded
behmd that prevailing over th
given wave-surface by au amount
corresponding to tlie distance c.
The intensity of th disturbance at P
depeuda upon the area of

112

VARIATION

OF INTENSITY.

[284.

thc first Huygtiens' zone, md upon thc distance c. In th case of

svmmetrv. wohavo

'n*~ 7r\7i~
'c"+c'

which shows tliat tl)0 disturhancc is less than if R were innuitc in

tho ratio J~-t-c J~. This duninution is thoefcct of divo'gcticy,
and is the samc as Avouldbc obtfuncd on the supposition that the
motion is !unitt;d by a co)nc:dtube wttoscvertex Is a.t the centre of
curvature ( 2CG). Whcn th surface is not of rvolution, tho
v:due of ~"p~M c may Le expressed in tGi'msof thc pnnc'ipa.!
radii of curvaturo 7~~and
with which
is connectud hy the
relation

so that tlie amplitude is diminished by divcrgcncy in th ratio

~/(jf~+ c) (7~+ c) ~Jf~, a.rcsutt which might bc anticipated by
supposing tite motion hmitcd to a tube formed by normals dt'awn
through a sn~U coutour ti'ttcedon thc wave-surface,
Although we !)ave spoken liitherto of divcrging waves only,
th preceding expressions )t)~yalsobc a.ppHedto waves converging
in one or in buth of the priacip:).!planes, if wc a.tta.chsuitable
signs to
and7)~. In such a,casetlie arca of the nrst Huyghens'
zone is grea.tcr than if the wavc were plane, aud the intensity of
th 'vibration is correspondingly increased, If th point jP
coincide with one of th principal centres of curvature, the
expression (2) becomcs innnite. The investigation, on which (2)
was fuundcd, is thon insurhciolt; ail tlint we :n'eentitled to afirm
is that the disturbancc is tmtch greatct' at ~'than at othcr points
on th samc normal, that thc disproportion incrcascs with thc
frequeucy, and that it would becoincinfinit for notes of infinitcly
high pitch, whose wavc-tcngth woutd be negtigible in comparison
with the distances coneemed.
285. Huyghens' principle may also bc applicd to Invcstigato
th reflection of souud frum cnrvcd surfacctj. If thc materia.1
surface of thc rL'ncctoi' yicidcd so compictely to th arial

385.]

REELECTIONFROM CURVEDSURFACES.

1]3

pressures that the normal motion at

were the same as
.1 won!.)have been in the absence of every point
thc reHector, then the sound
waves would pass on undisturbed. Thc
retlection which actually
ensues when the surface is
unyielding may thcreforc be re~arded
as due to a normal motion of
each lment of the
renectorerma)
and opposa to that of the
primary waves at th same point, and
may be nwest.gatcd by the formuh.
to p).nc surf~ecsin tho
propcr
manner of thc
preccding section, and subjcet to a simiL-u.limitation to the relative
magnitudes of th wavc-Icngth aud of tho
otitcr distances concerncd.
Th most
intcrcsting caso of reflection occurs wbcn the
surface is so s))apc<!as to cause a
concentration of rays upon a
part~tar po.nt (P). If t),,
ori~naHy from a simple
source at (2 and th surface be an
eHipsoId of revolution havL
and ~t].e concentration i.s
compfete, the vibration
itsfoc~aW
re ected from cvery eicme.t of th
surface being in the samc
ou
arnval
Pb se
at <2. If <? be
innnitely distant, so that th
nc,dcnt
are piane, t),c surface becomes a
paraboloid havin~
its f.cus waves
at P and ~ts axis
para)!cl to th incident rays. We must
not suppose, however, that a
symmetrical wave diverging from
O.s converted hy rencctioa at th
cHipsoi.hd surf.~e into a
wave converging
in fact it is
symmetrica!Iy
upon
P,
spher~cal
to
see that th
casy
ii~tensity of th convergent wave must he
~erc.t
Nevertheless, when the wavelen.th
very small in con~parison.vith t!~ radius, th dinereut
parts ofis th convergent wavc hccotne
approximately indepeudeut
of one another, and their
progress i.s uot materially aneeted by
th faiture of pcrfcct
symmetry.
Th mcrcase of Juudne.ssdne to
curvature dpends upon the
area of rd)cct.ng surface, from which
di.sturba.K.csof unitbnn
phase arrive, as comparud wit). thu area of t!.e hrst
Muy<d.ens'
.one of a piane rcH.ctor in thc samo
position. If thc distances of
t).e rencctor from th source and
fro.n th point of observation be
cons.dcrahfe, and th wave-Jcngth bc nul
very sma)!, tj.e nr.st
zone
is atready rather large, and tbcrefore in th
Muygbens
ca~
of a reOector of moderate
dimensions but Httie is muned
by
it concave. On t!.e other
i~uaL in laboratory exprimenta
~d~
when the d..stances are moderatoand
th .soundsen.ptoycd are of
..=h p~ch.
th ticking of a watch or thc
of etectric
cracking
concave
p. L.s
reHectorsare very enicient and
give a distinct cunccuttatton of somd on
particu):~ spots
R. JI.
8

114

FERMAT'S
PRINCIPLE.

[28G.

28G. Wc ha.vc secn that if a ray procccding from

passes
aftcr rcncction a.t a. plane or curvcd surface throngh 7' thc point
7t'atwhici< it. mcots t]tc; .surface is dcu.:r!nmcd hy th condition
that ~/?+-K.P is a minimum (or in sonc cases a m:).xinium).
Thc point Il I.sthcn thc centre oftitc systcm of Hnyghcns' xonc's;
thc amptitudc uf tho vibmiion at .<tc'pon)s npon thc arca of tLc
first zone, and It,s pha.sc dc))(;)n)s upon thf di.stmtcc ~~)'+ ~7~. If
thet'e bo uo puint on t)K; mn'facc of Utc runccto)', for which
()/~+ 7~~1s a maxitunm or a nlininunn, th .systf'm of Huy~tcns'
xoncs bas no centre, fmd tito'c is no ray prucc~din~ from Q w])ich
an'ivcs a.h aftt;r rcHcetion fron thc surface. In )ike tnanno' if
sound bo n'ncctcd ])iorc than oncu, th course nr a rny is dctcrmincd Ly tin' condition that its ~)H))u Itj-n~'t!)hctwccn any two
or a ininitnmn.
points is maxinnnu
Thc same prioci{))c may hc apnik'd toinvcsti~atct.hc r~Y/e~'o~
of sound in a mcdimn, whosu )ncchanic:d pr<~pcrt.icsvary gradua~y
fnnn point to point. T!)u variation is supposcd to bc so s)o\v
that no s~nsih)u rcHcotiou occurs, and this is nut incf'nsistcnt
~ith ducided l'tTractionof t))u rays in tra\'L'))it)~ distances winch
im.')udt.! fL vcry ~rL'at, !tnnd)(.'r of wavc-)t-n~t!).s. It is vident
that \vhat wc arc n<'w conf'crnod with i.s not. mcrr'iy t))C )cn~th
of th r!'Y, hnt a).sn tho yotocity \it.h ')nch thc wave trav(.s
a)on~ it, inasnnu'!) as this vetocity is no )o))~'r constant. Thc
conditinn to hc satisth'd is that tho time occuph'd hy a wa\'c
in tra.v~itin"' :don~ !t. ray )x't\ct.'n nny two points s)):d[ bc
bc th< votooity of propam!txi<nu)H'~r a mini!!]un) so that, if
:U)d (~' an L'iemunt f thu Icngth of t)~c r~y,
"atiun at any p"int,
Tins is FcDnat's
~hc condition ]nay ~c cxprcsscd, Sj' t~'t/==<).
tum;.
principtc of tcast
T))o fnrthor duvelopoment of this part, of thc suhjcct wou.)d
Icad us too far into t]ic domain of ~coniL'trica] optics. T))C fundi).mcnta) assu)npti"n of thn s)naUn(;ssoft))cwavL'-h'n~th,onw!)i<')l
thc doctrine of r:'ys is hui)<, havin~ a far wIdcrappHcation tu thc
than to thosc of .sonnd. th<j task of<h'vc)(~pin~
phcnorncna of t~'ht
ils consc'qucnces nmy propt'rty he tuft to thc cnitiv.'t.tors of thc
sudiolls thc
the m(;t))od.s
nwtJods uf'
a-rc
Il! flJlIIII\'ing
In tl
t)~'
fo)to\vin~ scctiotis
uf optics
optics a!'c
sister seic'uco,
SiHII'
sc'h'ncc. Irl
< tw<~isoiat~'d qut'sti~ns, \)to.s<; acoustica! intcrcst
an~hcd toonr'
issutnri~nt (odcm.'nh! th''irc~nsid.'r:Ltion io <))c prsent \vork.

287.]

WnrsPERINO
GALLERIES.

115

287. Onc of th rnost

strikir~ of thc phenomena conno~ted
with tbcpr')~tionof
sound withi'idohui!).
isL.h~
prescnted by ".vhispering gai.-ries," of which a ~ood ~nd
easity
access.b!e cxampie is tu be found in th cirodar
th ba~
of the dme of' St Pau)'s catt.edral. As to ~aicryat
th prcise modo of
act.on Mou.stic:d auD.oritics arc not
c-ntirc.Iy ~.rreed. In the
of
opinion
theAstn.M.ncr
Roy:d' thc eft-t is to"bc ~cribcd to
rcHcct.on fro.n t],c stu-faco of the dme
ovcrh~ut, ftnd is to bn
oh.scrvcd at th poh.t of
t.heg.-dtcrydiametricdty
opposKo to t).e
s.n-ee of .sou..<).
Evcry r~y procccding f-om a niduint point fmd
rcHcctcd from thc surface of a
sphcrica! rcHcctor, will aft(T
re~.enon .t..r.s~t
that diamctcr of thc sphorc which contins tho
This.)iain<.tcr is infect a d~graded fonnofoneof
'antpomt.
t))etwocin).sti<'surfais
touchai hy.sy.stemsofray.s
in~c-nora),
ht')ngt)~iodofthuntrcsof
principal cnrvaturcoftftosm-fac-cto
tlie
my.s a.'u nor.na). Th concentration of rays on one
d.amcter Uu.s cf}~t(.d, doc-.s not
rcquirc thc proxinn'ty ofthe
radtant point tu thc
rcHL-ctm~surface.
J~d~ing fro.n .somc observations that I ])!wc madc in St Pau!'<!
an) dispn.scd to think that
wh.spL.rn~ga)Jury,I
the principal
phunon~nonis
Tbca))~nnat jo,,dnc..sstobc.p)aint'd.somL-w)tat.tiff..runt)y.
withwhich a~hi.sp.ri.sh-.ard
i.snotconnn.-d
to thc position
diamctricaify opposite to that occupicd hy the
w)..spcn.r, and thcrL.Jorc, it won!.) appuar, doo.s not
donend
maturndiy npo.i thc symmutry of thc dme. Thc
whisp~.r secins
to crccp rom~ thc
~aHcry itor~ontatfy, not ncce~ari)y ainn~ the
si.ortcrarc. but rath~- aton~ that arc towards w)iich thc
whispcrcr
h'(~.s. Thisis
:Lco))s~ptun<-cofthcvcrynnc(p)atau()ihi)ityofft.
wh.spcr in front of' and bchind thc speaker a phcno.nci.on '~bich
mayca.sdybcob.scrvcdinthcopcnair'.
Lct ns considcr tho course of thu
rays divc~i.~ fro.n a radiant
.situated ncar thc .snrfarc ofa
po.nt
rcOcctin~ .spt.crc, and fut lis
dnote thc centre of t)tc
and thc dian~.tcr
spi.crc by
pa.ssin.
tt.n.ugb
by .ij', so ti.at~ is tho point on the surface ncares~
<)
J) we ux our attention on a
rny which issues fro.n P at an
t).e t~cnt
au~e ~v.th
plane at ~,we see tbatarierauy
nu.nber of renection.s it continues to touch
aconccntric sp).erc of
rad.u.-j ~~cos~, so that the wi.oie conicat pcncil of
rays ~hich
'Airy(;S',<x;t,2)H!(..]i)i~n.].s7!,p.lJj.
~t/A~.(.)ut.t)..l.').s,)s77.

116

WIIISPERINGGALLERIES.

[287.

originaUy makc angles \vith the tangent plane at ~4 numcricaliy

is over aftcrwfu'ds inf'tuded betwccn th reHecting
Icss than
Ti)o
surface and t!~at of thc conecntric sphre of radius 0~ cos
usuat divergence in threc ditncnsions cntailing a diminishing
is rcptaced hy a. divergence in two dimenintensity varying as
sions, hke that of waves issuing from a source situatcd between
two parallel renecting planes, \vitb an intunsity varying as ?'
T)ie less rapid cnfccbicnicnt of sound by distanco than that usuatty
expcrienccd is tite luading feature iu thc phcjioincaa of whispering
ga.nerics.
T)~ thidtness of th shcct included betwccn thc two spherea
becomes ]<ssund )css as -/1 approachc.s and in thc Ihniting case
of a radiant point situatcd on tnc surface of tbc rcHcct.oris
approxicxpresscd hy C~I (1-eos~), or, If h(; sn~)), ~Oj-1
]uatc]y. Thc soHd ang)u ofthc p(.'nci),whicbdct.ct'mincsthuwltoio
is
amount of radiation in thc shcet, is 4-n-~ so that as
diminishcd wit)iout litnit t)tc intunsity becontjs inrinitc, ilsconipa.rcd wit!i thc intcnsity at a nuitc distance from a shai!ar source
in th opeu.
It is vident ihat this clinging, sn to spcak, of sonnd to tho
Rurface of a concave waU docs nut dpend upon thc exaciness of
the sphcncal foi'tn. But in thc case ofa. truc sp)icrc, or rathcr of
any surface symmctnca! witit respect to ~1~1',thcre is in addition
thcother kind of concentration spoken ofat thc commencement of
th prsent section which is pccuiiar to t!)c point ~f dianietricaHy
opposite tu th source. It is pro~ahte that in thc case of a nearly
spherica! dotne like that of St Paut's a part of thc obscrvcdcuect
dpends upon thc symmctry, though perhaps th grcater part is
rcferable impiy to titc gnerai concavity of tttc walls.
Th propagation of earthquake disturbances is probably arfected
hy thc curvaturc of tbe surface of th g)obe acting )Ike a whispering gaiery, and perhaps cven sonorf)USvibrations generated at th
surface of th !and or water do not entirely escapc th same kind
of inHuencc.
In connection with th aeoustics of public buildings there are
many points which. still remain obscure. It is important to bear
In mind that t))Closs of sound in a singie renection at a smooth
wa is very smaU, wbethcr tlie wai) be ptane or curved. In order
tu prcvcnt rverbration it may oftcn be necessary to introduce

S88.J

RESONANCE

IN BUILDINGS,

117

J
carpcts or i.hangi.igs to
absorb the sound. In somc CMCsth
prsence of an audience is found surncient to produco Die d~ircd
e<!ect. Jn t))o absence of a)l
dcadening matcriat H)e prolongation
of sound may bc very eonsi.tcraldc, of' wfticii
perhaps thc most
striking ex:L)nptcis that ~urdc.t by th Baptistcty at Pis~ ~hero
th nuttj.s of thc commoti chord
sung consccutivdy !nay bc Lcar<t
ringing on togcthet- for many sccorKis. AcconHng to Henry' it is
iinporL-mt tu prcvcnt thu rcpcatcd rencetion of sound baekwards
~nd forwards along tJK!
of a h.-dt Int(.ndL.dfor pnbiic speakn'g, w))ic)i may bc acconpHshcd by suit~hty piaeud objiquo
surfaces. I~ tbi.s way tho munbcr of rcHcctionsin a
given tinie is
iucreasud, :utd thc unduc prutongation ofsouud is checked.
288. Aimost tho on!y instance of
acousticalrfraction, which
ha.s apractica! inicrest, is t)ic dviation of soaorous
rays from a.
rcctihnuar course duc to bctcrogenuity uf thc
atlnosphorc. Th
van:Ltionof prcssm-c at diffurent levcl.s(tocs uut of itsctf
rise
givc
to rufraction, since thu
vu]ocityofsou.td is indcpcndcnt ofdo.sity;
but, as was first pointcd eut by Prof. Osbornc.
Rcynoids', thc case
JHdtfTcrcnt witb tite variations of
temprature whicb arc usuaHy
to bc met with. Th temprature of
is determined
t)~cat~nosp)~ere
prineipaHy by the condensation or rarfaction, which any portion
of air must undcrgo in its
passage frum onc )cve) to anotbcr, and
its tiormal state is t.ne
of'convcctivecquDibrium' rathcr thau of
uniformity. According to this view th rchtio.1 betwecn pressure
and dcnsity is ttuit exprcsscd in
(U) 240, and th velocity of Sound
is given by

if r. bc tlie vulocity at thc surface. Tho

corrcspottding rcht.tion
'h))f'r.~Mo<t-oc.l.s;n,p.ii{)_
187t.
~7'n)<'f-~tH~H/<t.f<y.Yni.xx!i.p.C31.
T))omnu)), ~t f; r~fn'e~'M c~Mt;t&)'tMMor <emp<M
u! the (Kmo~/t~.
.u<)))e/tM<tT~m())'r.18(jI.C~.

118

ATMosniERic nnr~CTioN.

['288.

bctwccn tcmpct'n.turc!:U)dc)cvatiun obtainciJ by mcansof quation

riUlS2Ki!s

v'ho'e

is th tonpcratnrc

a.<-t))e sm'fncc.

According to (4.) tl)c fa]t of Ictupc-raturc wouH he about

]"C.~))t. in ~;{(H'c(.t,whit;t)<Iuu.S])(~d!Ht'r)nuc)t front t!)~rc.su!t,.sof
(!t:d.s))('['s b.dtoouoh.st'rv~tiuns.
W))L'ttt.hL'.skyi.s('JL!:L)-,t])cf:(.))of
mot'c t':)pi([t,))an\v!)(')tthc
t~)nj)c)'!du)-edu)'i)~ thedayis
nkyis
c-h)u'ty,})uLt('w:L)'dssu)ts<tth<'t('ni))~)-at.ut'<jbt'c<)))K's:tpj))'()Xnn:Lt(;)y
cu)tst.!))tt'. l'i'ub:tb!yot)c)uarni~tt.sitisuf):unw:tr;nurabuvc:th:).n
b~uw.
Thc cxplnna<)()n of ncnust!c:)t rcfrnctmn as (~-pcnth'nt,
upon a.
va.)'i:ttiun of tt't~pcrnture with !K;i~))t is :d)n').st
(jxactiy thu .s:unu as
thi~t of'thc optic'al phcnnniRnon f)f mirage. Thc euryn-ture
(o"') uf
a ray, v'husu course is appfuxhnatdy !)nrixont:i), i.s casi)y cstitnatud
by t)ic tnctimd givc'n hy Prof. Jiuucs T))om.son". Not-nud phmcs
dmwn at two cunscentivc points a)t))i~ thc ray mc'ct at, L))uccutre uf
curvature aud a.rc tangcnt.I:),! to t))~ wave-surfa in its two conStjcutivc positions. TtK'p')rti()nsufraynatt.c\'ations2:ands-+8~
rcspcetivciy int~rccptcd behvccn thc mx'ntat pianos arc to one
anutt~r in t)K; ratio p pa))d atsf), sincc tm'y aru dcscribcd
in tliu ~nitj tituc, in Uiu ratio r
F+ 8K Huncc in ttiu li)nit

In tbc nonn;).! st:ttc of thc attnrtsphc'ro a

r~y, whidi stn.rts
))()n/ont.:L)ty,turns~-mduidty upwn.nLs,!Ut(t :tt a suthcicut ()i.st:t.ncc
[):s~so\'u[' t,huhc:)df't'a)tuI)s'')'c't'w!)osL;st:()))i.satthcsfui)c
ic\'(.-fnst.i)c.suu)-cc. Ji:'L)K!suu)'C(3bu (;t~;v:i.tc(),t.))Csound isltc!u-(t
n.tt))c.surface ut t.)~c.'u-thby!)-n.).HSof:trayw))ichst:u'tswit)i
()ownw:u-() inctitt~tion; hut, h' ))uth thc oh.survcr and thc
suu)'C(jLn()))tht.surf:)cc,<.)tcrcIsnodi)-uct)'ay,:uidt)tcs()u)i(~s
hcin-d, if'ttta)),by]nc:u)sof'<]i<}'ra<"t!on.
'J'hc observer maythcn
bc .said to bu sit.uatcd iti a souxd shadow,
n)L)iou~)i th~rc tnay be
no obst:ic)c lu titc direct liuc butwccu biuMuIf a.ud th source.
Accordm~ to (3)
'A"(ff)tr<Sept.20, 1877.
'Su(!KY('rttt.,0;;i/tt~ttMn/'j1/)r<~< ~tf<));.v.p~.t~j.218.

288.]

C'ONVECTIVE EQUILIDRIUM.

H 9

or thc ramus of curvaturc of ft hnrixnntid

my is n-Lont ton timcs
thc ])(.'i~))Lthn'ug'h witich:), b()<)y)nnstf:dtu)hk'rthn
action off
:i, \'(.')(jcityc(j))~I to thevc~ocityof
~).vity ni onh'rtn ;).('<[))))'!
s<'u))(). Jt't)tCt'tt;v!iti<)nsut'<))()!).s(')-v('rnmtnf<hcmn)rccbcz,
n.~d .?.~t)H:)'e:).t.L'st())ntimcuatw)ti(;)i
thus~md
c:mbe ttc~t'd
cLhei'tYisct.itu.nbyditt'i'it.ctujnI.s

Tt is not to hc suppose') that thc condition ofthc fdmosphcrc

I.sidways.su('hthatthc'r(d:tti('nhL't\vcenvu!oeityandc]nY:).ti')nis
'))ui)thtj.su))
thatcxp)'L'.sscdin(~).
i.SM])ini)~,t,])c variation ot'
tt']))))(')'atu!'c upw:u-d.sis ]uurc]-:q)i(); ontLcot.ht't'Itnnd.as.Pt~f'.
]!~yno)ds h:ts rctnm'k~d, whcn j-ai)i is f.;).)jit)g',a niuc!t s)o\vcr variation istobu
t~xpccLud. Jnt)ica)'eticrcg'if)))S,))c.'rc'thc
t)ig!tts
tuut'c influence titanconnrc]<)))g'a))(I.sti)),ra()iat.i<~nn)ny])avc
vcction i)t(]ut(;i')ni)ii)i~ tJjCL'quiiibnum uf tcinpcratum,n.nd ifsothe
propi~itinu nf .soi)))t! in :t. )H))'i/out~t (Ht'cct.io))wou)d hc f:t,vourcd
ui' t.])c atino.sphcrc.
by MK't)pp)'oxi)n:i<.t;!yi.sotht;ro):t.I o~diMun
Tito grnt.'nd dift'm'c)iti:U quation for thc p:).t]fuf :t ra.y, wltu
t))C surfas off.~uat vctt'eityat~ p;u':dt(.'l phLH.s,i.sreadityoljt:)!ned
fron thc ]:t\v of silcs. It'~ bc thu at)g!tj of incidence, ~sin
is
notnttcrcd
titccaso
by:t. r('f[-a<;ti!)~su)-f;)ce, fmdthcrcfhrci)).
supposcd rcmain.s constaxt .'don~ titu \v)io!c course ni' n. my. If x
thc constant value of ~siu~
Lot.h(jl)0)'!xonta)co-urdu~tc,and
bu calicd c, wc ~ct

120

PATII OF A RAY.

[288.

or, on cfueting th Intgration,

in which Fmay be cxprcsscd in terms of~ by

(3).
A si)np!cr rcsult will bc obt:uncd by ti~ing an
n.pproximn.tc
fur)n ot'(~), which will beiLCcuratuutiuugh to
rc-prc.scnt tho cuscs
tiic square aud Itigitcr
ofpmctica.1 intrust..Nu~tucting
puwurs of
s, wu m:t.y take

thc ori~!n of .f buing takcn so as to correspond with ~= c, that is

at thu place wi~re thc my is Lurixuutat.
ExprusHnj"-~iu tenua
of~,wc~nd

Thc path of each ray is theruforc a catenary whose vcrtex is

9l/
downwards th liuear parajuctcr
aud varies fi-om
C
Is ~('y-i)c
to
ray ray.
289. Anothor cause of atmospheric rcfmction is to be found
in tlie action uf wind. It Las long bceu known that suunds arc
gencraUybcttcr Ituardto lucward th:ui tu windward uf thc sourcebut thc faet rumaincd uncxpjamcd utitit Stokus'
pointcd out thut
tiie incrua.si))gvetucity of t.huwind ovurhead mnst interfcru wit)t
thu ructilmcar p)-up:)gationof sound
i-ays. Fro)n Fcrtnat's law of
!ast tinic it fo)Iuw.sthat tlic course of a
ray Ifi a movin", but
/~t'<1~.

~'<

ma?,

22.

289.]

REFRACTIONDY WIND.

121

otherwisc bomogenoous,mdium, is tlic samc as it wou!d be in n

mdium, of which a.U tin' pa.rts arc :)t rcst, if th< '/(~ty of
propa.ga.Lioubc inr-rcased at cvcry point by th componcnt of
th wind-velocity in thc direction ai' th ray. If th wind bo
borizonta), and do nul vary In the s:uno hurixont:d p)anc, ttie
course ofa ray, wliosc direction is evuryw!icrcb~t slIgtttLyinclined
to t)):i.tof thu wiud, inny be c:dcula,tcdou t!ie sfune principlus as
wo'c app!icd in tlie prcct;ding scctio!i to th citsc of :),y;t.riabtc
tonpcraL'u'c, thc nonmd vu)ucityofpropagn.tion at any point being
inct'L'asod,or ditnini.slicd,by thc luc:d wind-vu]ocity,according n.M
thc motion of thc sound is to iccward or to windw~rd. Tbus,
w)icnthcwind ittcrcusesovcrttcad,whicitm~y bclooked uponasthc
uonnal statu of tbings, )torizont:d my tr:LVt.i][ing
to windward is
g)~du;d]y beut upw.n'ds, and at a moderato distance passes ovcr
t!~ hc'ad of an observer; rn.ys tm,vuHing \vittt t)ic wind, on thc
othcr hand, are bcnt downw!U'(.)s,so
t)<a.tn.n observur to Iceward of
thc source bcars by a direct my which starts witb a siight upward
mciiuatio)i, !t.t)d))as tho advantage of buing uut of thc way of
obstructiuMsfor thc grcatcr part. of ils courue.
Tbc law of tufraction at a horizontal surface, in crossin~ which
tbe velocity of th wind changes discontinuousiy, is casiiy invcstigatcd. It wiMbc suftcient to consider tbu case in which th
direction of th wind and thc ray are in th samc vertical plane.
If0 be th augic of iucidcnco, which is also tbe angle bctwecn t!ie
pliulc of th wavcand th surfileof sparation, be thc velocity
of thc air in that direction wbicb makes tbe smaUer an'de with
tbc ray, and F'be thc common veiocity of propagation, th vulocity
of th trace of thc plane of tbc wavc on tbe surface of sparation is

wLich qu~ntity Is unclianged by tho rfraction. If thereforc ~7'bc

t)tu vclocity uf tlie wiud ou ttie second sidc, aud be tlic au"e uf
rctru.ct.ion,

which dirfcrs frorn the oi-dinnrynpticnl !a.w. If th wnni-vclopity

v:u'ycoutmnou.sty,the course oi' n. my )~~y bL;c;Llculatcdfrom tlie
condition that th expression (1) re)na.Iusconstant,

t22

TOTAL REELECTION

I!Y -\vrND.

[:289.

If wcsuppose
that !7'=(),thc ~reatcst~hmssib)e
valueof
~'i.q
Atastmt.utnwhcrc

~7 hasthi.svi).]nc,t!)0(1ircct.mnf)fthumy
whi(')t.s):u'tudt~uia])~]u6))nsh(!('o)nt'
pamjiutt.~titcrufracti))~
mu'i'.u~.s,:)))() :).st.)'at)]))i~h(;)-u~'))as
ViLtuccannotbc
agr~~T
])t;tmt('(tat.!L)).
T)tusari)yt)'avc])i)!~))j)\v;))-(].si)).sti)tairata.)i
)tH')Hi:t.<)n(~7r~tt)t))(!hn)'i~<))iisr('fi('f'tc<)])y!t.win(tovcr))t'a(l
('t'V(')()C'ity('C('('<iin~~t;t.t.givti)in(~),a)t(]
ttti.siodupL'm~'uOyof
~'ttV(.]<)cit)(.s<.t':))<('nnL-()i:t<(j.st.rata.'i'ct.akcatnnncricalcxamph-,
aH ray.s w))t).s~-upwant IxctitnUion i.s Ju.ssthan
])", a.ru totaDy
r('nuc[L'() t'y ~i"d~f't!)('s!U)n'a/!in)uthnt<i)!~att!)c
moderato
Sj)Q('dofL') !nl)t;.s])ft-])(n)r. 'J'hc (-flirts
uf.sm'aa~indonLhc!
]"){)!ati())iof.s)m(t
Ovu)'t.)t(;
cnmtotfiLiitohcvt-ryintjxn'tiU)).
nurf.Lceoj'.stii! WiLto- ~nun) )))uvin~ tu )cu\v:trt),
ht'in~ con~nc't
~t.;t.\v(;('n paraitt') )'f(!(.'ct.)!)g p!:nh!.s, diverses in t\vo (HxicnsionH
0!~y,n))<!)nay (tn'r~t'cruhc ])c:U(tat (ti.stiUH-s f:Lr~rc:ttc)' U~n
wou)d')t,))(;rwi.suht.;pusst))iL'. Anot)K'i-p<).s.si))]L;t;t]L'cL<)t't!)(-reH(;ctor
ovcrhc.'nt
h' )~H()~r .sounds :m(ti))i~ w)nch in HtHt air wontd
bcititcrcujttcd
by))i!i.s (~-othuroLst.
Fur the
lotcrvc!)))~
pr()(h]c(.i()n'))'t.!ic.s(.;))))).)on)('nait, i.suot ticccssaryt.hatthL'rcbc
:d).s<n('cfwHK) n<,).)h.S(')u'cuufs<))tn<t)n(,:).sa))pc:n-s:Lt,(~)cc
fr~m tJ.c ff))m of(~), mc-rdy t!)at
\'L')ucitiu.s
U
t.hc(/<ccut'
at,t:)Li)ia,sunit'icnt.v:L)m'.
T))(*()ifrc)'t.'])ti;d
't"t.iuntnt)it.'pat.hufaray,w])(jLithc\nnd\'cIocity~i!jconLmuuusIy\u'i:dj)c,is

lu cn)np.-n'ing(.'i) ~-ith ~) of (Le

p)-(.c~)i))g section, which
tMthc Ct)!T<j.spHU(ting
(.~nation forordixary
rufractio)~, wc must
]-(;nic)nhcT))):~risn(.w
c..x.st.at)t. Jf,furt!m.Si~t'ufobtaitiin'ra
(t'imite ~su)t,wcsu))jx.su
that. U.c huvuf v;u'iatiuu<,f wiudat
(iitTt.'t'L'ntluvds is titat uxprcssud t)y

289 J
wMchisofthcsnmc
cour.st.'oi'a ravis

HEYNOLDS'OBSERVATIONS.

')23

formas

T)ic
(II) ofthcp)-HC('d!ngscct;io)i.
acc()r()i))if)y!W:)tu))ary"~ti)('p)-f's't('s")d.sn
t"'t.t.)"r(~.sa.most!ntp<)r[:))<Ldist)nc;U()n),~t\t'u)tti)ut\opr~btcms.
Wftunt))C!rcfracti<))ti.softhcor(hnary)u)n),dt;p~ndi))~
upon a
variab)c\'c)()(;it,yofpropa~atiu]),thu()ir(;cti~nufar!)y
!nayhc
n'vcrscd. lu timcasc ')fat)n.s]))tL'ri(;(!fr:n;t.it)jt,()ucto:)()i)ni)mtinno)'temprature
up\Ya)'(t.s,t.huoo))r.sc ('t'arayisa
catcnft.ry,
w))nHCY(;rtuxi.st)uw))wan)s,inw)tic)~crdin.'ctiunt))t'r!y)n:)ybc
prnp;)gatt'<). Whcn thcf'traction
isthtGtuwind.whosc'vcfocity
i)~r(;:).si)))wa)-t)s,a(~r<ti)~tot)K!!a\v
cxprcss(;(Ii)t(f!)wit)L/3
po.s!t)\'r-, t~L'))at])<){'ai':)y,\v))().sc(1i!-u('ti<)))i.supwa.)-t),i.sat.s.)a)nng
~catt.)..Nywit)) v~rtGxd-.wnwanIs, buta ray wh~cdin-ction
is
duw)<\van[c;u)th.t tra\-L-tai<.ng t))is pat)), h) thutatt~r case th
Ycrtux ot'the cat~naryalong
whicit Oturaytravci.s
is dircctud
npward.s.
~().

I"H'(!pap(ThyR..yn()h)sa!rf;a(]yr(;f(-)-rc<Ito,anacco))nt
Ls~i\'t'n')f.S())n(!it)(rn;stin~<Xpurit)tcnt.s<'s))t.;cia!)y()ircct(;t)t()tcs(,
thu theury of rdraction by \vin)!. If wa.s fount) that
fn tho
<.)irrcti()nof't)tcwi)u!,w])t;tt itwasstrH)~,t)t(.;sou)H)(~f'anc!c(;tric
bd))cout<) )'ch~ardaswL-i)witht)tuh~<]u)tt))c~ronrt(l;).swh~)i
]'ai.s~),(.'v<'nw))(.nina)iuHuw\Yi).htht.ihu))
)u~h~ fr.xn vtewby
t))cs)<)pc()ft))n~roun<);an.)
uoa(iva))ta~v!)atu\rw:t.s~m)cd
eith<hyf)s<'(!n()in~tna))(;]~va(in))o!-r:)i.sin~t!tc))<j]).
Tiu~,wit)t
tht.; wit)doi.'L'rt!)(j~r:~st)tcso))ndcuu)(.[
hc ))car<) I-t-0 yards, atxl
o\'crH))()W:i!f)<)yiU'()s,<jit))t::r\vitht))L'hcadiift(.'doruntI)c~rou))d;
w)t(TGasatri~))t:m~]<\st('tI)uwi)Klon:d!of'(;!)sionHtitumu<~cwas
observer or t)tu bu))."
cxtt'n<)cdl.'yraisi)igcit))crt)t(!
"KJ(.ation w~ fouud
t.<;ta'cctthcra)~cof's()un(]a~amstt))0
wm<)in!t.ntucht)U)rL'tn:u-kc()]n:L)jncrt)):m:d,r!'dtt:U)"'h's."
"Ov~r<))(;~r:tss))(..snn))dc(.)dd))ch(~rd\Yit)tt]tL!).~<[ont))c
thc !K.)),.t.id :Lt:!()yard.sit.w;ts!ustwit)t
~oundnt~Oy~-dsfruHi
thc ))~td 3ic(jt inon t))c gro)md,fmdi(.si))H
i))tc))Hit.yw!)s)ost
wht-n standing- cr~t;~
:!()yard.s. At7()yi,rds,)~nnta))(ting
crcct.titc.suund
int~r~d.s,:u)d was otdyfiuntty
w:Ls)nstn.t)ong
hc'f)rduvcn<I~c));))nt
thcc'a.r
iti)Cf-an)c'c<)]ttmu(jt).sn~inw)~-n
wnsmi~<tOfL~tfn.rnt)h.~uud,tUtditr~chcditsiu!tiDtu))sity
ata)tc]t.-va.tiu)it'12iL'<jt."
Prof. Rcyno)()s t)n)s smns up th rcsults of his
experimcnts
1. "Wi)oi th(!rc is oowind, soond
procucdit)g' ovcr a rou~h
surface is niorc iutcusc above than hulow."

124

TYNDALL'S

OBSERVATIONS

2.

[290.

'~A.s !nng as tt.e

velocity of tbc wind is ~cat~r abovc th.n
LcJ.w, .o..d
U~d
.)
~'~
~c~rd'h"
a~d hcnce its
~d,
range cxtc,.d.d ut thc surface of t).c gruuud."
At.nosph.ric r.f~tion
),a.s ,u.1 in,po,.tant
on thc
b.ri~
ud.L.h-.y
~j,
tlwj,J,
,)'U1lI'S
tt.c.att.nti.n of twoc..i.cnt
L.cur.J
,,hy.sici.ts Pr
~y.
~y
)~

it
L

~~ccul
T~
~Lfl

'T

7~

who.so .nv.sL.~tions

~r~?'
n
ue

havo

.smg

~sphcre
latter

b,.en

'o.
PLcn.nncn. which
wl.ilc
~-rvcr..
cuu~))y e.tcnsivc

Ly ~ccu~t
fr.r. uncqua!
hc.tiug or
in~hi.
~ti.g
~"L~d.
bas
Ty,M
~~tric bc!I

P o
by
~b~n~
7
~~s
d' Ilml, altllOlIgh it lnust hu
densitit,s;
ndluitted tllltt the al turllat()lIfo!
cOllSl,ll'I'ill)IC

illlll
IIIO)'(!
th an
ean w'c:ll !Je
ahrllht
SlipplJSl'd to occur ill tllc; l'Ull ail',
uxct'pt }JI)l'hap8in
ot' the sulicl
grullnd, SO111C
of the

LULc\j)iai]at)on
~i-~My to
~i"
nj <)ucsttn
'rh us it was fcnmcl tlt.vt
the last of a sirun
lal,vceclon the
of brlulually
dillliuiHl1ng ilitunsity, wllUSC rlurutiuu
solllotilrles
mllch

ob.sc.rvcd

"v-].n

~OHH.s.

t)

l),).s

phu))u)nc)ton

was

aud
SlJ!oothllcss,"
cannot
11'h'`~1'clltlyIJUo,ttrihutc(1 tu any otlter cause tluln
tlmt asSI~IIC(1to
Tyn~H.
It is ti~ref-L
prob~
acoustic.1 opacity arc bot),
concorn d i ho
of
fub-sigl,vls,
l1?wiuo we slloulcl cert;lillly bu
tlisposud to attacll

suluc of '1`ymi<lll'sovll
oservatiolls mlluit of
explallation 1111011
~=~=~=
t]Jis
l'Jril. L'rmts,l~~l. S'uun~l,
8rc1c<litiuIJ,
C'h.YII,
~Z:?,~ltJ~

290.]]

ON FOG-SIGNALS.

125

A faihn'e in ?'ec~)?'oc:7y can only bc cxplained in

principe.
accordance with thcoy Ly t.hc action of wind ( 111).
According to thc nypothcsis of aconstic c]onds, a difforcncc
might bc cxpcctcd in the bchaviour ofsounds oflottg and of short
dnration, winch it may hcworth w!iiic to point out hre, as it docs
]tot nppcar to hve becn notiecd by any prviens writer. Since
cnergy Is not lost in rcficction !).nd rctraction, t]ic intcnsity of
KuHatu'n at a given distance from a continuons source of sound (or
Ji~ht) is not a)tcrcd byan cnvcioping c!oud of sphuric'n.) form and of
uniform tlensity, tlle hjss due to t)iu IntcrvL'ning pa.rt.s of thc ciond
hcing compcnsatud by rcncction froin thosc wbich lie bcyond tlie
sonrco. Whcu, howevL'r, thc sound is of short duration, the
intt'nsity at a distance may bu vcry t)U)ch (H)ninishcd hy the ctond
on ac<'o'mto)' the diH'crcnt dista!icus of its rcfk'ctin~ piu'ts and thc
conHcqncntdrawing out of thc sonnd.ait.hongh thuwitotc intcnsity,
ns mcasnrcd ))y thc tirnc-intc'gra), may ))c th samc as if thcrc ))ad
ht'cn no c]ond at aH. This is porhaps t!tc cxptanation ofTyudaii's
fthscTvat.ion, t)u(t dincrcnt ]\i)x)s of signais do not aiways pruscrve
tho sa)nc ontur of ct'cctivuncss. In sonic statc.s of th wc-atitC)'a
howitzcr nriog a !))). charge connnandod a ta.rgcr range than tho
witistis, trnn))~ts, or syre!)," \v!nh.ion othor days "thc htfo'iority
uf tlte gntt to thc syrcn \vas (~nionstrato) ill thu cicarcst tnanncr."
It shuntd bc noticcd, howcvcr, t))at in thc sanic scrics of cxpcrinn'nts il, \vas fonnd that thc liahi)ity oftin; sonnd of a gnn "to bo
qnonchcd or doncotcd by an opposing wind, so as to bu practicaily
~sctcss at a vcry short distance to windward, i.s vcryronarkab))!
Th refraf.'tion propcr mu.st bc t!)c saine for all kinds of sonmts,
lmt for the reason cxptahn.td ahovo, thc diffraction round thc cdgc
ofan obstacle tnay bc h.ss cf'cctivc for t)ic report of a gun than fur
t)je snstained note ofa sircn.
Another point cxatnincd l)y Tynd:dt was th inftncnccof fug on
ti~c propagation of sonnd. In spite of isotated assertions to t)i0
eontrary', it was gcnera)!y bciicvud on thc anthorityot Dirham
tbat t])C innncnee of fog was prcjudicia!. TyndaH's observations
prove sati.sfactority ti)at tbis opinion is crroncous, and that tno
passage ofsomtd is favoured by thc homogcncons condition of thc
atmosphre which is tbc usnal concomitant of' foggy weathcr.
Wben thu air is satnrated \vit)i ntoisturc, th f:dl ot'tonpel'atxrc
with ctuvation according to thc law uf cunvectivc eqniHbrimn is
~L'efur oxtuupluDcsor,/r<it;/trt'<'
f~'r~/

xt. p. H17. 1SP5.

~26

LAW

0F

DIVERGENCE

0F SOUND.

[29 1.

mnch ]c.ss raj.id than

in tho case nf <!ry air, on accent of thc

FroniL
Cf'nd.~sati<.nof'vapuur~])i.~t)K.nacc<.tnpanicH(~p:u)si..n.
<t.
~t'<t!atio))hy'r))<))))sf)M'.ta.})p~at.,t)):)t!).,
f.~rh.
"f''vap()r;)<it)na))()<~nth'))saHo)two)dd hctodiminish
thc fa)) of
.,).).n)f:
t.p<.)-tur(.Ly
T)u.acu.s<i.
n.fr..t<.ti.m(]uetut.np~Ltun.wn.idthus
'!f)..ss(.u.),;t)H) in <,t1.r rMp..ot.'j no<)o)tbt
')'n~tLiun<.ft),c:tirwuuh)bc.f~'uur!th)u~thu}u-<.p!~Lion.'f
soutx), pruvid~) nu obstruction WL-ru(.n'd
hytito Hus)~n<)cd
rartic).st))(.t)).st.v~.
h.afuh)~c!)~)<.cr~.sh:)i)i)tv~ti~.t.-thc
d)st))rba))ofj));mu s..u(,r<u)s wavcshy a .smati u)).stac!u,an(t\VR
(tiatnctrrof
.shatI)ir)(tthaL(h..un~-)(t..{,n.!st)))(.nt)t<.n~io..f<.)iC
thuoh.stacit-tuthcwavt.-)~))!)])
of <).<.s.Huni.
lh~)'t'adt'r\\h").s(h..sir<)us(~jn))'sui))~t])is
sultjcct tnnycfm"Ou ti~cKt.tractio)) ni' Sound hy thc
su)ta]):t)~r)<y]~.yn.,)().s
as th..au<)Hn-)t.i<s att~tdyn.~rrudto.
At!n(hsp~.rc'a.wc)t
Tt
t)mt R..y)i..)~)~cswith
"):tyl~n)(.tin).t.)
H~nryincf~.sid~
n' rc.f)-acth.))tu!)u()mrc.d)y important cau.s~ofdishtrhan,
but
fm't)tL'r<jb.~naL[u)).s:tn.'tnuc)tn(;cdu(). Suca).soi!')4.
2f)).

0" IL''assutnptiott

());)(. thudiMturhanf'cat

an a~rtorc
n)!inc)'L'('niMt))(~ain(.a.sit.wf)u)<t)tavc:))(;<.natt))('san)rpiaouin
t)K'a~)K'(;ut't))(..s.'rr(.n,w~]nay.su)v..va)-i~u.s))n)t)!~t)i.sr.'sp..t-t.i)~
t))~!<)it))-a<').)nn(,f.s<iU))<)
by t)n-.sa)n.'))h')h~d.sa.sarL'cn)pj.)yo()f(.r
t))~currc.spundi,rn).)c.siH},y.sirai
phrs. l-rcxat..p!)i.o
<))~turha)H'(-atadi.sta)~~<)uthct)))'t)tt..rHi.i.(jfat) in)it)it<-pia)tR
~itjt a<-irr)))ara))urh))-(.on
wa~.pK~ed
wavusot'
whidtpianu
Sound itnpit.~dir~Uy,
thu a).:d.~ot)s
n.ay hu cafotiah.dasi.i
probk.Utui'Lhc(Htrr.K-ti~)tpath.i~furn)udaLt).cfocu.s()f'acin~dar
object-~a.s.s. Tt.usi..tf.c~.sc
of'a.syh)mu!.nc;dspL.a)dn~tnu,)})~
t)ics..und Isa maximum
a)u)~t))u axis (d't))ciu.sL)-mncnt.v!t(jrc
aH thu (.-)u))h'n)a)'y
')i.sturbaaru.si.s.s))i))~fro)n thc'varions puints
of't!)up]a)tc <t)tL;m.)))t)tarcinon<p))asu.
Inohtifpu-tHroctions thc in~n.sit.yi.s ).s.s; ),nt itd~.snotra))
m~Tia))ys)t<.rt
of thc rt.n.xi.nnm vahtu unti) tlh;
<.)j!i<juity is suc)) that thc
difr.-runcc of distants uf t))c ucar~t fu..) fm-thc.st
points of tho
n~uth :L..tu).nts to about h:df;L
At ns~n~vh~
wavc.-tf.n~th.
thc 'ttout!ttnayh.divid~[intotw<)
gn.atcrubiiquity
parts, of
whieh ti~ ncarcr .-ivcs an
L-t~ct cquai in magnitude,
n~r~atu
~.V.;)tr/n\<f..).)/)H.,fr.<.1;ir,]-(:
'A~7't;j.<.Yu).
!<!< p. :;j.
)n7t'
i:jl~,

SPEAKIJSTG TRUMPET.

~1-]

127

but opposite in phftsc, to th~t of t1~

fnrthcr; so th~t the intensity
in t)d.s direction vani.sht.s. In (Htions sti!!
tu~-coLixjUf. thc
s'~nd vi~.s, !,).)~s
tu ..n ixt~ihitye.jtia!
te. ~<7jff' ~~i,
to xcro,:Lnd sunn.D.u .r~tions
!L!t.hc:Lxi.s'ain.)i),.i..is].cs
c..)T<spr,n.ii,~tot).cb.~))tnud<h.rk
rmgs w],ichsn.-m).nd tl.e
t).<! in~f~
ce).tndj)atc!t ofJi~M
.sh.r. If~'d.-notptho
md.u.sut' <hcniouth,th~u~it
is
~i.ichti.L.fir.stsik.aoccur.s
W)K.ntlK-(]i:unctcr.,f<hctnout)t.)oc.St)otcxccc~
sin'C!0~.
~X, t)tc c!<jmu)~t:u-y
di.sturhtUK-c.sc..)n))ine witjx.ut any con.sido~bb
thc intc.sityi.su..ar)yu..ifurn)
anta~.ti.stnnt'pf.aso.nnd
in~)l
'inactions. Itnpp~u'sthit.tconcctttndi~naj-s~mtd
:).!ot)~)JK.is
t)~t
thu
ratiu
rc.~i~.s
:s).ui.[
~oo.i'iti..n
ncb
).
u.su.Oy.sati.fiud in <k. ordinary us<jnf.sp~ku~tru)n)K.t.s,wh.se
cHi~~nryd~nd.snLtfh.ruponan
vohunc
in<T~cint).uun~i~d
<~s(nmd(~S()). AVh.)..nvcv.
tt~vitu~t.ion.s f.n/of~.ry.short
~vo-)..ngt)),fth-mn}.d,ofmdc)-;)tGsixt.i.sc:)}.:d.hoft.ctm.rn. v
<Lsidt.ra~~cm.~ntrationa)~~ti,caxi.s,a.s]I~vcmy.sdt'~nficd
J"t)'cc:isu(jf:t.)iis.s.
2f)2.

sn(-I.c;L)cu)atmt).s

AXhnx~h

;).s

p~c~di)~(iu)):)rcu.sd)t)
.,f

]')),K.)~

it

diftract.i.m,

:u)xiii:ny:t.ssnmptiun

..nwi.idt

.stn.-Ny:u.d~n..)~)jy
J'x'idd.t
up~)

<

ap.'rtum

a.s<-n.n

<hc

ns

ih~c

~ivi))~
tnust

rd~n~dtoin
..fthc

~u.di()~
nnt

h.-

~.r~.tt~n

~Hy:n~fmmdcd

!shy

t1)at

thc

).),)c:m.s

Thn.si))t))(!(-;t.s~i'i~v:LVLidi)-t)y
n~n~d
in thcphmcofthu
v..)<xity
mcn~sc-st'rotn

is)~.t.<uthst;mt,:Ls).ash.-cn.supp~s.],hnt

ti)C(~)ttr<jtuw:)n).st!t<'o)~hc.(.<))nmL;-inf!nih;

at

.i"<.rd~rt.)i.tn..shn~tct))(;(~)tditi..)i.shywhi<-h

t)n-tu:J')uf-it.y

~d('~nni))('(),tt-tnsiu)-H)<')n())!it'))ts)JppuHuth!ttt~(j:)p(.rt)))'eis
Ti'e
i).(-i<k.).t.
w~vu
")'.
<=eus(~<)
~t)..ctud,nndt)tu

thc

vclcity-putu)itl:d

<jn tj)u

t)t~i~(;itsc)f.

i.s th~)

p..rftiy

hc~ttivu.siduut'tfic

sct'L~n(.<;=0)is

= co.s (/<< x~)+ cos (/;<+ ~.r)

(1),
S''vi"g-,whcn.-c=(), <~=2cf.s~.
T)tisc.))T<sp.)nd.s t.uthe~mi.shi"g 'jf<))cnor))):d v~tocity m~r thc:))'
t))c
~fth~~p~~m~;
cc)npf(jtiuuf)ft)m pr<)b)c)nnv[Hir(.;s tt.s to tk't~rmi)t(! :). \tn:Lb!c
J'u)-m:d v~).K'ityovur tLu :q.ht~
such that ti)c poUttit~I <Inu to it
(2~')).s]!:dti))cr(;a.s~-):)y t))uc~n'!t:u)t(p):mt[ty~cus/~iitcn).sni!)"'~T~t,7.r~');<t~t~f~t.).)~u(!.

128

DIFFRACTION

TIIROUOn

SMALL APERTURE.

[392.

from the ngative to the positive Ride; or, sincc the cros-sing
involves simp)y a. citangc of .sign, to dtermine :i vfdue of the
n'nialYi-).i)yovcr
tlif.irc:a.f{h[)ap.-rtu)'c'whi.f'hsh;~) c~n
tlie positivt. sit)o <p=cos?;i' ovcr thc samc fn-cn.. T)te resn)t of
RUpo-pnsingt)tc two motions thus dL-Hncdsalisses a tho conditions ot'thc prob)cm, giving thc s:nno vc'tocityimd pressure on t))o
two sidcs oft))c !),po'turc,aud avanishilig norma.lvctocityovcr Uio
rcmaindur of thc serceu.
If T~cos(<~+e) dnote tho value of

at tlio varions points

of tho arca. ()S*)of thc apc'rturo, tho condition for detcnuining

7' and e is by ((i) 27M,

whcre ?' dnotes t))G dista.ncc hctwcen t!)c cloncnt

n.~cdpoinLhttI~capcrtnrc.
piL't.t't)uuuf<~ fur .mypuint
i''ivcuby

fLS*and a~y
Whcn~fH)<I<;arck)-to\vt),t.hecomon t]tc positive sit)cuft)tc .sc'rccuis

Th expression ofZ'and e fora fixit i)pertn)'e,e\'cn if of circuler

fot')n, is pn)))!d))y beyond tbLi poWL'r(jf kouwtt mcthod.s; but in tho
cah!Cw))L-)'uthe()in)~t).si()))s:u'uVL-tys)n!).))inc<))np!H'Is()nwith titc
W!t\c-I~n~ththL;.s<))ut.i(j))('ft))tJj))'ub)(.ti
m~yLuL'H'tjctcdiurtitc
If )'bu t.)tc<)ist:mcLibutwu(jtit.wc poit~s,
ct!'c)u amtthu~ilipsc!.
bot,h of wbictt :).)'usitu:~tL''t in Uic npci'tm'~ Kr tnny Le ncg)L:ctud,
ahd wu tttc'u obtain frutn (~)

mfi.ttcrwhicit must be
p i.sthc dcusityofthc
shuwin~
o that. ~7r
distribntcd over ~S'inf'rdRrto producc thcrc thcconst:).nt potcntift.!
uxity. Atn, distance!from thcnpc])ing on thu positive sidewe
j])!)LycotJHidcr?'asco)ist!Lnt,n(!t:)L)\c

~-J

HLLU'TXJ APERTUK~.

u'hbrc
1
r

~t.'t.l!
"n

:uu~
n.

future-

t)to

dcnotiog

~=-~J~'S',
L-

)~

.u;.}.f.~i~

p)g-ct)nLt

fur

:ut

totd

(ii.iLtKud.

]29
1 :J

<~fUttity
ir.

wiH

of

natter

oc

s)icw)t

axis

cHip.scufsotmnajnr

ff,!u.d

ccccntrit'itvc,

l'esult,
s1101IId
~Lsr(-su)h!s.juit.cdirfL-rcntfn,.nU)atw).ic]iwcHLou!.)aLt:,i,)
'hi:
is fjuite
tile
]las 011
[ro!l1 tIl1tt which
Z tt"']'y)"jtitosisth:(t,t))on(jr)n:dve]f)(-ity
i))thcnpL')'<u)-cha.st.))c
tlie
\duu prnpcr to t))L-pri)n:u-y w:tVL-.Li thaL c:Lsc
by (:!) 2~
jl
j

T))C(]:m-;t(.ti())]~sm)hd i.s~.snhj~.t.which
i~s nttr.K't.cd!)).).
!'tt)' !)t.~))ti(.n <.it,])(.rfrum
")nth~)j):)(i(:-i:uiso[-cxpfTi)nL'nt:d!.s<s.
t.ndt.rAi<))0)~htt)c~t)u)-:))c).:u-~(~uft).t-p!K.no.n(.n:Lisw~tI
:md thcr~'orc- no vcry
st'd,
<)).<r<(.ncs :)r<! to b(!
shrHh~
'j['<'ct..< tl)~ cxnct t]H~rcUcit)s<)htth.nora
f.of'Utu.sin)j))cr
)"-D)))u).s,w!ii<.)tt))u.suhj.~t.})n..scnts, wo).)JLc i))tc)-~)i)~n.,),
<
witit tix; pt-t'.scot impci-frct. )))(.!)ju().s,
sr~))(..t))ii).L;'}))'<.haL)y
"ng]tUj(-du))L.inthc\i)y<jt't.p(..ri,,)t_.t)t;Lh'))h):)tiu)).
TLc'va)tH'r)f;Lfn))c)!()))~w]H<-j).sati.stics\7'=<)t))n)U~))titc mh-ri~t- of'~ .si)u)']y-(~)))]<<t(.~]ch~~
sp:n'c ,S' c;))i''h(.
c.\).rcs.s~] as t,).~pot.<~t)i!t!<.i'nj:)t~rdi.sfriL)ttut[nv~r<)tu
surfa~
"t~.
I"!tccrtai)).sd)scth!s].s:)).~t)'))cof'th(.'d:(s.so['f)))tcti<)j).s
~~L whx'hvc
iu-(; now occ))))ic<],w1)!<-)t
.s:)ti.sfy \7'(/)+A-=().
i.s
i'~uwixg
Hc)).t)t~)~spr<~r'.
Dy(:t~).'stitr<.n.))i,
if~
!t!dcn(.)t(-:)nytw'jft)j)<t~)).sf)f;<
't

77<f'f))')'fy,jr.),/7.<r/;t<')')<H~<');t))~<r''))Mf<f.f'H.'f<7'f~;)
j'.i. ].sf!f).

):. H.

r,'(!i..

H.) ,y,.

130

EXTENSION

0F

CREEN'S

TIIEOREM.

[293.

whcrcp-~prcsott.st)~
distance <.rfn)yp..in<.fr<.)nf).~x<~n)~in~
within
At a)I points, cxccpL (~ (1)
va)u.s)ic~; :md t)t0 ]~st, tcnn
I't(l)buco)nL's

"i~c))]t~v:m].s)),WtiIia\-t.uipxprL-s.sion
<<.rthcv:))ucof~at
anyia~riur
point 0 ~tcnnsufth~
snrfa va)ucs of-~ andot'
~l~

1n
the c;isc uf llm conmam potl.'lItial, uu wllch We l'ail back
lr,lrl:
-Int]icc.LM(;oi't))tje<)mnt()npotc'nt.i:tI,0)twl)ichw(;f!L]t

by putting A:= (),

w~uld bu ~utc-Dniocd hy thc surface vahtu.s of

~tc, thi.s Jaw ccascs to bc imivursalJy truc.

<t
Fur a givon spacc <S'thcrc is, ns in thu ca.sc
iuycstigatcd in 2(i7,
sorics of <)chTtni)tatc vaht~.s oi'
corrcsponding to thc periods uf
thc possible ]no(]c.s <.)'simplu ijannonic
vibration, winch may takc
p]a<'('wit))iHa<-]u.sct]ri~i() cnv~)up(;Jtav!n~tL(.f.,rm<)f<S'.
AVhh
a)iy<.f'tht'.suva)n(\s<d'/<itiso])\-i.nhstIi!)t-c:m)H)t))C<1('tL'rn)incd
hy its t~nna) variation ovcr <S',n.)xt U~- fact i))at it satisfis
<),u (~uation ~+~=0.
thruu~x.ut
Bntij'tLc supposa!
va]uc ui/t du nut coincidc wit)i (juc of t))u scric-s, t)<L'uthc
prob)~;n

]
~1_- L.

IIELMIIOLTZ'S

THEOREM.

];~ 1

for thc .Mercncc of

~s dctcrn~tc
any two possible ~nti.ns if
hn~c would sat.sfy ti.e condition
~i~g
a
o~
w!nel~by hypottic.sLs~uuot be ..tis~d
with
cond~ou
th(.'assutncd
value of/
Jf tho ~i.nen.i. ,f ~c
sptcc .?be very sm.]I in cc.np.n-son
~7~/7' but !.tt!o
t.t
.hf).
fro.n Aniction winch .s~i.fics
tia-ou.huuL
thc o~t~tton ~7~=0.(),
20.). On hi.s oxf:c~.sionof Grcun's th~rcm
(J) Hd.nh.
ionn.Lsh..sproofofthc i.nportant ti.~rc.n
c<it..Linc<)
in t.)mfotfnwin..
s~tcmc.nt:
~A
.cc~
M ~r~ &
~M
is
r< ~.y point A.
B
~ee~ A, /~<~E ~ec~~e M~rce
<6- ~~w~.
If tlie quationi

to a sp..co
cncl.s. hy a ri~i.) ~.m.Luy a,.)
c~j~tdy
any nu,nbcr of-dutac). r~i.) H.~ h.dics, and
-nng
If~
bu
vcloc.ty-potcnttafsdue to so.u-CL-.s
witJ.in <S'w~
]~.cd

t32

IIELMIIOLTZ'S

TIIEOREM.

r20-J.

itfo!!ow.stha.t
~,=<(-).),
which i.sthc symbuticat statcnu'nt ut' Hcimholtx's thcorcm.
<S'<'xtend to intinity, thc surface Intt'~rai sti)t
Iftbespa.cc
vanis)tcs, and thc rcsuit is thc samc Lut it is not ~cccssary to go
intodct!u!!)t;rc,asthisthL-oretnisinchnh:'dinthcvast)yn)on'
T!)o
gner:).) prin('ip]of'rcciprocityc'stab)ish(.'<)
inChaptnrV.
investigation tiicrc givcn s))cws titnt thc principe ronaitts truc in
t)ic prsence of (iis.sipativc ft'rcc.s, pruvidcd th~t, thcsc :u'isn from
resist.ancL's \nyi!)gas
thu fn'st puwcr ofthc vetucity, that thc
ftuid m;cdtiot bu Ihtinugcneons, nur t)tc t~'iglihonritu'' h'xhc.s rigi't
r fix~J. Iti t)ic ;).pp)ic:Ltion to infnutc sj):)('c, ~)) obson'ity is
avnitk'd hy suppo.sing thc vihr:tti~')is tu bc Hio\v)y(hssip:).tcd aftoh:tvl))g (/sc.)po~ to :L distance (j'ont ~t :md 7~, th sourct.'s undor
('ontooptfdio)].
Thc rcadcr must ca]'cfuHy rentonLcr titat in fhis thcort'm
introcqunl.sources of soundaru thnHcpruthtcctIhythL'pL'riodic
duction andft.h.strftctionofcfptat
<p):mt.itic.s of finit), or HonK'thin"'
who.sc cr'cct, is thc .s:uno, and th:d c(ptal .sonrocs do not ncccs.stu'ity
ti;ne.s. For instance, n.
uvu!vucqn:t.) amountfjoi'L'nc)'gyittcqu:d
source cdosc to thu .surface oi'n. hn'gc ubst~ch.' cmits twiee as much
cnc'rgy as an C(;ua! sonrecsituated in thc opt'n.
As an cxiunpic <'f tlic n.~c f this t))corcm wc )nay takc thn
r'asf ot'a hcaring, or.spea]\ing, trtunput L'onsisting of a Ct~ni(.d tubt.
whusc cHicic'ncy is thns St'cn tu be thc s:))ne, whcthcra sonnd pro(Incud at a point onLsidc i.sobscn'ed at titu V(.'rtcx 01 t]tc cne, or
a. source ofcqua) strengt]) situatcd at thc VL'rtcx is obsct'Yc'dat th
cxterna! point.
It is in)pcrt:mt atso to bprn' in )nind that IH.dmImhx'.s fonn nf
thu rcciprocity theorcm is apptic.tb)L;oj))y to &/t~/e souro'.s ot'sound,
wbic)t in thu absence of obstac]cs wuuld gcncrate synunctricat
wavc.s. As wc sh:d] sec More ch-arly in a. sub.scqnL'ntchaptcr, it is
possibh' to ha.vc sonrccs of sound, \vhich, t))ongh conccntratcd in
an intinitdy sma)) rf'~ion, do not sa.tisfy t))is condition. It will bo
.snfficicnt hc-n' to considcr tho case of~/o~g sources, for which thc
modined reciprocal thcorcm hn~ a.n intercst of its own.
Lot us suppose that
is a. si)n])lc source, giving at a. point
thc potentia!
fmd that yl' is an equal a.nd opposite sonrcc
sitnatcd at a nf'ighbouring point, who.sc potontial :)t 7/ is
+ A-

29-Lj
AtTLICATION TO DOUBLE SOURCES.
13.-3
rr) 1 .)
botb sources bc in
operation simultaneousiy, the potential at 7,'
is n~.
New let us suppose that tbcrc is a
simple source at 7~
whosc mtcusity and p).a.sc arc t).c san.c as ti.ose of
t)ic sources at'
an.! J'; th
is
rc-sulting potcntial at
and at .r
+ A~
It thc .h.stance .Lr bc <)cnotud
and )~ support todmuni.s),
by
~it .ont hnnt, th<v<.)~.i(.y <,r t).c fiuid nt .f ill the direction ~L.r
..s thu ,nut
ufA~
H.~c, if vu ~.n,,
~),
as tf.u imnt oi t~o
~)
(,p~
,)~
is dnnnns).),
~nd who.su ink.n.sity i.s incrcascd v'it).o..t
""in.
suc), a manner t).~ thc
j.rudncb uf ti.c intcn.sity ,u.d
t 'u di.stancc ,s t],c .sa.n. a.s for t~-o unit
.sin.ptu sources j.kcud
t''D unit distance
npart, wc .nays.y ti.at t)ic vulocity of- thc fh.id
'Lt. iu .hrcctKm J~f' duc to u..It
7.' is numc.risimple source
caify cqua! to the potuntia) at duc to a unit
source ~t J
wf.osc ax.s is i.) t!.o dir~Iou .L.
Tiu.s t)icr,rcrn, l;c it observed'
Ls truc .n sp.tc of auy u]).sta<'tcsor rctiL.ctors
that may cxist in th
J'e~tdjnurhood oithu sourucs.
~iu.

ifJJ'aud

7.

reprospnt twn u!)itdoub!c sources of t)~

direction 7~' duc tnt)iG source
.sa.ncp)..s<tLcvu)of.ityat~i,,
is tho sainL.a.s t)ic
vcjocit.y at .4 in dir.-ction .L~' duc to thc
source 7~7~. Tj.osc and othcr rcsutt. of ILlikc
ctt.-n-~c~r may also
bcobtamcd on an inuncdiate
application of the ancrt principte of
IUS. ijtcsc cxampiM will hc sufHcicnt to
sttew t!~t ill app)y,n<r
tho
is ncccssary to attend to the characte'
pr.nc.pfeofrcr-iprocityit
oi the .sources. A double
source, situated h) an open spar~ is inaudibfc fro.n .-u,y pon.t in its
c.juatonat phu.c, but it doL.s not
o'low tl~at a sunjde source In the
e.p.atoriat plane is inaudible
from the position of tbo duuh!c
source. Ontllis priucipte, 1 beiievc
'~y bc c.)]auied a cnrious cxpcri.ucnt
by Tyndal)~ in wjdcb'
ti'erc was an apparent faifurc
of reciprocity'. Th source of sound
c.nptoycd vas a recd of very high pitch-, mo~ntcd i.La
tnbc, aion.r
than iu oblique
whuseax.stlieinteusitywasconsidembfygrcatcr

dn'ectio))M.

surface

Tbo kinctic
cncrgy T of thc motion within a c!oscd
is cxpresscd by

e<ul.
etlilioll, p, ,105,
A.

.Su<

0" -S~

't""OnthcApj.)icationofthorri,)cip]cnfJ!MipMcitytnAc~stif.s"
Vu), xxv. p. 118, 1876, or r/.<
~~c~r/
J~. (o) 111.p. 3u'

3rJ

VARIATION

0F

TOTAL ENEHGY.

f295.

'~wbicbt))cfi!sttcr)nt-(-p)-c.sc-nt.st.bnw)'ktr:u)Stnit(udacn).sstbc
huundary A',and t))c .secondn'prc.sunt.stbu wurk dune by Intcrn:d
source ut'.sunnd.
If tbu bound:n-y 6' bu :).Hxc(! ri~id
cnvch'pc, nnd (hcrc Le no
uttern:d S(un-c(j.s,
.rf;t.;m)s its initiai viduc t))ruo~])outt.bc motiou.
T))is princip]e !):LS
bec!) npp)ied by KircbboiP toj.rovt! tbc (!cto-of tbo motiox )-c.s~)]tingfrom ~ivun
tunut~ocss
:u'bitr.ny initit
cunthLiuHs. Sincc every ctcmojt of7~'is positive, tbcre can 1jc no
nx.ti.m witbin
if
bu zuro. Now, if tbcrc werc two motions
possibfc corruspondin~ to tbc sf~mciniti:d. conditions, tbdr diffcr(.'nccwuu)t[bc n, motion for w)neh tbc initi:d Y:ducof ~was
zcru
but by what itas jn.st bcen s:ud .sncba. motion aumot cxist.
')7<ii)<~f'tt<t')'.l/tff/j/(~p.3n.

OUAPTER XV.
DJRTHER

A1THCATION

0F

TUE GENERAL

EQUATIONS.

-o.

Wf!~ n. train of p]anc ~vcs, othc.-wi.sc

uni.npcdcd
"npn~.s upon a .spac~.ocu.piud hy natter, w].o.sc ~cch~uc~t
proi.t.L..sd.ncrfrmn
t).<<,fthusnn-.n.i.!i,~med:un,,
.sccun<hu-y
~av..s..u.u
tu t)..ch.cn~iLruwn~~vhk.].~ayLcr~c.das~)i.st,u~,Lncc.)uc
~u.~Hrcoft).nc.)in,u-~
point of~rc
~hcn
thc 7-~
c.sp.<iy.,ppn,pr.atc,
~ce,a.swc]I
.s thu a!tcr~t,un of ~nccham~.1
pn,)x.tics, i.s snin!). If thc
~u.uuu
.[
~)..st.c)e h. nuid, t).e ,nod.n:c.)
parties
sp..)~n nf ..u-c twc-thc
and t)i~ ~y.
~7-7~
no
acc..n,.t ,.s h.rc t~kcn of' fridiou or
vi.scu.sity. In thc cil~tcr on
.si.hcr.c~) hannonic ana)y.sis wc .s)udl consi.ter the
proh]om i.crc
p.paso.! ou t)K..suppositiun t.hat t),c .b.stac]ci.s
.sp).cnca), without
rcst..ct.uu
tu
the
~"y
s,na)Inc.s.s cf thc c).u,~ of ~nccl.mica)
!up<.rt..H; Juti.cprc.scntinvc..sti~atiuu
t).c furm of t)~ oLst~cJc
a.s.su.nc t)..t i).c s.juarc.s aud
arb.tr.ry, but
higbcrpowcrs
is~t)ic
of
chfuigc.s ofjncciia)]ic-al pt-opcrLius may bc onutted.
(ienotc thc (U.sphiceincnt.s
If
parniid to tbc .i.xc~ of
co-ord.natcs of the p:u-tic)c, who.sc
c<)ui)i)jriu,n position is dL.fh.cd
and if~ Le t!,e norn~U
~y,
tho constant
dcnsity, and
of co.npre.s.sib.hty .so HuLt
=~ t!.c quations of motion arc

.m.) ~vo .smuhu. cqu~.on.s

~n
Oa thc a.s.sn.npCun
~ud
Li.at titu witoJu inot.un is
proporti.nia) to e'-<, w)iurc as usu:d
anL) ( 2-l..t) ~=?~o=~7r\
(1) m~y bc writtcn

1~6

SHCO~DARY WAVES

!29(J.

Die rctu.Lion Lctwcen th(; condensation s, nnd t!)o

d!sp!:tcements
ubtainud
??,
by mtograting (3) 2:!S witli respect
tat,hctiin<is

For thu System of pmnfu-y w:t,vcs:u!vancing in thc direction

"f -A', ?;:U)() ~vimi.s)); if~ ~hcHn.v:).)u(-.s<-)f~:ut(is,:m(t
~o-~bct))(.' )nQc)):uuc:t!con.sta'tt.sfur thc H)t(HstnrLcd))i(;(Un)i)
wc!ta\'(.'asii)f2)

L"t.s')'~nt.snt,i.sfy(~):Lt,t))cr(~K'nof'di.stu)'h;mc<<jn:K;(~)U!tt,
"i't)nj\H'iatiu))
i)i?~a)~[o-,w]tit;))onrsth<jru.
f~;t.u.sas.smnu
O'at (.h(-(.~)np)(.Lc!v;i!u~ ;u'u~-).~+~ ~~j .su)~t.hnt.u
'"(~).
')'i)(.'nt:).kin~:)~'<~)ntof(~),w(.'gct,

!t. ).s tu bf <.))scrvc<)t)i:i(

~o- vanish, Gxccpt
:).
Hn-m~])
sh.:dt .sp.
w).i.-)L i.s n~k-d
as Un. n~-iun
d..st.ur)):~(..ci
?;, ~.s',I"))t~<))crc.su!t:~t'tttO(!ist.urh:LuccfL)-ct~betrcatcd
us s).)! .[nnntki~rthu
<r
su that m onr.i.p~w,A~;
!xiin:dn
anajy.sis ))) Viu-i:Ltio).sof~iu.
cr intite fir.st~
~Hs.,f(.)..u..t((;)a~h.Lc-)~.t(.d,1,u.~
thfTcmuft.iplicd
'sn':t)!
.t)):.nti)ics. ~'cthu.s~htainfm.n
(.)a)~((!)by(1itfct-'tiatiu.u..):u)ditiun,wiLhu.seoi'
~), astl.cdt~i-cnti.-Ucunation
'n.~

296.] J

DUE TO VARIATION

0F MEDIUM.

137

mw)nuh

tho intention
cxtcn.Ls ovcr.t volume
co.npietdy indud.thor..g.nof<ti.stu.~Lr.ce.
Thc mt~!sm(M)
mayLc
t~u.sf..rmc~ withtt.c
aidot'Grcun'sthcorcni."
('aHmr. tlie tw.<
parts j-cspuc~V(j]y aud (), wc );;Lvc

wt.o

~dcnotu.s tiie.st.rf.Lce ut- thc

spaco thro~h whidi thc trit~
cxtcnd..
i..t.tioa
Nu.' oa ,S', A~ ..u~
v.uus).,
~~(A~)
su U.at, buLh t)i&surface
int~r;t!.s (fisappL-iu-. Morcovcr

<IeuuLL.s
thc cosinc of thu nn~)c ),etwucn
~hcru
and
TLc
fmcardnncn.sn)n oi-thc i-~iou of di.stnrbaucc M
nL.-dcct.cdm
and
is nc.~ccted in cornpariso~witli ?..
c~~mrison with
If 2' bc th volume of t!.c
spacc throu~h whic)i AM, A<rarc
writo
.sc]).-i))jfc,wcmay

1~8

LAW 0F

DEPENDENCE

ON WAVE-LENCTII.

f29G.

if on Lhc r~ht-hand sidc.s A~, ~<r rcfer to tite

mean
thc\'n.ri:).ho)tHH)question. TJtusfromCS)

o
m terms of ,vo hve from
(3), =
exprs
L'").s,)t the e~~ot.s~i,,)) foi, thc
prim~ry wavc.s bc
and (12) may Le;put Intu t)iu fonn
=

of

~d
=e'<+.~

thc condc'n.sation oft).c

~vh.ch~dL.notcs
wa.vc.sat
r"y
t''cpiaufdisturbanccattimu
/,and~dL.notus<.).c
condcn.sa<'onnft).c.su~.n)ary~v~att).L..sanu.time
atadi.stan~o~-frn.n
<1.ud..s<urb.-mc.Sinceth.dif!crc,ph..LScr~u.su.~dLythe
~c-~coiTc~nd.s.simp)ytothodi.sta.tc~w~nayc.,nsIdcr
t.aLa.snnpiorcvLT.satof
phas.ocursatt)..
l.taccnfdi.st.u.hancc.
J'an.p),Lu.k-ofthc.
s~n~uy
~.sisinvc.i~)y
i.rupurtion~
<c d..s(anco r, ~nd
t"t.)ie~~reuft))cw~vc-)..n.t),
of
"'c twotcrn~
(.xp~.sscd in (J3) t).c first i.s.sy.mnct~~tinnH
'chunsr<,un.)Lhup)~ofd~urbancc~vhi~t)~scc.-n.tvar!c.s
~sthc
and thcsccondary
e..s,ur<h.a.hvc.unthupri,nru-y
r~s. iLus~,h~~t~Lic]t~v;u-!c.sbc.avcsa.s.~
.9~
.source
~"d a ],)acc
~Y),icho- v.u.ius hd.avc.s as a duuble .source
( 2!).t).
T!'at t).c .sf-cnn.L-uy .ii.sturhanc-o ~u.st
vm-y as
miybc
pr~)..amcdia<.L.)ybytLc)nL.thod
uf.)im~.siun.s.
A~ a..<)AoU.c atnpHt.td~ i.s
L''))~n,
nc-ccs.s;u'i!ypr~~rti-.nal t. T .utd in
~eurd.mcc
,j,
wiL),thcprincIp)u.,fc..K.r,ry,t
~yinvcrsdy
a.
No~t).co,.]y<tu~nt.tic.s(dcpundcnt,
u).,n s)~ti,nc,a.nd
u)
~h.ch
~ss)
c;ui bc afunctio.),
t!.cr~ioofa..npI.LudL..s
arc
"c'ty<'t'sound),
an.) <T,uf~hid.thota.stcannot
occurluthexp.~iun
ut'a.sim~fumt:.),
asiti.sU.co.dyoneof
t'.uhvc ~iehinvutvc.sa
rcf~i.ceton~s.s.
Oi-ti.ercmah.i,~
f.)uant,tic.s
7',
thc )~t is t).c on!y<,ncwhi~
am!
')v~
a
.~re.tcc to ti.nc, and i.s (.h~-cf.).-e oxdndGd
Wc arc
of wllicll ().c n.dy cu.nhmatinn
.-u~t
'cf~wHi.
varyi,
as 7'
at)(tjodcpL'ndcnt, or thc uni of Jcn~h, is 7'r'' X' 1
~i. it.kTcstit~ application c.f thc rusu?ts of
this sec'ti<m may
bemadctu
cxph.i.i what ).av~j b~n ca!)ud /<</c
cc/~e.
'f~nt~i~~i~e~
I'{!t'nut))]',trti<.)(..s,"7~J/Jut)t.,lb71..
'f<t<lH7:),v)tt.:)H).

of

296]
]

ROUNDS
ALTERED
INCIIARACTEH.139

Jf the primary sonud bc a

co.npound nmsical note, thc variuus
cntnponcnttonM.u.- ~tt~d!ui.}.
~)-t~
Thuocta.'u
<")-cxa.up)c, is stxtccn timcs
strongcr rcJativdy to tl.c fund~'
.ncntat tonc in t)tc
sccondary t).an itw.~ i.i thc prunaryMund
hcre is thus no dirriculty in
undcr.stan<)i.~ ho~ it may f~apnen'
that cd.oe.srcti.rncd fro.n .snch
rcf!.ct:ng bodics as ,o,,p, of ~es
.nay Le nused .~n oct~e. Thu phenu.n.non Las ai.so a
conujiun~nt~y si.I.. If a nu.nhcr of-smajl bodics lic in
thcp.~<,f
w~v~su) .oun<),t],c vibrations wl.idi i.s.sucfro.u
U.c.u in dl dircet.un.s arc at tlic expcn.so of f]..
cncrgy of thc m~in stro~n, and
whcrct).o sonnd is co.npuund,t).e oxatt~tion of
the hi~her h-u-n~n.c.s H~thc .seattcrud wavus involvcs a
pr~orthmal ~r.L.icncy
ot thcm iu th direct w~vu art.cr
passif thc ob.stae!c..s Tins is
pL.ri.apsth cxpfanati~ ~certain cchocswhicit arc .said to .-t..t.,rn
sound g,r
t)h.n th .n~ina], fur it is ~-no~n that thc
of
pitch
to.ic
I"o
.s~ptto he c.stimatcd too )uw. Duttheovid.ncc
is cunH.ct.ng,and t!.c w)..de
.suhjcct n~nirc.s f.u.th.r carc.fuicxno'nta)invc.st!,ration;
.se ~).o mayhve th i~nayhuco.n.ncndudtott.o~.ntionof
n..c..ssaryupport.uni~.s. W).i)ea.t altre
"< m thec~~c~-of
aso.md is ca..siiyi,itu))igib!u,and~nst
".deed ~,n.nJ)y j.appcn
al:,n:t..d c.xto..t, acha.~m
thc
ut
;b
]" cft
~,n),)c tone wuu)d Lu a viofa~nn .,f t).c iaw'of furcud
vib~tto.Ls,and i):u-d)yto bc ruconeifudwiLJit)~uretie:Ll id~s.
In obtnini,,g (1.3) wc),LVo
n~ctud thc L.f)-uctuf t)ic variable
"~urcui thc tucdimn
~f~~r~cc.
~Vhc~thc di-sturbancc ou th..s .supposition is
th~-o.~b)y !<nown,wc mi~ht appro~i'~atc a~
t),c .s:uncinanuer. Tha ad.Htio~a!tur.n.s .su
uhhuncd
w..uhl be i.cccs.sariiyuf thc second ordcr i.i
A~ Ao-, so t).it our
cxpre.s.s.un.sarc ili ail casesc.jrrcct as f.r as thc fir.st
powers of
Uiu.sc(~)a))titics.

EvL-n wfioi t)io rcgi<j)i of disturbanco i.s not

.sma[I in conpanMfmwith
ti.c .s:uno .~ti.od is appticabic,
prcvidcd thc
~.arcs nf A~,
bc rcai)y nc~i~i)~. Ti.c tu~d cOuct of
any
ub.stactc inay t)ic.nho
ea!c.datcd by int~rati~i fro.n tho.se of its
P~t.s. hi tinsway wc inay trace thc tra.nsltio.i f~.n
as.nnit
of
'-cgion d)stt)i-banccwhose ~<?y~ccdocs not corne into cut)sidcr:tto a. tbin p)atc of a. fcw or of a
grcat many square wavc'c'~ths in area, ~inc)t wi!! u)timatc!y rcHcet
accord: to thc
r~u!arapLl(..a!
t)~
Hntift),c.,b.stad~huatai)d.atcdin
'hrccHonofthc primaryrays, t)ns inethod ofcaiL-utatiuu suon

~0

SECONDARY

SOURCES.

f2DG.

couses to bo practicaDy available, because, cven

atdiough tbc
change ofmechauical
properties bu very s)n!).)L tl.e iuteraction
of tifc varions parts of t)ic obstac'te c.umot bc )cft out ut' accuunt.
Titis caution is moru espucia)iy ueodet) In
duaiin~ witb t)ic case ut'
f!ht,w!ture t)tc wavc-)cngUi is so cxcce<!ing!yHmidl i)i cujnpun.suM
wit)i t.hc dinicnsi~tis ot'urdin:u'y ubstacic's.
2!)7. lu sumc (]c~rcc .si)ni)n.r to tbe cftcct prudnccd
by n.
ch:u~c iji tlie mGc)t:Uiic!d
pt-upL'rLtG.sof~ sn):)U r(.-gi(.not' t)m tinid,
i.stb:)Lt.w)tichcn.suc.-iwJ!ut) thcsq~u~
ufthutm~hm
i'I.-n.\s:u)ywtiurc to .suc)) i)))port!tUt!c th~t. it c;m bu un
iun~crn~uctcd.
Lhun :tC(p)irL'M:t.fi))itnvaiuudcpendcnt
\7'~+~
upott thc square
of thc jnotiot. Such ptaccs tftct-uiurc act !ikc som-ccs uf
sound;
thc pcridds ofthc sourcc.s inc)ndu)~ thu subtnxttfph'.s of th ongnmJ p~riod. Thu.sn.nyjtart of.spit.cc,
:ttwhic)t th intcnsity
nccunndatc'.s to a MufH<;iuntcxt(;)~t, Lccotncs itsc]fa
s~coodary
tonus t~t'th p)-im:u-y .suond. If
source, c)tuttingt])<:h:u-!)~)nic
thc'rebGtwopt-ini:nyso)mdsofs)tf!icicntittt(;nsi<y,
thcsucundfu'y
vibr:tti(nts hin-c t'rc([UC)tH)(.\s
which arc thc .smn.s fmd din'urcnccs of
thc ft-L'~Hcuc~softttc prinKH-ics( GS)'.
2!)~. Tjic pLtdi of a sound is !!ab)c to modiiication whcn thc
som-L-caod thc rccipicnt nrc ili !-c:]:t.tivemotion. It is
c)c;u', for
]').st:(ncc, th:tt an observer approac)d)ig a (ixed source will tnect
thc wavc.s with a
frcqucneycxccodin~ that propcr to the sound, t)y
tlio numhcr of w:wc-]cn~ths pa.s.scduvcr in asocond of timc. Thus
if v bc tbc veducity of titc! observer nnd M t]):tt of
sonn<), th
freqnotcy is n.tturcd in tbc ratio Mi
f<, accor<)in~ f~ tho motion
:.s towardaor fro)n tho Mourcc. SInccthca!tcratio)i
ofpitch i.s
constant, a musical pc'rfonnaucc wou]d stiH bc iioard in tune,
:dthough in the secondc:LSc', wnoi ft and v arc nc:u-)y cquid, tl(c
fid) in pitcb wout(t be so ~reat as to
dcstroy a)l musical c))aracter.
If wc coutd su])posc! to bu greittcr than M,a sound
prodocot aftcr
th motion bad bcgun. wodd ncvcr i-c'ach thc observe')-, but sounds
prcvious)y cxcitcd wotdd bc graduaHy overtakon and itcard in tbc
ruvcrsu of tbc natural ordcr. If u=2(t, titc observer wouht ficur
a musical piccc m correct tinic and tune, but ~<fc/~w~.
Corrcsponding resn)ts cnsuc when tbe source is lu motion and
tbc observer at rest thu altration
dcpoiding only on thc relative
tuotion in tbc Hne of hearing. tlie
source and th observer movc
JI10VC
\vitb the samo velocity tbcre is nu a!tcration of
frcquoicy, witethcr
1
Helmholtzhor
Hotmhoitz
iiborCombiul1tiollstnc,
Combinatioustuuc.Pogg,
AlI/l,B
Bd.
d, X<IX,
xcix.H,
Pogg.H.
,HJ7,
H..107.
18.C.

j
i
t

2~8.')J

DOrpLER'S

PR!NCirLE.

14t.t

thc mdium be in motion, or not. Witb a. rcJativo motion of

is 'jry c'iit'ptcnous,
40 rnUcs pur bonrthc
i~t.crationofpitcb
amountin~ to a,bo't)t a. semitonc. Thc whistte of a, loconctivc is
from an
hcard too hi~I; as it .ipproaches., and too Io\v as it !'GCC(Jcs
observer at a. station, (.'han~'in~ rathcr fiLubteidy at tho tuon~cnt of
pass:
T))L' pritipipic ofthc altration ofpitch by relative motiol) was
first <unnciat(.'d ))y Doppler', and is of'tcn caited Duppto-'s prinStr:u~'n]y (jnou~h its Ic~itimacy was <sputud by Put/va)'
cipe.
whosc ot'jccttun \vaH t!tn rcs~dt of a. cont'usxm bctwcct). tw')
pcrf~ctiy distinct casc.s, that I)t which titere is a rciativc motio))
ut' tt)L:sunrce and t'ucipiunt, aud tha.t in \vhic)i thc medhun is in
tnotiun whiJu thu .suurcc and th rcipient arc !).t rest. In thc
bttc)' ca.SL'thc circmn.stfUK'csa.rc mechanic:dly t))C satuc as if thfi
nu'ditnn wcrc at rcst and t)'c source a)id thc rcipient had
connnon motion, and thcrcforc by Doppicr'.s principfc no change
ut' piteh is tu bu c'xpcctud.
.i)opp)crs principic bas bccn cxpcriinc~taHy vcrinc'd by Unijs
f!at)ut" and Scott Ku.sHcH,who cxantincd thc .'Llturati"n.s of pitch
of ]nusica[ instrnmcnts carricd ou. fucomotivos. A iahoratory instrument for proving thc cbiu~'o of pitcli due to nmtio)i Ijas becn
invcnted hy Mach~. It consists of a tube six icub in icngth,
coitru.
Atone ond is
capabtu ofturni]]L;'a.houtanaxisn.tits
bhjwn bywindforccd
orrced, chichis
p)ac-t;d a.s)na]lw)nsth;
of
ah)))"- thc axis of t))Ctnbe. An o))Mt'rvcrsituatcd in tt)G p!anc
rutatiou !)c;u's a note of~uct)U).tin~ pitch, but if )tc p)accs ]n)nscff
in thc proion~atinn of th a.\is of rotation, titc sonnd becomes
.stuady. P(-'r))ap.s tlic si;nplL'st cxpc'ritncnt i.s that dcscrihcd by
Two c" tuning' furks mountcd on rL'soxancc casc.s are
Koni"
rfthc
prcparcd togivc '\v)t!)(-:u-hcL)tcr four bcatspcr second,
~)'avfjrofth(-'forkshc)na()L!t('app)-oac]tthccarwhi)etbcothr'r
ru!nain.s at rust, one bcat is ~.s'<fur cacb two feet of approach if,
howcvcr, it bc tbe more acute of the two forks ~'bich approachf'H
in thc samc distance. A modification
)hc ca.r, onc bat is~t:~
ThnnriGdcf!ffU'tji~']I~ichtof)(h'r D<ipp(.')stcrnp.
rmK, 18)2. Sec l'isko, /)/<'
\Viun,18C~.
;t')'<'M.-(~ftrff<<<'r~
vnr. lf!7.
H'/fx.
vm. Ut. M52. ~'or~f~t- ~f)'/<
t.xvt.p..T~l.
~'r~g.<
'P<tf.;g.cxn.
p. nf!,lMt, and cxvi.)).):M.
"]~'ni):sf'(;~)yf)y;f

</<<..t/)'~f/Jr~tt.<f)')~.

rnt-is,1RC.

DOPPLER'S

PRINCIPLE.

['39 g.

of this cxpcrhnent duc to

Maycr' may atso bu notiecd. In this
case onc fork excites th vibrions of a second
-in unison wit),
Jtscif, tho cxcitation bcing madc apparent
by a sn.aH p~dui~m,
whosc bob rcsts against t))c
extrc.nity af une of thc pro.s. If thc
cxcitmgfork be at rest, the etfcct is apparent np to a di.stn.cc of
('~ct, but .t ccascs ~.).cn thc cxciting f. j, ,novcd
rapidiy to
o'-iro m the dm.cti(jn cf tjta Jinc
joining t]ju twu furks.
Ti~creis some di~cutty lu
troating inatbcmat:c~!y the p.-obic.n
of a movu~ source,
.-u-isingfro.n thc fact th..L any practic.-d smn-cc
~ts a so as an ob.stacie. Thus in tbc
case of a bc)I cm-ricd
t'.rougb tho an, we s!.ou)d rcquire to solve a
probk..n dif)1cu)t
ouougb without incfuding H.e vibmtion.sat a)]. But thc so)utio,i
of .suchaprob]cm, cvcn if it couJd bc
oLtainod, woutd throw no
on
particuJar i.ght
DoppJcr-s law, and wc ,nny tbc.rcforc advanta~cou.slysnnp)ify the question by idca~n.g thc bc!! into a
snnpio
source of sound.
Tn l.t7 v-c considcrc-d th prob]cm of a
moving .source of
d..st..rbance Lu t),u case of a strc-tchcd
string. Thc thcory for
aenal .vavc.s m o,,o di.ncnsion is
prcci.sdy simDar, but for tlle
ancrai case of thrcc di.ncnsiun.s son.c extension i.s
ncec.~ry in
ontcr to takc account of the
po.s.sibiiityuf a motion acro.sstt.c
d~ct.ou of the sound r~ys. From
273, 27(! it appuars th.t t!.c
efrcet at any point 0 of a .sourof sound is
thc samc, ~hcth.r t),c
source he at rcst, or ~hethcrit. ,novc in
any ~nncr on thc surface
of a. spi.crc dL.scn))cdaLont
as centre. Jf thc source inovc in
.sucba manncr as to
change its distance (~) fr.un
its e~et i.s
aKcrcd in hvo way.s. Not
o.dy is th ~~c of tho distnrbancc on
arr.val at ~a~cted hy t).c variation of
distance, Lut H.c~
a).soundergoes a change. ThcL.tter
co.nphcation Lowcver may
he put out of account, if wc limit oursc.Ivc.s
to the case in ~hich
tlie source is sun.cicntly distant. On thi.s
understanding we may
assert that thc enuct at 0 of a
disturhancegcncratc.I at time nnd
at d.stancc ?-LSthc sa)nc as t!tat ofa .siniilar
distnrhancc
at the time t + and fLt t])c distance )-- a~. In th ~ncrated
case of a
penodic disturba..cc a velocity of approach(r) is cquiva~ent to an
mcruase offrcqucncy in thc ratio ft f<+r.
20f). AVc~-i)[ now invcstigatc tlie forccd vibrations of
t]jc
an- conta.ncd within a rcctang.dar chambcr, duc to
internat sonrccs
of sound. By 2(!7 it nppcars ti.at thc rcsutt at
timc < of an
()) xt.nt.p.27~.1R72.

2M.']
B
g
g

S
B
g

initia

RECTANOULAR
condensation
'3

confincd

CHAJMDER.

to ttic ucig!.bour!)ood

143
of t!t point

f.-n.n wltich tLc c.Hcct of .tn

i)n)~rcsscd force m~y bo (Jc.)uc(~
a.s ..) ~7(!. T).c
di.sturb:u.ccjj'j'
co.umu.ne;LtL..[ at'
Un.c Lun~dcnotcJ
or ~,(~
byJJJ')~.<
th
ru.suhant di.sturbancc at tituc is

T)'c .symmctry of thi.s cxpt-CMionwith

respect to
y, a.td
'.s
an
?/,
cxamp]c of thc pnuciptc of rcciprocity ( 107).
.In fJtGca.c f h.-u-mntncforce, for which
(0 =~ cos?~
wc hav'uto considcr thc vn)uc of

thi.s mt~al
StncUy spc.ing,
bas !ia <tcf.).;tc va)uc- L..t if
wc ~s). for thu c.xpre.s.si.mof (hu f.n.ccd
vit~ions on)~ wu nmst
~t
t).c u)tc~tcdf.mr.ii.,natt).c
)~.cr ii.n.t.
,n.yb~cc.a
.y supposm~ t),u ii,trud.,ctiou of vury .sn~)I
<)i.ssip~ivc forces
Wct,h)t.suut:ti)i

As m.~ht !c bccn
prc~ictctLthc cxp.-cssionsb~eomoinfinitc
in
of a comci~~cc LeUvce.itho
pc.ria.tuf t)~ snurcc an.! one
c~e
oftho .tnratpc.nrKf.s ofthc c).mLcr.
A.yparticuf.-u. normal
Y.b,t.n will i~ot bc cxci~<), if tl.e source
bo situatcd on onc
of its luop.s.
Th cffuct cf a
nu.p)icity
''ysumnnattonori)it(.raLi(jn.

of .sourcesn~y r~dify bc infcrrud

U~LIMJTHD

H.i

Tt'UK.

t/

the
When sound iscxcitcdwithma.cylindne~pipc,
is by th forced
si)np)cst Mnd of excitation t.h~t we cun suppose
vibration of n piston. Jn ttu.scase thc w:u'(js f~'c ph~c irmn
th hcgimung. But it is Imp~rLnnt :dso to irtqnirc w1ta.bh~ppcn.s
over thc
\Yt)cn the source, inst(:'n.d ot'hcinL; uniiorndy'Hftuscd
ofit.it. Ifwc:~sumG(wh:tt,
section, is conc~ttr~tuditiotn.'point
sufficient distance
tn~) t]):itat:L
howcvoi-, Isnfjtu)n't;r\'<jdty
from thc source thu wn.vcs Lccomc plane, thc law oi' rccipt-uclty
inf'orxKLtion.
issuHitcicntto~uidcu.stothcdusirctt
300.

two
L(.'t~). bc asitnph' source in nn un~n~cd <uh(~/)',
points oi'tttcs~Tnc normal St.;ct.ioniu thHrc~ionot'phincwavcs.
7/~h~ tn thc .source J
thc potuntiaisut/~and
~.<- A~fC.
~nd accordin~y
arcthusHmc,
hythch~v of rcciprn(;i(yenu:d
sourcos:<.t 7~'funi 7~' wou)d ~ivc thf .s:unc potontia! nt ~). From
this it fo!!uws tlutt the cn'<'ct of :my source is t))c sn.)nc n,t a.
distance, as if thc source wcrc uniforndy (hifusL-dovci' tho section
whidt passes throu~h it.
Forex:).m])ic, if~:U)d7~werce<)U:).)
sources in opposite phases, thc di.sturb:u)ce at /1 wou)d hc ni).
T)te cner~y etnittt'd Ly a, simpit' .source situated within a,
tuhc ]na.ynow be cah'nhdcd. If thc section of'ti'c tuhc he cr,
the potentiel due to itt
:md th source sueh thatin
theopcn
woutdhc

th \'u)ocH.y-p()t.c-))ti:Uat :L (Hshtm-c withio thc tuhc 'i)l bc

U)e s:n))e:iLS if th C!U).scofthc dist-m'hanec wct'c thc mnt.iou
t))cs:i.mc total displacemcnt,
of:), piston !ttt)tenri~it),~ivin~
!U)dth cncr~y C]nittcd will nt.sobe ttic MmnL'. New i'rum ())

~1
I!cncc,~sjn2.{.i,
t)~ .sourceisgivc'nhy

I~ERGYEMITTnn.

145

<)'cc!)p)-gy(~r)Rmitt~o)!C(7c/~t'(~

If thc tube Le stoppcd hy an innnovit.bte

{ti.stunp):).cc()dose tn
thc source, tho \vho!e cncr~y is onittcd in one
direction; bnt
tins )s !i(-)ta)). In con.scqnc-nccof th (ionblcd
prossurc, twicc
as n)uch cno-gy f).sbcforc is ()cvc)nped,an.1 t))us in this c:tse

Thn)i;))-r<-)wpr
th tubc,Lho
gr~tcri.stiicono-~yissuinrffroni
?L~)\'L;)tsuut'cf;. It i.s
into-c.sting to compa.rc tjtc cfHcicncy of
a,sou)-cc;)Ltt))cstopp(-;(!en(tofn. cy)in<)ncn.ttubowiththatot'
:") ef)))~)source situatcd at t))e vcrtex of a, cone. From
280
wujtavcinthclattf-rcnso,

Thc Rncr~iM omittcd in thc two casps are t))C sfune whcn m=
~'o-,
t))at is, whoi thc section of thn
cy!int!(.;t- is cqu:d to tho an;it
eut otT hy thc cne: frnni a sphre of )'a(]jus x'
301. '\Vc ))nvc now to examine hnw far it is truc! that vibrat)f)ns w)t.I)in a. cyimdrical tube bccomo
approxhnatdy p)anc at n.
suuicicnt ()ist:mce from tbcir source.
Taking <,hcnxis of~ pa)'a!)c)
to thu gcnurati))~ Jincs of th
cyiitxicr, lut. us invc-.stigatc t))t:
)notinn, whosc potcntial varies as e" nu thh {.ositivc ~dc of a.
-soxrcf-,sitnatcd at ~=(). If
bc t))C potcntial and
stamt fur
r/'
tlie cqnatton oftho ]not.inn is
.+
j'

If'

bt! I))()upct)dcnt, uf
it t-cprcscnt.s
\')))!-atKm.sw))o])~
li'tite potentiel bcHto.
tmnsvt.t-sGtnttiCfLxi.sofUiccytindur.
prr-pnrtion~dtde'it
ntnstsatisiy
!()

14 G

VIBRATIONS

IN UNLIMITED

TUBES.

[301.

as well as thc conditioti that over tlie boundary of tho sectioM

In ordcr that thnso quations n~y be compatible, is rcst.rictod

to certum dufinite vtues con'usponding to th pcriods of thc
D~tu)':dvi!)r:Ltiuns. A xcro value of~) gives ~)= constant, wh~'h
sotution, t)tougii it is of no significancuin th two dinicusion p)'f)bk'm, we nha)l prusentty bave to c<'usi(!(.'r.rur oach athnissi~G
v.duc of thcrc is f).dcfinite norma.1fnuction Mof and y ( 0~),
such that a sutution iH

in which

corrcspondin~ tn ~= 0, is constant.

[n thu actun] prnb)cin mity stiU bc expandcd in thc samc

sries, provide~ t)):~tJ~, ~1,,&(. bc t~gardeJ as fnnctions of
Hy substitution in (t) wc gct, liaving regard to (~),

Thc solution of the gcnc'rai cq~ation in ~t nssunn-s a di~crcnt

is punitive or ncg'~tivp. K'thc furccd
f'urin, aecordixg as

DISCRIMINATION
0F CASES,

t~

vibrion be graver in pitch than thc

gr~vest of th purdy transvcr.scn~ura) vibration.s.
cvery fimtc value of;~ i.. g,-(;aterth~
Ls<m.Hi,(j~at!vL-. Pm.(,n)g

~w un.). th L.ireu.n.sL.u~.s
suppt, it is C'vidc'I1tthatthp
.notjun ducs not b.co.nc infinie
so that a)! t).o c.ciicicnts
with
I'su,ncwi.atdim..rc,.troason
thc.~ncistn.cof.
u,
.sh. t).c
..s
can bc. no wavc in t),c
ngative dirccLiun. W~ ,nav
t.Jt'jrctut'eta.ko
e
111

+
.(1_),
(~~
a.. r.p,-ossi.n which rc.ducc.s Le its
n.st tc.n. wi.en i.s
suf~iu-Uy
Se~.
A\ c c.ndu.)c that. in ..]! ca.sus U.c ~vc..s
ukin.d.c.)y bccunK.
~c~
!o,
Jrctr~est cf tlie
?<<?-/ ~-(~~c~e ;<M).s-.
<~=~e')+~+~+,,

v.

~cg.u-c.st transto 2.1.S.H

=3..H.~
If
~ca
thc wavc-Jc.ngth nf the forccd
33J)
vibrion
e..c~
~'o~!T!
i' t?hat
uh.n.atdy bc.cun..p)..n.,n)t.].ou~.
tho w.vc~u
i.cwav~
iaH
.s
.t
of tl.c aL~.c limit. F.
~.1.
exa,.p)e. if thc suurcc
ot v.bmt.on ho .sy.nn~t.ca) with
respect to thc axi.s cf t),c tubc
<7. a ,unp), source .situai) on the Hxis
it.c.!f, t),c ~avc..st trans~
vci-.SG~hrat.o.j ~.ith whieh wc .shouhi ).ave
to d.at wouid bc .nor.
an octave
than in th. g.ncra!
ca.sc. an.) H.
~n
J.i~cr
'c .)t uf thc forced
v.brat.on n,i.r).t ),avc Jc.s.sthan i~f t),c. abovu
valne.

inasmnch
a.
r/<r,

A.c.,~) vani.s)~.

It appcar.s
accr,n!mg)yt).at t)~ p]anc wavos a<,a dist.anco a..
'esamcaswonM
bcpro.juce.] hya rigir) pi.ston at, thcm~m,
K)2

t~8

REACTION OF AIR

[!

cxists. Any
normal
mcan
actuaUy
th
samc
vdocity
~iving
"orm~ motion of w~ich t.)'f-~-t~tivc ~n~ positive p.~ts arc equaL
produccs uitimateiy no cr'cct.
Wilcn thcrc is no restriction on thc eharactcr of tLe source,:ujd
are graver t1':m
whcn some ot' thc transverso natnrat vibrations
are positive and thon
the !tCtn!Uonc, sotne of thc vah'cs of
tcnns enter of th form

source arc tiropagatc-dto

Indicating that ttic pcculiantic.s of tlie
au in~n'tc distance.
Thc prubt~m hre consido-c~may hc rog~'d'~ as n.gcncraH~For thc cfiLSuoi'n.eirc-nhu-cyHn'io-Itmay
tionnfth!),tof26M.
but
t,<-wcn-l.L.dout complutdy with Htt. aid or BL-sseFsfunctiun.s,
this tnust be !uft tu tlic rcudu)'.
302. In 27H wc )):u'e fuUy dutentuncd th !noti<'nof th'~
of
au- duo tu thc normal perdue motion of a bnunding p)anc plate
dcii~nitc cxtci)t. If clo bc thc givcn ~on~al yc-Ioeitynt tlie
mcut~.

7' distant ?' from f~ Tho

at
thc
point
nny
vdocity-potcntia)
gives
thc
~maimter of this cha.ptcr is devot'~ to th cxfuninatum of
whioh crises whcn the nnrn~l
of
thi.s
case
probicm
pjirticulai(-ircuhu-:u'ca of ra-hns
nvMrL
vaine
constant
a
Itas
giv~
Yulocity
it is zro. In particnhtr
thcof
rcmai)xk.r
t)tc
p!anc
over
whiln
7~,
th air will
wc shan invcstig:itc wh~t forces<1ucto thc raction of
harmonie
act on a rigid circ~n- p)ate, vn.raiing with a snnp~
motion in an (-qua) cirodar arcrturc eut out of a rigid ptanc p)atc
extcndin~ to intmity.
wn
For tlie wh.~e variation of pressure acting nu th plate
l.avc ( 2-~)

3~.j

UN A VmKATfNG

CIKCLJLAJ~ PLATE.

wherc o- is Ui<jnaLurat
dunsity, !u.d

~f)

varies as <<"<. D~us by (1)

~.ch we hve .,ow to

value, cac)~ pair of lments is to bu
takeu once o.dy, and ti.c
product is tu be sun.med after
n.ut~pi.c~un
bythc~tor..ti~
ti.d~nutua!
distancu.
1 he bc.s mcthud ~.t
s~cstcd
by Prof. A[~xwe[if..r th co.n.non
Ihc qu.ntit.y (.) i. rc~rded a.s thcwork
potcntiat
that wou)d bu
con.su.nc.!
t).c comj.~c di.ssuci~tion of thc
ni~tcr composn,
O.c dLsc, t)~ is to .s.y. iu tlie rc.n.vd of
cv.,y c.icmcnt from ti~
mfh.enco ot .v.ry othcr, oa th
.supp.sit:on tlmt the putu.iti~ of
two clement.s is
proportio~d to r-'e-- Tj~ a.nou.it et- work
rc.[.urcd ,v)nch dpends o.dyo~ t).c initia ~d ~d
state.s,,nay
bu c~eui~tod by
the
.suppc.sin~
operatiou pcrfbnned in any way
L.t may Le must convc.ni.nt. Fur this
purpo.so ~e suppose that
th d.sc is divided ,ut.
elu.ncntary nn~, ,~d th..t caeli rin.r
~~d
away to in~ity bcfuro any et- the Intenor
ring.s arc ~i.stlll'ued.
first stop is th adcnhLt.on of thc
potcntlal (~) at th
of
dise
of radius c.
c~c
Takmg potar co-ordinatc-s (p~ .yith
any pomt of th circmnfuroico for ple, wc hve

Dus
must bc nu,]t:Iic.! hy
a.~d aftcrw~d.s
2~~
quan~ty
with
ntte~ate(t
respect to c bctwcun thc Ihuits 0 aad
But
it will bc convenieut ih'.sttu cn'cct
transfonuatioD Wc trn-c

whcrc

is writtca for 2.c.

(.) is thc B~c!~ onction of .cro

ThcoryofJtesouanco.J7t)7.yMfx.1870.

150

REACTION0F AIR

[303.

order ( 200), a.nd 7~(~) is a. function deHncd by th cquaticm

Frox) this thc <ot:t! pressure is do'ivud by introdnctton of thc

!<'f/0-f/~)
factur
so
j.sothat
7r K~
TT

Thc re!K;ti"nof thc air nu thc (fisc may tho.s hc divulud into
two p:u'ts, nfwhict) H u-~t is proportional to tlie vclocityof tlie
dise, and thc sccun't to thc accolcr~tion. If dnote th disof the dise, so
wc hve ~= :t
=
pt!LCC<ne))t
that
=
and thercforc in thc cquation of motion ofthe dise, th raction of
thc air is rcprospnted by a frictional force ao-. 'n'-R".
tj"~ )
( 1
retarding thc motion, and by an accession to the mcrtia c~ual to
'7TO'
r~
r)\
~~(2~?).

~02.]
When
(~2()().

ON A VIBRATINGCIRCULAHPLATE.

151

is stual), we have from thc ascendinc. sries for

iz
,7

From thc nature of the case thc cocMcicnt

must be
of
positive, otherwisc thc raction of tlie air would tend to au~mcnt
i't.stc~J uf to retard, thc motion.
That (~) is in f~ct dways Ic.ss
t))an x may he vcrincd as Miows. If lie betwcen 0 and pr, and
bc positive, sin (.: sin 0)
sin is negaLive,and theruforc also

Is ngative. But this

intcgm! is J, (.:)
ncgahvc for a!! positive values ofz.

z, winch is aceordingly

When
is grt, ,7,(2~) tends to v~nish, and then t))C
fnct.cnat tenu bccomes
This i-esu)t might
si.npiy a<r.7r~
hve becn expeeted; for w!ien
is very large, t))c wave motion
the ueighbonrhood of the dise becotnes
~pproxi.nate)yp)ane.
h~e then by (G)~nd
in ~hich
(8) 2~5, ~=~
is the
density (o-); so that th retardmg force is 7r7); = a<r.7r~
We h~vc now to considcr the term
rcprcsGnting !ni itration
ofmertia, and among otiter titings to prnvc t)~t this altemtton Is
an increase, or t)tat
(~ is positive. By direct Intgration of the
asccndiug sries (5) for ~(whieh Is aiways convergent)

This part of th reaction of th ah- is tlierefore

representfd by
tlie
snpposing
vibmting plate to carry with it n. mass of air cquat
to that cont:uned in a
wliose base is the plate, ~nd whosc
cylindcr
Q7~
hcight is equal to
so tliat, whcn tlic p)~tc is s~ciently sn~H,
the mass to be added is
independent of tlie period of vibration.

152

RHACTKJX UF AHi

[~U~.

an Intcgrn.1of which cvo'y douent is positive. When s is very

l!).rgc,cos(3sin6) nuctn~tcswith ~ruct.tnLpit.ht,y,:t)tdtims 7~(~)
tends to thc i'orm

is grcat, tho aacemhng sries fur 7~aud 7~, thougb atways

n)ti)n:~te)y convcrguut.bccomG uscicss furpract,ic:).t c:t.)cuh).t,ion,a.nditt
Is nccu~sary tu rcsort to othcr pruccs.s(;M. It will bc ohservcd thui
thc diMcrcnti:).l cquaLion (1C) sa.tisHt.'d by 7~ is th saine as th:Lt
be!onging to th Bcsscl's function < with tho cxccptimi of thc
When

Thc funetiou A'is thereforu

terni on tho right-ha.nJ sidc, viz.
TT~
included iu thc form obt~nned hy iutding to th genenti sulution of
Hcsscl's c<(U:T.t[<)n
contiUtnng two :n'hitr:u'y constauts :U)y particular
soiutiun of (!()). Snch a particuhn' sulution is
~7r.A''(~)=~+l'3'3"l'C'=.~+l'.3=.5~.7'(21),
Dt scrics on thc
as may bc rcadi)y Ycntled on substituttun.
right uf (~t) notwithMtandin~ its utthnato divcrgc'ncy, tnay bc
is g'n'at.. It is in fact
usc't) succu.ssfuHy for computatum whcn

302.) ]

UN A YlBRA'nNC!

CIRCL'LAH

thc anatyticd c<iulv~]e!ttuf~e'(s''+/3'')-

PLATE.

153

and we nii~ht.taku

dcteDnining thc t\v<j fu'bitt'iu'y cnn.stiUlts by !ui cxa.nunation of thu

furms a~su!U(;dw!)cn .? is very grott. But it is pct-hups simpicr tu
foilow th method uscd hy Lipschitz' for Bcssefs fnuctions.
By ('i') wc hve

fe"f~w

wltcrc ~ i.s a complex vfu'mLIcof

Considcr tlic intcgral ) .-=
~vl+~'
th fn)'m ?<+!
Rcprc.scuLing,as u.sua!, simultn.neoua p:u)'s of
values of a)i(t u Ly t,hc co-ordinats of a point, wo sec that tt)c
value of Die intc~r:t.) will be xero, if thc intt.'gra.tionw!th respect
to M r~UT~~roundth rcct!mg)c,hosc angutar points are ruspcctivcty 0, A+~ , w!ier(;Ais any rcat positive quautity. Thus

Thcri!sttcrmonthjri~htiu(24.)
Iscntirelyunnginiu'y; M
therctorc fuDowsby (22) that ~7rJ.(~) is thc rc:U part of titc
.secondterni. By expanding tho binumial nndur t)ie late~r.d si~n,
and aftcrwit.rdsiutcgmting by th fui'tuu)a.

154

0F AIH
REACTION

[302.

By stopping thc expfUisiut) n.ftcr any (tcsircd number of tcrms,

and forniutg thc expression <u)' thc l'cmaindt;)', it inny ~)c .~vcd
th;~ titC c-rt-ut' c~tunuLtud by Hc~icL'ijtn~LiiC rcma.mdci' 'cU.aoL
cxcccd thc i:~t tci'ni rctainct) ( 200).
In )ikc m:t.nncr tho ItUi~inary part of thc nght-hand
uf (24') Is thc e~uivak'ut of ~t'7r/~(2), su tha.t

mcmbcr

It nppears then th~t 7~ docs not. vanish when is grea.t, but

approximates to J. z. But ntthough tlie accession to th mcrtin,
Aswn.f!
to ho cxpccted,
tho ficrieH
wititinbrac~ets
arethosamoas thosothat
occurin thoexpression
of thofuuctiou'~(~).
).

303.]]

ON A VIBRATINO CIRCULAR PLATE.

155

wh:ch is pmpnrtio~al

to 7~, bccomcs !nH)]ite with

it vn.nishcs
')!thnat(;iywhcnc()rnparo()wit!)th''a)-(':),r,t't))(i~isr;)n<)wlt.!)t.!)n
nt.h<jrtcrm whicii rc'prcsotts t))u ttisnip~Lion. And t!tls
~t-ucs
wit!t whiLt \vc shouht nnticipitte f'rom tlic
thuory ofptane \vavcs.
If, Im)cpcHdently of thc rc:tction of Lhc nii-, tlie m~s uf tito
p]atc bu jV, :U)d t))u furce of restitution bc
th cquatiott of
jnotiu!) of t!fu pJatu whun nctcd u!t by mi
imprcsscd furcc
proport,iot)!t.)to e" will bu

Two p:u-t!cu]:u'cases of' this problem (k'servc notice. First let

.Vand v:mi.s)),so t)mt t.hc plate, itscifdcvox) uf !nass, is subject
Lono ut.hcr forces tha.n F a.m! t!iose arisin~ from acrifj
pressurc-s.
Smce ~=~(~,thc friction:~ tcrm is reta.tivctyncg)igiblc,andwc
is very smail,
~et wh(;n

Ncxt !ct
aud bc such (.hatt!te n:ttund penod of tlie plate,
wltcn subject tu t!)e react.Kjnoft)ie !ur,is thc same us that
imposcd
it.
Undur
tljcse
circumsta.nccs
'tpoa

Comparing with (31), wesec tha.t tlie amplitude of vibration is

grcater in thc case whcn tlie incrtia of th air is bahtnecd, in thc
ratio of J C 3~
shcwing !a.)'~oincreasc whcn
i.s mnalt. In
thu Hrst case tho phase of Lhc motion,is such that
compa-rativ~ly
)itt)c
work
is
<!one by the force
A'cry
w])i]c in th second, thc
incrtia of thc air is compcnsatol by th spring, and t)tcn
bL-in'
of tim same phase as the vcbcity, docs thu maximum a.mount(~'
work.

C'UAPTEU XVI.
'nmuRY

;ill,`3. Wth, pipe If,

UL'R).JSUNATUR.S.

(),le end alld open nt tllc ot]Il~I'

wc luul

ln cl'l.taill clafillite
vilrratin~
lmriruls to
to itsco/f il IJI(JI'(~ur Il'ss
COIl1Jdl'tl~ i(JUI)L-11(lellc("
of tho extl'rll;d
If tlll! air Icyolld
atlliusphere.
thu nuutiul witllin tlm
pil: \l'ollld Ilavu 110
tu usca)Jo, :tilt! the
to
cuutuiuecl coIlIll1n
system 110t sII1dect to
lit aetllal
di.;sipatiou"
tilt!

illertiu

tllC

C'xturnaI

air

~t?~?~

uf tlw pipe
.?.r~~
bu
ilJsiglli/iilnt,
:lI1d tllull vimtions ou ce
caciteti itl tlie pipe have Il
ul' }Jl'I'.;i:tulll'C, '1.'llu
Ilarl'O\or tho ch:lIInd of co1,),,
"f
~.)M,.I
ti,a
extc.n.J
'~di.
'c.
L~n,
s.
'<
wvvities
'c.
U~
cOllstitllte
ill
tlle
l'usonators;
lresellce uf <tu extu l'Il a1
of
sound, tllu colltviuuul ,ic vibratus in
ullisou, alld witll ail
and
fOl'ced llcriml,,
~EE=~
l'iliing ts .gl'l!at intcllsity iu tlie of
appl'Oxi~S~~
rI 'SUlla/uryiclds tliu
vil)l.~Ltioli.4
111)
IJel'UlIling
~t.m,~s.e..n<i,u-y
th~rv
~t.tu~
i,
sll10t.

as it wel'e witllil it,

.source T),c

.s'yH~

case of
<
<~
t c.
<' T~
~i<!
of
i"c.~i.s.
8o
tlutt tlm 11l'I'SSllr!!is
absulut.dy COIIstallt. If nosv the
is cli:r that
F:
t.lm CUI[aillt'd air
will lm aL :111)'tillle
vury nuarly Il tlm (~tlIi~<
dc:nsity) concH!wlldillg to tllr;
-Ic

303.]

rOTnXTIAL

EXKRCY 0F

COMPRKSSIOX.

157

momentary positon uf th piston. If th tna.s.sof the piston bc

vcry considrable ni coniparison with that of the induded air, th
natural vibrations resu)ting from a displacemc'nt wiHoccur ncarfy
~s if the air Lad no inerti:).; and in dcriving thc poriod frotn thc;
kinctic and putcntifd cnergics, tLe former nmy bc ca!cu)ated without aHowanccfor th incrtia of thc ni)', fmd tho taticr as if t)ic
ra~factio)i and condensation Wt;re nnifonn. Undc'r thc c'ircnmstances cnntumptatcd the air acts meruiy as a sprin~ in virtuc of
its rsistance tn compressionor dilatation; thcform ofthc.containing vessel ia thcrcforc Immateria!, and t)ic p(;ru:)d of vibration
rcmains tiic samc, providcd t)te capacity bo not varicd.
W))cn a gas is comprcsscd or rarL'ncd,th<;mccitanical value of
the rcsuiting displaccmcntis found hy rnu[t!p!ying cach infinitsimal Incrcmcnt of vo]n)nc by the corrcsponding' prcssurG and
intcgrating over thc range rco~uircd. In tho present case It is of
course only the diffrence of pressure on thc two sides of t!)C
piston -winch is rcatiy operative, aud this for a smaU cha~~e is
proportionat to the altration, of volume. 'Thewhoicmechit.nical
vaJuc of the sma)l change is th samc as if thc expansion wcro
that is )):i]f tlie nna), pressure
opposcd thronghout by thc ??;<?f!;?,
thus corresponding to a chitng'! of vobfnn; froin ,S to /S'+8)S',
since ?)= f<

Lut us now Imagine a vcs.se'tcontaining air, ~vlioseintcrior

connnnuicat.cswitli thc cxtrnat tittnospLcro by a nan'ow apurturo
or nc'c~. It is not t!if)icu)t to sce that this system is capablu of
vihmtiuns smiilartot.lto.sc j)tsbc()nH)')u)'cd,tt)e air in th nei"'hhnm'hoodof titc apcrturc suj~dying the p)acf! of th piston. By
sufMcicntlyincreasin~ th pc-riodof thc vibration may be madc
ns lon~ aH wc ptc~se,a)td wc obtiun rinaUy a state of thin'rs in
(!nmrfu-(!
(1'~)3iH.

158

KtNETIC

ENEHGY OF MOTION

1-

which
thc~nct.cc.c~y.fthc
nation n~ybo n~cctc.)
c.~pt
t'.c
.f U.c apeur, and the
potcntial c,.o~y
le ~hourh.
c.ic.fatc.d as if th.
the
~-0 c~~
T.
ou t).c two si~.s, or in virtuc .f'it.s
own
<L~ aftc, .suc). p,-<ss.n.. ).,s
~asc.]. thc. air ,vcs
npproxin~.tefy
"v ;~d'r
d

..st.)y,
mwd

~s~
that t.).. sj.<
r"
D.ruug). w),id. t),c )<i,ti,.
i
~T.y
ill
w).c.L
arc ..h.ut t.
pr.cc.cd
n.t of
t.)
arc

r~i"?~
<
"J!
s < < accm-ntc
su~<t)y
ca)r.,)ati.~ of tj.c r,iLch
~'Jn-v ),
is
illdufinjt('!y
grullt io eUlI1pal'iSOIl
witll fllo dilJH'lJsions of tlm
=:
is
,`~()~1. 'l'lie l:il'tic
cll('rg'j' tlie
motion uf ail
111(Y71171)!'CSSI))1C
nl;ly lle uXpl'l'ssl'd il terms tlie
dellsity p~ tlie
rnte or tI'aIlHfel',or
cnrrcnt, l', fur mulcr the cil'=''=:-?.~
tllc illotioli is
tll(: S:UIIl'.
al \nj's
Hi 1)('('

l' 11('('I'SS:ll'i]y

varies

as

p and

as

.2,

11'l~ Illll.y

put

~~T'
~y on U.c ..t.<f,
nlt;mm~l,i, a Ji'll'nl' I~Il;ltltir,y,;1~
lllay 1)(tiufurrcnl t'l'III 11tllCfilet flmt
:3 in SP1l<'l'alJd-1 1 in
timc. 111
be tlie
il'O
\'('!c)('it,y-pntvlItial,

l'y GrL'Oll'S1111'01'('11),
w](.re tlie lltl'gmtion i; to
hu l'xtf'lllll'} 0\'(11'
i,
sensihle. l1t a sld1il'il'nf clistanr~c~
01}l.illll'r sicll' uf tlm
~F=~
becOIIIOSC'ollstnllt, :lIlt! if t1~
apl'I'IIII'I'. rp
constant. \"allll's lm )PIIClt(.cI
lu- ~y1 nncl
tllat Ilaif nt' fCl\Y<lrd.
wlJieh
the'
{J'liC]
wc'
:EE~
I,W(!
flo\s,

304.]

TIIROUGH
NARROW
PASSAGES. 159

Now, since within

or
vah.cs,

is detcrmincd littearly by its surface

is proportiona) to (~).
.Y ==c (~~ ~), wogt as bcforc

If wc put.

Thc nature of th constant

cwitibnhnttcrundcrstondbyco))t)ie
(dectricid
si(b;)-ing
probicm, w))osc conditions arc mathonatK-a!)~
ideatica! with thoso of that undcr (Hucossio)~ Let us
.suppose that
theHuid i.s replaccd by
unift)nn]ycon(]uctingmatcria], an.! t)iat thc
Lun)](]!nyof titu dtanne! or aporhu-c is rcp]ad h~ insulators. Wc
know that ifhy batk'ry powcr nr utho-wiso, a di~rcnco ofc.Icctnc
tx'tcntial bc maintaincd on thc two sicles, stoa()y currcnt
tlirou~h
tho apo-turu of propottiona)
jna~nitudc will bu gt-nf~-atcd. Thc
ratio of the total currcnt to t)n. ul~ctrontotivc forcu i.s caik-d thc
C(~!</M~/< of the (-hiulnc), and tLus wc sec that our coxstant
c rcprusunts si).)p)yt))is
cnn<h)cti\'ity, on 1)~' supposition that th
sp~Hiccoid.tcti)~
powcr of thc hyputhctira! sub.stancc is unity.
Thc sa.tnc t!)~ may bu tbcnvisccxprc~ud
by sa-yin~ tii:Lt c is thc
sicle of tbu cube, w)K)sc rc.sist.ancc bcit\vu<;n
opposite f:)c<~ is t)to
.satncast))atoft)icc)unmu[.
Lttbcs~~tut
wushidioftcnavaii
m))-.sL'lvesof't))cc)cct,ric!t,)ana]ugy.
~Vhf-n cisknuwt),
theprnpo-tnnc
c'asi)y (tuduecd. Since

of tb~ rc.sonatorcanhc

1CO

NATURAL

riTOt

nnd
w.H ~ncsd.rc~iyasthchncar
ho observe.
is functiun

OFJ~OXATOt{S.

f'~o-t.

dhnc.n.sio..
ufth~sizc.

The~vc-k.n~h
h.
and .shape ofth~
al.so upon thc natur.
'natoron)y,whi)ot!,efreqncncyd.pcnds
<.i U.c ga.s a.id it ..s
important <.orc.nark that it i.son t).c nature of
g.s n. and ncar t).c ci.a.nc.I that D.c
pitch deponds and not on
that oecupymg the inturior of th.
vc.ssd, for thu incrtia of tlle air
in the latter situation dr.cs not
cne int~ p)ay,whilc the eompres.siblilty cf ai ga.scs is vcry approxunatciy thc .sa.uc.. Th..s In
the cascofa
pipe, thc substitution ofhydrr~.n
f.)r mrin t).<~
nc.,g)t],ourhood oi- a a~jc wo.dd ,nakc 1)ut )itt)c
di~.n-nco but its
~rtcct in <ho neighhourhuod ofa
!nop wuu).! hc c<.nsidL.rah)e.
Hithc.~wchavcsp~f-u

r.fth.hann.)ofcnnunun!ca)!onas
sn~h',hmdt!K.rehcm.~than..n(..d,ann..),t).c
pn.Demis nnt
os.scnt.attyaltcn.d.
Thc.sam<.for.m.)a
~rt).c~.pK-ncyi.sst!)t
ifa.s
npp).cah)G,
hcforc~
u.)d~-s(a.,d hy c <hc whoie cnnduct'vlty hctwccn the intL.rior aud cxturh.r of the vessc). Whcu tho
channct.s

are

.s~uatc-d

sumcicnt.fy

far

aj~-t

tn

act

indcpcndonDv
onc ofanothcr, thc
i.s
thc
rcsuXant conductivity
simple .su.n of
thosc bcinng.ng to t].e
~.paratc cha.u,c.]s; othurwi.sc t)ic rsultant
!.sJcsstha)ithatca]cn)atedhymcrcad()ition.
If tho-e be two
prccisciy simitar c!)a!)nc].s, which do ])<.<
iiiterfc~, nnd ~vh.~ conductivity takcn
.sc.pa.-ateiyis p, wc ].ave

~Oi.J

StJrEfUOR

A~D

I~FERIOR

L~n'rs.

ICI

shcw.ng that th note is ).igher than Ifthcre

wcre only o,
channct in thc ratio V:~ =], or
by rat.h<.r Jc.ss t).an a f.ftl~-a. ]~J
obser~cd by Sondj.au.ss and
pro~.d tj.eo~.ticaiiy by Hdmho!t/ in
t''e case, whero thc
d.anndsofconunnnication
consistofsimntc
Lofes
u)t]tcntnnitc)ythi;)sidc.soft))<.ro.s<.rvoi)-.
305. Th invcsti~t;
of thc
condn<.tivity fur varions Mnds
ofchanndsLs an important
part oi-the th.cryofrcson.tor.s;
b.,t
m a!I c-.xccj.t very icw c.~ thc ~ecur~tc
so)ution of thc nroLi~
t).e po~.
bcyo,
of cxi.stin~ ,,mthc,ties.
So.nc gcnem[
r.nc.pic.s throwin~ )ig).t on thc ~.e.sthu.
jn.~y howcver hc Jai.)
~d
,n
<!own
in.-u.y ca.se.sof inturc.st an ~pp..oxi,n..Ltc
solution, .suHicicnt for pt~cttca! purposes,
may hc obt~in~
Wc know (~
2~) th~ thc c.nc~y oF fh.ij
fio.vi..
channel aumot bc grcat.r fhan th..t of
tin-o.~h
.~ny ~ctitiou~
motion g-vm~th~nc
total cnn-cnt. Hcnco, if th
channct Lo
r'arrowed in any way, or
~ny r.b.s~ruetion hc intn,.)uccd, thc con<ct.v,<y,.stt.crchy
<)unini.sh~,)~use
tho attention is ofth.
nature oi an additions) con.strai..t. Hetor.
thc change t).cilm.!
~freeto
adopt thc distribution ofnow finat!ya.s.su,ncd
rn
cases whcrc a rigorons .sutution cannot be
chtainod we may use t),e
muninnm property to e.sti.na~ an inf.rior
limit to tho
con.iu<-tivitv.
i.c cncrgy ca)cu!atcd fn,.n a
hypr,t)H.tica! law of <)<nveau ncv.r ho
c.ss than th.truO.
aud nu.st cxc..c.d it un!cs.s )hc
hynothctica!
a.ndt))eactua!niot)ou
concide.
Anothcr ~cncra! principh, Avtuch is of
frcqncnt u.so mav !.r.
more
.statcd
in
convc.nientiy
.k.ctrica) iauguago. T).c
qua~ity
w.t). wh.ch wc arc concc.-ncd is thc
conduc.tivity of a certain conofmattcr
<ctorcon.po.scd
ofunih spcifie
Thc
con~.n.tivity.
,.s
that
if t).c conductivity <,f
rrn.c,pic
of t).c condnctor
any
part
'Je jncrca.scd that of thc who]c is
incrca.scd, and if t}.c conductivi~
of any part hc (hmini.sf.cd that of
thc who)e is dindnishcd
Mccption bcmg ma.]e..f certain vcryparticn)ar
cases, v-hcrc n.'
.-dterat.oncn.suc.s. In its pas.sa~
Hn-o,~h acondnctorclectricity
'L.strd~tc.s itscif. se t],at t).c
encr.~y di.ssipatc<! i.s for a givcn total
cnrrcntthc]ca.stpo.ssib!u(.S.').
-[fnowD.c.specinc rsistance ofany
be
P~-t
dumn.8t.cd, th total (fissipation would bc !cs.s t!.an
heforc
even if t),c di.strib.ttion ofc..rrent.s
rcmaincd unci.an~d
will this he thc case, w),c.n <he currents
redistribua the.n.sctves so a.s to makc thc
dissipation a minimum. If an innnite)y
~.ir.

162

SIMPLEAPERTURES.

[305.

thin

lamina ofmathT
strc-tc-hiog across thc channd bcmade
perfcctiy conductin~. tin; rt.-sist.tncuof f])e wholu will bu diminishcd,
nn]css t)tc iarnina coincitk) with onc ofthu undistorb~)
cqui~nttal surfais.
In th excc-ptcd case MOcH'cct will Le
produccd.
HOG. A))]f)i)~ (1iffu'(;t)t k)n(f.s of cl):u)nc)s !m
Important place
!nu.st))C a.s.s)g))C()tot))0.su
cunsiHt.in~ofsim~tcapcrturusinun!m)it;c(tpi!uuj\v:).!)sofin<tnitcsin):dti)te!<))L'ss.
Inpractica.):q)]))tcatiuns it is HufHciunt thitt :). wa)! bu
vury thin in proporHon tu thu
'-hn)(.-nsi<)t)sof t.)tu a;)urt.urc,
aud:t.p[)t-uxi))):ttc~y piano wiUutt a
(hst.Lncu frum thu apurturu lar~u in
propurt.iun tu thc i~mc
quant.it.y.
On account of (.hc;symmct,i-y un t)t<jtwo sides of thc
wa!), tho
tH()ti())t()t't!icf)nidint!~p):m(j()ft)tcapcr<,))!-emu.stbc!tonHa),
aud tfturef'otc t)~ vu)(jcit.y-}tf)t.cntia.Imust bc
c~st.imt; over thc
n-)n:undcr<jft.)t~ p)a.!)uthu mution mosL bu
exc!))M[\'c)j tfn~ntia),
so that. tu
on one sidc of thc piaoc ~e h:tvu thc
dtjturtninc
coudrions (a)
= 0 over
= const:mt ovcr tl.c ~pertuj-e, (/9)
H~
thc rcst uft))c phmc cft!.c waH,
~) ~== constant at ititinity.
Since wc arc conccrncd on)y with thc di ~renes
of \ve may
-st'ppf'.so tt.at at itifinity
vani.s].c.s. It will b~ .sccn Htat condit.i..u.s
(/3) and (y) arc satist.ud hy supposing
to bu t)tc potoitfat of
natter diHtnhutcd ov(;rt))capc!-turc-; t!te rcmaindcr of
attractif
thc prubjon cunsists i)i
dcterni.ung th distribution of mattcr so
that its putcntial
may bc constant over th .satnc fu-ua. Tho
proi.icm is m:Lt]t~tn!Ltic.i)tythc saine as that of
determining t].c
df.stnbntiou of cluctricity on a
chargcd eonducting pjatc situated
"ian opcnsp~cc, w))osc fcnti is that of t))c
~pcrtm-c nndcr constdLration, .nd t)iu cooductivity (,f thc apt-rturc
tnay be cxprc-sscd
'n ~i-ms oftj.e e~x~
of tiju piate of thc statical
prohicm. If
dc-untc tho constant, potcntiat in th
a.pL-rtnre, thc ulcetricai
rsistance (fur onc side oniv) will bu

th intgration cxtcndmg ov-crthc arca of tlie

opcning.
AI
ff~!
qu~ntity of mattcr Jistributcd),
jj~~=27rx(whu)c
and thus, ifj~be t).c
caj~eity, or charge corn-spon()i,,gtounitputc.nti.d, thc total rcsi~ncc is (7rJ/)- Aceordiugly fur th coa-

~1 J

ELLrr'riCAPERTURE.

T~~

ductivity, whieh is tlie rcciproc..Jof t).G

rsistance,
So hu. as 1
aw.ro, tho eitipse I.s t)io o.]y f.,nn of
apertm-c
f.r winch cor Veau be
dcter.ni.ed thcorctijy',
in w!n~
'.c rc-.s.k is ~ch.d.1 i. t,
e.Iip.conduct.r.
l'.u
1~.?' n th. h et th.~ a shujt Loundcd
by two conccntric, si.nihu..ud
s..n!arly
.H.p.suids exert. uo f.rcc on .n i.tcru.)
~d
p.t..ie
~.s c..sy to sce t)..t t).u
sup~ci.I
at
dcn.sity
.uypoiu~f. ne .?-'
.i~
j~
~j.,
P
~dnece.ss.uytu
Pc peud.cui.r ~) let <)
~~J
tallgent
Le pc.nt in .sti.u.
Tj.us if
bo th dunsi~,
;=
whole qu~L.ty of
is givcn by
nu~cr

.iwenow
suppose t~tci.sinfinitdysma)!,
t.cuhu.c..se of' an citiptic pi.to, and if
Lutwccu tlic t\o surfaces, wu
gct

wco~unthcparuo lungcr disth~ui.~

Wu !,avc nc.xtto find th va!ue of t]ic

const.-mtpotcnt.al (P)
thc value of
at tUc ccutre of thc
p!~ wc sec
~con~dcring
tliat

(c~IsnTT"
LyHoi~h.itz
~ruiL, j~t. ~7, l8),0), whnHo rusutt is
cquivn)t;nt tu (.S)
b~ fur the momeut tho thirj principe axis of thu
sHipsoid.

11-2

1C4

nsthcf)na]

ELLIPTIC APERTUHK.

expression f()rt.))cc:)pf)c!tyofnne)]ip.se,wh(~cscn)it))!)j')raxi.si.sf!nth)cpc~)~ricityi.se.
Jnthcpnrt.icolarcnsbofihc
=
=
e
circle,
0, ~'(e) ~7r, and Ums for circ)Li ~t' radius 7)',

['30G.

COMPARISON

30G.]

WITH

CIRCULAR

APERTURE.

1G5

From this rcsult we scn that, if its

ecccntricity bc smd), the
oonductiv~y of au eiiptic apcrturc is very ncar]y thc .samc as
titfLtof a cirotiiu- aperture o/'
e</?<~~-e< Among various furms
"t'apcrturc ofgtvcu m-cn.titcre inust bc ouc whic)t has a nnnimutu
conduetivity, a)i(], t]mugh Aformat proof ]night bc <)ifHcu)t,it. is
c~sy to rc~ni.sct)iat this eau ho no ot~cr thati thc circle, An
itifurtor limit to th value of c is thus
a)wa,ysailui-ttcdhy tho cou'Livlty of ~e circle of c.~al arca, that is 2
and ~hcn
V \7!
t)ic truc furm is
ncarty cu-cnt!u-,t]ti.sH.mb may bc takcu M a close
approximation tu th rcat v~nc.
Ti'c vainc of

thon givcn ))y

In ordcr to shuw iiw

sti~ht)y a moderato ccccntnci~y~cc~
t).c value of c, 1 hve c~cuktcd th
foHowingsl.ort tab)c wit)i thc
~d of L~en.h-c's va]ucs of
7-).
Putti~- e=si.i~ wc hve
eus as tho mho of axes, tuul fur the
condnctivity

e==shi~.
o"
30"
5~
~0"
70"
~0"
90"

-ooooo
-~204
-50000
-C~79
-rcGOt
'SGG03
-939M!)
-98481
1-00000

~eoH~.
i-ooooo
-!)39<i!)
.8GG03
.7~04
-C427!)
0
~0000
-3.~03
-173G5
-00000

7r-27''(<')(l-e~.
i-oooo
l.ooon
1-0013
1-OO.it
1-0122
1-0301
l
r07i~
l-l'J5-t
co

Thc vainc of t])e last factor ~-ivoi in t!ic fourtii column is

tho
ratio of thu eonJucLivityof thc
e))ip.scto
o/' ci')-c/eq/' ~)<f~
f<e~. It appears that cvcn whcn tho
cl) ipso is so ccccntric tha.t
thc ratio of t])e axes is 2:1, thc
conductivity is incrcascd by
onty about 3 per cent, which wou!<)correspotut to an attention
of littic more th~n a comma
( 18) m tho pitcli of a reson~or.

166

CALCULATION

BASED ON AREA.

[306.

Thcrc Hccmsuo rcason to suppose that this

approximate in<!cpendence of shape is a property pcculiar to th cHipsc, and we
may condudc wit)t soinc conHdcncethat itt th case of fmy moder~t,c)yc!o))~tcd oval aperturc, th conducth'ity may be calculatod
from thc arca alouc wittt a co))si(tcrab!u
dcgrce of accuracy.
If thc arca bc givcn, therc is no
snperlor lunit to c. For suppose the arca o- to be distnbutcd over Mcqual circtcs
su~ciuntfy
farnparttoact iudcpGndcnDy. Titc arca of cach circle is Mand its comh.cttvity is 2 (~)
Thc whole ccuductivity is )~
titues as grcat, and thercturc incrcasc-s
indefinitcjy with7;. As a
ruic, thc more thc opcning is c]ungatcd 01-Lrokf;u
gc-]iGt-:d
tho
up,
grca.ter will bu thc cunductivity for a givcu arca.
To find a supcrif.r linut to thc
conductivity of agivcn apcrtm-e
wc may av~il oursctves of t!.c
princip)c that any addition to th
apcrttu-c must be attunded hy an incrcaso in th vaine of c. This
in thc c:MCot'a square, wu inay be s..rc that c is
ie.ssth.ui for t)te
c.rcutnscnbcd circle, and wo ),ave
ah-cady scen that it is <r,.e:iter
than fur thc circlc of equal arca. If be thc side of th
s<, ~u-c

fhc tones of a rcsonator with a

.square aperture calcntatcd from
thcsc two )nnits wo.dd difcr t.y abont a. wi.otc
tone; thc ~-avor of
t))cm would <toubt)e.s.sbe muett t!ic ncarcr to thc
truth This
sl.cws
t).at cvcu w)icn anajysi.s fai)s to
exampic
give soutien in
t).c niathe.nat.c~ sensc, we ncpd not bc
a)tugct),cr in th <hu-kas
to thc magnitudes of-thc quantiLics with whicii wc
are deaHug.
In t!ic case of si.mhu-orifices, or
sy.stcmsof oriHccs,c varies as
th hucar dtmcnsiou.
307. Most rcscnatnrs uscd in p~cticc hve nccks of
~-Gateror
Ic.sslength,
cvcn w),cn thc.rc is
nothing that woutd be ca!)ed
neck, thc t)ne].ncss of thc sidc uf t).c rservoir cannot
hc
a!ways
nc~ccted. Wc .s].~ t!)erefore examine tf.c
conductivity of
channe! formcd by a cytindrieal
horing tfn-ough a.i obstruetin.
plate b.~undcd hy para))d planes, and, thun~h wc fait to solvc t)~
prob)cm r.gorousfy,we sh..djubtain information .sufHcicntfor most
pract.ca! p.n-pn.se.s.Ti.u thick~ss of thc p)ate wc sha!! eall and
Z
thc radtus oft!(c cytindricat ch:n)!)d /)'.

307.]

CONDUCTIVITY

0F

NECKS.

IG~

Wh~tcver tho rc.si.s~nceof th c~nnc!

will bc lessened by tlic intr.xtuetiun of mn.y bc
Iufinitc)y
Dnn d)sc.sofperfucb
co~tnctivity m n.n,)
~)
TI.e ct-cctof tlie dises i.sto
proch.ce constant potJntud
ovcr thon- are~, ~d t].e
pruDcm thu.s inodiflud is
su.sccpt.b)c cf rigDrou.ssointiun. Outsidc J .~Kt
thc motion is thc ~ne as tL~t
prcviou~y invcstiwh.n thc obstn,cting plate is
~tcd,
infim-tciyt.hin.
hctwucn ~ud
thc ftow is unifunn. Thc rc.~t-'
a-ncojti t))ercfurc ou thc whoic

Th.s correction is in gnrt undcr thc

mark, but, whcn Z is
cuntp..u-isonwith~ thc a.ss.nncd motion coincid~
~ysn~i,ri
an<hnurc n.arjy ~.ith tho .ctual
nation, ~nd t)n,s t!,o va!ue
n~-e
oi
a:m (2) tonds to buuotnecorrect.
A snperior limit to the rsistance
mny be c.-dcu!.tc<!from
motiun of th Huu). For t).is
wc will suppose
pm-pose
I.ypoth~c..t
t!nn p.stnns introduecd at Yl
and t!.c cr~ct uf which
n~n.tety
will be to inakc t)~ normat
vdocity coa.st.nt nt those p].s
Withu, t).e tube th Howwill t,c. unifor.n a.s
befo~ but fur tlie
externe space wc Ih~'c a ncw prcb)em
tocon.si.tcrTo .L-ter.ninc
tbe mot.o.1of a fh.id bounded
byan i~nitc phu.c, thenor.n~
vclocity over a cireur, ar~ of t).e ptanc having a givcn constant
vaiuc, and over tbc rc!n;undurof th p)anc
being ~cro.
Thc potential may sti)t bc
rogardcdasd~tomattcl-distribntcd 1
ovcr tlie di.sc,but it is no
longer cotant over thc arca; tt.e ~6~
of thc mattcr, I.owcvcr,
is constant.
being proportional to
T)ie kinctic cncrgy oftttc mtiun

I<jS

CONDUCTIVfTY 0F NECKS.

[:~7.

If thu dcnstty uf th(; tnatt<r Le takcn as unity, ~=27r,

thc ruqxir't!

rath) t.s~xprc.s.hy

-t,

wta'rc ~dcnote's

a)ul
t)tC

p')tc))t!:ttt)))it.st']t'<'ft).('h'(.'u!;)t');)y(j['L'fj)):t<tei'(jfuniLdL'nsitynnd
cf')'!h)i)[s7t'.
T))csi)))[)k'.st)nL't])<)(l<)f(.dun].L(,it)~<)o))(')KLsup")it)icc~nsi~crnt.iuit <,)).')titruprt.'MCtitMthewo'k i'c'([U)t'(jdt()Lruak un thc
(Hscintoiufmit.L'HunaIc'iL'mcnt.s a)i(t toi'L'movcthL'tnfruni
cncit
othcr'.sinOucnct'
Ifw(.!t:d<:cp')]:n'co-r)t'(]it):itc.s (/?,~),thHp")c
)tavc for t))C
L)L'i))~).tt))u(;~Q"t'<tL'tU.scwh'jne)'a<n))sisH,L!
t]K: !i)))its<ji'p bum~Oand
putent.ia.t atthu
put(. )''=;)'~6'<
H(tc<js~tnd<)tus<of~bc'in~7rand-t-.t7r.
T's
r=~(:).
Now )ub us eut ()<)':).sh'ip uf brcadt.)t
f)-o]ntitc cdgo cf t))c dise.
Th work j'c<[uin'd tu rmuovu titls to an inHuite di.stanco i.s
~7r~t.4H.
If we gr;u)u!tl)y parc th discduwu touot])ingfu)d
canyidi thc piu'ing.s to infixity, wc fhid fur thc tot:d work hy
intc~r~tij)~ \Itft rc.spL'ct tu from 0 to /<

!i<)'Jis r<'jtcntcd hnre for thc nnhc of thoso who

niay wish to avutti
tht'(im!cutU(.'softhM)t)ur(.'cf))nplt;tuin\-(.'i,t)Ht~H)n.
TJiiti tncthoj uf calcult~tiug
was st~cstej
to thc outhor by Professer
(.'tf'rkMnxwci).
part of

307.j

CORRKCTION TO LENGTH.

1~9

It must bc observai th:~ a hore dcnotcs thc con-cctiun

fur one
end. T)te who)c rL-sistiUtce
cotTfjspottdsto a. !cn~th Z+2a of
tube Iia-vin~the section777~.
Wftcu Z is vary grcat m rci.Lt! to 7~ ~'c
may tf~ slinpiy

Tbe correction for an open ond is a fonction

(~
of eo.ncidin~WithU.e !ow<.r!imIt,vi~w))unZvani.s!h-s.
AsZincrca~
mcrc.a.seswithit; butdo~r.ot,evenwhcnZis
infinitc.attain
thc snpci-iorlimit
7~. For consi.lcr thc motion
going on in any
n'iddie piccc of thc tube. Thc kinctie
cncrgy is grcatcr than
cf.rrcspuitd.sincrciy to thc lo~tli of t).c picc-c. If therefore thc
piucc be ronoved, and th frce cnd.s brougttt togcttier, the motion
<jthcrwiseconthmmg as bcforc, th kinetic
energ-y will hc diminisited more th:ui con-e.spond.s
to tt)c fength of thc pice subtmcted.
~~?'i't'M~ will this bc true oftite real motionwhich would exist in
tbe sitortcned tube. Th~t, wlion Z = ce, a docs not becomo
is
evident, bccausc th normal vc-tucltyat th end, far from
being
as
was
nssumcd
in
thc
constant,
calculation of titis i-esn]t mu.~
increaso from tiie coitrc out~-ardsand becomc Innnite at the
cd'~e.
A furthcr approximation to tlic value of a
may be obtained by
assuming a variable vclocity at thc plane of tbe mouth. Th
c-alculationwill be found in Appeudix A. It
appears that in th
c;tse of an Innnitdy long tuLoeccannot bc so
grt as '82422~.
T!tCreal value ofa is probab!y not far from -82
308. Eesidcs th cyHndcr there are very few forms of
channci -whosc conductivity can be dctcrmined
mathcmatica!)y.
W))cn howevcr tlie fonn is approximatbiy cylindricfd we
may
obtain limits, which arc usefut as altowiug us to cstimatc thc cHect

1~0

TUBES 0F REVOLUTION'.

[308.

of snch dcpartures from mathema.tica.l

accumcy M must occur I[i
practicc.
Au infurior Hmit to
thorcsistanccofn.tiyc)nngatRdn.n<t~pproximatc~y Htt-:u~))tc(jn()uct,(jr tnay be obtait]~) ini!)ic()in.tc)y by the
i))ia~itj!uy introduction of ait mftnite numbcr of piane po'fccDy
co)i(h)cti))g' inycrs pc-rj)nn<ti(j))iarto th axis. Ifo- dototc tLe arc;).
of thu ficction at nny point y, tho rcsist:u~c LctwL'ca t\vo
layurs
di.stunt (~ wiH bc o-(~and
thuruforc th whotc actu:).! rsistance
is ccrtain!y g)'cat(jr thaii

un]ess indccd tho conductur be truly cynndricn.1.

l'i or<]er to find a supcrior Innit wc
n~y c~cu)atc tlie kinctic
cncrgy of t)tc cun'cnt on t))c hyp~thcsis tii~t t)tc vutocity pM-~Ocl
to titu axis iH unifurm over c~c)i .st.-(.-t.!on.Thc
!)ypot))eticaJ motiuji
tH thi).t whi<;)t wou)(t fuliow frotn th intrnttuction of a.n inHnito
munbcr of n~id pistons tnuving frecty, :),nd dtc calot~tcd rcsu]t i.s
ncce-ssiu'ity in exccss of th truth, untess thc suctiun he :d).so)utc!yconst;u)t. Wc shaH suppose for thc sakc of
simplicity that t)tc
channol is synnnutricfd nhuut H)) axis, in which case oi' course tho
motion of'thc nnid is synnnctricf).! :d.so.
If
(IcnotR the total cun-cnt, wc hve M
n.xial vc)ocit.y a.t any poitjf rc

for tho

308.]

SUPERIORMMIT.

1~1

This is tho qnMtlty which gives a

superior limit to tlie resist~nco. Thc first to-m, which corrc.spondsto thc
componcnt velodty
i.s
thc
s:m)c
a.s t)tat prcvious)yobtamcd for the lowcr
?<,
Junit as
ttavc
b~-cnforus~n. Th difcrcncc butwecn t)ie
!~ht
two, wh'ich
givcs t).c utmost en-or invutvcd in tddng eit)ier of titGH/:ts tho
truc vatuc, is

In a ncariy cyHndrical d.annd

is a sma]! qnantity and so

thc rusult found in this manucr is

closely approximatc. It is not
ncs.s:ny U~t the section .simnid be nca.r)y consent, but
only that
it shouid vary s)uw)y. 'l'lie success of the
appi-oxi.nati..n in this
iu.d snni! cases dupends
upon thc fact ti)at, tfto quantity to Le
~tHuat~d is at a ttnnitnuin. Any rc!tsonab!e
appmxim.ttion to thc
r~) motion will ~ivc a ru.suft vcry uear thc truth
accordinrr
to thc
of
t)m
din'urentia! calcuhts.
p)'mcip)cs
By mcans of thc propertics of thc potcntial and strc.tm
ftmctions th prsent probJou adtnits of actual
approxitnatc
suhttion. If
find
dnote th values of titcsc ftitictioiis at
any
point
7'; M,dnote tlie axi:d and transverse vclocitics

173

APPROXIMATECALCULATION.

[308.

If 7'\Jcuote tho vn.tucof as a function of x, v'hcn = 0, thc

in ternis of
gnral valuesof <~ a))<! may bc expi'CM.scd
by
m('n.nsof(7)a!id(H))nthc,(.'ric.

i.s thc quation conncct.mg y and 7~ lu tlic prc.soit proUon y is

gtVtjn, au<t wc ]iavc tu express
hy ])tc:m.s of tt. By succcsijivo
approxnn:itiu)iwcubtainiro]n(l())

Jhc expression for th resisLmccn.dniits of considcrn-bicsi)np!Iiic~tiou by intcgra.tinn by pru-t.sin thc case whcti tho channct is
oft!)u !iniits of intcgrat.ion.
truly cylindrical iu Utnui~hbotu-hooLl
I)i tUs way we Und fur tlic fi)i~ rusu)t,

eoc~n'iolts ofy wiLhrespect to a'.

dcnoting thc ()iir~rcni.i:L)
It thus nppc:u-s th:it thc supcriL'r ]umt of thc prcccd!n~
mvestign.tion is in f~et tho cun-ecLi-L-su)tto L)tcsecond ordci-of
f'f~/f;/.<))t)yott
~/Mf/)<Htf;<;<:()/
r~'OCCCfh')~
Vo).Yfl.Ne. f~.
.S'OC't'/y,

COMPAmsON

308.]

WITII

EXPERIMEKT.

1~3

npproxun~ion. Jfwc regard 2/as a fonction of~herc

~isa
s.n:dt
'tu~tity.n~)IsMrr<.cmsf.i.sf(Tu~cont~,n,
30.'). Our ]<n<~v)..)gc
c,f t)ic j.vs on wf.ieh t).o pitch of
resunators dcponds,1.duc to tl.c i.hour.s of scvcnd
cxpcr ncntor.s

:mdm:).thcnt:t.t)ci:ms.

T).e oh.surv~iun <h~ fur

.nouthpieec H.c pitch of.t

~uuiy upun thc vu)mn. is duc to
rosun~or
dupcmLs
Li.scuvius
..1.0 f..und th.t t],c pitch of
~)cd wit]. watc.r
M part)y
MtaUcrcdwhcnt).ci)n.sk
w.-LSiuciincd. Thi.s r.suk~scon
f!nncd hy Sond).aus.s\ Ti.o )~~r observer f.und
iur~cr, th..t i,.
thc case o<r..son~or.s witf.out neck.~th
influence of th aperture
its
d~pcndcd .nandy upon ..u-<t,,dt).ou~i ~hcn th ~po
vc..y
chjng~tcd, a certain ri.s~cf i.hc]~ u.s.ted. Ko g. t].c formu!~

thc unit uficn~th bcmg t) mii]imctrc.

Thc ~Mry of Uns kind of rcsonator wc
owc to Hdmholt~
whorieiornuu~ts

ln

pmct.cc
).s so long that
ns
itsdf

(4)

it

doc.s

thc

.supposes,
bc ncg!cctc.),

m,

not

oftu.i

currc~iun
on

th

us suppose.!

dt).crt).at

h.~pj.un
fur t).L. opui

ends

othcr

so

hand,

i~ (~.

th

ucek

ean bc )K-]cctcd
short
U~t~b
c.u

\Vcrt)R.i,n<

~s

t).c

first

'r cubi~hon I.feif~.

l'npll. :1 rrn, r.xxxr.
7-T.
=
Crc))c, Bd. n'ii. 172.
1S(!0.
U.ber dk. Seha)I.~hwinK,n
~r Luft in c.rhit~.n Glasri.Lrcu
u~ in .edeck
tpn l'f~fun von n~ichcr
Wuitn. 7'),t.y~ix.
l.S~f)
Mcn.uire Hur )c.s vibrations sonorM
I',tir..)

174

HELMIIOLTZ'S

INVESTIGATION.

[309.

to shc\v that titc cffc'ctof an open end codd bc reprcsoited by

an addition (~) to tim leugt,h, indcpundt.'nt,
~arh.' .o, of
aitd\.
Th approximate thcnrctica! dutcHnination of N is due to
Hu)t)dK')tx,who gavu 7r7~as thc correction fur an opcn end
<itt.edwiLhan infiititu Hangc. His mcinud consisted In invcnting
i'ormHof tntjt! for which thc prubfon was so)nb)G,and scluctin~
tt).t.bonc w])ic)iagrcud most m!:u-)y wit)t a cy)indcr. Tite cort-ccti~))} 77- i.sri~orunsty applicable to a tnbc whosc radins at th
opc'n end and at a ~ruat (Ustancc ft'otu it is 7/, but whic!t in tho
ofUte opun end bulbes .slightiy.
HL'i~'hbu)))'])um)
From t))c tact that thc true cy)indcr !nay bc dcrivcd by int)'tj()ucin~an obstruction, wou)ay infur tbat t)iere.sult thus obtanicd
is too smal).
It is curions that tbc proccss foUowcd in this work, which was
Ht'titgivcn in th menioir oa rc'snnancc, leads to cxact)y thc same
r<sult, thon~h it would be dif~cult to couccivetwo mcthuds more
uniiko c'ach othcr.
Thc correction to tbc Icngtb will dcpcnd to some extcnt upoo
wbeLhcrt!iC itow of air front thc opun end is obstructcd, or not.
Whcu thc ncck projccts into opcn spacc, thcrc will bu less obstrnctiun than whcn a backward Ho\vis prcvcnted by a nange as
Sttppo.d in onr approxinmte caicn!:<.tions. Howcvcr, th unc<rt,ainty in<rodnccdin this way is not very important, and we
may gou;)'a))y take a=~'7r~ as a sufncicnt approximation. In
practicu, wht,'tithc nccks arc short, Die hypothesis of thc Han~e
a~rccs prc'tt.y wc)I witti tact, and whcn thc uccks arc Jong, tlio
curnjcLionis itSL'tfofsuhordinatc itnportaucG.
Thc gnerai formula will thuu run

whure<r Is tlie M'enuf thc section of th ncck, or in numbers

A formula.not dtffuring mneit from this was given, as th emLoduncnt of tlic rcsults of I)!s )neitsuru)ncnts,by Son<n~uss' who
'?..i"cxL.53,219.

1870.

30D.]

MULTIPLE
RESONANCE.

175

at tlie .samc time

cxprc.s.s.d a eunviction t).at it was no mre
".np.n~)

~mu)aofinturp.)atiM,L~t!,o.u..n,
pf r~sonators
ti,uury

Inc

,,atn,
wib).

~.nc tnnc

t.ucks

wa.s

give-a

abo~u

th

lu a mc.noir un Rc.s~nance
pub~hud in ti.c
fur 1871, fru.u w].icit inust uf thu JasL
icw
durivcd.
p~cs )s
:310. Tho .simph .nethod of
c.~c.uh.tingth. pitch ..frcsouators
w.di ~.ch
),~c bccn uc<picd is
.j.),)iu.bfc to thu ~vc.st.
~cof
v,br.tion ou)y, ti.o ch.r
.,f ~.ch i.s.juitc
distinct.
'ho ov.Ttu.cs ofrc.s<.n..Lur.s witll
eontrac~d ncch.s ~.o
n.Iativdv
~y h~h and tlic c.,rrc.spon.]i~ un.des of vibr.~iun arc
1)y no
'nc.a,.s .ndupcndcnt of t).c inertie of the
..ur in tho intcrior of the
l),c ci.a~ctcr of thc.sc mo<).s wi)) be
more ovident
r~rvoir.
K.n wc comc to considcr H.c
vibrat.-ons of air within a cu.n-'
]. c y c!d
as a sphorc. but it will
v~d..such
nu.dy h.ppcn
tllat tho p)tch can be ca!cu)atcd
tJ.eurcti(;:diy.
Ti.crc arc, howcvcr, cases uf
,n,dtip)c rcson~cc to which our
is
t'.cory
app),cabJe. Thcse occur wf.un two or ~orc vcs.sd.s
c..n"atc
channcfs
~th each ot).cr and wit), thc externat
by
air.
.d .u.c rcaddy trc.atcd
Ly L.rangc.
n.cth.d, pro.idcd f course
ti.at th wav.tc.~th
af thc vihrati.n is
suf!icic.ntjy I.rge iu comp.-tn.sonwitht)~di.Hc.iu,).suft)tuvc..s.sd.s.
Suppo~ ti.at thc.rc arc two rc.s.rv.irs.
con~nunicatin~
~it!. cach othcr aud with tl.c c.Lcrnat
air by ~arro~
passage, o~

If wo wero to con.si~r
a. a sin~c rservoir an.!
ncc~.
~pp]y
wc shouM bc !.d to
~f prenons
cn-.n.ou.s rc.snit fur
furmuf~
formula is fOlllldcdan tlle
:~tluLt
witlia the rescrvuir
aS~lImptioutll1Lt
th
u.crti.. of thc air j.~
,~hatiormu)~sf.,unded
bc~u.nption
Jcft eut of
mayth
t).atw~h:n
ac.our~, ~hcr~
tho rc.sc.'vcir
it Is
is
cvnicnt

pa~gc

that

iuay

th

hc

of

cne~y
a.s ~rcat

~')-uc~t'

a.s

the
through

motion
th

(~ </te ~(~(~ .S-~i(.

t)n-o~h
two others.

th

Koy. 2.i, 1870.

conncctin~
Ho~cv.

a~

17G

DOUBLE

RESONATOR.

[~~0.

~'
P~"
invcsLig~ion on th Hfune K'c~l
Hnid1
.Y,, A\ t])C totat tr:u).s~.rs of
perfuctiy. Dcuoti)~ by
Lave as m (2) 304. fur thc kiu(jtic
wc
thc
Un-ce
pa-s~gcs,
tlirou~h
cncrgy t))C cxpi't's-iion

as thc cqu:ttirm to dtermine thc nutund toncs. IfJVbe thc

~e two values of~" bcing of
frcqucucyof vibt-~tio)),~=course rca.1and ucgativc. Thc fonnu]:).simptifles considcr~Myif
~=c,, ~'=<S'; but it will be inore inst-ructivc to wurk eut this
case from tlie bp~inning. Let = =?"~ = ~'c.

~J

noUBLE RESONATOR.

~y

~'hich rc.juirc.s
= 0. Tho motion is thei-cforc th
th~
.sa.nc
t.:Lkc
'night
p)acc wcro th Ct))u)nun!c:tt:on bctwccn ~'an<~S"cut
ofT,aud bas its fi'Gfjuuncy~ivcn by

~1).cvibmtion.s~ro tl.us opposcd m

phase. T!.e ratio of frcqucncics
is ~vcn Ly~=~+2:
s~cwin~ th.tt th second mode
h~ thc .shortcr penod. In this mo.Ic of vibration
t))c conncctinrr
passage acts m .somcmcasurc as a.second opcning tn hoth
ve~c]~
:md t!)us nuscs t])e pi~-).. If thc
pesage hc contractcd. thc intervat
of pitc)i bctwecn th t\vo notes is stna)).
A pfu-t!cu)M-easc of thc ~cncriL)fonnuia
notice is
ohtfuned byputting~=(), ~Lich amounts to woi-t))yof
one of
t))c commumcatiuns with thc cxtcrna) air. Wcsupprcssins
ttms obtain

n. ir.

178

PAUTICULAR

CASE.

~310.

It appcars that thu into-v~ from

JV, to A~ I.stnc samc as from
or ratbcr more th.in a fifth.
~to~,namc)y,(2'(JlS)=l'(i1.s,
)t will be fonnd that wh~~vcr th vainc of w
]nay hc, <Lchttcrv:d
hetwool thc two toncs caunot bc 1~-ssth:in
2'4.).i., winch IsaLout
:m octave !uid a minor titini. J'hu
con-c.spunding vaiuc of?/t is 2.
A sunilar mcthod is applicable to
any combination, howGvcr
f'otnp)icat.u(!, of rcscrvuirs aud cunnecting puisages uado- thc
~in~c rcstricLioH as tu thu comparative magnitudes of t)tc rpservoir.s aud wn-c-Icngth.s; Lut Hie
examp]c just ~ivcn is sumci~nt
to iliustratc titc thuory
of']nu)tip)c rsonance. A few mensure)UL-ntsof th pitch of duubtc rcsonators arc dctailcd in
]ny rnc-moir
on rcsottanc-c, atrcady refon'L'd to.
:].
Thc quations winch wc ))ave
cmploycd Intticrto t:~c
no account of thc (-scapc of
encrgy from a resonator. If' tho-c
wcn- rcnHy no transfcr of
cnurgy ~ctwcon a rcsonator and thc
cxt'jrna) attno.sph~ru, t)tc motion won)d bo isoiat.cd and of
IItUc
jn'acticat iuterc.st; nuvcrthcic.s.s 1)10 characteri.sticcfa
rcsonator
ru.i.sists In it.s vi))rations b~in~ in
~rcat mca.surc indcpendent.
Vibrations, once c.xci~d, wii! c-onti.mc for consiuL-riLbicnumber of
pt-riods \vit!tout nmoh ]oss of rncrgy, and HK.ir frc.mcDcy will bc
at.nost cntify indcpc.ndcnt of t).L. rate uf
dissipation. Tbe rate
of

dissipation

is,

howhvf-r.an

important

fcatm'c

in

tLc

chamcter

~-1

COMMUNICATION

0F ENHH(.Y.

l~:)

~f

rcsonator, on whic). it.s bcL~io.u- undcr certain circ.tm.stanc~

.~tcnaHy <!e,x.nd.s. It ~j)!bc n~crst~!t).~t
t).c<]i.ssip~ion
'.cro spoken uf n.e~ns
tii..
c..)y
cscape f cner~y fn.m thu vc.I
aud its nc~),bour)tood, and Its
ditl-us~onin thc .s.n-roundm~
and ~ot. t).c tr.sfu.-nmtion of
o,-di,.a,.ycncrgy into ),~
~n~hmn,
Oi such tran.sforn~tion (nuc.t..atio.).s tako no accuunt, uniuss
spcc.at tcnns bc imroduccd for t)u- purposo of
roprcscnLi, t].
cfucts ofvi.cosity, and of t!,e cond~ctb.i and
r;u)iatiouof hcat.
Cl.

In ~rcvicus chaptor (
278) wc .s.w).owL. cxp.-o.ssth motion
<'n thc r.~t ,f Lh. i.~nitc
H.ngc (1~. (n). in tenns oi-ti.c ~rn.d
Ycjoetty of the <huJ over tlie di.sc
We foun.), 278 C:~

~cre~iHp)'op(H-tion!(I<oc'
If r hc t),o distant bctween
any two p.,ints of thc .)i.sc, ~<. is
.s'n-tU qna~.ty, an<}6--=1
appruxi.n.~civ.

T~. nrst term <)op.n.).s

npon thc <!istribut:on of th c.n-cnt. Jf
wc suppose th.~
is constant, wc obtain
u]ti,natc!y terni rcprc~nting anincrcasc <~f I.c.-tia, or a correction to'thc
Icrgth.
oqua! to
Thi.s ~-c j,avc
n)rc.~)y considorcd, undcr t],c
s"p!.os,tion of pi.f, ,t
r~~
'"s.s.pat.on ~.pcnd.s,i.s ind.-pc.ndcntof thc distributiun
of current,
')~_2

180

t':nt.

.RATH 0F DJSSl'ATrON.

hc'in~ a fonction of<.hctct,a) cxn-cnL (.Y)nt))y.

at-(.t.'ut.i()H~)<.)n.st.t.')')n,W('i)ave

Coxfitiin~our

Th ('f))')'(;sp"n()Ii)~work ()om; (hn-ingit.LriULsj'ur of ()))!<)~Vis

fm~ ssincc,
ftO-t.hc
iacu, as
(J7I~ ~V;
j'. ~Lllll
30-t, th
;~s It)
in :31)~
tlie cxprc'sston.s
exlressionsfnr
t~e Imtent,i;~l
potentat
nn(ikinetict;)iC)'gics:n'c

in phtcc of (3) 30~. In thc valuntiun of c nu aDowancc nut.st hc

i))C-t!)(i(~fu)'t))uin(.ti:).(~t't,h(' <!ni(tont))cright-I):U)(I
sittoof-~t,
<,otim tcr)n oxtittcd m th~ expresMon fur ~).
<'o)'rc.s])f)))(!in~
-K<)na<,io)i(;'))is of thc stamLtntfoDnfurt.hcfrccvihmttons
f)f' (1i.ssip!tti\'c sy.st~ms ofnnu dc~rcc afi't-cc~om ( 't5). T~e
M'
itithu r~tic e:l 1
:m')phtm!cv:uiL'sa.sc't'l)~i))g(Hmmished
af'tcr:). tnoc&quatto

Ift]tcpit.L-h((]ctcrminc<!

by?<)bc

~ivcn, )))(' vihratiotLS hve thc ~re:ttcs<,p<')'.si.s<onccwhc)i c in

mn!~fL'st.,<.h!).t.i.s,w))L')tLhc )tcck isjnoutcontractL'tt.
If )S't'c ~i\'ct), wc!i:tV(.tun subsLitnting fur c its vahtc in to'ms
ui'~and;

shc\vi))gt))a[.tm(L'rthnnccircuni.st.;uiCt'Ht.])U(h[['~tiu))('fL))L')t)<~ion
it)C)'tsc's)':q)i'Hy:t.s/t(HtninihihcH.
Jnthcc:).sc()fsin)i!art'e.sot)at('r.sex7<):UKnLcn

forceiscfttcu'K.)Uftti(.n(.)it<<~)]ynpprf.xin)tt~nrLsmuc]tnHthodissi[)ati\-c
Ltt('dont)jt;H))p])(~iti~nt)t)ttthavi))rati(~)isj)ermnuentiLntthiswi)Ucndtono
!n)ttcritdo'rur\lK'nt.I)()
dissipationiiiHmftU.

3U.]

NUMERICAL
EXAMPLE.

181

which shows that lu fhis case t)te samc

proportional loss ot'
n!np)it.udc atways oecut-H :tftu)-thc )apse uf thc satnc numbcr of
]'criuds. T))i.s rcsu)L may bc obtidncd Ly thc mcthu<t of d!tncnsions, as n eonscqncncc of thc priocipic of dyn:unica.)
.si))ii):).)-ity.
As an ex:unj)) of (.), I may rdc!- to thc c~G of a
g!ubc wit)t
ncck, mtu))(]c<) fur hurtti))~ pho.sphorus itt oxygol gas, wltosu
capaeity is -251 cuhic fuct. It wa.s f..u,x! by (-xpcri.ncnt t).at thu
note of tnaxhnu))). rusntianGu )nath 120 vibrations
pur sucom),
so that 7t=12()x27r.
Taldn~ thc vch.city ofsuuud (<;) at )i.20
f(.'ut pur sccon!, \Ye mn) fron th(.'sc(]ata

Jm)~i)ig from tllC Sound pro.htccd wLcn tlie g~bo is struck,

] t))ink t))at this cstimn.tc m)t.sf, bc too
!ow; but it .shoutd bu
ohs.j)-vc<)U): tho ab~ncu of t.hu iatinite
fhmgc itssorncd ia t,)tc
th~ury nmst infjucncc vcry njaLcriaUy thc rato of dissipation.
Wc will oow ~xanur~e tlie f~rced vibrations duc to n,
source
of somu) externat tu tb<; t-eson~tn)-. If the
pressure 8;) at thc
jnout)) of )))(' r(;sQi)!t<,orduc t,) tlie source, i.c. c;t.!cuktc<I on
th
supposition t,)):).tthe mouth is cioscd, hc 7''c""<, t)ic quation of
motion corrcspon<)it)~ to (: but ~pplicabic to thc forcod vibra.tion only, is

w))!ch agrccs w:(.]i thc cqo~tinn ob~inc<!

by Hutm])o!f,zfor tho
case whcru thu commnnicatiott with thu cxtGr!)fitfur i.s
hy n.
simple !ipcrLuro(S()(;). ThcprL-sc)itproh)cm i.sne:u')y,butnu<,

1S~

r<JH(JKD \')I!i{.ATIO\.S.

j~L).

(putL', :). c~so uf t.])fit tru~tcd iji .K!, t)tu diffrence dcpcndin~
upou thc fact that, i))(i c')'-)H<io))L(.)t'()is.sip;).tiu))in (7)i;!it.sr!f r
:Lfu'ict.~)t~ft))Lipc)'in(f,a)Hl!)ut!t.n:L))S()!ut.(.')ycnn.-it.:u)t([um)tit,y.
tf'tiu' p'')-h)L),()(!t('nttit)t'dby /c,and~'hn
~i\'(-n,('))s))c\v.sU)at
th<)in~Tii:([\'n.ri:diunofp~.s.sur<~(<)!s!Lnt~in])nnw!)cuc=/<
d)aL i.s,\vh<'n t))un:)t.m-;d n"tc()f't.))uro.son:tt,~(f~1cu~(:(;<)wit)t<.)nLfU)u~it!ci()t-tIi.s)j)at.iun)i.st)t(;s:))))L':)st)t!~ot'thu"-('t)c!).tu)"'
.sutm'). Tt'c maximum
\i))r:)ti()!),A\'))()nth)ic<'inc)()('))cu()t'p(,'['if)ds
i.sjx.-r~ct, varias invt.r.~ty as <S; hnt,if'A'bcHm:L]),!tvcrys)i~ht,
int'')')atityi)tL))c
pcn~i.si.s.s)tf!!ci(..nt tn cause a )narl<L'(tfa)nn~
ufl'inth~
mtcxsity uf'ti)u~so)):Lnce
(-)'!)'). h) t))<!p)'~cticat
us<-<d' rcsonat~r.s
iti.sn()t.at!v:utta~'uustocrn-)yt))C!'cdnct.[u~
t)hS':n)(t<cryfar,pr''ba))iyh('c:Lu.st!thc;u-ra))~(;)nc))~)](TL'ss!u'y
t'orronm'cti))~ thc inttior w:~ t!)cci)r oroLhtjr.~nsidv'ia.)))):n':tt.usi))\-uh'ca.t]rj)a)'t)n-cfn~n
tttDHUppo.sit.ixnso)) w)nc))thn
cahi)hth()))s :).)'(-fu)))h)(;(],w)Nchb()t-())nL'sn~)ntant) !n<)rc
impurtanL
:L.sti~c()inte))si.))ts:u-t' rr()uc~). W)~at!m
scusitivc nppfLmt.n.s
i.s itct.ineotmuuti~n~ith
Lhemtc-rior.it.sInthccxpcmnL-nLut'
r(;i)tihrci!)~thc n<)un(i()t'~tUHin~-i'nrkhyn)c:U).s()f!L
rsonant)'
(-t))ur ck'ntoits Gnt.cr Intu thc (tuuiiLiuu, :md adist-inct
mvcsti~Uon
i.s nt'ccs.Siuy (;!)')).
)n\irLnuuft)tupt'it)cip!u()frc(;]p)-()cityt.cinvc-sti~th)nofthn
thu ci)'cctof:L
prec<din~pamgr:)p)t )nnyb(;a])pijc-dt<jc.d(;ut:ttc
soun'uof sound mt.uat.t'd in t))cintu)-ior(!t'iL)-('.so)i!tt.or.
3]~. 'now
p:)s.son tu thci'm-thcr discussion ()ft]tcpnd)]rtn
\Vu sl)!))[m)pj)f)sc t))!t.t t.hc opcncndofthc
ofthu.~pL-upipt'.
])ipc ispr\'i(K'd wiL)):m intinit.c ();u)~c,:t))d th:).t.its dmmc~r
i.s sm;dt in C(j)np:u-is(j)~vit.)t t))u
\v;Lvcie)~t.)i ot'thu vibration
um)o'cot).sidr:Ltit)n.
As :m introduction tu t))C())K.stion,wc
winftn-thct-suppnst;
thatthemoutiiofthc
pipe i.s htt.<-()with~fr<dyi)h)ving
pi.st.un
wiLhuut thic)<n(.'ss inu) )n:t.ss. Th~ p)'ccc<)in~
pn-)!)tons, froin
w!)if)) thL'pt-L~ntdiH'cr.s
i"rc:)titybutiitt)u,)):LY<j:d)'<u!ygiv(;n
usn~son tuthink
t))at Du; j'n-s~t)C(i.)t't))(j piston wiH~u)s.;
nuitopurtitttt )nudinc;((ion. Witiii)ith~t)))jc~csup])u.sc(2.5.'))
t)t:t.tthL;VL')u(.'i(y-pot.u))ti.di.s

oi'E~ rjpE.

31~.]1

On th right of thu pistou th rchition bctwccn

t on~i
js by 302

183
and

~)
\n

v
buing thc radius of th pipe. Froni this t)t solution of thc
]"'oLh'm tnn,y hc obtrunod without :H)y restriction as to thc
.smuHness ui'
sinco, Imwcvur, it is ody whcn /c~ is smid)
tf~.Lt t))Li prsence uf th piston wan)d nut
nt!).tcri:d)y mudify
thc ([estion, wu nmy as weU )):Lvethc hoiefit of thc
sitnpiification
aL uncc by taking as in (1) 3)1

Nuw, sinc:e the

N~w,sincc
tlie vaincs
valllos of
pistoll occllpil's no RpaCl',the
thcpist.onoccuntcsnospa.,
ot'

Ol.c>
\.c/ '1
)nt).stbo<))Cs:u)tC()nbothsi()(;i()fit,n.n<lsinect)[C)-cisnont:Ms,
thc Itku must bu truc oi't)tc values ofj~f~o-. Thu.s

In thisexprcssum tlie te-rmcuntfunin~Hin~ dpends upouthc

dissipation, :ut(!is th sunic as if thcrc wcre no piston, whitc tha.t,
`~"Il
cfti:ct uf
cft'L-ct
f']' t)tc inertie of
f. th
involviu~
1.
1 cxtcnmi
caterual1
1
rcprcsent.sttto
air in thc ncigtdjuurhoodof thc mouth. In ordcr to comp:u-cwith
pruvious rcsult.s, !ct a be sucti th~t
~7.'

184

THEORY
0FOPEN
ENDS.

[312.

Thse formu~ -showthat, if thc

dissipation be ]oft out of account
thc vcioeity-potejttia) i.s thc sa.nc M if t)iu tube
v/cre Jun.~hcncd
of thu radins, and th opcn end tlicii bchaved as a
by
loop.
Tho amount of thc cnrrcetlun
agres WLthwhat prev:ous investissons wouM Ii:LveJed us to cxpect as H.c rc.sutt of thc Introduction of thc pisto!). Wc i.avc sccM rcason tu know that
t!.u
true val.ic of a Hcs
and
~nd U.~t thc prescn
bctwccn
~7~,
of t!~Gpiston dous not aHcct t)to tenu
rcpresonting thu dissipation.
But, b~forc discussing our rcsults, it will bc
advanta~cous tu ii]VL-st.~tc thon afrush by a rathcr dincrcnt lucthod, which Lcsides
bcm~ of somcwhat grcater gcncndity, will hdp to tlu-owligtit ou
thc tncc!):u)icsofthc<)ucstio)).
313. Fur tins purposc it ~iit bu convcniont to tiltift thc
ori~ta
in thc ngative direction to such distance from t)ie
jnonth t!at,
th wavcs arc thcrc ~pproximatcly
plane, a disp~eonent which
aecording to our suppositions nccd not :unount to more than a
sm~Haction of thc wave-Icngth. Thc
difHctdtyof thc question
consists in finding thc connecta betwccu th wavcs in thc
pipu,
whic-hat n sunicicnt dist:uice from tt~ mouth are
p)ane, and thu
divcrgin~ wavc.s ontsido, wi.Ich at a modumtc distance may bo
treatcd as sphcrica). If t!)c transition tako
ptacc within spacc
.sinaH comparcd witb thc
wavc-Ic~th, whicit it must evidcntiy
do,tt tho dmn)ctorbc.s.na)Icnough, th prubtumadn.itsof
solution,
wh~tcvci- .nay bu tiic for)n of thc
pipe in th ncighbourhood uf
thu niouti).

.Lpuint,7~,wiiosc distnncc from ~1is

t)iu l'elucitj,n.~k-mtc, tliu
vducit.yput,cu(.j:dis(27U)

3J3.]

TMEORY0F OFEN ENDS.

185

Lcb us considur thc bch:n'iour of t.hc nmss of air Inchu~d buC:md n,hcnusp))<ric:dmtr{~ccw!tosu
twnthcp~ncHecL~H~t
centre is~l,iuid
radius ?',)'bcin~I:n-g-tj in conipari.so)iwit.)t t)tu
d):U)tct.L-rot'tho pipc,b)tt..sma)linc(')))p:).risonwith<L)i<j
\t\'L')';))~tL. Wit)tD) tins sp:L(:ct))u air nu~tniuveitpproxitni~cty~s
:).n.
incun)}))-c.ssib)cOuid -\vuuiddu. Nuw t.hc uun'cnt ncru.s.sthc hcmisphcric:dsm'i'n.c<j

This is tlic first condition; tlie second is to bc fuuud from t))e

cunsidorattonthat thc total currcnt (wliose two v;ducs hn.vcjust
buen c(tun.ted)is proportionfdto thu di~crcncc of p(jtcuti:d at tiiu
tun)H!t:).!H.Thus, if c dcuotc th conductivity of thc pussa~c hct\vccn ttiu tertninal surfaces,

fn tins expression thc sccotidtcrm is ncg]i~!))]cin compariso)~with

thu first, i'ur c is at niu.st. 'tunat.ity ot' t!~ s:unc ordor as thc radius

18G

CORRECTION

TO LHNCTH.

[313.

If/t! bc thu radius uf t,hc tube, wc

ni~y rcp!:LCCo- by 7r7~.
Whoi tlie tube is a. simple cylindo-, and thc
origin !ic.s at :t
di.st.utcc A7, iruta Litu rnouLh.wuknow that
who'u
o-c''=AZ,+~,
i.s :).n)[)n))o- i-at.!tcrgrc.itcr thau 7r. In such a case
(thc oi-i~in
t:d<~ti
Luing
sxtHcicuDy uc:u- titu inouth) ycKis :). mnaU qu:mt~y,
tmd titL-rcturc fron. (10)

At tllc .~m~ Litnc cu.s~ nmy bu iduntiticd whh

unity.
Th~ principe) tcnn itL
~invotvu~ co.s?~, )n~y t.)ten heealotif t.)iutuhu wct-c prolon~cd, andi))ct-c wcru :t,
latud,
luop :tt ,).
i~mL.situat.udaL:n)Istancc~ huyont the actual posidun of ti.u
hiunL)j,iu nccord:Uicuwitit w).at wc t'ound bufurL-. Th(jsu rusult.
appruxitnatuiat-urdimn-ytubes, bucomu ri~-uruusw)iun thu diatnutut'
i.-jrud~ced \vit.Luutiimit, fi-ict.Ioii
bcin~ nc~cctud.

3)3.')

RATU 0F

DISSIPATION.

187

If tho-c Le no n:U)gcnt J, the value of c is

s)ight)y modined
bythu)\'niova)ofwhat
actsasan
obstruction, but th principal
cH~ctis o)) thc tcrtn rL'prc-sontingtbc
dissipation. Ifwc snpposfa~ an approxitnation t)iat thuwavcs
divo-gingfrom~ arc sphcricai,
wc
~veiiiiist
nmst titl~e
takc 1(irtite
fur t]te ettri-ciit
current 4-n-r

instcad C)r
iIlSte~t(I
of
Thu0
f/
~Trr'~~y-'
u!ti<nate<t.ctoft))<t-at,i<jn wiH )j2to hnjvc thc cxprcsHionfor
H'c V(.-)<)city-p()tcntifLt
uutsi()(; tho muuth, tis wcU ILSthc corrcspun<)ing second tunn in
(invoivi))~.sinM~). T)tu :uuount of
is t!ms .scctLtodupcnd ))):ttcri:).t)y
')tMs)]):Lti<.))jL
ontitcdcgrecitt
which t)m \v:Lvc's
arn fruc t<j divo-gc, atK)onr a)i:L)yttc~cxprf.'s.~iuus
must nut bu r~mk'd :L.stnurcthan run~h c.sthnit.t~.
Tiiu c~rrt.'ctthcory <'f thc open org.'ni-pipc,
including cqu~tioxs
(H) :nid (i2), wfm di.scovcrcd hy Hut)n)toltx', w)K)SG
method,
f~nvt-vur, di'rct-.s cot)side-:).b)yfrom t)mt hcru adoptc'd. Th~
c:n'!ic.st suintions uf thc probicni by L~t-fin~L-,]). BcruunU), and
Etdc-r, weru foundcd on tho a,sst))nptio)ithat~ta~
opcn end
thc pt-c.s.sut'o
cou)d not,vary ft-om thi).t of tho ~urrutmding atmoMphcru,a. principic w))ich )nay pcr!)aps cvu)i now be consi()crc()
apphcahic to an ond wliosc opcnm-ss is ideaUy pcrfcct. TIiu tact
ttiat iu au ordiaary casus cncr~y cscapus is a, proof that tttcrc is
nut anywfturuin thc pipe au absolute
)oop,nnd it might hve bccn
('xpectud t))at the ino-ti~ oi'thc air just ontside the tnouth would
ha.vu the eUccIof an iocreasGin thL: Iu)gth.
Thc positions of tho
nudcs in a soun()in~ pipe \cre
invc.sti~at(;()cxpcrimcntauy hy
!S<).vai-t"andI[o])!dns",wit)tthc r~uit that thc intervai bctwecn
thu mot))a)td thu ncarust
noduisa.I\vays)c.sstitim t)tu h:dfof that
s~pa.t'atingconsccutivcnod~'s.
31~ Expo-imcnta) d~'tDi-nnnationsof t!.c correction for an
opun on) )iavc gcn(.'ra))ybeun )na.dcwittiout t)ic usu of a riangc,
a)))) it t)tL'ruforcbcconics itnpurtant tofortna.tanyra.tGarough
u.stitnatcofitsctrect. No ttK'orctiea.tsotution ofthuproDt~nof
:m unOangt'd opcn cn<)))as )nthurto hccn givc'n, but it is
casy to
scu t)i:tt t))o rcmovalof tho ffangc will ruducu thc correction
tn:Ltcria)!yhc'!uwthcv:dnc -~i27t' (Appcndix A). In the abscncu
ul' tticory I hve attcmptcd to dutcrtninc Hic innuoice of a
naxgu
'Cn'))t',]3~7,r.l.

18(!t).
~HcchL'r(.-)tL'isnr)MYibn)ti<'))St))jt')ti)'t~f/o'xt.t.xxn'.l.S'

-'Acnftt
)')1.

vihrittiuns
1~

incyliudncnit.ubt;

c'ftHt&ro/~f

j~Y~

~83

INFLUENCE

0F

FLAN(!E.
[314.
L.

cxpcnmcnta.Hy'. Two organ-pipcs ncariyenougb in unison.with

onuanot!)L-r to givc countablobcatswerc
biownfrom an or'~an
bdiows; tho cifect of t)~ nangc was (icdnccd fro!n titc dift'cr~cc
in thc frcqucucics of t).c be;Lts
according as ono of tho pipes was
nitugcd or not. Ti)c correction dnc tu th Oangc was aboit -2/t'.
A (prohaDy more tmsLw.rttfy)
~}K-t,itio)i of' this cxpuritnt.nt ).y
Air Jj<.s:uu)uct ~:Lvc -2.')~. Jf wc Huhtt-act -2~7t; frntn
-S27~ wu
ct'tain -8/
w''ic)tninybcrc-g:u~(~asabntttt)hJprubah!(-v;dnL-of
thu conwt.ion f<m
unft!.)~) npt.ncn.), ..))t])e.s))pp)siLi.)t).at
iti cunip;u'i.-iunwi(.h thc diani~Lcr <jf thc
thcwavc-IcngtitisgrcaL
pipc.
Attcmpts t~ (~to-muic th cnrrcction cntirdy from cxpo-inx-.nt
Lavu not )cd !.iL))crto to vcry prccisc rc.sutt.s.
ALua.suronent.s Ly
Wurthcita' on doubfy opoa ptpcs
gave as a. muan (fur cach en.)')
winiu ~r pip(j.s opcu at o).c end
-~i:
onjy t))c muan resutt was
-7-tUA'. In two carctut uxpoi.ncnt.s
Ly Bo.sa!X)u~' on duubiy
('pt'n pipes thc correction fur onc end was -C~7~, whcn \= 12
7/
an.) -5.):i w),en =:~U.
Bosan.,nct Jay.s it duwn as a gcn~-at
i-)t)c tLatt))ccorr~iun
i.t'crcn.scs
(uxpr~.dasafr;Mtiunof7.')
wit]i t).c ratio of diatnctcr tu
wave-fcngth; part of this I.icrpa.sc
inay Ilowuvur bc d~c to thu .nutunt rcaction of thc- ends, w].ich
causcs thc p!anc of
.symmetry to behave likc a rigid wa)!. Whc.n
t).c pipe is oniy
modcratdy long in proportioli to its dia.netcr, a
statc of things is appro;Lchcd w!ac)i
rnay bc more ncarly rcproHcntc-dby tho prc.scnec t!)a)i
by thc absence ofa ftangc. TIjc compan.sun of Hicory and obscrvatiort on this .subjcct is n mattcr of
-soinc dimctdty, because whcn tlic correction is
sma)), its va)u~ as
calcutatud frum observation, i.s aH'uctud
by uncertaintics as'to
ab.s.dntc pitch and t).c vclocity of sound, wtnie for thc
ca..so,wbcn
t))C correction is relativdy iarger, w!tic)t
expcrimcnt is more co.npctctit to duat wit)), titcre is at prsent uo thcory. rrobab)y a. more
accumte v~luc of th correction cou!d be obtaincd from a
re.sonator
of tbc kind considcrcd m 3<)(i,wh~-ro thc
communication witb
t)ic outstdu air is by a simple
aperture; th "k-ngth" is in tbat
case ~ro,and thc correction is
cverytiung. Somc mca.suroncnt.s
cfthis hmd, in whic]), ))n~-cvc.r, no
grr..at accurapy was attemptcd,
will bu fonnditimytncmuir
on rsonance'
'r/;<7..v~(.))!iS(;.
]S77.
=~t;t).(;/NM.(:t)t.x.\xt.p.;);)~
~~t/1A;f/.(n))v.p.~)u.
is77.
''7'/<t/.7'nu<.<.lH7].
~uutcrcjuarki,tiiermjx~ S~a]sn.S<~t])i)tuss,r,)~t.ltn,21U(lH7<'),(ind[
],ymyself(/u,
Hrj.t.1870).
f~vnr't'imfntnUt)'

~t\.r~

314.]

EXPERIMENTAL
METHODE.

183

Varions mcthods hve bccnu.scdto

dtermine thc pitc!)nf
re.sooators cx])cri)n~i)ta1)y. Most frc;()ncnt)y, perhaps, tl)c rcsonators
havu b'jcn madu t.o N/)e~' after t)ic manncr of organ-pipcs by a
Kt,t'(?:)ni.
ofnir bh)\vn obhfjnc'Jy across thci)' months. AIthoufh gnod
rL's~it.s hve bcun obtaincd lu this way, onr ignorance as to tho
modt~ of action oFthc wiod rc:)](tL')'n
Lho mbthod uns~tisfitetory. In
Busnoquct's Mxpt'rimeiits thc pipes wcrc not acLu~Hy mn.dc to
spc:).k, but sitort discontinuons jets of air wurc btowu a.cross t])e
~))t;n end, thc piteh bcing usti)nntL-d frum thc frcc vibrations as
thc sound di(.-d awa.y. A )n(.'thud,simi!at'in principte, that 1 Iiave
.sonn;titncs cmpioycd wit!t a.dvn.tita~'cconsists I)i uxcitingfroei'vihra<io)tShy !ncans uf a Uuw. In order to obtain as we)] dufincd n. note
as pnssibic, it, is of importance te accomtncdate t))c hardnc.ss of th
substance with w))ich thc rcsonator coincs into contact to th
pitch,
a loAvpitch rcqninng a soft, b)ow. Thus thc pitch nf a tcst-tnbc
may bG dutcrmincd in a mumoit by striking it against t)~o bout
!\u(;c.
I)i nsin~ this mcthod \vc onght not cntirdy to ovorloo~ tho
fact that thc natnra) pitch of a vibrating hody is attcrcd
hy a
<(iDn dpend Ing npon t))C sqnarn of thc dissipation. With thc
nntfLtion of 45, tho frcqucncy is diminishcJ
from ?t to
or if A' bc thc nurnbcr of vibrations t'xccntcd whi!e
?'() c~r''),
tt)C amplitude faHs m thc ratio e 1, from M to

Thc correction,
itccnuut:.

howcvcr, -\von)d

mrc]y be wortit

tahin"'

illto

Th mc.isuroncnt.s givoi m jnytncmoh'nn

rc.sDna.nce wcrc
co)uh)ct.f;<l npotin.dii'f'L'rt-'nt, princi~tchyc'stimii.t.n)~ the note of
tnft.xImnniLrt.'son.'utcc'. TItcearwaspLLCctIincommonicn.tinnwith
thL'[)jteri()rofthccaYit.y,Y))i)ct])ccI)ron)!'ttie8c:t.)cw~.ssoun()c!d.
tu ~)us way it was found pns.stbtc with !i HtUe practicc to o.stiinn.t.e
t)K' pitcl) of a gnod rcsor):Ltf)rto about tt qufu'tcr ofa scnnt.onc. In
the citsc of'.sman n~s~s with Jong noc~.s,to wL!c')) thc n.bovc mct,hod
W()ut<]not bu ap])!ic:d))c, it wfts funod sumocub nu'rcly to ho)d thc
0;
nea.r thc vibmting wires of n. pianofoj'te. Tito resonant note
:)~))~o~)ncu(~
itsutfby a.'nuvu)']))~ ofthn body ofthc f)as]e, e:~i]y npr<'cptil)tu hy titcnngcrs.
]nunH)~t)ns)nc't))odit,i.si))ip(.)rtantt)):t.t
thcnittd
.s])ou]d bcfrccfrombiasin
snb-dividing' the intcrv:).!
bchvcoi two conscutive .sonitone.s. Whcn thc thcorctic~t rcsu)t

~0

DISCUSSION
OPMOTION

[314.

i.s known, it is atmost, impossible to an'ivc at an

Indcpcndcnt
opinion hv'xp('itiir'nt.
31.'). Wcwi)) no\v, fu])uwingHu]mhohx,examine more
c!ose)y
~hc nature of thc motion within t)ic pipe, rcprcscntcd
by thc
Wchave
funnn~(H)3)3.

whcrc M is a. posttivc jntcgcr.

T!)C(1ist:mcchctwccu conscutive m~xim~is thns
a~)(~thc
v:duc cf thc m~xnnu))!is sec~a. Thc miniinnmva)uu.sci' Z/' occur
npproxitnatcfy wlieu (a; a) = M<7r,

ORIGJNAT1NG

315.]

WTHIN

AN OPHN l'IPE.

191
t

Th fipproxin~te tuagnitudc of tLc maximum is ~sec~x, a.nd

<hnt ofthe mminuuu A:o-coH'47r".
It appcars that ti)C
tnaxima.of velocity oceur in the s:unc parts of thc tuhe as thc
nnnimit of condcnsit.tio)!(:md rtn'cfact.ion),and the nuninm of
yclocity in thc samc places as <.I)enmxi)))~ofcondun.sfitiot). Thc
scries ofloops nnd nodcs:u-G:uTa.)tgc() if thc in-st
loop werc at a.
(listancea bcyond th month.
WIH) rcg'fn-din th phnscs, wsec t!~t bnth
and arc in
~encrt s)naH and thcreftjre with t1)Cexception of thc ph~ccs
whcre7,' and ,7' arc ])ofn' tileir rninima. the w]io!c motion is
.synchronous,as if there werc no dissipation.
Hit!)crto we hve considercd thc prohtcm of th(; passade of
p):ui0 wavcs fdong thc pipe and t)~Ir gradu:d dinnslon frojn tho
month, -\Ylthcnt regard to thc ori~in of th plane -avcs thcmsctvcs. A)l tliat wc hve assumcd is that t1ic origi)t ofthc motion
is somowhcrewithin thc pipe. \Vc will now suppose that tiic
mution is duc to the known vibration of a. piston, situated
at A'=- tlie origin of co-ordinaLcsbcin~ at thc mot]!. Thus,
whcn

==

n.nd tins imist hc !n:~c to corj'cspom)with titc expression for thc

p):inew!).vcs,~cncmlixc<)
bythc it)tro(h)ctionofarbitniry ampjitude
:uidp)ia.sc.
AVcmayta~c

by which tind e arc detcnnincd.

In :tccoi-d!uiccwiL)i (12) 313, th corrcsponding divergent
wa.vcis rcprcsc'iited by
Tt

1 QO
Il.

MOTIONDUE
AIOU'IO.
DUE TO
Iv W

f;~

Jf<?)M givcn,
is
~.en.

,s ~.catmt,
whcneo.,K(<+.).n
In

t)~t h

of t).c ~t.i~
.Ibr~i.n
v.y
71
.t.,
n.,
th.n~h
n~mtc. sincc cos~can~t~is),.
Wi.cn
hun~uU..much
cont.rac~, oos.y
bcco..c .sn..)., h.tt
in t!s c~
is ncccssa.y that t)~
a.)ju.st.,ncnt of ncrio ho
c.
in
onicr thut thc Hr.st te. cf
very
( i'.) ,n.y b. n~
le
tlie
:~r:
.;ecolH1.
CC)SAC2
i. lie.-ll-]Y
~p~c.i: il
l'quaI to unity.
Tho ininhnum of vibration occur.s
whcn
snch t)~t
tLc pi.~n is ~cd
at r~p
tlmt Case

T.cv.br.ion
out.si.jc tbc tubei.s tben,
accorda te tbo value of
c.,ua)to.r.s~a]fc,than th vibrion ~.ich there wouU
be
't!~
"S
P~~
tlie
hlane,
316. Onr cqu~ions
may ..J.sobc ~ppUedto the investigation
.f tl.c motion cxc~cd in
tube by cxtcrna! sources of~nnd
Let uyuppo.sc in the first
place tliat tho ~out). of tbc tube i.
c).soJ by a ~.cd pL.tc fo~ing part of the y. p]ane, .nd tl.at
the
<h.c to t).c cxt.n~I .sources
(approximatoly constat
p~cnha]
ovur
t!.c plate) .s undcr tbc.sc c;rcum.sta)icc.~

~-he.-c

Is composcd of thc
potcntini due to each .source and its
.mngc .n th
pi~no, as c.xpt.h~d in 27H. Inside t!~ tube lut
t)~ potcutiat be

so that <~and ils difrcrcntia! c~mciont are eoutinuons

ae.-osstho
).:UT.cr. Th<y.sical
nK~in~ofthi.s
Is.simp)~. Wci.n~me
~OnntLctuhc suc), a ~r-t.ion as is extermine.)
hy th conditions
ti.atthcvdoc.ty at thc n~nth is zro, and that thc condensation
at thc moutli is th same as that duc to thc sources ~f
sound who)i
thc.nouth inc~sed. h. i.sobvions that
undcrthc.scn.-f-u.nstancos

~nc.j

EXTERNALSOURCES.

193

t)ic elosing plate may be rcmovcd withont any altration in th

motion. Now, ijowcvcr, tficre is in gnral a finite
velocity at
.T=- and Htcrefure \vc cannnt suppose the pipe to be t!icrc
stop])cd. But \vhen therc liappois to bc a nodo at a; == that is
t" say wiicn is Huefit))at eus/<-(/+a) =0, a)l thc contiitions :u-e
sattsficd, a))(t t)tc actual mution withiu thc pipe is that cxprcsscd
~y (2). T)iis tnot.ionis cvi'K'ntiythc same as might obLain,iftttc
p)pc wcrc c]osQ(lat LoUtends; atn! in cxtcrnal spacu thc potcutiat
is thc sa.mc~s if thc mout-huf th pipe wcro ctusud wit,]itho
ri~id
p)n.te.
lu tljc gcncra! case in ordcr to rcftuccthc air at =
to rest
we must superpose on tho motion rcprcsentcd
hy (2) another of
tho kind invcstigatcd in 313, so dctcnnincd as to
givc a.t a; = i
a vclocityequat and opposite to that of th first.
Thus, if th
second motion bc ~ivcn by

It nppcars, as might ]~vc bccn

expcctcd, tha.t tito rsonance is
grc~tcst wiieu thc t-cduccdlngth is an odd multiple of
317. From thc principtc that in t~to neighbourhood of a node
thc mortm of thu air docs not conic much into
play, wc sce that
in snc)i p]accs thc form of a,tube is of little
conscquencc, and that
cnly th capacity need be attcndcd to. T))is considration ~lows
us tu calcu~tc tlic pitch of a
pipe which is cylindrica.1througb
'nost of its leng-th (~ bttt ncfn- the closcd end
cxpands into a
L"]b of small c~paeity (~. Thc rcducc<t
Ict)gth is thcn cviJcut]y

'Hd)n))ott~,C)-18<!0.

R. H.

I!)4

EMARGEMENT

AT A CLOSED END.

1317.

whero a is thc correction fur thu op(;n (-n<],nn<) o- is thc Mrcfi of

the tr:).nsve)'.se s(-c),io)) of thc cy)in!))'ic:d piu't. 'i'his fm'muta is
-c).u~L~~y!ph~i~h~~ih~dc~aU~uf~~)~~
cy)[n(h-ic;d f)-)u <)ucsnot take th .shapc ot'on enitu-~emunt.
AVhGnthc oikrgGmcnt rcprcsunted by <S'is too )a)'~o to a)]o\v
of tlic abovu trcatment, wc )n~y procccd as fuitows. T))c
dissipation buing ucglectc~, Die vuincity potential lu the tube
!n:).y bc
takcn tu bo

is the cqufttion dctermining thc pitch. Numcrical

cxamphs nf
thc application of (3) are given in
jny mcmoir on rsonance
(P/i~. 7'm7i.9.1871, p. H 7).
Simil:u- rca.sonit)~ provcs tl~t iti any cn.se of
sta.tion!i)-y vibrations, for ~'hich tlie w~vc-Ien~t!) is Hcvcra.t thncs as grt as the
duunctor of t)ic bulb, th end of t)ic tube
adjnining thc I)u)b
bchitvcs approxitna.tcty as an opcn end if ~,9 bc inuch
grcatcr
Huui o-, and :ts :i stnppud od if
Le mncb luss than o-.
3~8. Thc actio)) of a rcsonator whcn under thu innucncc of :t
source of sound in unisoi with itsctf is a
point uf considrable
duUcacy and i:npor~ncu, aud onc on w])ic)t t))crc ])as b(~n

318.J

ABSORPTION

0F SOUND 13Y HE.~ONATO!!S.

1S5

good dcat of confusion among acoustical writcrs, thc autbor not

excepted.
Titcrc are cases wbcrc a rcsonator absorb.s Sound, as it wcrc
attracting thc vibrations to itself aad so (tiverting them fron
rL'gions whero otherwisc thcy would bc fe!t. For cxampte,
suppose that thcre is a simple source of sound
sitnated in a
narrow tube at a distance
(or any odd muttip)c thercof) from a
cfnscd end, and not too ncar tho mouti): thoi at
any distant
oxicmal point
its cfuct is ni). This i.s an inuncdiatc consu'ptenco of thc principtc of recipt-ocity, bccanso if ~1 werc th
source, t])Ct'e cuuld bu no variation of potcutial at A Th
restriction, precludin~' too grcat a proxinuty to thc mouth, may
bc dispcnscd with, if wc suppose th source
to be din'used
unifornily (jvcr t))C cross section, instcad of conccntratcd in onc
point. Thcn, wltatcvcr may bc thc Hizcand shapc of thc section,
thcrc is nbsoiutuiy ne disturbanco on thc furthc-r si(h'. This is
c)car frum thu thuory of vibrations in onc ditm.-n.sion thc reciprocal form of th propositionthat wbatcvcr sources of' distmbancc ;nay cxi.st bcyond thc section, jy~-rZo= 0tuay bc
provcd
ft'om Hc))nh()ltx'sfurmuia (2) 2'): hy inking
for t]te vclocity
potcntia! of thc purcly axial vibration ofthc saine period.
It is scarcc)y ncccssary to say that, whcncver no
cnen'-y
is cinittcd, tho source does no work; and this
reqnircs, not
that thcre shaH bc no variation of pressure at th source, for that
in thc case of a simple source Is impossible, but that th variable
part of th pressure sbaU bave exact)y th phase of thc aeccicr!
tion, and no componcnt with tho phase ofthc vclocity.
Othcr cxanip!es uf th absorption of sonnd by resonators are
an'orded hy certain modincations of Hcrschers interfrence tube
uscd by Quinekc' to stop tones of de~nite pitcb froni reacitin'
tlie car.
In th combinations of pipes rcprcscntct) in Fig. (!3, th soun()
c'ntcrs frcciy at
at
it nnds itscif at th mouth of a resnnator of pitch identical with its own. Under thse circumstanccs
it is absorbcd, and thcre is no vibration
propagatcd a]on~ 7?~.
H is cJear that th cylindrica! tube ~C'
may bu rcp)aced by any
othur rcsonator of th same piteh (7), without pn~ndicc tu t)tu
'{'{;f~.cxx\'))i.!77.tn(;f!.

L;

l~G

QUTNCKE'S TUBES.

[31g.

1
action1 off'mthc
apparatu.s. Thc ordinary cxplanatiou by intcrfcrcnc
(so ea-Hcd)of direct t~udreffcctcttwavcs is thun luss applicablu.

Thcsc cases ~hcrc t]te source Is at thc mouth of a i-csonator

innst net bc confuscd with othcr.s wbcru thu source i.sin thu
intcnur. If
be :i source at t)i huLtotn uf a.
stuppcd tubn w])oso
rcductjd Icngth is
t])e iatcnsity at fm cxtet~id point
Lu vast)y grL-atcr t)i:ui if th~u h:td bccn ])o tube. lu factm~y
thc
potential ~t duc tu t)ie sour :Lt is thu s:unc as it would Le
t).t wct'c thu Muurecat J..
31~). For a do.soruxatninatiuti of thc mcc)h'micsor
wo shaH oblain thc proDcm iu a !'L.rnidisutnbarrasscdrsonance
of unue~
ccssary dif!icutt!cs by suppusing t)ic rcsuj.ator to consist of a
smati eircular ptatc, bae!.c-<Ihy a
spritig, and hnbeddcd in a!i
inde~nttc ri~id p)anc. It was provcd iu a
proviens chaptcr (:30)
tl.at if J/ bc th ma.s.suf t)tc
its displaccmc.t,
ptatu,
. titc furcc of
restitution, 7i' thc radius, aud o- thc dcnsity of thc
air, thc quation of vibration is

where Fand

arc propoi-HonaIto c"

If tl.c natund pcriod of vibration
(tlie rcaction of cxtcrnal air
indudcd) cojncide wit), t!.at i.nposed, t!.c cquatiou rcduces to

319.]

RKSONATOR CLOSE TO SOURCE.

197

Lct ua now suppose t]iat F is duo to a.)i externe source ot'

sound, giving whui Dm plate i.s rcst a poto)ti:d
which ~ill
bc nc~riy consta,!i.tovci-th fu'ca (jfthc plate. Thus

so that 2-n-A:i.s tfie w.ivc-icngth of thc )in.tum!noto of thc rcson~tor. If ~bu writtcli for ~/+!jo-7~ t.)ic cquation con'cspondin~ to (5) ta.kcstliu i'ortu

from winch wc may infer us bcforc that if /<=/<: thc

cfDcicncyof
thc rcsonator as a..sourceis iii<)cpe)x)cntof ./< W)tcn t!tc
ndjttstmcnt is impc'rCcct,tho ]a.w of f~Hing oil' deponis
upu)). J/
T))U.sifj/' be grcat fm<t
Tt;MmaH,
atthou~h thc mnximum cfHcicncy
of the rcso!]n.toris no le.ss, a, grc~tcr
ficcuracy of adjustnicnt is
rcquircd in ordcr to approac]) t)io maxunun) ( 4')). lu t!ic case
of rc.sonators wit)t suup)c npcrturcs J/'=
so t)iat ~)/
'ovaries as jf)' Accon]iii~)y j-esonators with s)n:LH
apcrtoi-c.src'p)irc thc grcatcst pt-ccision of ttUiing, but thcdi~-rcnco is not
t'nport.-mt. Front a comp:n-i.sonof Hic prcscnt investi~tiou with
th~t uf :ni it appcnrs tl~t thc conditions of
cOicifjncyarc difiuront a.ccordin~as inturnal or externat cffuctsa.ruconsidurcd.
Wowift no\v rut.m-nto the casu of isucfn-omsmfmd
suppose
f'n'ther tftat ttic extornal source of Sound to which thc rcsonator
yl rusponds, is thc motion of a simifar plate
w]jose distance
c from ~1 is a quantity I:u-c iti comparisou wilii thc dimcusious

198

REINFOnCEMENT
OT SOUND

[319.

of tlie phLtcs. T]tc

intensity of 7~ may be suppo.scdto be such
tim.titsnutc)iti:t.[is

Th relation of phases
)nay bc reprcs~ntcd hy rcgarding th
Jnduc~d vibration
as proce~hog frojn
by way of J, and as
bt-ing suhject to au additionat retaniation of
su that D.e whoie
retar(!atiun betwccn 7~ axd
is c +
In respect of amptitudc
13 ~i-catcr t))au
in thc ratio of 1 ~c.
Tf~ts when ~e is .sj))n!),H.c Induccd vibration
is much H]c
greater, nnd thc tuta] soun() is much hn~cr tnau if
wcrc nut
permittc.! to po-ate. lu this case thc phase is rct:u-dcd
by a
quarter of a pcriod.
It is important to hve a c~ar i<)e!t of th cause of
this
aug-mcntation of sount). I)t a prviens ch.-tptcr ( 2.SO) wc saw
that, ~heit ~1 is ~xed,
gives ont much ics.s sound tkui n.ight
at ilr.st hve been cxpcctcd from thc
pressure duvclopcd. Thc
expiu.nat.ou was tJ.at t))e
of thc prsure was
unfavourah)c
thc hn~cr pfu-t of it is conccrncd
ouiy in ovcrconung thc incrti:t.
of t]ic surruunding :ur, and Is incHectivc towards
thu perfonuanee
of work. Now thc pressure which sets
m inutlun Is t)~ who)e
pressure, aud uot inuruiy thc insignifiant part that would of itscif
do work. T)ic motiuu of
is duturnu.tcd by thc condition that
that cu.nponcnt of th whoto
prcs.surc upo.i it, which i.as t))c phase
of thc vulocity, shai) vanish. But uf' t!m
pressure that is due to
thc mution of~, thc largcr
part bas thc phase of thc acclration;
aud thcrcforc thc prcscri))ed conditiuu
requires a)i cfptaUt.y
hetwccn th stu:dl componoit of' thc
pressure duc to ~t's niotion,
~ud a pressure comparable with t)te
large conponeut of thc
pressure due to ~'s motion. Thc rL-.s~h is that .i becoines a
much tnorc powerfui source titan
Of course no work is donf
by the piston ~1 its efTect is to augmcut thu work dune at 7?,

319.]

BY REHONATOUS.

1~9

by modifymg' t))c othcrwisc unfavourabic relation betwccn t)to

phasesut' tbu pressure aud of t]<e vciocity.
Thc inHnitc ptane in thc preccding (hscussion is otdyrcquircd
m ordcr tfiat wu may nnd roon) bchind it fur oui' toachincry of
sprints. If -wc arc content with still norc hig)~)yidcabzcd
sources it.u<!t-~unatu)-s,wc ni~y dispoisc \viL)tit. Tu utn'h piston
niustl)ca(t<.)cd a. dupUcit.tc,vibrating Inas!n)i)ar tnfmncr, but in
thc cppo.sito<tir~ctiun, t))e ufluct oi' which A\'i)lbc to tttal tho
nonuat vulocity of th nuid vanish ovur t)tc pl:tnc JA Under
thusc circmn.st:utC(jstho plium is without in~ootcc fmd may Le
runovcd. If thu sizc of thu pintes bc rcduccd witLuutlimit tlicy
bcc'oxH-uh.imatcty ('(jnivajcuttoshnpie sourcps of <]uid; aud wo
co)ic)udctlif).ta simple source will Lccumc more efficientthan
bc'furcin thc ratio of 1 /cc, w))cn at a sma)l distiuico e from
It thuru is a!)uwcd tu oporatca. simple rcsonator (as wc
may cali
it) of I)kc pitch, t!)at is, a sourco in whicit t))c inci-tia of t))c
iinmcdiatuiy surroutiding Hnid is compcn.sated Ly sfxne adquate
imd which is set in motion by cxtcrna) causes
m:).chine)-y,
ordy.
In tllu prsent .statc of our knowtcdgc of thc incchanics of
vibrating Huids, whiic titc difHcultics of dduction arc for tho
mo.st part still to bu ovurcutne, any simplification of cotKJItionn
whie!)afiow.s progruss to bumade, wltitout
~Lo]]y dcstroying t!to
pr:).ctical charactur of thc qucstiuu, tnay bc a stcp of "-reat
importaocc. Sucit, for cxamptc-,was thc introductioti by Hchnhu[tz of th idca ofa source concuhtratcd in onc
point, rcpruscnted
anulyticaHy by thc violation at tliat point of th eqnatiou of
eoutinuity. Pcrhaps in Hkc manncr t))e i(!ea. of a sin)p)c rcsonator may bc uscfut, althou~h thc thin~ would bu stiil moru
impossible to construct than a simple source.
320. Wo havo sccn that tburc is a grca.t augmentation of
snund, when a suitabty tuned rcsonator is close to a sinip)o
source. Much more is this thc casu, whcu thc source of sound i~
cumpound. Tiio potential duc to a double source is ( 294~,324-)

200

RESONATOR

AND DOUBLE

SOURCE.

and thcrcforc thc potcutial duc to the resonator

at

[320
distance

is

If~, Vtuush, thc rcsonator is without effcct; hut

whcn
=+ 1
that :s, wi~n thc r~on~or Iles on
thc axis oi- th double source,
wchnvc

Thus wo mny consur that th

potentiat duc to ti)c rcsonator
isgreater than that duc to thc double .source in thc ratio
1
th anguiar variation
bcing disregarded.
A vibrating rigid
.spiicrcgives thc same kind of motion to thc
surroundmg au. a. a double .sourcesituatcd at its
centre; Lut th
substitution suggcsted by this fact is
on)y pennissibic wl.cn th
radius of th sphre is .smaHIn
coinparison wit!, c: oLhcnvise
tlic presse of tlic .sphre modines th action
of thc resonator
Ncvertitdes.s th prcceding Investigation, shews how
powcrfui
in gcuM-d th action of a resonator is wt~'i
placcd in a suitable
position e)ose to a compound source of sound, whosc character
is suc)t that it would of itself
produce but little efruct at a
distance.
One of tlie best cxampics of tins use of a
rcsonator is an-orded
by a vibrating bar of g)ass, or mtal, heid at thc nodcs A
strip
of plate glass about a fout
and
an
inch broad, of mdium
]ong
thiekness (say inch), supported at about 3
inches from th ends
mcans
of
by
.string twisted round it, an.swers t!.o purpcsc
very
wc)i. WI.en struek by a hanter it
gives but Iitt)e .soun<texcept
ovcrtones; and e.cn t)K'se mayaimo.stbe got rid of
by choosi~
a hammer of suitable .soft.ic.ss. This
denciency of' sound is a
consquence of th sn~U di.nunsious of t).c bar in
comnarison
t).e ~ye-Jength, wiucfja!tows of th
ca.sytran.sfercnccof air
~th
from onc s.dc to ti.c other. If now t!.c mouth
ofa resonator of
thc nght pitch' be I~dd ovci-one of th free
ends, a .sonad of conToKetthobo.stefeet.t!,c mouthof thorc.s.nnt~.c.nH'.tio ho
pr~tv e!o~ to
hc .r andthcuthcpitchifsd~MIy
t.~u it ..o.d b. 1~
fmaiadju.tmcutn~y bom,.l. by
v~yi~.thoamouutof cb~ructiuu T ~usc of
reBotiators
)gofgrcatnutiquity.

320.]

TWO OR MORERESONATORS.

201

sidcrabic force and purity may bc obtnincd

hy a. wcU managcd
bio\v. lu this way a.n irnprovcd )j:u'tnuiuc<jn
may bo consLructcd,
with toncs much !owcr thau wouRILe
praeticab)o wi~tout resona.tors. In tlic ordina-ry instrument t!ic
wavc-IcngLhs a-ro snfHeicntly short to pcrnut ttic bar to communic~Lc-vibrations to tlic
a.ir indepundently.
Tho rcinforcemGn),of th soutid of a LcU in n, weti-known
oxperiment due to Snvart~ is an exemple of t)ic same ~odc of
action, but peritaps tlie most striking mstancc is in thc M'rangomcnt ndopted by Hchn!)oltx in his cxpcrimcnts rcqnirixg
pure tonc.s,which a.re obtaiued LyItolding tuning-forks over thc
menthe of reson~tors.
321. Wlicn two snnpic rcson~toi-s
separtitely in tune
wit)i tho source, arc close togcther, the eMbctis Icss th:m if Lhcro
wure ouly onc. If thc potcutials duc
rcspcctivcly to J~ ~(~be
wu nn).ytn.ko

Lot
aud
i-cprescnt tlie distance
tlie potentiaLs
th:u, would uxist at ~t~,
if t))crowct-c no rcso~toi-s; thon the
couditiuns to dtermine
arc by (5) 31:)

bincc
is Hmd), thc cfTuct is much !css than if there wcro
r'n!y onc rcsonn-tor. It mu.st bc obscrved howevcr t!mt tlio
dnninishcd cUbctivcncss is due to titc rcson:).tors
putting ono
anothcr out of tune, tuid if thi.s tcodoicy be
conipcn.sn,tc(!'bya)i
ftftcration in t! .spriog-,any numbcr of rcsonn.tors])c~tto'thcr
)mvc just t]te ccct of one. This point is iHnstratctt
hy :{()~
whcrc it win bc SGcn(32) t!)!it titougit thc rcsuuiLncedocs not
depond upuu thc sixc of t))Cptatc, still thc incrtia of thc air, which
bas to Le couipcnsated by a spi-in~ docs
dcpuud upou it.
J~t. ff.C/t!'m.
t.xxiv.1823.

202

FORMATION
0FJETS

[322.

322.

It wiH be propcr to say a. fcw words in th!s

place on
an objection, w!:if-h ha.s bccn
brougbt forwa.rd hy Busam~tct' .t.
possibiy inva!idnting thensual caicutationsof
thu pitch of resonators and ci' thc correction to th
]cngth ofor~an pipes. Wfien
itnid iiows in a. stca<Iy strc'atn thn)))~h :(, hoie in a t]un
phtc, thc
motion on
no ]nca))softhcc)):L!-actcithcfowpn.s.susidcisby
invc.sti~Ltcd in ~OU. ]nstc:ut of divo-~in~ :Jtcr p~sing Htc hu!c
su us tu futiuw t)ic .surfaco oft))c
p)atc, L).c ftnid sh~pc.s itself into
forthuca.scot't~o
fmappn~i)nat(.')ycy)in')rK'a)jt.'t,whn..cfonn
dimensions canbccahutatcd
ironi formuki~ivcnbyKirchhnf'
On Htc high prc.ssurc .si()c thn motion dues not dc-viatu so
widuiy
fro)nt))atdctunnincdbyt).cdcctrica!Ia\
InlikumantK-i-nnid
pas.sin~fn)t.\v:u-dsfrom
pipe continues toinovc in :icy)indrical
stremn. If thc extcrnal prcssnru hc thu
~rc~tei-, thu ci):u'actui- of
t))u tnotion is di~urent. In tftiH case tilL- .strcani Hnt.s
convoie
from a)) directions to thcmunth of t.])C
pip(., aft.c-rward.s~tin.ri~
t)tumsolv~ into n. pnr.diui ))nnd)c, whosu .s~tiou is
considcrabty
IcsH than t))at of thc pipe. It is ck-ar that, if tite formatlun
ofjets
took pla.ce tu aoy considuraDu uxtcnt
dnring t!ic passage of air
thrung-h H)C niontits of rusonator~ our c~IcuLtions of pitclt woutd
Itave to bu scriousty modinud.
Thc prcise conditions nnd~- winch
Jets arc formod is a. snhjcct
ofgreatd.dicacy.
It!nay(iV.jnh.()onbt~Iwhct]K't-t]toyw,,n)doccur
at aU in irictiuldcss ituid movin~ with vdocitios so sntal) that thc
cor)'ospondin~prL-.ssun's,whie]t nrc pn~portional to tho squares of
thc vc)ocitk's, are incon.sidcraoh'. Hut wittt air, as \vu
actuaity
action nfthp pressures ta Le fonnd in
!)a.vcIt,moYingurH!cr<!)u
n'sonators, it must bo admitted that JL-ts n):t.y sonnjtimcs occnr.
\Vhitc cxptjritnuntin~ about twu yuars
ago witit onc of Kuni~'H
brass i-csanatur.s (jf pitclt c', 1 noticct! that whcn thc
norrusp<.n()iu~
fork, stnjngiy cxcitcd, was ]ietd to thu moutt), a witid uf cunsid~
~bic force issncd froni th nippie at t))c
opposite side. T!iis ~ft'uct
jnay rise to snch intcnsity as to L]ow out a catidio npon whosc
wick thc strcam is diructud. It docs not
dupcnd npon any pL'cniia.r
tnotion of th air nL-ar thc ends of thc fork, as is
proved hy
thc
fork upon its rcsojiance-box an() prcscntin~ thc
inounting
opun
end of thc hox, instcad of t))C fork itscif, to t)tc mout)i of thc
rcsonator, whcu thc cilect is ohtaincd with but sligtitjy dinunishcd
'J"~f;VM;Aun.lH77.r.l2S.
l'hil.
Dec, 1S7~.

322.]

OURI~G

SONOROUS MOTION.

203

Intcnsity. A simitar rcsult wnsobtaiijudwith a forkand rcbou.dt, ot'pitch an oct,a\'(!jowcr (~. (:[oscr examinatioa npvufdcd
thc fact that at t))~ .sidcs of t.]te )npp!c tI)G outward
nowiu-T
strcam \vas rcpiaccd ))y onc in thc opposite direction, so that n,
to enter
tondue of n:nno frotn <).suitabtyp)~ccd c:m)HuftppO!a-C(t
t))c nipp]n
thc sanic Urne thn.t annt))cr c~hHc situatcd
imm<-([i~tc')yin front was )))o\v))nw.~y. Ti.o two efFoctsn.rc of
cour.suin rc:L!itya!t(.!rnatit)g,:u)donJy :tppe:u-to bc sitmdtanerms
in conscqucncc of th in:d)i)ity of thc oyc tn foUowsnch
rn.pid
Thc
fortnation of jct.s mxst ma.kc.1.scrious draft on thc
changus.
Otur~y of thc tmjtion, an() thi.s is no donbt t)tc rcasnu w))y it is
todosc thc: nipptc in ot'dut-to ohtaitt ;). po\vcrf)n Sound
nncc.s.s:).ry
fru)n :).t-c.son&toruf this furm, witen suit:ddy tunud f'ork is
prcscnted to it.
At thc same fiinc it docs not nppca.r prob;tb]c tt)~t jet fonnfttion occurs to any appreciabie extcnt !).t t)tc )nont)ts of t-csonators
as ordinariiy uscd. Thc ncar agrcenK'ntbctwccn t))CobHGrvudand
thc c:dcu!atu(t pitcit is ahnost a sntneinnt pnx'f of this. Anutl~r
iLr~umcnttonling to thc samc concluston ninybu dt'awMfrum th
pcrsistenccofthu frcc vtbmtions ofresoun.tors ( :ni), whoscdurat)on scoms to excludo nny hnpurt.fmtCt).uscof dissipa-tiou bcyuad
the commuuicatiu)iof niotion to th surrouuding air.
lu thc case of organ pipes, wt)crcthc vibrations are vcry powcrfu), ttjcsc arguments arc Jcss cogent, but 1 ncc no reason for t!unkingt!<at th motion nt the uppcr&pt'nouhHO'crs greatlyfrom thn.t
supposcd ni irchnhottx'tj cfdcuhttion. No conclusion to t))u contnu'y c:u), 1 thhik, safc)y bc dra.n from the phenomena.of .stcady
tnotion. la thc opposite extrme case of impulsive motion jets
certainjy ca.nrtct bc fonnod, as fot)o\v.sfrom Thmuson's pri~icipic
of least cncrgy (7!)), and it is doubtfu) to wilich extroue tho
case of ponodic tuotiou )nay with gruatcst plausihihty bu assitniJatcd. Observation by thu mcthod of intermittent illumination
( 42) might lcu.dto further Infortna.t.icnupou this subjuct.

C'HAPTER

XVII.

APPLICATIONS0F LAPLACE'SFUNCTIONS.
323. Tim gcncral cqn~tion of n, vcioeity potcntial, w!)cn
rcfcrrcd to polar co-ordin~tos,takcs thc foi'tn ( 241)

If /c vanish, wc hve thc quation of tho ordinary poicnUa.1,

wttcrd dcuotes thc
which, as wo know, is sa.tisHc<I,if '=~
sphcric:Usurface harmonie' of ci-dur?;. On substitutiuli it appc:u's
is
t)mt th quation sati.sficdhy

whcrc

will satisfy a.!i.cqun.ti<msueh ns (2).

Comp~ring (1) n.nd (2) wc sec that to dtermine

function of?', wu hve

as a

1 On tho tttcory of thoso fnnctinns the If~tostEnf{)ish \vnr]is nrc

ToJhnntor's
' ~'M)t<-<;uMLaplace, JL~t/tc,fuif~~t'M< aud Fei'rcfs' ~/terf'cat TYan/tottt'ctf.

333.]

SOLUTION

IN LAPLACE'S

FUNCTIONS.

205

In order to solvc this quation, we

may observe th~t when r
i!,c ).,i(!ji(; tcnuisrelativcjy
i.n-~).u,
ncgligiblc,andthnt
LhcuLhcsolution is
.nu h:ut~ Mnn nmy Deassumer ta itoid
good for tho complte
cquatujn (4-),it' we look upuu !Utd uo lungcr as constants, but,
as funcLit.n.sof ?-,who.so!i:Lturcis to bc dctcrmined.
Substitutin~
in (-~),wc <iudfur 7)',

.Lhc symbois ~1, and~, though indcpcndeiit

of?-, M-efunctions
of th angular co-ordin~tes in thc most
gGiiei-~ case, they arc
any two spherical surfMo harmoniesof order M. Equation
(a)may
thcrcforc bo Avrittcn

<i.

On thc Communication of Vibration ~om a

Vibratiug Dody to a surrouudiDc
~<t!.rr<;))s.l8C8.

20G

EXPRESSION

FOR RADIAL

VELOCITY.

[323.

Thc forms of thc functions F, n.sfar as 11= 7, n.rc exhUjttud in

th n.ccompa.nyingt!t.L]c

Jo m'dur to intd thc ]!u)ingturtus iu 7'~(~)') whcuMris srn:t,[),

wc It:u on rtjvct'sm~'thu surius in (!))

32t-. An important case of our gnera.! formulai occurs when

rcprcsents n. (iisturbance which is propagft.tcd w)tol]y oM<~(?'
a.nd
At:tgrc:).t distance i'rom tlie ong:)),(w)=~(-~)-)=l,
thus, ifwc rcstorc tlie time factor (e~), wuhnvc
of which thc second part represcnts a. disturhn.ttcc travcDittg
iuwnnis. Under the circumstanccs contemp)a.tcd wc are thcrc= 0, and thus
fm'uto Utkc

which rcprcscnt~ in t.hc must gner:).! manncr the ?)"' hiu'numic

c'))u])t'n~ntcfndiMturb:uT.cc
oft)ic~ivcnpcriod<.]ifru~i!igit,sc)f
out.a)'()s]ntoi~f))))t'jsp.icc.

324.'}

DIVERGENT WAVES.

207

T))c m'igin of tlie (hsturbfinco mny bc in n.

prcsci-iLed normal
motion of tho surf~cof
n, sphre ofmdiu.s p. Lf't. us suppose
<.ha.t :)t !').nypoint on thc spitere th outward
vchjcity is reprcsuntcd hy ~'e'
bcing in ancrai n,i'uuction of t))e ROHitionof
thc puitit consi()e]'c().
Lt'

bc cxpfmded i)i tlie sphcric.'L) hiu-mnnic scries

whcrc the sunimation is to bc extcndcd to all (nitcgra)) values of

M. Tlie rca.! part of this e({uation will give thu vulucity potcntiid
duc to thc normal vuloeity f/cos/< 1 at thu surface of thc
sphre )'=c.
Prof. Stukcs bas apphcd this solution to the cxp~natioti of a
remarl<ab)ccxpcrinicnt Ly Ij(\stic, ncc-onhngto which it n.ppea.rud
that t!ic sound of:L bu)! vibrn.tingIn f).part~Hy cxh~ustcd reccivcr
is (bmini.shudby the i)it)-oductiouof )iydrogcn. This p:u-a(tuxical
phcnomcnoji ha.s its o'i~in in t!tc n.))gmentc'dw:ivc-!cngth duc to
the addition of hydrogcu, in consquence of whiuh tlic hc)).loscs
its hold (so to spc:d~)ou the surroundinggas. T!)c gc'ncnd exp!anation Cimnotbe Lutter givcn than ni titu wordsuf Pruf. Stukus
Suppose:). pet-son~ojnovc his han()to andfro through n.sma.U
spaco. Ttic motion w]neh is ucc~sionudin t))c air is ;dmost cxactty
thc stLmuas it would Itave huunif thu air ])!n)hL'cnnu incompressible fhud. Thcrc is mcrc iocrt.!reciprocaLingmotion, in which
thc air unmcdiatdy in iront is push'jd forw:u-d,:md that Immcdiittuly bchind itnpcDd aftur t)tc inoving hody, w!n!c in thc
antcriursp:Lcc gcucr;d)y thu :ur rccudus from the eno-oachmcnt of
t]ie moving body, anditt Die postcrior spacc gcnoraDy f!n\vs in
from :in sides to supp!y tlie vacuum which tends to bc crcatcd so
tha.t in latral directions thc now of the fluid is backwards, a.
Thcassuniptiou
of fi real vuluefor Uis cquiya]out
tu limitinHtho normal
t~Luht th Htunn
t!)nf<]))u')-o
Yc)~ciLy
!'=r. To inc)~ thonio.tt
phitscai)f)v<'['
t;cncM!
uuriutiaotion wouidhfLYt'
t') hntt'c:it,f'~
ftsco)i)j.)cx.

208

0FSONOROUS
FORMATION
WAVES.[324.

portion of th cxccss of ihud in front going to supply th dcncicncy behind. Now conoive t))c po'iodic time of th motion
to bc contmuany diminishcd. Gradu:dtytho altcrnation of movGmcnt beco;ncs too nLpidto permit of th(.!fu!I Gsta.Dishmcntof thc
mct'c-)yIoc:Urcciprccn.tin~f)ow; th air is .scnsibiycotuprcs.sedand
rarcRcd, a-ndi),sensible suunft wa\'c (or wnvcof t))e samu nature,
in c:t.set))c po'iodic tima bu bcyoud t)ic )i)nits suit~bicto hcarin"')
ispro}~giitu()to f),()ist;u)ce. TLc s:unu t:(.kcs p)!).ccin a.nygas,
an<)thc niure rapi<)bu tho pro])ng:LtiunofeondunsatKjns:md rfu'efactions i)i t)tc gas, tbu niurc ncarly will it approacb, in rctation to
t)ic motions wc ]t:LVu
undur considcra.tion,tu thc condition of a.n
incompressible nnid thc more ncariy will thc conditions of th
dispincmnunt of t!)Ggas n.t thc surface oftiie solid bc satisncd by a
murcly local rcciprocating itow."
In discussing the solution (.), Prof. Stukcs gocs on to say,
"At a grt distance from thc sphre th function~(~?-)' bccoincs ulthnatulycquat to 1, and wc hve

"It nppcars (from t)ie va.Iuoof

) that thc couipcmcutof the
velocity a]ong thc radius vcctor is of tlie ordcr ?' and that in any
direction pcrpcndieular to thc radins Ycctor of the order f" so
that tho latcra! motion may bo disrc'gardcd cxcept in tlie neighbourhood ofthc sphre.
In order to examine the influenceof thc Interd motion in thc
ncighbourhood of th sphre, !ct us compare tlic nctual <.1isturba.nccat a grcat distance with what itwcuht hve bccn if a-Il latral
motion had bcen preventcd, suppose hy inHnitdy thin conica)
partitions dividing tlic ftuld into donontary canats, each boundcd
by a couical surface having its vcrtcx at th centre.
"On this supposition thc motion in any canal wouldcvidently
hc thc samc as it would bc in a!l directions if thc sp))crs vibratcd
by contraction and expansion of the surface, thc same a round,
and such that th normal velocity of th surface was thc samc as
it is at t)ic pa.rticuhu' point at which thc canal in question abuts
on tlie surface. I~uwif~werc constant thc expansion of ~wou!d
1 hn'vemaneaomcsliHtttchMRca
m rrof.Stores'notation.

324.J

EFFECT

OF LATERAL

Le reduccd to its Hrst tcrm

shou)d i)ftvcfrom (.~),

MOTION.

and seeing tha.t

20~)9

(~r) = 1, we

Thi.sexpressionwi)! apply tn a.ny p:.rticu!ar cana) if we

to
t~c
doiotc Lhunormal vctocityat t.hcspl.crc's sm-fafur t,hn.t
part.icJtar
and
canal
thcrufure to uLtiun tm expression appHcabic at H))cc
to aU tlie e!Lna)s,w~ )):tvumcrdy to writc ~'fo!'i\) faci)ita.to
a- compari.sonwith (~ and ((!),I sha)t, howevcr, writc
for U.
Wc hve then,

Tt must be rcmcmhcrcd that this is

merciy an expression applicabto at once to ail thc canais, t)ic motion in each of wincft takes
place whollyalollg thc radius vector.and accordin~tyt!je expression is not to be differcntiated wltit
respect to or Mwith th view
of fmding tlie trans versovelocities.
On comp~nng (7) with the expressionfor the
function in
th fK;tu:d motionat agrt distance from th
sphre ((!), wc sec
th:it tlie two arc idcntictd with the
is dividud
exception that
by two different constants, na.mc!y~(~c) in tlie former case and
7'(!<-c) In the latter. Tlie sa.mewill he true of th Icading terms
(or those of the order r'') in the expressions for the condensation
and velocity. Hcnce if the mode of vibration of th
sphre be
such t)tat the normal velocity of its surface is
expressed by a
Lap~ce's function of any one order, the disturbancc at a grea.t
distance from t!iG spiiere will vary from one direction to another
according to th same law as if latral motions had been prcvented, th amplitude of excursion at a givcn distance from tlie
centre varying In bot)) cases as the amptitude of
excursion, in a.
Jtonnal direction, of th surface of the sphre itself. Th'
on)y
dinerence is that cxpresscd by ttie symbohe ratio ~(~e)
7~(<).
If we suppose (t'/te)reducecl to th form (cos + i sin
a,,
a ),
tlie amplitude of vibration in the actual case will be to that in the
to
and th phases in th two cases will
stlpposed case as
differ by o~-a,
"If th normal velocity of th surface of th
sphre be not
expressible by a sing!e Lapkce's Function, but only by a series,
imite or Innaltc, of sucli functions, the disturbancc at a
given
R.IL
14

EFFECT0F LATERALMOTION.

210

[:24.

grt distance from thc centre wi!)no h)nger vary from one direction to another according to th saine Ia\v as thc normal velocity
of thc surface of th sphre, sincc t))C moduh~s ami hkcwise
thc amp]itu<tc ofthc imaginary quantity 7~(~c) vary with thc
ordcr of thu fttnctiou.
Lct us now suppose thc disturhunce cxprcsscd hy a La.pIaCH's
'fonction ofsomo ono ordcr, and scck thc numcnciLl v:duc of thc
ahcration of intcnsity a.t a. distance, produccd by th latcrfd
motion winch actutdly cxists.
"T!)e intensity will bc measurcd hy tlic vis ~< produecd in a
.~ivoi timc, :md consc(jucnt!y will vary as th density muttiptictt
hy t]ic velocity of propagation mtdtipHcd by th s<)uaro of th
amplitude of vihratton. It is thu )ast factor atone that is diifercut
h'otu what itwouhl hve bccn if titerc had bcen no latral motion.
T]te amplitude is ahered in tlic proportiou of/~ to
so tliat if
thc quantity by winch thc intcnsity that would
/~u''==
ttave existcd if tho nuid had bcea hindcrcd fTomlatral motion
lias to bc dividcd.
"If
be the Icngth of thc sound-wavc corrcsponding to th
that ~c is t))C ratio of thc
period of thc vibration,/c=27r\so
cn'eumfo'cncc of thc .sphcrc to thc lengtb of a wttve. If we suppose th gas to be air a))d to bc 2 fcct, which '\vou)d correspond
to about 550 vibrations iu a. second, and tlie circurnfL'rcnce27rc to
be 1 foot (a sizc and pitch which v'ou)d correspond with thc case
ofa connnon houfic-bc))),wc sh:dl hve A'c=.
Tho fuHowin"'
tahtc givcs thc values of thc squares of tho modtnus and of th
xc

;t=0

?;=!}
l

4
1
(~

17
5
2
l-ax

;;=!
1(!'25
S
5
1(!

M=2
l.l'87n
<)';)lL!5
89
l:):i<)'2

0'5 l'()<~5
0~5
1'2:' M'()t!2
1B'2i ~()87H
W:\O'2
4
2
1
()-~
0"

1
1
1
1
1

0'M58~
1
'J'C
1:<
(:U-2'.)t

0'87.T~
]'8<J
41~
JDfii~
J'JHM

t;=t
--l.

l!8-t8
M
3i)'
2:~91

ll.Sii789t)
12:!IJl!H
0-81.15:)
16
1M~~
18Si).'<:i
muCS~l)~

_1.
20-177
l.i!);8
30()):t7
720H)::)71

720Hlj:J71
181f;()xl0"
lo~
l-tuf::)
2!)U-1C
1;-it)<w:n
57':<)')!)7
1?(J!JJ:<]"

?
S
SS.

F
S
2.

ratio 7~ for tlie ftmctions.7~(t\'c) of thc first ih'c on]t;)'s,for e:Lch

ofthc vaincs~, 2, l,[md~of~-c.
JtwiDprescntlyappcar~y

324J

STOKES'

INVESTIGATION.

211

the table bas bccn extendcd further in the direction of values

greater than than it lias in tlie opposite direction. FIve significant figures at least are retained.
"Whcn ~c=cc wo gct from th ana!ytica! expressions =1.
7'
Wc sec from th table tluLtwhen ~c is
somewhatlarg-o f,, is liable
to bu little !css titan 1, and consequentty the sound to Le a little
more Intensethf).n if IntGrn.1motion }i~d becn prcvcnted. T)tc
possihility of that is cxpin.incd byconsidci-ing that tlie wavcs of
condensation spreading ft-on thoso compartmcnts of th
sptto-c
w!uch fit a givun moment arc vibt'ating positively, ~.e. outwards
~ftcr the hpsc of a hatf pcriod n)ay Iut.vc sp)-ca.dover the
nci'dibonring cottipiu-ttncnts,wnich arc now in t)icir turn vibrating
posittvcly, so that Htcsc latter compartments in theh- outward
motion work ag:u)t.sta somewhat grcatcr
pressure than if snch
comparttnont )iad opposite to it only ttie vibration of tlie bas
winch It Lad itself occasioncd; and th sajnc
cxp]anailuu applics
MtM~~t&-);t!<i~~ to the waves of rarfaction. Howevcr, th incrca.sc ofsoundthus occasioneclby the existence of Jatm'at motion
is but snitdl in any case, whereas when /cc is somewhat small
In
itrcreases enormous)y, a.nd tho sunnd hecomes a mcro
uothing
with
what it would ])a,vebeen had lateral utatiua heen
comparcd
pre-vcnted.
TIiGhigher be the order of th function, the grenter will he the
numbcr of compartments, alternately positive and ngative as to
their mode of vibration at a given moment, into which the surface
of tho sphre will h dividcd. We sec from t)ie table that for a
g)ven periodic timc as well as radius the value of 7,, becomcs considrable whcn ?t is somewhat high. Hovever practically vibrattons of this kind are produecd when thc elastic sphre excutes,
not its principal, but one of Its subordinate vibrations, th pitch
corrcsponding to which rises with the ordur of vibration, so that /c
increascs witti that order. It was fur this reason that the tab)c
was extended from /fc=0'5 further in tho direction of high pitch
than low pitch, namely, to three octaves higher and only oue octave
lower.
"WIien the sphere vibrates symmetricaHyabout the centre, i. e.
so titat any two opposite points of th surface are at a given
moment moving with cqual velocities in opposite directions, or
more generally when the mode of vibration is such tha.t there is
no change of position of th centre of
gravity of the volume, there
143

212

LESLIE'S

EXPERIMENT.

[324.

is no term of order 1. For a sphre vibrating in thc manner of a

bcU t!tc principa.1 vibration is that exprcsscd by a terni of thc
order 2, to which 1 sha.11now more particu)a.r!yattend.
ruttinc. for shortness. ~c" <7.wc hayo

so that the utmost incrcasc of sound produccd by latral motion

unounts to about 15 pcr cent.
"I nowcome more particuhu-Iyto Lcsiie's expcrimcnts.
Nothing
is stated as to th fonn, sizc, or pitch of lus bcll and evcn if thse
had been accuratdy describcd, there would have bccn n.good dea.1
of guess-work in fixing on the sizc of th sphorc which should be
considercd th bcst reprsentative of the bcU. Hencc a)t we cn.n
do is to choose such values for and c as are compare with the
probable couditions of thc cxperimcnt.
"I posscss a bcll, belonging to an o!d boU-In-airappn.ratus,
winch may probabjy be somewhat sMnihtrto that used by LcsIIc.
It is ncarly hcmi.spheric:d,t!t0 diamctcr is I-9Ginch, and Hic pitch
an octave above t)te middtc c of a piano. Taking titc numbur of
vibrations 105C per second, and th vcloelty of sound in air 1100
feet pcr second, we hve \= 12-5 inches. To reprcsent thc be)l by
a. sphre ofthe samc radius wonid be vcrygreatly to undcrra-tct!ic
influenceof local Ctrcu)ation,slnce ncar the mouth the gas bas but
a little way to gct round from thc outside to th Insidc or the
reverse. To rcprescnt it by a sphcre of hait th radius would stiU
apparentiy be to underratc thc cnect. Neverttte)ess for tho sak
of rathcr undcr-cstimating than cxnggcrating thc innuence of titc
cause Itre investigated, 1 will make thse two suppositions successively, giving respectivcty c = -98 aud c = -4.9,~-c= ~D2(~ und
~c = -G3 for air.

324.]

NUMER1CAL

RESULTS.

213

"If it wore not for latral motion tlie intensitywould

vary from
to
gas
gas in t))o proportion of tlie density into tho velocity of
propagattot~ and thci-efore as tlie pressure iuto tlie square root of
the density under a standard pressure, if we take tlie factor depcuding on tho devoiopment of heat as sensibly t)ie sa.mefor tlie
gasus and gascons mixtures wit!i winch we have to deal. lu the
ibHowJug Table the rirst column gives th gas, the second tho
o
O*

~-j
o
p

i-)

w
o

co
to

o
e5

M
o
c-<

'p

$000000
~o~So~c~o

CID
~n

sgg~ssp:
'<'

co

00
t~g~ts~cacaf
~Q
'p!~
O

f-

<f:

Ip
Q

oo
OD
B
'-t
0
<p

0~

[~

o
.O~
~f
<-<
0
<p

f<
rl
p

r-<

G~~1

UQ^

pp

<p

'-<
M
0
~-<

m
eo
ro

tg<g<DO<C<OM

))

t_"

&<

mBfoBmMM
t~W
'-<-}<~0f-<9<
ob
e~
c~
~<
t~

rt~
M

0
[~
o

t.

c<

L~
r-

t~
tr
0

f~
f
Ot

c)

r-<

h
tM

e?
Q
0~

o e<)

.-i
<?

d
Ot

f~

co w
o
o

?<

CO
o~

M
m

o)

M
00

M
CO

t"
o

r~

.-<
<-<
'o

r~
<?

S
t~

P
rd

r-) *-<

t,n.

.r

11

s
.ri

g
u
g

:a

.h

E-'

g
S

'o

<i

214

DEFI.CJKNCY
0FTERM
0FXEHO
ORDER.
[324.

pressure ~), in atmosphres, the third tlie density Z) undcr tho

pressure referred tu thc dunsity of thc air nt th atmospht.'nc
w])a.twould hve been tlie intotpressure as unity, t))e fnurth,
sity h:ut t))0 motion beeu \vhol)y radial, referred to tho iutcnsity
in iur :).t attnosphcric pressure a.s unity, or, in othcr word.s,a.
quantity varying as x (the dcosity at pressure 1)\ Thon folin\vthc valuesof (/
a.nd th Jast Lciu~ thc aetual iutensity
ruferred to air as htjfurc.
Anit~pcction ofthc numbers eontaincd in the-columns headed
wl!!shcw that thc cause hre investiga.tcdis amply sufRcIcntto
aceount for tho fa.ctsmcntioncd hy Lcsiic."
Thc importance of the suhjcct, and thc master!y mander in
which it lias hecn tre~ted hy Prof. Stokes, \viHprubably bc thought
sufHcicnttojustify tins long quotation. Th simpiieity of thc truc
cxphufation contrasts rcinarhabiy with conjectures that had prcTiousiy bccn advanced. Sir J. Hc'rsehe), for cxampic, thought
tbat thc mixture uf two gascs tcnding to propa~atc snund wit.h
din't'rcnt velocities might producc a.confusion rcsutting ia a.rapid
stining of t))Csuund.
~25. Th to'm of zero order

where is a, complex constant, corresponds to the pntential of a

~'w~ M!t?'eeof arbitrary intensity and phase, situated. at the
centre of thc sphcrc ( 27!)). If, as often Imppens in practice, tlic
source of sound be a solid body vibrating without much change of
vulume, this tcrm is relatively dcHcicnt. In the case of a rigid
spliere vibrating about a position of cquMihrium,the deficiciieyis
absotutc', inasmucli as th whole motion will then be rcpresented
by a tcrm of order 1 and whenever thc body is very smn,)lin
comparison with titc wave-length, th term of zero order must
be insignificaut. For if wc intc'grate th equation of motion,
ovcr the sma,Uvolume Inciuded between th body
\7'+~=(),
and a sphre closely surrounding it, we sce that the whole quantity of nuid which enters and leaves this space is small, and that
therefore thcre is but little total flow across th surface of the
sphre.
1 Thccentreof thospitorebeingthe originofcoordicateB.

325.] ]

REACTION

ON RIGID VIBRATINCt

SPHERE.

215

and ~is proportiona! to thc cosineof t]ie angle bctwccu tho direction considcrcd a.nd somo iixcd axis. Tliis expression is of thc
same fonn as thc potcutial of n,(loitblesource ( 2D4-),situated a.t
thc centre, a.ud coniposcd of two cqun).i~ndopposite simple sources
]yin~ on thc axis in question, wtiose distance npai't is innnitcly
S)na.t),andintensi.ties Such t)iat tho product of the intensities aud
distance i.s fiuite. For, if ? bo tlie axis, and thc cosinc of the
angle hutwcenx and r bc it is vident that tlie potential of tlie
duubtc source is proportional to

It appca.rsthcn that thc disturbfmeoduc to the vjbrutioMof a

sphcro as a. rigid body is thc same as tha.t corresponding to n,
douh)e source a.t tho ccutrc whosc n.xiscoiucidcs with tit Hue of
thc snhcrc's vibratinn.
Tho raction of thc air on a small sphci'c vibrating as a, rignl
body with a. harmonie motion, may bo l'cadily catculatcd from prcceding formula. If dnote thc vcbcity of the sphcrc a.btime <,

21G

INCREASE0F EFFECTIVE INERTIA.

[325.

Thc opration of thc a.ir is thcrcforc to incrcase tho effective

incrti. of th spho-c by~) timesthc inorthlof
t))cairdisp!nccd,
and to ret:n'd t))G motion bya. force
proportiond to t.))Cvctocity,
:u)d cqu:t! to ~-n-pc"
thse eifcetsbeing in gcucm] fmicuon.
of thc f)-c<}ucney
of vibration. By introduction of th vatucs
of/
:md 7~ we nnd

Whon xc is sma)), we hve approximatu)y

~=~, <y=~V.
Hcncc th cucctivc incrtia of a sma)! spbcrc is incrc:iscd oncby
ba]f of that of tho air displaecda
(p)antity indcpcndcnt of thc
frc(]uoncyand thc same as if tlio nuid wcre mcomprcssiDc. T))c
di.ssip~tivcterm, which corresponds to thc cncr~y cmittcd, is of
high ordcr in ~c, and thcrcforG (the cfccts of viseosity bcing
disrc~rdcd) thc vibrations of a smait sphre arc but slowly
datnpcd.
Th motion ofan etHpsoid through an
incompressible fluid bas
bccn investigatcd by Grccn', and his rcsnit is
applicabto to thc
c:).icu)ationufthe inercascofufFt.'ctivcint-rti~duc to a
compressible
tbo
nuid, providcd
dimensions of thc body bc smiJt Ht compariso!!
with titcwavc-JcngtIi ofthc vibration, rur a snndt circulai' dise
vibrating at rigbt an~es to its p]anc, th increasc of cneetivc
incrtia is to tbo mass of a sphre uf nnid, wbosc radins is
optai tu
tbat of th dise, as 2 te 7r. Tbc rcsntt for tbe case of a
sphum
abovc
was
ohtaincd by Poisson", a sbort tinic bufore tbu
givoi
publication of Grucn's papcr.
Jt bas bccn provcd by Maxwell' that thc various tcrms of
tbc harmonie expansion of th connnou
potcntiat may bc rcgardcd as duc to p~if~t/~e~o~~ of corre.spoading dpgrces of complexity. Thus

is proportionat to

~.hcre there

arc t differentiations nfr'' with respect to'thc axes

7~ &c.,a)iy
in
numberofwbicfnnay particutar cases coincittc. Itmigbtpo-haps
7~
Dec.]R, 183H.AJsoGrccu's~ti/.w<tffct
~H.~tctt-oM~,
J'~x.
cthtedbyFurrers.~LtoniDat)
&Co.,lti71.
.1/<?w.'n'.<
(/c /f(~f~)~
~< .S'c''))cc<,Ton. xi. p. 521.
M)t)LWt;H's
A'~c~-;)'<y n)M/.)j<if<t!M<, Ch. !x.

335.]

MULTIPLESOURCES.

317

hve bccn cxpectcd that a simitar taw wou)dhohi for tho velocity
for ?'
Tins howcver
potcntia) with thc substitution of r"
is not thc case; it tuay bc .shcwtitliat thc potentia.1of a quadruple
(~ e
dcnotcd
in ccncnU uot to thc
source,
by
corresponds
?'
M/<
io'm of thc second ordcr simpiy, vix., e" ~(!'x)-), but to a
cu)nhin:diunof this with a. term of zuru ordcr. Thc ana.h)gytherei'urc hutds uniy in thc sing)c iastim uf tlie ~6
point or source,
fd'tcrany nuniber of dii'crtliough of eoursu titu function )-e"
cnUations continues to satisfy tlic fuudfuncntal quation

It is pcrhaps wort,h notice that t!ic disturbancc outsidc any

imaginary sphre wltich comp]ete)y encloses thc origin of sound
tnay Le reprcsentcd as duc to thc normal motion of the surface of
auy smaHcr concontric sphcre, or, as a p!U'ticn!ar case whcn t!ic
ra(hus of thc sphre is innnit.uly sina!),as due to a source concentratud la one pohit at thc centre. T!ns source will m gnerai bc
composcd of a combina.tion of multiple sources of ail ordcrs of
complexity.
32G. WIien thc origin of thc distm'hancc is the vibration of a
r!g)d body p:t!'aUel to its axis of rvolution, the varlous sphcrica.1
Itarmonics rcduce to simpte multiples of th zonal fia-rmonic
(~). which may Le durincd as thc coemcicntof c" in thc cxpanmon of {1 2e~+e~~ in rising powcrs of c. And whencvcr thc
sohd, busides bcing symmetric:d about an axis, is a.Isosymmetrical
with respect to an equatoriat plane (whosc intersection with thc
axis is takcn as origin of co-ordinatcs), the expansion of thc
rcsultingdisturbance in spho-ical harmonies wlllcontaintcrmsof
odd order ouly. Fur exampic, if thc vibrating body wcre a circulai'
dise moving pcrpcndieutarty to its plane, tho expansion of -t~
would contain tenus proportiona! to
(~), 7~ (/<.), (/u.),&c. In
thc case of th sphcre, as we hve S(.tcn,thc sries reduccs
absn!utc)y to its nrst tcrm, amt titis tei'))iwiHgencrally be prpondrant.
On thc othcr hand wo may hve a vibrating System symmetrical about an axis and Avithrespect to a.ncquatorial plane, but in
such a mannur that th motions of th parts on thc two sides of
thc plane arc opposed. Undur Uns !ica.dcornes thc idal tuning

218

ENERCYEMITTED

[32G.

furk, cumposcd of cqual spitcrcs or para!t<jlcircuhu' dises, wbo.sc

di.stattCGapart varies pcriodiea!]y. t"!ym)nc;tryshcws titat thc
vutocity-potcutiaLbcit~gthc satue :).bany pointant} :).Lits imii.gein
thc pLnu of nytntnuLt'y,must bc an <;v<j)]
functionof ~u.,and ttrcforu L'x)))'essihicby a, scries c<jnt:uriin~ouly thc cven funct.Ion.s
~c. Titc second fhnc-tion ~(~) wou!d u.sn.dfy
7~(~),(~),
prcji~ndcra.tc, though in particidar ca.scs,a.s fur cxn.mp!cif thc
Lody ~o'c cutnposcd of two (Usesvery cioRCtngctitcr in conpfu'ison
wit)) thci)' dia.mctcr, th symmctric:d tcrm of' zcro ordcr nu~ht
bceome important. A conm~-isonwith th k)iownsotutioti fortho
Hphcrcwhosc surface vibrtes a.cconh)~ to any ]aw, will in most
c:t.sc.sfttD~sitmatcrmi fur an c.stima.t<:a.sto thc relative i)nport:m
of thc various tcrms.
327. Th tt:).! mission of cucrgy hy a vibrating spitcrc is
round by )nu)tip)ying th variable part of thc pressure (proportioual
tu ~-) by th nonnal velocity and. integrating over thc Hurfacc
(2-t.). In virtue of tbe conjugatc propcrty th varions sphcrieal
harmonie terms tnaybc takeu scparatclywitttout lossofgeucrality.
Wu bave ( :)

327.]

FROM A VIBRATJNQBPnERICALSURFACE.

219

Now, since titero eau Le on tho whol.c no accumulation of

cno'gy in thc space mciudcd bctwccu two concentric spherical
surfhcL's,t))C rates of tr:uis)~is.siouof cner~y n.crosstitesc surfaces
inust bu t))c same, that is to say ?- (:['/3/3'ct)must bc mdepcndcnt
of?'. It~ order to dctunninc t))Cconstant value, wemaytnkctlK;
p!).rticu!arcase of 7' mdcfinit.dy gi'cat, whcn

It may bc obscrvcd that the !eft-hand membcr of (5) when

multipticd by t is thc imaginary part of (x+z/3) (a't'/S') or of
(~r)~ (-?'), so tha.t our resuit may be cxprc.sscd by saying
tha.t th iniaginary part of j~, (~r)~, (- '/o') is '/cr, or

In this form we s!~l! hve occasionprcscntly to mako use of it.

Thc samoconclusion may be arrived at somcwhat ]norcdircct!y
by a.)iapplication of Heimhoitz's thcorcm ( 2!)-t), i.e. that if two
functions Mand satisfv tlu'ouirh a.closcd sDa-ceS the Gnua.tmr)

220

SOURCE

SITUATED

[~27.

It will bc more instructive tocxhibit ~a.safut]ctionof

tho
no)')n:).lmotion at thc suri'ace ofi), sphurc of ra<)iusc. From (2)

Tins fur)nu!a )Mybc vcnfied for tho partieular cases ~=0and

M=1, tt-catcd in 280, 32;')respectively.
~28. If th source of disturb:uice bc a normal motion of a
~)n:U[p:u't of thc surface of tlie Hphurc(?'=c) iti Lhe mnne'Imtc
)K'i~)tbour))Oodof titu point /t=l, wc must takc in th gcncr:d
HulutionappHc:t.btcto divergent waves,viz.

ON THE SURFACE0F A SPHERE.

328.J

2211

We will nw cx~tnino t)tc probluni wjtun/<-ci.s not vcry smn.)),

ta~ing fct'simplicitythc cascwhr'is
!'?<))'<t
n ~)'c'
distance Ot~iy,su th:).t/,(~)-) = 1. Tj:f! ihctor on windi tite relative iutun.sitic.sin various dirucLiuns(tcpcttd is

Thc foHowing table givcs tho mcans of calculating J~ and

fur any value of
whcn /<:c=~, 1, or 2. lu thc last c~se it in
ncecfjs~ryto ~o as far as ?:= 7 to gct a tolumblyaccurate rcsult, a.nd
for !arger values of /vc tho calculation would scon bceome
very
Jit-bonons.I)i a.)]prubtcm.s of Uu.ssort. tbc harmonie analysis sccms
to lose its powcr whcn thc wavcs are vcry small ni
comparison
with thc dimensions ofbudic.s.
/<c==~.
2ft

0
1
S
3
4
C

+
+

2~
2

C4
~(:(i
+I.t')()~
+175CU2

1
7
35
8M
+
8M1
.)-:}~1~H)
+

(M+~a~.(~+~)

+'4
+'J8td.)
--()6()J:i!)l
-'()();i-lM7
+-n<)(~()5!)
+'UUUUlii

(,t.~)~-(~+~)
+-22
-)2307n8
--0:M88M
+-()OC:M()1
-)--n<)<)~~
-'UUUU2)i~

222

NUMERICAL

RESULTA.

[328.

/<-c=l.
r'
a

(N+t)ft-(a.).j8")

();)~)~-T(n')

~J
i
2

n j
i
2

-t.
+

i)
't

n:)
tBi)(!
d~t
.)
.)())!):)
-U:}(::i.i()

3
n
7

-<-

+
+ -I

i
i
H
3l

:il7')
f'h~it
+'!<)1~17

f/

a
0
1
i<
4
.'i
G
7

1
2
!.(2
1
+
S!'5
1'75
8
4
IC-IM73 + 35'm.~
-)-]8<(~i;
~t~ 8;')'4:;7.ri
.t
+!i:!H'H()
-U77-H~iG'8
-8M1-7
-)+
+

~~a
+'<:
-J.tD.ttO
--f)tr.7H.t
f'()(~t.t:tH
.f.'()')')7.s7
-'0()<~))7
--ooouOf!

-)-~5
-1
-'a
-t7]0
.)'o:t0t))~ a
-)'()()(i!))~
-'()<))).();
-'<'()))t)7:!
-t'ooooo.~

2.

'(;).~)a-(a.)-)
-)-I
+-<:
+-tn'.)80
-(.') i
--t))H70
-)-f'l:<tf!
.).'()'U
--(JUU7'3

(t))-)~(ft=+~)
+-3
+':(
-'(.7114
--173
-t'IOM?
.)'()1)!~
--tHtt.f!
--oou;)~

Thctnost interesting question on which this analy.sis informs

usisthe
influence which a rigid sphre, situatcdcfosc tot))C
source, lias on tho i)itcns!ty of Sound in dif['c')'cntdirections.
3y ti)c principlo of reciprocit-y( 29'j-)thc som-CGand thu p]~cc of
"bscrvation may be into'chnngcd. AVhcnUterefore wc know t)m
relative intcn.siticsnt twn distant points 7~,7?', dnc to a. source ~t
en thc surface of the sphre, wc hve tdso th relative intcnsitics
(tncasm'cdby potcntial) at thc point ~1, (h]0 to distant sources a.t
:md 7~ On this account thc problem lias a. doub)' Intcrcst.
Asa. nnmerical cxatnptc 1 hve c~dculatedt)ic values ofj~-}-'<7
aud J~+ C'~fur th a.hovcvtdues of ~c,whcn ~.=~=J,~=0,
that is, louking from thc centre of th sphre, in the direction ~f
th source, in thc opposite direction, and IatcraHy.
Whcn ~c If!zero, thc value of F"+ <v~!s '25, which therctorc
rcprc~'nts on thc samc sca)e as in th t:d))e thc intensity due to
an unubKtructcdsource ofc'fpia] magnitude. Wc may intcrprct ~c

NUMHIUCA.L RHSULTS.

328.]

323

as thc ratio of tlic circtnuibreuCLof t)tc sj'hcrc to th \vavc-)ci~th

0
oi'thc sound.

7'f'

J~hC=

I._p
1
-1 1

1
-1 1
0

-).():!)-'L):H17'
'J.Lt!)-)~))t!);
'iH~ii-~K:i:!i)t

-(!)i7H.8) '2:i.-<(T.!)<
-t<)(~)(~);():)f
-)-'H'!li)Uii-'i)()JH)7.i;

-7')'!)-)-2:i)21t

o0

'lij:HH
--15:)M1-~7(!<J;

'2')tH')l
-H;);)7~)
-}1U!~)

-(M1(;1 1
-s.')~j'o
-CM.S

-(!M!)8
-;i.(;~
'ajlj~

'[j7(jIW

In lookingat thse figues tllo nrst point \v]nch attracts

attottion is t)tc conparativeiy shgttt dviation from unifunnity
i)i thc intensiticsindifTerent directions. Evenwt~n thccircmnfercncc of t!)Csphre amounts to twice thu wavc-It-o~t)),tito-c is
sca)-(;e!yanyLhht~ to bo cancd n. sound shadow. But w)).tt is
pcr))~psstill more uncxpcetcd is t])at in tl)c fh-st two cases tlic
iatensity behmd tlic sphre cxcccds t))at in a transvcrso diroction.
Tias rcsult dcpcuds !n:LHi!yon tl)c prcpondcra.ncc of t!)c tcrm of
thc first order, whic)i vfnushcs with /n. Thc ortler of the more
importfuit tcrm.s Ino-c'ascs with ~c; whc-)iA-cis 2, the pruictpu.l
t(.'rtnis that of UtCsecond ordcr.
Up to a certain point the augmentation of thc sphre wH!
incrcase tho total cnorgy cnnttef), bccausc a simple source cmits
twiecas mueh ellcrgy whcn close to Mrigid plane as -whcnentirely
lu thc open. Within the limits of thc table this effect masks tlle
obstruction due to an ino-casing sphre, so that when u.=l,
the ihtensity is greato- whcn t))Ccircumfo'cncc is twicc thc wavcJcngth than whon it is hatf thc wavc-Iength, thc source itscif
rcmaining constant.
If the source hc not simple harmotlic with respect tn timc, thc
rebdive proportions of thc varions const.ituent.swih vary tosonic
cxtcntboth wititthcsixc of the sphre, andwith thc direction
<jft))Cpoint of ohservati~!], illustrating t)ie faudatnc'nta! cbaractcr
of thc analysisinto simple harmonies.

224

KFFECT 0F SMALL SPHERE

[328.

WhcnA-~ i.S(]('('i(]r)))y!(;.ssth:u)onc-ha)f,t))cc!dru]nti<'n
o~y
Le con'iuct.cd with .s))f)ici(.'))tapproximatiu))
a)~uLr;uc;!))y. Thu
!su)t,[.s

It:ipj)(':).)-s<))f).t.st)f:)r:tst)K;tL']'tr]in~t))cint(;nsityis;ut
('vc))f)))ict.iu))<)fjU,,vix.t)tc.s:L))K;at:tHytwn
points ()ia)nc(t'ic:d!y
")'))".s(!<L F"rt.)tupti))dj):)) directions ~=+~ur(),thc))U)))cri(-:).)
c'!L)cu):).t.h)t)ut'L))Li('<)cf)i(.'iL'nt.<)f'/f'c;'isc!tsy<)))!tcc())mtufU)Csin)p)u
Titus
v:t.)uc'.sthL-nn,ismnt:J!)y<t(j fonctions/

\Yhcn /<:c" catTLbcncgicctcd, tlie

Ititcnsityi.s ]c.s.s iu a h~c~!
direction t.hn.n nnmcdi:).tc]y i)i frunt, uf or buhind thc
sphre. 0)-,
by th reciprocalproperty, a Kourcu at a (H.st.:mccwingivca.grctd.cr
inten.sity 011 t)ic sm-~cc of a small Hphere ab titc: point fm-t.hcst
frum the source titan in a hLtcra.1pusitiuu.
If we apply t))c.so formu).c to tho case of ~c = wc
t,
"'ct

wmcn

f~rcc

prctty

ctosciy

witti

thc

rcsuiLs

of

thc

more

complote

catculit.tlun.

For othcr v:ducs of th coc~cient of ~< In (~0)

nil~ht bc
CiJcuIn.tcdwith t!ic aid of tables of Legcndre's functions, or fi-oni
t)tc fullowm~aigebraic cxprc.ssionin tcnns of~

Thc f~o'~ce of iutcnsities in th directions /~=+1 1 at)d

= 1 inay bc very simply cxprcsscd. Thus
1 For tLc forma of th funetious
7~, Hoc 3M.

0F SOUND.
ONA SOURCE

328.]
]

9'~
-j*7

At thu .sanic timu thc totat v:L)ucof J~~+ f< nppruxiin:t.t~s tu

'wh(ji)~<)SH!un)).
~'ttCsct)))h'1"sh:).(.'at)mt,L!rc's(i))~))u:u'h~('Ht))Ccx~]a:)ati"n
pcrc(.'pLtun<)t'tiiu<[t)!U'tcr
()f't.hep:trt.p):)ymt)'yt)~t:\vou!u'.sinthc
f)'U))TL~')Hc)t:L.S()))~dp)'('CL'C()H.
jft shouht ))uohHurvct1that. thc Viu'i~tions (jfiotcnslty in (Hffui'f.'nt
diruct.iotts :Lbuutwhic)i \u itavcbucn spL'!tkit)~:u'u <)uc totitc
!L.s:t))<)Lst:Lc!c,n.n<l)n'ttothcfactt)):it
prc.scncu t't.hespho'c
ofat
Lhc source i.s n titc circtu~fo'cucc ot'titCHphuroiusLcfut
thu c(.'])trc. At a ~)'cn.t (tistfuicc :i small (i)sp):).cc'mc))t cf
finurcuof.su)m')wHt:Ln'L'(.'t Lhc ~~<-sc but )h)t,t))(ju~c;i.uia.))y
direction.
In onh')'tf'fhxl

t1)~:L)t''r;(ti<~t tjfpim.se

wch:LVcfo]'a,s)n:dl

.s))!teru

iront which ~'c tnay infcr <.)~t,thc pitasc ut (1!.st:uicci.sth .satnc

{t.s if thc .sotn'cc Lad hccu sit.ua.tcd at thc puint /t=l, )'==:jc c
no obstacle.
(in.stc:).d(.)t'~=(.'),a)idthc't'c)tadbcun
32!). Thc fm~ctiona! Hy)n))ols ~nd
tcrms of 7\ It is !\nown' that

C'unsidcr))f)W

th

sytnMicopc')'))-

niny bc expresscd in

t
P,,(!),fmd)ctit.
(/Y

f'pcratf'nn~
'Thnmsnn)U)dT~it't,.Yf~.7~7R~(~u~t<'[UnnM~fnr~~yt.
H. 11.

)5~)

22G

ANALYTICAL

EXPRESSIONS.

[~20.

330.]]

MOTION CONTIN UOUS THROUGII POLE.

227

MO. Wc havo ah-cady co]).sidcred m sonie dtail t!tc form

iLSsumcd Ly om- gC)~r:Ll exprensions whcn thcre is no source at
infmity. Au ci]u:L)!yimportant eta.ss of cases is dcnncJ Ly tiic
condition that Lttt'c be no soarcc at t.hc origin. Wc 8h:d! now
investi~'atc wluit l'c.stt-iction is t!i(;rcby itnposcd on. our gcncra!
cxprc.ssion.s.
Rcvcr.sing t.hc scrics for~,

\vu )):t.vo

Sincc thc fmictiou P,, is cithcr \vho!)y odd or who)!y evc']),thc

is whuDyrcid cr whoDyhn~ii);u'y.
expression for
Itt.o'dcrto provcthatthc~'n.toc
of~in(.))'cma.!nsfinitc
\hcn t- vani.si~s, wc bogin by obscrving that
~53

2288

ANALYTICAL

EXPRESSION!

[330.

asis()Lvi()n.sw~c)iiti~consn)crcdtI):ttt))ccfT~ct()f()itTt'L')tti:)t)nL;
c")))yitun)))<~t)rtin)(!.s\ith!-('.s~~ttow
ist'))n))itit)!yi<.h~
thccrr('spu)n!i)~p)W(;i-<.t'
!tt-u)i)!U)tst.jcx)):H)'[t)nj~i))~.s~cha\u u
Miun()nt))eti~))Linn.sccndi))~pu\vL-r.s<jf'7'.

Nuw any positive intgra! po\vt.r of

such ns
c-nn hc
cxpnmtcd in a ter)tnn:~ingHc)'K-soft))C fonctions 7', H)cfnncticn
uf itigh~st. ontcr buin~ 7~, Jt ful!o\~ th:it,
if~ < n,

by knowu pn'po'tic.s of Utcsc fonctions; so t,I):ttthc lov'c.st powcr

/-n
~(/<.)c'
i.s(/~)". K~ainingon)y Lhc )cadmg
of~-inj
tenn, wc may writc

~0.]

FOR VELOCITY-POTENTIAL.

32!)

~'hich shcws that

ViUti-ihcswith ?', cxccpt when ?!= 0.

Thu cctttplutc sries fo)'
whcn t))crc is ho Snrcc at thc
p()t~,i.s)u'))'e cot)vc))K'nt.!y obt:um-dLyt.ht':iLi<t oftttcttt~oryoi'
iiL'.s.sL-1'.s
fnuct.ion.s. Ti)C(Hn'ut-CtiU~eqm).tion.s (-) ~00, sat~Hud
byt))c;.seinnct.ioMS,vix.

i.s thu Bcs.sul's fmtction ofordcr ~i.

Wt)0) ?.is mtc~rd, r (~~+ I) = 1 2
?~ but hcrc wc hn.ve
tu do with )?t fractiona). :n)d of' thc fonn ): + :'i, being' mt.cgra!.
ih t.]tls CiLSC

230

DESSEL'SFU~CTIOXS.

Now tlie function

satisfics (.i.) 32:3,vi~.

[330.

wlt.!i whichwc :u-cat prc.soit conccmu'.),

3:~0.'j

PARTICULARCASES.

231

It wHIbc convctucnt to write duwn fur rcfcrcticc tlic formsof

-Jr~nd ,f"r<.h'Htt.t~cor<I')'

33!. 0))c of the most intorcsting n.pplica.tionsof thse rcsulLs

is to t))e investigation of th motion of a, g:is within a, rigid
sphuricid cnvulopc. Tu detcuninc t)tc frcL:pcriod.swc Imve on]y
to suppose

va.nisttcs,when )' is cqua.1to thc radms of t!tc

tha.t
cnvclupc. T!tus in the case of tlie symnictrical vibrations, wc
havetodutenninc~
ta.n~'=A:r.(1),
{m cquntion which wc hve n.h'e~t~yconsidcrcd in thc ehaptcr
on mctnhnuics, 2()7. Thc first nnitc i-oot (/<= l'-t3037r) eon-csponds to the symmetrica.!vibratio)i of lowcst pitch. lu the ca.so
of hi~hcr root, Uio vibt'i.Ltionin quc.stiun !ms .sphcrical node.<i,
wh.scmdli correspond to t)tc infurior roots.
Any cne, w)tosc vortex is at the ori~'in, m~y bc nuulc rigid
without ~n'ccting thc conditions of thc question.
Th loop.s, or places of no pressure v;u-i:).tion,aro giveil by
or /<-)'=W7r,w])crc M is any intcgct-, except,
(~)''sin/o'=U,
zro.
T)tG case of )t=l, whcn tito vibnttion.s may bo ca.Dcd<U;T.tnct)':).],is pcrha.ps thc most intcrcsting. ~S' bcing a harmonie
ofordcr 1, is proportion:d to cos 0 whcrc is the ang)e bctwecn r

L~2

DIAMETRAL

VIBRATIONS.

[_L

an<tsomc~xcd()h'cct.ion ofrcfcrcncc.

Si)ice~va.uis))fson!y
it.tthcpi)lu.-i,t!)('():u'))oco~ieal)iutfc.-i' ~vithvL'rt.cxn.tt.'iiccentre.
Any jne]-i(H:ui!dpiiLnc,!~o\vcvo-,is nod:d, n.nd tnay be supfoscd
]'!gi(L AJong nny spccificd ra.dins vectur,
fim)
vanish, iuxl
vix. wLon i;u) /<-r=/<-r.T))c
d~ngc sign, Avithcos
C
(~)' sin
\it)) thu nod:).)surfaccH
loops in thc' prc'.sutttcase t)ierctut-ucoinci<tu
ofthe r:u)ia) vibraLiun.s.
Tof!)x~))csph(.'ri('a]))H:)c.s,\v(;]);tvc

Thc fh'nt root is ~'=0.

Cidcutatin~ from Tri~nomct.ric:~
T:')')(.-s Ly t.riat :m() c'rmr, 1 fi)idfurt))u
))t'\t,)'(~)(,w)(K')tco)'rc~punt.Lstu tlie Yibr~thm orm'j.st.intpo-Latiwitttiti
f)..sp)tcrc,
~-=ll!)'2Gx-

1 ~tl

s<jtlt:tt. )-: \='3.313.

T))cairs\vny.sfrun)sidut<'si(k'in]tU)ch
<.))C.samcmnnncras
n) a.(1ou))!yc)<).sc(t pipe. Withunt-an~ty.si.swu
nug'httmticip~tc
tt)att.)iL;pit(.tWonhtLo
h~~u' fur thu sphre titan for !).ciosc(l
pipe ofoqua! Jung't)), hecausc thc spho-c may bc (turivcd f'rotn thc
cylin<)(.-rwith c]osc<! C!t(ts,hyf)))i))~t)pp:irt'.f'tt)u]:itturwit.)i
u1)St)-)K-tin~)natcr)!)],thccnL'ct,of'w))ic))]nu.sthtoMlt:)rpc~thc
t))C))i:)s.stohc mo\')'(I rondins but.iit.Uc c));H~(.'d.
.v])i)c
lu f:K;t,fur a. c)').sc<)pipe of'lungDt 2/

Tf'csphcrci.sthn.s])i~))crinj)it.eht))ant)tecy)i))(IerLvab(.)ut,
aFourth.
T)"vibrt

ion !)ow)i))d<;rcons!(Ic')'f)t!onist])C~ra\'cstofw]ti('))
<)'<sp))(;r(-i.sca))a.b)<iti.sn)()rut.)ta))a)t<)('ta\-c~rav(;r<.))ant))c
~vc.st, radia) vibration. Tf)un('xt\')))r:diuHuf'i)u!jty})e
i.ssuch
tli;Lt
nn
~>,i
t)~tA~=~4()~or
or
Jo~

:HKttst,J)L'ruforch)~hc)-lh:mU)u)ir.sLra()icL).
'A)h~).!i~asut'~cp\\i!it'hn)i;;ht))CHt)pp()~Jnj.th),v.x.('m.aL'r!H\\hi(;i)Uterc
i)-!)o)nutt~t).

VIBRATIONS

331.] ]

0F

SECOND ORUER.

233

Whcn
is gre:<.t,t!n; rocts of (2) mny be convcnicutly CtUcu);'t''dbv!"f'f'~sof'n.Hn)'n:s. It'/<'r=)/t7r?/.t)t(')'

fn'tn which wc mny .SL'tcctfur spccud con.sidL'ratu~n

L)iufuHo\vh)~
nutabtu cases:
(a) thuxonai])fL)'tnonic,

'Hcrc;

is proportiunal

to s!n20,

and thcrcfui~ YtUii.shes

whcn ~=~7T. Titis.shcws U~tthcc~hd

p1a))f isn.!)f)(1:~)
surface, .so t))nt thc s:mic niotiou might tak~ pt~cu within a closcd
)tu)t)).sptx;rc. Ai.~u.sinec duu.snot invulvc N, :my muridiatud pIcUte
n):ty Le rcg:u'()cd :m rigid.
(~3)

tlie

.sccLon:d

H'crc again
nodat.

But

I~n'tnonic

varies as sin 2~, and thc cquaturial

plane is

varies as sin~M, and tito'cfcr ducs nut vanish

it)dcp(.:ndcnt]y ui' 0, cxcupt, whcn sin 2M = 0. IL appcar.s accordi))'dy

that h\'o, aod hut.two, inuridiana! planes an;' uoda), aud th:tt thusu
are at ri~'ht angles to onc anut))cr.
('y) thc tct-scral )iai')nonic,

Inthiscasc~~v[H)t.shcsindcpcn<Icnt)yofMwithcos2~t))at
is, Avitun0=J7r, or .7r, \v]nch givcs a nodal cne of i'cv<j!utlun
mitose
1
vcrtic:d1 a))g]u
1.. i.s n rig-ht
1t :mg)u.
] rl~ v.u'ic.sa.s sinM, und1
M
<hust))Ci'cisu)K'n')f')'idi:)n:)tno(hdp'):mc',nn'1))))tonu

234

AVAVE LEXTHS 0F

VU3RATIONM

['331.

t]i:).nany of tlic radi:t) gr~op.

gtvi)]g a Loncg)-a.vctLi t)tc ca.suof thc gcn~ral hannonie, thc
cqu~tioti givit~r tLc
toncspo.s.sibtuwidun:), sphcrc of radius ;'jf):tybc ~'riLtL'n~))
~(J

corrcspnmfH.g Lu n.einuru important .uod~ofvibr.i..u.

In~iJ
cxhihitu.] thu f'rc(j))(.-ncyof thu varions vibrations rcfL-rn~
to tho
of thc whoc My.stchi. Thc 'i~bic i.s extcndcd iar
~-avcst,
cnou~h
t
includc two octaves.
TAni.nA,
Civu~ tho Yalnesof for a at'bcro of unit radius.
Orderof II~naonic.
0
0
3~

~
c '3r
ICI
gt~

l~:i

S'OJSf; l.MOO l.:)!)3

-81:).~

-f;7(i~

-C8~51

~~U8
'0'*8' -c~.i8

-.il<:7i)

-SOM:i

-.15:~0

-:if!

..10;);~)

'n<)8:M
3

-;);)C2;<

1~77

-80!).~

-7;~<)

4
].ll:t
.c;).~

C
.o:ioo

C
.g~

wf'niiN

33L]

sniEnicAL

ENVELorE.

2:35

TA!<M;n.
I-H.chofcaeh
tuoo, ruturruM
tuno.r.ferrud
toh'raveat.
_`_

l'OOOO

Ordc.r
Or.h.r
~ritchofc.~
1 ~')
N.UI~lher
Il
of of
f.l1Itlll)(H'
1
tnoe
of
CllC\
01',1(,1'
of
"t..ue,~fcn-cd
"f'"t~
HMinuine.
IlM'manic.
tuH~vcst.
"tlIUI CH..
1101OH
III
1
1

2'85.10

J-C056

3~~8

~'l'~S

:<)21

2-KiU

!)

!711t

3-713

!i'773

Ii
il
II

!2.
If wc drop mmcccs.sfu-yconstants, t))o p:u'tictd:n-solution for the vibrations ot'gns\vit]nn~sphcric:d e:)LSCofradius
uuit.yi.src'prc.scntedby

In gc'ncra.ii.singthi.s, \vc must rcnicmbcr t.])n.t mily bGcomposcd ut'scvcml terms, corrcspondin~ tu cach of winch there may
c'xi.st:),vibration oi'm'bitnn'y ampHtndc :utd pha.se. Furtiicr, each
tcrm in ~S'may bc associatcd witil any, or a)!, of thc vaincs of
(tctcrnnncd by (2). For example, nndcr t)iu Iicad of M= 2, wc
might hve

A]iy two of thc constitncnts of arc co))jugn.tc,i.e. will vanihitt,

w))L'umu!ti])iicdtogctiio', and intcgrat.cd ovcr th Yolmneof th
sphre. This fo)!uwsfron thc propurty of t)ic sphurical harmonies,
whcrcvcrthctwotGt'ms considcrcd co'rcsputidto diffcroitvtducs of
or totwo din'ut'enLconstituent.s ofA' T)~couly ca,scrcmamin~
for considura.tionrct~uires us to .shuwtitat

23GC)

CASE OF UNIFORM

[:}32.

which!snnimni~)iatcc<mscfpK'not'a.fn))(huncntaipn)p('rty<)f
<)K-S(;functit)n.s(~0~). Ttn't'cisthcn-forcno
(Hfticuityhtadapt!tI)<jgu))ut';UM())))ti()n tu pr<\scri))(;(!initit cij'cmn.staaces.
L]0)-d<'rt()i)i)).st.t-;dcthissuLj('ctwcwiHt:).kcthcc;tS(\AvIicro
in its position ()f'('()uiiih)'i))mh))t.is))')<)vin~wi(h
nntiaDytheg'asis
constant. \'c)ocit.ypa)';t]!t;)t');r.
Thi.s condition of'thin~-s wuul't bc
:)ppn)xi)n:)Lt<iyrca!i.s<),ift))t;casc,h:u'in~
)~c])pruvi)[.s)yi)iuniftn'm motion, w~r~.sud~chly stopper.
Sincctnct'C! isno

Inititnc'ondcn.sntionot'nu'ofactiun,

(jnn.))titiL'.s~v:u)i.s)). ]f ;~)'u
~'hicit nh(~H tlhitth('.S!))ntion
o)'duri)isp)K')'ic-at Itar~tonics.
forn]

nHthc

i)uti;diyu)tity,wch:i\'('~==.c=~,
ff~tt.'titisoxiy tc-nxs of t))c fir.st
'J'hcsotution is t))<jru)'urcof U~

33~] 1

iNrrrAL
fl

Thc (.'ViUttation
of )'<

YHLoci'rv.

237

(/<r)~ )n:LyLe cH'cct.a!hy t.hc !ud of

:)~cuur:).tt.hcurc))i)'c)atingi.othc'sc fonctions. HythcfutKhuncnt.iLl

ttinL't'untiaiL'quat.toB

233

Sl'HERICALSHKLL.

shcwin.~t)~t LhoHi-sttcrm m th scrics fur

ijnpurt.nmt.
Jft may )jo wcit to j-ccaithpt'ut)t;tt

[332.
is by fiu-t,!m most

333. In n. sunihu- manncr ~c m.~ytrcat t))C

proDcm of th
vibrattons of air inchtdcd bc-twn ri~-id conc~ntric
stj)uj)-ic;d
surfitccs, whosc mdli arc ?-, nnd
For Ly (J3) ~3, if
dl'
vani.s))for thc'.scvatucs of)',

H33.]
When thcdittbroice

l'LANE WAVES.

239

botwcct~ )-, :md

is vcry snudi comparcd with

citttcr.Lhcpnjbtum "tctittftL-siLscH'witttth~Lot'tJx.iviItrationcff~
i
is bustnuh'<diu(k'pt.'udL'Ht.iy.
.sphct'ic~tHhuut,uf:ut',n.nd
ln(l)
as itis vident, tim.tiLntu.st:).!):!2~f-~bc]ndcpu))duttuf;
proxim~t.ciy bu iu t)tc c~ML;
snpposcd, -\vc)):ivu

Thu jntcrv.d bctwecn ttmgravest tonu (/<=i) !Uid Um ncxt i.ssncli

thitt t,woof t,hcniwou!d nuLkca hvuti'Lh(octave
+<ift,h). Thc probtum of Die .sphcnc:d .sliectof g!M will bc f\athcr con.sidcredin
L)iui'uUuwingchapter.
~t.
TI)next ~ppticn.tio)tt)i~ wc st):t.!i n~kcof Litu
.sphcnctU
hat'niuuicantJyMLsis tu invu.st.i~tc tLe (Hsturb~ncc whidi unsues
w])(;)).pttUtuwn.vesof sound hnpingc on aiL
Lstructinf spitere.
Tnkin~ th(.'centre of th sphcre as origi.nof po)<u'co-o)-din:t.tes,
and
th direction frum -\vtiic)itijc wavcs corne as thu :ixis of
tel 6
hetttcpotcntktofthuunobstructcd
plane waves. Thun !e:u'in"'
ont !t.nnnn(!Ct's.s:uy
co)np)excoeincicut, wc havo

nnd (:hc solution cftiic proHon rc<juh'cs L!t c.\p:msiou of c" in

sphcrica! hat'moruc.s. Ua !Lcc(jnntof tho sytHi~ch-y tlic tnn-monics
roducc thonscivcn to Lpgonh'c's i'uncttous 7~ (~), so t)iat wc
may
tahc

\v)iorc J.
:u'c functions of~ but not uf/t. Frorn w))at bas bccn
t!~t ~t, conHidcreda.sa func:drc.tdy provcd wc mn.ytmti(.-ip:Lt.c
t.i<uiof)',m.ustv~ryas

but Lhc;s:imc rcsu]t may f~sHy hc oht;unt'd

dircctiy.

M)))tip!yi)!~

~0

SI'IIERJCAL

C~) ''Y
M, i~.d intcgrating
~=+I,wc~hu)

OBSTACLE

wit]i ro.spM-t.io

f~

fron)

~=

1 to

I" thcpn,L]<niin).an<!
th ~vh<.)umotion
nut.si.L.O.c.spl.cre
Le d.v.)u.) into <wo
parts; th. f,that,
'Y
n.pn.su.h.d ),y<&
to un.li.sturL..) phm.v..s,
.-u.Xhc.s~
c.o,)i,
~s<ha..rc.hn,t~h<
prince
~thusp!.<.rc~ndmdi..Lti,~ct,tofU.. ]aU.r ,.u-L
h~, ~havo
~I~H.i..d
on
(-~
rq))acmg t])c ~oto-a! )t:u')no))ic ,S' bv

~ovc!<~t.ypotcnt~ofthow)m!c
.notion isf.n)n.t Lya,)it,on
r~a,.i
~,t~
c-nsta.ts
hci~do~ni~by
c.n~it.ons. ~)H,scfonn .Icp.nds ..pnn t).c ct.u-actcr of<).c'h<-n.y
t).c ob~-uc
t.r.nh.dbyth.s).hcro.
Tbc.sHnp)cstca.sei.st)~of~"ri.ri.l
an.
hxcd .sphc.rc,anj t),cn Ll.ucundition
to bc sati.silcd whcn 7. lec
i.sthnt

334.]

rnc![D

spHH)m'L

ousAC'~H.

241

At a.su~dott distance from thu source of ()is<,U)'')):mcc

wc m!iy
Ta o)-t)t-)-t.o
h)k~ ~(~r)=l.
to th mitution of i-L~)
probton, wc m:tys(.'par!tt,o thc ruid au<t imn.~in:L)'ypa.rt.s, and
t))nnv!).w:tyth(.'):).tt.('r. (Jti this .sn))posit!o)~thu p!:).nc w~vus arc
)'cpr(.sc))tc'!hy

Cunfhnng our.sch'csfor simpiicit.y'ssa.ko tn p~rts cf spa. a.t ft

<ho sphre, w)u'rc/~ (;<)=], we pt'ocpcd to
~)'(j:Ltdi.st!).ncci'n))n
<xtrt).ctthc rca.t pa.rb of (8). Sincc thc functiona 7~ arc w)io)]y
cvcn or wIioUyf)d<1,

As cxa.mp)c.swe may writc down th ternis in [~], invoiving hannonic. of ontcrs 0, 2. TIie futiuwing tab!e of the
futictinns 7~ (~) wi)) bc uscfu).
1C
L:.tl.

DJSTUBBANCE

DUE TO

f334.

Ihc solution oft~e probicm hcre obtained,

though ann.!ytica))y
is
of
quito gnerai, ]iard!y pmctical use cxccpLwltcn ~c is a sm~
'{uautity. la this case we may aJvatit~eousIy expa.)idour results
JU rising powo's ofM.

PIOID

~d

SPUKRCAL

OLiSTACLK.

343

Itnppcfu-.st!i:tLwhi~ [~,] and [~,] aro of t)iG s~mc orJcrh)

t!.p.srnfJIquant.ity/ce,[~Ji.shvom-dur.s in~ier. Wcsha)IHn<i
prcscutty t)t:it thehighur )):u'mn))iccronponoutsin [~-] dpend upon
stii) nxn-ce]cvatcd powc-rsof A-L'.For n first
approximation, thon,
wc may confinecur~tvc.s to tho ctcmunts of ordcr 0 aud J.
AIthou~h [-J conta.in.sa cosino, n.nd [~-J t).sine, they neverthetcs.s(.liH'min ph:LSchy a. sm:Ut fptit.ntityon)y.
Comparlug two
ofthe values

of~"

in (2)) 330 wc s~c that.

Whcn Mis at fi))high, tftc expressions tan /<:)'nnd /3

very ncady idcnt.ica.)for mojcmtn vaincs f~r.
'Whcn Mis (M,wc gct in !i ncar)y similar mnnncr,

s<bccon')

~44

I\TE~8ITY

OP SMCON))ARV AVAVES.

'i')ieYc]ocit,y-pot('n)iaIof<))cdisturb:u)ce
f'.n.ni:\e(!n"~).'<tit~i\-t~r.j.})r..Aj,,)ith'!v,

[~34.

duct~asm~tngid

For a givcu obst:M-]caud a gi\)i (Hst~)icethc ratio of H)C

n.mp]it~d(jsof t)n; sc~to-cd and th ttircL-twavcs i.sin ~cnen~ p)-oportiorjft!to thc ioverso s~uiu-cuftLc wavc-Lngt.h,and Lhe n~iu of
intenshicH is proportiona! to t)tc inverse i'ourt))
pwcr ( SOC).
In order to compare th intensitics of thc
pruniny nnd
scatto-cd Rounds,we may suppose t!ic former to
ori~nH.te in a
simple source, providcd it, bc surMcient]ydistant (/<') from 7'.
Thus, if

It must bc wc)t undcr.stood tL~t In ordor that

ihis n..su!t in!ty
nppty must bc grcut compiu-cdwith th iinear dunc~ion of
aud must bc grcat
compared witb X.
To find th leadhig term in th
when is
expression for
sma!], wc !)avp in tlic ~rst place,

334.] ]

FURTIIHR

APPROXIMATION.

2455

wLiie if M bc o<td, we hve mcrc)y to replace t" hy

t-esnkbch~thcustiltre:).

thc

Byn)C:msof
(;!J) wc nmy vo'ifytitc rn-st two t~nnsin
In ()7),(t.S). ). T.t))cc:tsc<.t'~=-.(),(:)
(~pressions ir,r[~],j~],
<k)~.sn(jt;(pp)y.

th<:

Ag!un,hy (3)),

C~nbihin~ (17), (18), (~3), (:~t),wc !ubvcthf!v:dnoof[~-j

comptute as fur a~ t!ic ternis which arc uf thc ordur x"c"cuinparud

~C

i'R.SS(JR.S

ON OUSTAt.'LH.

{'334.

witli th two leadiu~ terms givcn ui (21). la

conpounding thcpfu-tialexpressions, it is a.s ncssary to Le exact with respect tu
tLu [jhascsoi' thc cutuponcm.sus with respect, to their
amptitmtcsbut forpurposes rc()uir)')!gody nue !)n)-)nouicclment at ft
titu~
th phase is uften (jf suburdinatc
importance. In sueh cases wc
jnay takc
Ft'om (:H) (,r (~!2)if appoar.s ~hat t)tc
riscs
tuadin~tcrtu iti
two ontur.s in
wlt)) cac]t st~p in the ordcr of the Larmoxic; and
that
is itsuU' cxprcsscd by ;).serins
co)it.:uning on!y evct], or (jnfy
udd, powers of/cc. But bcsidcs huin~ uf hi~hur ordcr in A-c,Utc
!mt)n~ ict-m hectucs )'apid)y smitHur as ?! incrL'i~c.s,ou account of
thu uthur factors wh!ch it contnm.s. Tins I.s
c-vident, becausc fut:Ui values uf ?t and
jf\ (ju.)< 1, thu .sa)nc i.s truc uf
+
\vhiK- t" oniy aiuct.s thc phase.
lu particuiar cases any one of the harmonie clments of
[~ l
may vains).. Frum (11), (12) sinco ~+~
cannot vanish, wc
havu in such a case

tLc f~mo crjuatinn as th.~t which

givus L))c pcriud.s of thc Ylbratious ufort.k'r?! in n.c)o.su.)
)'i).dius c. Aiitticco))s~hcof
sutL'f-ationwtttshuwthat
t))isrc.s)))tmig])thn.(3bccn
(..xpcctcd.
Tho taI))L:(' ~:U is app)ic~b)e t.n
<.)ns(~<stiu)t a!)(t sl~ws, :unon"cUtcrt)!H)~,t)j:tL\))cnA:cissni;t)t,no)iarmotiie(dcn~cntiu[~]
cati\i).))i.~)t.

In conscqucnec of tlic ncrial pj-essurcs th

on
sphnro is f~f;tfi(~
by a force p:)ra)tcl to tlie axis of
w])ose tcndency is to set tLc
Mphci-cinto vibration. Tito m~gnimdc of Dus furcc, if o- bc thc
dcnslty of the uuid, is givc'n by

iii which, by th conjngate

propcrty of Legcudrc'.s functions, oii]y
thc tenn of tlie first ordcr aH'ccts t)tc rcsuit uf tlie
intgration.
~ow,when~=c,

334.]

SOURCE AT FINITE

DISTANCE.

247

which cannot be sattshcd by any rc&lvalue of xo. We conclude

that, if th sphre be frcc to movc, it will alw&ysbo set into
vibration.
If InstG~dof bcing absohtt.clyp):me, the primary waves have
their orig'itiiu a. unit source at a grea.t, though dnitc, distance J~
from tlie centre of tlie sphre, wc hve

SYMMETRICAL

KXI'UH.s~ION

r~S~.

~'ichisthesfuneasifthc
source h~Lconoi
th spito-c.fmd
th point at w!)ic]i t!).' putcnti.d is requij-cd ~t gr~at (hstancc
i.s;n)cx;unp)c<jf-t)K.gnera! rr:neip)c
(.3~8),!U]d
ofReciprocity. J!y
n~m)iing tlic prineip!c,and n~kin~ )t.sc oft)tere.suJt(~) cf'~ :~)S,\vc
Hect).at if thu s..))rccuf tlic
pri)n!uy w:ivu.s bc a fi..itc.ti.sta'ticc
li, thc va)nc of~.e tot.al potcntiat:Lt nny p.,int on
t~.sphcrc i.s

li .4 and 7~ Lo any two

points extcntal to thc spi.crc, a unit
source at ~1 will give th sanic total
putcutial at
as a unitt
.sourceat would ~ive at J. In (.iH.crcase th total
potentiat is
or
two
madcup
parts, ofwi.ieh tf~fir.stis th same as ifthcrowpro
no ob.stac)cto t]tc fn.c propagation of th
wavcs, aint tlic second
rcprcsGnts tlic <li.sturbance<))tc to tlie <J)Kt!tc)c.Of thse t\vo
parts th nrst i.s obviou.siyDie sanu-, whictiever of thc two
points
Lu ~~u'ded as source, and thercfore thc other
parts mnst a)so be
cquaL ti.at is tl.e v:.htc of~- at 7~whcn i.sa .sourceis
to
thc value of~at .1 when is an cqua) source. Nuw e~ua)
when th
sourcu
is at a grt distance
t).e vatue of
at a point
whose anguiar .Hstanc.. fro.n
is cosand iinuar distance
h-ota th centre iM?', is (30)

and accordnigly On.s 1~ aiso tLc value

of~~ta
greM di.stancc
w!icn t!~ .source i.s 7~. Hut
Is disturbancc
since
radiati,~
outwnrds from the sphcrc, it.s value at
finitc
distance 7); n,
ai)y
Le intcn-cd frum t).at at an limite distance
in't.,
hy intro.hci.
cach i~r.non.c tcnn tho factur
Wu ti.u.s obtain t~ fut(~.
luwtng- symmctricn) oxpre.ssi]i

which~iv~tl.ispartufthc
uthct')S!)unitso)))w.

potentia! ~tcith~r point, whenthp

334.j ]

FORSECONDA.RY

DI.STUHBANCE.

2-H)

It shodd be ohscrved tha.t th gnral pat of thc argument

ducs nut dpend upon thu obsta.c!ubunig eiti~er sphurical or ri~td.
From thc expansion of e" ut sphet'ica.1Itarniooics, 'WGmay
(tc'htce t)t:tt of thc pot.eutia.lof w~ves i.s.sumgi'rora a )nut simple
som'cc-'f.fhiitL~ydistant (/') frnm thc origin of co-ordnm.tcs. T!te
potcnti.d at :). p<j!nt 7~:).t a)i itifinitc distance -Kfrom th origin,
:md iu a dh'ectiun making au auglc cos't with )', will bc

f)'omwhidi wc pnss to tlie ca.sc of n. imite Ti*by t)tc sit'uptc inLroductiu)) ofthu i':tci<jr/, (t~~).
T)ms tlie poteot):).)at fi)tit,L'ty(KsttUitp~jint uf :Lmut.sotu'cc
at ~1 is

Ha.ving considcrcd at some icngth the case of a, rigid

spherica! obstacle, wc will now skctcit briuf!y tho course of t))c
investigation ~vlieu th obst:).c)e is gascons. A!t.hough In ail
hatura.1 gascs thc eotnprcssibility is nearJy thcsa.mc, wc will suppose for th sakc of' gcncndity th:Lt t))c m:Lttcr occupying thc sp))urc
ditTcrsin comprcssihiHty.as \ve)l as in dcnsity.froin tttu mdium in
which th phmc waves advanf'c.
!).

is of
Exterior to the Hphcrc, <~ is thc sanie cxac~y, a.)td
t))L'.sfuno furm as bcf'x'c. Fur thc motion inside thc sphcrc, if
~=27r-X'
bc thc internai Wfn'c-k-ngth, (2) 330,

satisfying Lheecudition ofcontimnt.y throug)i thc nti'c.

UASEOUS

OBSTACLE.

r33~,

cxpressu~ re.spectivdy th equalitic-.sof th normal motions and'

of th pressures on th two sidc.sof th
surface. From
thcsc Gquat.ons the coni~Gtc soh.tion bounding
be wor~d eut; but
may
we will hre confino our~.fves to
rinding tite Vfdoe of thc Icadinc
torms, w))o) /<-c,
A:'carc vo-y sina!
lu this case, when ?'= c,

335.]]

CASEOUSOBSTACLE.

25 L

as thc expression for th rno.st important part of thc disturbaucc, corrcspondixg to (~1) 334 for a nxed rigid sphre. Ibt
as inight hve bccn expecto), that thc term of zero order
:ipp(;n.)'s,
is ducto thc variation ofcomprcssibility, and tt)at of ordcr one to
th variation, ofdettsity.
From (13) wc tnay faUhach on thc case of a rigid nxcd sphcrc,
by maMug both o-'aad Mt'inrinitc. It is not surucicnt to makc o-'
by it.sdf infinitc, apparently bccausc, if ); at th same time
rem:t.i))cdhnitc, ~e'cwoutd not bc smaH, as thc investigation bas
assumed.
Wlicu

-w, o-' o-a.rosmal!, (13) bccomcs equivalent to

correspoudmg to <~=c<')S<~at the centre of th sphre. This

agres with th resuk (13) cf 296, in which thc obstacle may bc
ofnnyform.

KQL'AL C'OMPRESSIUILITIES.

t):~5.

Inactu;dg!LSG.s~'=~,an.tthct(-r.nof~)-oon)crdi~ppcars.
If tiic
~as occupyin~ t)tc sphcric:d .sp;tcobG inca)np:D-ab)ylimiter
tti<mt!tC!ut,)iL'rgits,er'=(),an(t

so th:Lt in thc term uf onh.r onc, t)te effect is twicc th~t

of a n.ri()
Lody, a)td )in..st)te ruYurscsi~n.
Tin.-grcfd.L-tis takcn froa two paper.s by
]~:u-tof Uns c))!~ptG)th author "Un thc vibmtiuns
uf~ ~.s contimicd within ri~<!
.sphcricfd cnvuinpc,"funifui "Jnvesti~~ion ofthe disturbancu produecd by ~spi.urica! ohst.adc on thc waves uf
sound' aud from
tho pa.pcr by Profussor Sioke.s
atrcafiy refut-rcdto,
'S'uc~~2'~w<MarehH,

1872; Nov. 1.1, 1872.

XVIH.

CIIAPTER
Kt'HHRICALSHUETS0F AIR.

~OTLONIN TWO DIMENSIONS.

~30. IN former cbnpter ( 135), we s:w tlu~t a proof of

Fnuncr's tlicorummight bcobtaiucd hyconHuleringthe mccha~k's
(~t'n, -vibmting strmg. A sunilar trcn.tmcnb of th probicm of
:). spl]erical shect of air witl. Icad us to n. proof of Luplace's
t-xpunsion for n. function which is nrbitrary ~t evcry point of
a spbenca.1surface.
is thc vclocity-potentia], thc equation of
As in 333, if
contiuuity, rcfcrred to the ordinary polar co-ordinates M,t~kcs
the form,

Whatcvcr may bc tho chamctcr of th frec motion, it cnu

Le iina.lyscd into n. sries of simple harmonie vibrations, thc
nature of which ia detennined by th eorrcspouding functions
thc
considorcd M dpendent ou sp!K'e. Thua, if -<xe"
quation to dotenuilie as a function of and N is

n~y be, it can bc cxpanded by

Again, wh~tcivcrfonction
Fourier's theorem' iu a sries of sincs and cosincsof thc multiples
of~. Thus

Wc hre iutroducc th condition tho.t

Bphnro.

recuis af<cronc rvolution round th

~.J-t

(:H\E);AL

DJFJ.'KJf.EXTJALJ.Tjcx.

r.3:;G.
L.

<).. c.~c.c.nt.s
f,ti,
,f
i
.v~-c
an.) hy
of t),e <ircu)ar ft)nct.ioi]s, cnch
titonju~atcpn.pc.-t.y
<crm of tlie .scnc..s must
.sati.~y thc c~u~~u
indcpcndcut)y
Accot'dn~ty,

is thc quation frf.m wt.ich thc

is to bp
c).a)-ncter of r.)<!<-te.-mincd.Tin.s cqu~bn n~y bc vmtten m vanuu.s
way.s~
Jn tcrms of (= cos B),

\\hen th on~na! onction

is symmctncal with
respect
to the pole, that is, dpends
upon Jatitu.Ic on!y, vanishos, and
the equations simplify. Thi.s case we
.n.y conveniently take
first. In tcrtns of

t hc sofnt.cn of thi.s cqnation In~)vc.s t~-o

arbitrary constants
t.vo
dcHnitc
mn!t.p!y.ng
fnnction.s of
and rnny be ch<ained
in th o.-d.n.ry w.iy hy
as.snmin~ ~n asccndin~ seric.s anc) dct~rnmnng t)if. cxpnnonts an.) cne~cient.s hy .substitution ~rhu~

in which .1 and H arc

tu-bitrnryeo))stantH.
Lct us now f.nt.Lcr
.supposet)..t
round thc ple a~ .sy,nn.e(.-ic.!
(~<.h is ~.n).,g!y

t~.siJcs bc.i~synunct,ie.]
r.sp.ct t: tL
a
th;it

33C-]

coxDiiox

o nj.; sATL-sf.'m;)AT po!s.

255

ere~ function of thc sine of thc )atitudc (~). Undcr thcse circun)stances it is cicar that 7~mnst vanish, and thc vrduc of Le
cxpresscd si)np)y by th nrst seriez, multiplicd by t))c arbitrary
constant ~1. This va]uo of tho vctucity-potnLia! is thc
It~icnJ
of
t)tc
consquence
onginal (Uifc)\;tit.i:dc-quation {uid of thtwo
restrictions as to symn~try. T))c vainc of A'
might appcar
to be arbitrary, but fron what we know of tho mcchanics of t!ie
pru'bicm, it is certain befurehand that /r is reaHy iinnted tn a
sries of particular values. T])e condition, which yet remains
to 'bc introduced and by which /i, is dctcrmincd, is that thc
original quation is satisned at thc po)c itsc!f, or in othcr words
that th pole is not a. source and this rcquires us to considcr
t)ie value of th scrics when ~.=1. SIncc .thc scries is an
eveu function of
if thc pole ~=+1 bo not a source, ncithc-r
will bu th ple /~=- 1. It is (-.vidcutat once that if ~bo of
thc fonn ?t(M+l), whcrc is an even integer, tho scrics termintes, and therefore romains nnitc whcn ~=1; but what wc
no\v want to provc is tliat, if th sries renmin unit
forjM.=I,
is neeessarily of th a.bovc-mentionedform. By th
ordinary
rule it appears at once that, whatevcr be the vahto of /t"
th ratio of successive terms tonds to the limit
and therefore thc scries is convergent for ail values of;u. less than
unity.
But for th extreme va)uc /<-=]f, a highcr method of discrimination is neccssary.
It is known' tliat t!tc infinit hyporgeometrical sries

is convergent, if c+~&
be grcatur t)mn .1, and c1ivc)'r''cnf.
if c+~-H~
bc (~(~u~Itn, or )css t)tan ]. Jn thu Ia.ttcr case
th va)nc of c+~
itttbrds a critcrion of t)jc dc~rce of
(Uvcrgcncy. Of two divo-~ent scrics of thc abovc form, for
w))ic]t thc v:)hu'sof c+~
a.)-c(tiffcrcnt.,t))atot)cis?-c~~c~
mfinitc for which titc value of c +~- a- is t])C smaDcr.
Our prsent scries (7)
may bc rcduc~) to thc standard fonn
Ly taking ~=7! (7;)-]~ whcrc )t is not nssutocd to bc intpf-nJ
Thu.s
nn<i)c''f~'t')))'/n')t''f.<, ]).7f).

('RtTHRtox

OF D!VEnr:CV.

f'33G

Accnrdn~y, sincc c+~l,

tho serins is divergent for
~= 1, ~x/CM ~-)~i';)f~c; and it to-minatus oniy wtien is an
even intcgcr. AVo ~)-c tf.u.s ]cd to thc conclusion that
when
tho pote is not n. source, an<! is an (.ven fnnction
of~u. must
bc of th form ?;.(); -}-1),whci-)t is an cvcn
intcgcr.
In !ihc .nanncr, ~-c juay prove t!~t w])cn is an odd
function
ff~, !Utdth poles :u-c not sourcc.s, =(), attJ
mu.st bc of thc
ibrm ?:(?t +~), bcing au o~(/
intogo-.
If H bc fraction~, both sc-rics arc
and
divergent for ~=~
although a combundion of (hon may bc ibnnd which remfun.s
~nite at eue or othcr po]c., t).crc c.-ui bc no cojnbination
wbich
rein~ns finitc at
p(,!e.s. If thcreforc iL be a condition tbat
"o point on thc .sm-faceof
thc..sphcre is a source, wc hve no
atten.~t.ve but to makc
int~-al, aud cvcn thcn wc do not
.sccut-chnitenosa at tho poh.s ui.Icss we
furthcr.suppo.se ~=0
whcn js odd, and ~=(), wh~n 7~i.s cvun.
Wc concludc that
<or a. compote .sphcrical
Jaycr, t).c only admissible vahtes of
which are function.sofj~titudc
on!y, and proportional to !.arnionic
fmicttons oft))c ti)nc. arc inc]tn)r.d.m.~n..
w.K-rc
(~ is J~cndrc's funcLion,and is any ndd or even
Th pn.ssibility of cxpanding an
'n~e.
functio.i of
I~.tnde in n, .sries of Legcndrc's onctions arbiLrary
is a nccc.ssary con.sc.jnence of w)~t I.~s now b.-cn p.-ovcd. Any
pus.sH~ motion
of thc laycr of~.s is rc-pt-escnted t)tc so-ics
hy

3:} G.]1

TRANSITION

TO TWO DIMENSIONS.

257'

;ind tho vatuu of ~r who) ~==0 is au ~?7~<tr// function of latitude.

')'))<nct])()<tt))at\(!]):)V(']m)'(ifot)(j\vud)ta.s!t)s<)tI)(.ulv:uita.gc
ofpt'uvm~thcet~tjx'atcpr~ncrLy,

who-n Hand ~t!U-c<1H~;)-('))ti))tt\n\'r.s. Furthcfunctions

-P(~)
arc thu ))<t(~ functious(')4)for
thu vibt-!).ti))g System nudm('onsitturation, fuut acct))'()in~)y thc expression for titc ]unct,ic
('no~'y c:m ot))y lHVu)\-).;thu A'<<c.s ut' thc ~Oto'n.lixcd vctociLics.
tf(!~) ttu nut. hotd ~uud, t))C~(jf~<c~'a).su of th \'c!ocities must
L'nter.
Thu vaiuc of i~-npprf~x-iit.te to n, ~/)!e )aycr of vibmting g:).s
cnn ofcom'.sc! bu ()u<)ucc(l!Ls:).pa.t'ticuiar c'n.sc of Die gcnend .sotution :L))])Iic:)b!cto a npherical I.~yor. Cottfmixg oursutves to U)o
c.'LScwitcrc t)H'rc i.s nu source at ttiu pote (~.= )), wc; h:(,vcto mt))C )i)nitin~ fumi nf -~= C'(~),
whcru
VL'.sti~iLte
(~+ 1)=/<
whcn c'' a)t(t /t' an.: infinitc. At th s:L!nc tune
1 and a.rc
infmitcsitta), attd c~ passas into th piano poiu' radiu.i (/'), sn
that ?u<=M'. Fur thi.s purposc die jnosteonvcuient funn ofj~(~)
is that uf Murphy'

shcwing that thc BcssuJ'.sfunction of zcro'ordcr is an extrme case

of Lcg'cndrc's functions.
W))0i thc sphcncal IfLycr is not complte, thc probicm rcrjun'cs :h dif~'rcnt tt-fntmcnt. Thus, If thc g:ts he bouxtied bywa.))s
.strctctting' :dong two paral)c)s of.t:t.titudc, thc compote intgra)
tuvlving two :u'bi.trary constants witi in gc'ncral be necessary.
Thom.'i<mnu<t'nut.'s~<t<.P/f!783.
[/-Bin=.uotf!iitt~P.]
~~)yf;t'2''NN('<))i!,
Ii).
n. t).

Todhuntcr's
~7

~58

VIBRATIONS

0F

A SPIIKRICAL

.SIIEET

[33G.

Tt'c ratio nfthc constants and thc admi.ssihh.

v.~uc-.s of/~ nrc tn ho
~tcnnuicd
:dtLc
LyU.L.,um!ary~.n)itinn.S(.xpn~h~'],;)<.
p:)LmlIu).sin fjucst,i<.iit)tc motion i.sw])(.!)y in
Th~v~u~
]<it,u~
of~ bdns'thrm.g].out
'ncricd~!c.s.s'tnanu~<y,t)ic.suncsarc
:mvay.s

convergent.

If ~c portion of)]ic.surface
Le thaLindu.]~
ocL-pi~])y~s
hchvcont~o
parafK.I.s <.f )atitu.)c ai. c.,ua) distants
from ~hc
~nator, thcqu~iun],cc(,)np.ssi.n)~r,si.jCf
D.cnf.ncarothur.-f
t'.c constants
and 7~in (7~ vani.si.u.s in t).c case of c-adi
nor.ual
lunctiot).
337.

~)'L'nthc.sphcrica)arcac<.]t~int)]ato(tindndMap..]c.
we hve, as
I.i thu case
uft).u.up)utcsp),L.n.,tointr..ducct}.u
a.source. F~th~puq~s~sducond.tionthatthcpulei.suot
tioniMtcr)nsof~i.c.sin<?,wi))Lcnior<convc))iunt.
If wc restrict oursetvcs for Otc
prc-suut to Hic case ofsynunctry
wc liavc, putting = 0 in
(~) ~3G,

Onc solution of tins

cqu~ionisr~diiyobtaiHcdinU.cnrdinnry
way by as.sn.ning ;,n asccndit.g seri~ and
su)j.sLit)tti).~in <).c
.hfturcntiai cq..aLion to dtermine L).c
cxpunpnt.s and co~i~nts
J
Wc~ct'

Dus value of~, is thcmn.st

ancrt so)u~,n of(1), suhjcctto
tl.e condition of (hntcnc.ss wh~.u ~=().
Thc c.,np)<tc .suiutJ(,n
".voh'n~ two arbitrary co.,stant.s prnvi.tc.s f<,ra
snurccofa.tntra.-y
intens.ty at U.c p.,Jc, in which case H.e value
nf~ is innnit~v~n
~=0.
-A.t)y.soiuti<.nw]tic)trc)namsfiHitc-wfion~=()!,n(]mvo!vo<
onc arLitrarycunstant, is t)..i-L.f~-ut),. ,nost
gc.ucra! pos.s:)<)ouudcr
titc ru.sL.-)chun t!,at thc
pulo bc n~L a sou.rc. Aecnn]inrr]y it
uunc.s.sary f.,r o,,r pu.-po.sc to .npjc.tc th suh.tiun. T)~ ..aturc
of Lhc second funcLion
(inv<.)ving a )o~arit].ni of~) wiJI bf. iih.stndcd in th particu!ar case of a
pianc Jnycr tu be con.si.tc.)
prcscntly.
'n('in).'sA';f';f~.h)M(~/f~),'?t,{;28.

337.]

HOUNDED BY A SMALL CFRCLE.

By writin~ )) ("+1)

259

for/~ t.ho so-ius wiH'in ~rac~cts bccomcn

Sioco c + f7 <v = j,)h' so'ics convo'n'cs for al) vo.h)cs of t'

from 0 tn 1 i)K'h)si\'p. To values of <?(= sin'' ~) grca.tcr than ~r
<hcso)ut.i(~nisi!t;)]~))ic:d))c.
is nn int<)', thc scrics bocomcs i~oitica.! wit))
'h('t)
Lc~cn()n''s functinn 7~ (~.). If t])(; iotc~nr bc cven, th scrinH
<'oi'ninn:)h"but,()t))C)'wi.su)'cmaiit.s)))nnit(?. T)tus,w)iC)iM=l,t))c'
HrriL'.si.s i([u))t.ic'al\vit.)) thc <j\j):u)sion uf~, viz. \/(1
in powurs
of~.
'J'hc cxp)'f'ssi<i))fur
in to'tns nf~ mnybcconvononttyapplicd
to thc invc.stigat.iot) <~ft,h(;;froc synuncLncat vihratioo.s of n. sp))crical inyo' uf !m', buuttdcd Ly n. small cit'c)e,vhoHc radius is Juss titan
the quadrant.

TtK; condition tubcsatisficd

i.s simply .~=0,an

(IV
or A:V, :u'c connectcd

uquatinn hy which th possible va)ucs of

with thc givcn boundary Viduc of f.
Certain p:n'ticn!:Lr cases of this proLIem may bc trcatcd hy
mcaus of Lc~'ndrc's f'unctions. Suppose, fur cxa)np)e, t))at 7t = G, Hf)
that /r=/c'~=42.
T)tc con'cspunding suhttion is -~=ytjf~(/~).
Thc ~rcairs!. \'a)))o of /n for ~hic)) '=()
is ~.=-.S3()2, corrcf jl,
= ')0f 37radians'.
.spnnding to <9= 3:)"):
If wo takc r~=?'so tha.t,r is thc mdius of the smfdl ch'c!c
)'nef).surcdaiongt)tcsp)tcr(\wc~'ct
which is tht- f'quft.Lionconncctin!~ thc v:duc "f'/c (=27r\) with thf
eurvp<t radius ?', In th case of :). suiatt ci)'c)< \v)x'.s(;a~gulfn' rn,(ii))s
is :;3" 5;}'. If thc; )ayci- wcrc pl~rtc ( 339), th \-a)uc of K)-wou)()
hc 3'~3')7; so tha.t it makes no pci'CL'))t.ii)!u<]ifrcrcuc<in thc pit.d)
of tlie gravest tono wl)cthcr thc radins ()') of given tengtil h~
T))~r)).<t)ftn
inthc nnit.uf cu'cnhu'mca.snrG.
)73

~GO

UNMVMMHTRICALT. MOTION.

strai~ht, crhccun-<.d

['337.

to an arc nf M".

Thcn~tofM~c~.npar)sonw<nt!<tjHnv~Y<.rJicniah..ri.)ny (tinrent, !fwcwu)-(..t.otak(.

t.hc ~)]gt!i ot' thc circutnfo-cnce ns t)m s:nuc :;) L).u hv.~
cases, Otat
=
is, rcp)acc c~ ?' hy c~ = ?'.
Inord.;rto(]L-ducoUH.sym.not.-i<-aI.s<.)uti.,nr<n-~p]:c
],,ycr
.Us<u,)y~.ssaryt~cinfini)<w),i).c.,nain.s~.itc
On
.~untnfthc
infinie va)uc~r,t),u
.s.ut.i.,nas.su.uc.si)K..simple
)~)!1
a

An ")<iq~)).h.ht!))V..sti~t!<nta)~.so]))ti<.hfot-tI)c
wi)ihugi\'('np)-cs(.)tt)y.
~{S.

~<

,sis<]i~.n.nt

phtnc.praLbm

r~n.

~<).)incnnti!dcquati<t
s!ittMi)c'(H)yt!t<L'()('nicK')tts()f.sii).s'f<),c()H.~u,i.s

'.c~utinnn.n.vh.

').yt~)it!~nufa~.o~.Wun..tinn~.riv.'J
~n.~)
~<n~n~u~~(~n~y~h~
.s
~itiY.i,,t~r.
Th.nu.th~of
n.
..)~
r.\(')~j~hcd)U'(ht).yt))<ji(.<c,,ft)j(.),hm(.);).,

procure

will ).c

UMYMMETMCAL

338.]

MOTrON.

2G11

Wc havo ])o-(! t.!)cco)np)(j<c so)))ti')n t'f

t)~!prf<)!ein oftLc
vibrations fa ~)))(.')'ic;)! taycrof~]~))))))!)'')
))V!tsn)!)])circ)c
\vt)'r!u)iusisI(-<,st.)Hmt))~(jU!t(h-))t.
~o)'(':n-!)\)!H('()t'.s',t.)n'r(j
arc a St't'ten of'p~JHsiUe\L)n(j~uf
~,t)ct.ci'<))i)!~d Lyt)tUt.~)tdi~=<w'Hy()ft.)tcscvduesof~thL;t'nnctmn

un Oif:

)-i_L;ht-)):n~.si<)rof(2),\vI)~nn)u)tipn(.-d t)ycr).s.?M or.siu~M, is~

n~rnmtnxH.~iun ())'<.))<.syst.utn. '.i'!hj:)~r~!tt.(.~)t':dtth<-n<)nn:tt
fu]j('(i())is<'orr<'spn))<]H)g).(jt-vcryfn!))ti.s.si))!uvatncof s :)!)(! ;vith
;m a)'))itr:))'y cocf)k'ic)it pn~xud tu L'acii, ~ivL-s :m
cxprcssio:)
capable (.)t'))(!i))jL;'it)t-t)tific(twit)) t,hui)titi:L! vainc of-i.u.vit.hn
i'mtctiungivu)):u-))itrat-i!yuvL'rt.hc:u-u;).()fUnjM)U!dici)'c)u.
W)[uut.h(;ra(1i)~()ft.hc.sph(!rcci.sini)))it,ctygrc:).t,isiufi))it.(j.
.').!)()~)))L'L'U)ttL"j
Jt'C~=;
~=A~~

:).funcLtunof )- prupnrtiomd to </ (A:~).

In tunns of'/t, thc di)'e)-(jnt..ial
cquatu~tt sa.ti~fiet!by tLc cocfHcicuLofcus&'M,(jr siu&'ct),
id

a.ud su)).st.it,utL- in (~).

is a(x-1);
so tliat
:t

f~

Thu

corfticicnt~ftht'
Inwcstpo~rof
T)t(.'r~)ati<j)) but.wct.'a
:(=(), or ?=).

f'~und by c~)):ttn)g U) xui'o t]~! Ct)'j(tic'iunt. ut'

iH

~2

COXDITIOXS

TO BE SATFSFIKD

[338.

Wuha\~]i.)\vtu})ruvc
th:ttt.).ucut)dit.iunt.)t!Lb)~.it.)H.Tpu]cis
a.source ru<)uirc.st)i;~ M- hua
positive i)it~r,m
w)u'c)i~sc
une uruL).r
ut' thu .suriu.s m titcuxpr~.s.siu.t fur
~t~ninatu.s.
Ft'rthi.s}n)rj)(..s~ it~iii nuLhc cnou~!i Lo.sftcw Lhi~thusurn.~
(u)iI~sturnun:~in~)arL-in)i.ntuw)tun~=il;it~i)jh~s..uy
tu pruvu U~L L),y ronnitt div~r~-nt :d't.t.r
nmitipiicfLtifnt by
ura.swc
(1-~)~
J'):Lyj)ntiL))H)rL;c.)nvutuunUy,t)i:(tt))'y:u'u
ItwHibu
innn[~whL'n~=il'cf~~<<r~?;A(lsuf)iciunt to cousider iit dtail t)iu casu uf t.hu fli-tit scrics.
Wchavc

:~8.j

WIIEN THE POLES ARE NOT SOURCES.

2G3

On th otho' h:md, tlic bmomnU thcorem gives for tlie exn:u)siunut'(l)"*

Sincc s 1 > .~s- 1, it. nppc:u's U)at thc so'tcs iu thc expression
:).ud Lhurcfor<~ aru inHnitiL's of ]ti~ht'r ordur t.)):m (l~)"
f()rurc)naiui)ifmit.u:Lft<t'ntuttipi[(.'i).t.U)tt))y (!)~.
Accordingty
canitut bu finitu at Luth potc.s utituss onc ur ut.itcr of thu .scries
).~)')nin~tu,\v)H':hc:m)))yii:tp))c)t\vI[cn~s'isxL;ro,c).'a.))o.sitiv
L'vc<),wch:Lvestillt.t) suppose/~=0;
h)(t.~cr. If Lhuint.)'bu
:unt it' thc intu~L'r be odd, ~i=(), in order to sccm'c linitcuess at
UtUpufL'.S.
Jn ciLhcr c:)sc t)) value of
put inLo L)tCiur))l

for t!t0 conpietc

sphcre may bc

whcrc thc constant mi)I(ip!icr is omittcd. Tho complete cxpresHiun l'or t)):Lt part of wtuch contains cos~'Mor sniS) as :),factor
li-ithurcfot'c

For most purposes, ]]owcvcr, it is more convcnicnt to group

the tcrms for w)uc)tMts thc s~mc,ratticr tliau thoso for whicit s
isthc!:i:).!nc. Titus for any vainc ot'~

who'e cvcry cocfHcicnt ~1,, 7?, ma.y bc rcgn.rdcd as coutaining a

Lune factor of tfte funn (10).
is ati fu-bitr.n-y function of ju. n.n(~M,and tLcrcforo
ItiLtiaDy
any suc)~ funct.ion is cupalde' of)jeu)~ roprosente') i)) tt)c ff)r)u

2G4

FORMULA. OF DERIVATION,

which is Laptacc's
Froin
(':),itis

expansion

in .sphoical

f'3~8.

surf.LCC )):u')no!)ics.

thc difTurcnt.iid cquation

(;~), or its

c;)..sytop)-ovGth:)t~i!-i

nfthcsii.muforHia.s

t)i:).t.\vctn:)v

g~n(.')-:t) solution
<)!),so

writc

Equation (13) i.s a gcncralixaticn

iunctiunsu.S(j(iin(~!).

of tho propcrty of Laptacc'.s

Thc con-(-H))un(ti))g]-~]at.in))s furthc-ph).nc

probfcmmnyh
()c(h!f'cd,!)s ))tjfo)-c, byat.tachin~an
inHnito va)nc! to ?<, wh~h
in (1~), (14.) is :u-bitra)-y, am!
= 1,
+
writing ?~=
Smce

~n bcing rpgnnh~ as a functitm of n. In Ujo ],it.

(~cn
()t'))]~h .s))))JL-etta ')ii'['cuti;Lt.iu!t)]nayl)eidu)ttitiL-dwit.)tt))tity,
n)idt!mswujnaYt:(.ke

\\hen ihc poic is not a source,

is proportiona) to ./(~-).
T)'u constant c<)c)!)cit.'nt,,)cft. utxtctcrmincd
hy ()5), )n;)'v ho
ru:uj)fy foundby
aco]))p:u'isun ot' t)n: i<utin~ ter)n.s. iLthu.'j
:U)peartith:Lt

a wcil-known propcrt.y ofUcssc!

Tho vibrations
casilydcaitwit)j,than

.s functiuns'.
1.

of a plane iaycrofgns
iu'cofcou~c
t))usuofaJay(.T<)i'ih)itceurvatu)-c,Lut,
T(~hn))tt.')'<~f~~r<<t;(');

;)!)().

more

VIBRATION

:~38.]1

IN TWO DIMENSIONS.

2G5

1 h:tvf prcfcD'cd te cx))iLit thc indirect a.s wc)! as thc'. du'cet

!netht)(1nfit)vc.sti~!(,tim),h<)t.hfo)'t.})CS!)1\C()t't.LcHphut'ic:))pn)b)cni
it.sutfwiUtthccorn.'spontii))~
Liiphtcc's explosion', :ux) bucause
thc cotmcction ))(;t,wucn I!c.sHct's:t.))d L:tp)iL(;c's functioosappeiu's
nut to bc ~oto'n.Hy undL-rst.out]. '\V<j tnn.y !)o\v~howuvur, procucd
to t))C in~(.!f')i<)L!nttrc:).ttnLiuLof t))C phuic }u'ol))u))).
33f). Ifi't. thc gcncml cqna.ti~n of simple aermi vibraduns

This quation isofthcsamcformasthat.'withwhichwchadto

d~d in troating ot' circult'tnoabraocs
(~ ~00); thc principal
]na.t)icmatic!Lt dit~crutTUCbctwcun t)i0 t\vo questions IIus iti tlic
tact tliat w)ntu m thc case of jnonbram.is thc cuudttitja to bc
sati.sncd at tlie botun)ary is '==0, i't H'c prusunt case intct'cst
attaches

itsuif ratticr to the boundary

spondmgtotitc
cnvclopc.

couinement

coudltio)i

ofthcg~sby

<

'=0.

ci'rc-

ft.rigtdcylmdnca.I

Tlie polo not bcing a source, tLc solution of (3) is

1 1 )m\'o hccn t))U(;h assista L.y Hcinu's /~M~tt;/t ~<T7~),(M~

Derl!n,
1801, und by Sir W. Thonison'H papurH ou L~ptaeo's Thuury of tltu Tides, ~<t/.
~Yot.)v.m75.
2 1 itre rcou- to tho HS)m)
natation, but tho rG~lnr will undot-stand that n cnrrf"~pnndHto the uf prcccdin~ Hoctio)~. Th )t of Laj)tacL''f)fnitctious isnnwmfthitt'.

~GG

HIGID CIRCULER

BOU~DARY.

[339.

Thc !owc;r vahu\s of /< satisfyin~ (.)

a)'(; ~ivcn in thc f'uH~witig
t:d))u', whidt w:)..sc:t)fn)a,L-tliront iL'm.smt'.st~tjlc.soi'thu ruocti.tu.s
t/ by tnn;u).s uf <.)tc ruht.t.lun~
tu bu ~)['cSHcd in LurutS
:d)<j~i)tg
of ~and
fsntnhrrofm-

h'rnn) circu)!H'))~(h'S.

tt-0 p

-/t 1

0
1

:}-.S;!2
7' ~i

t-.S!l
5-:i~~

I()')7~

S-)C

3
4
3

l:t
I(:')7[
(1
H'-Ci':

H-7<)<!
]t-.S);j
].s'u!(;

M-~2
:)~.t
C-7"~

i't-.33
.j.~oi
8'()L~

u.U!;5 ll~).i

Thup:n'ticuhu'm)htLiuH)i):~yLuwnt.t.ut

whcrc ~1, 7?, 6', D :u'c :u-1)itnu-y j.,r

cvcty !utn)issib]c value of
M.a)id/<
As
i'iLL)n'um'.spum}In~pru])]c)n.siurtItosphut-en.))<t
)nust.
uIrcni:u-n)L-inbranc,Lhu.suiti.)t':tHthcp!u'tic))int-Hu)utK)tis
bc ~ctio-id enou~it tu rcprcsoit, w)icn
<==U, arbitmry v.ducs of
and
As an exampJc of conpound vihr~tio~s wc
m.iy suppose, as
iu ~32, that thu initit coudit.ioi ci' thu gas is t)):~ dchtiud by

'X..trs~t)tc.s~r.sL'nuc~u:

/))')/,<

Kuv.l~.

339.]

CASE

OF

COMPOUND

VIBRATIONS.

2G7

an C()U!tt.ionwhich may bc vurificd m)tncric~))y,or hyan a.n:))yi)t t))u casuof()'~ :{:j.2.

tic:d pt'uct'.ss siniit:n'to tit~tapptiu~
~Vu)nny}'t'uvuL))aL

Frnm this (12) i.s d(;rivnd by pntt.in~ ~=1~ and ha.ving'regard

to Liiu i'undameut~l diit'L'rott.iaIcqua-tiuu s~tisticd by
which
shuwstii~t
IIitherto Avchavc supposcdthc
that
cyl!ndcrcomptetc,so
rucm's ~ft.cc cach rvolution, witich l'cf~Hrcs that
bu mtL'gnd
but ii'instc.ut ot'titu conphtc cy)indur wo ta]\c thc soctor iuciudud
tjut.wcoi 0=0 ~uiJ ~=/3, ft'i).ctiun:Llvtues of~ witi lu gcuul'a.i present t)iemsc!ves.

Siuco

vanishcs at both IInuts of

must

beofthefonu
whcrc !=!7T~f
beinguttcgnu.
Itpbu~nahtptot
partot'
7r (oi'Tr itscif), t))<jcom~!utc so~utiou mvulvcs on)y mt~nd values
tjf
as tni~ht )i:).vc Lucu iurcsL'ot but, iu gnera)) fnuct.tuns oi'
tractional urdc).'must bc mtroduccd.
Au ititct'estuig cxampic occurs \Y!)cn ~=27r, wfuch corrcspftuds tu t.))c casu of a. cytindcr, travct'sc() by a rigi't watt

2G8

ANALOOOUS

PHOHLEMS rOR

WATER WAV).

[339.

st)-f;tc))i))g f)o))i thc-o('ntt't't.()t])(-circni))fum))(c<))tip.u't'2n7).

T))L-L-ih'<'t<)ft))c~-atiis<.<)n;)).)~)-j)('ssi))!c:t()ifr(-)'(;))L;(~)f}))-t'ss)))-(i
no.such (Hff't'~tjccoccurs, thcwa))
un]tst.~omdns;h)tt,w]t(_'u
!n:)y ))C ]~nYovu(), ~nd t)<c vibrations itruinchnh~) undL-r tLn
<.)K'.)ry of :L cu)u]))(.h; cyHndt..)'. This .st:t<c oftLu~s
oconr.s
~L~tt~i.scvc!).
But wttcn~is (.ni.),
~is<)rt.h<'f.))-n)(int~(.t-+~),
anttt)i(.!])r~su)~s (.))<.))(; Lwoni(tt'.s ut t)tt'wa!):u'<<)i)1c)~nt.
ht
())'' ):)Lt.Lct'('iLSc~ is cxprrs.sihh- in finimtunn.s.
T)n; gr:).vc.st
(.uHuisubta!)K!dbyt.:Lkin~~=J,orM==A,whL!ti

Tho
""<)t])cadniisHih)o\ah)t'S()f'~a)-uthn)-outs(jf<:)n~
fu-.st. rout, (;dtur /<-=()) i.s K=.i-l(!
curn~jx~htin~ (n a <))(!
~L-ci()c(I)y~nn-crt]):))ia))y<)n(-,of\\)ticht)n;c..)Utp)ct.ur\)Ht(tt.ri.-j
C!)p!Lh)c.
Ttie prcecding

an:))y.si.s bas an intt'cstit)~ application to

thc
~n!Lt,))e)n:LticatIy:m:).)ugut)n])rub)uniuft.hu vibrations ut wa.Lcr
in HLcyHndricat vcssc! of mn~nn
(L-pth. T)tc r~dut- may
consulta ])apcr on wavc.s hy t)tc- authur in t])c
7V7~)/<r</7
fur
~/f~c
Prof. (it)())ricto\vhi.t
Apri),l87<J,an()papGrsby
rcfcrcnccisthL.rc
!na<)c. Thc()1).s.;rvat.iu))ot'thcpcri~]icti)n(j
is v'')-ycasy,!mtIint)nsway)naybc()htai))L!
ai) exprimentt
so)))t.iu)iut'pn)b!e)n.s,wl)u.su thuoruticnl trcatincntis
farbL-yuntt
tlie power uf knuwn tnuthud.s.
340. Rct.)n-ning to Utc compL'to
cytind~r, Jc-t vts sopposc it.
closcd by rigid tran.svcr.se waHs at ~=0, and ~=/, and rctnnvu
thL')-u.sti-ict.i()ntIjatthun)ut,I),ni.stu))(jt])c.s:unuina!i
transvo'.su
suctiuns. Tito gtjnerat dii'crcntia)
c<)ua.tlon ( H) is

whf'rctitccocfifu.nts

Tiusfonn
7~ mayhc fonctions of?-:ut(~.
su(-ur<jst]tcf)[tfi!)nc))tr.f't))L:1)u))))d:n-yc<jm!iti())).s,w])(.'ns=0,~=/, I,

3K).]1

VnmATOXSIN A CLOSRDCYfJNDHR.

2G9

.'m'! c:n'h tcr)n)~n)st sati.sfy thc difrt.')'c))ti:dc(p)nt.innHCpar!itc;!y.

Ti.n.s

winchis<'f'<]h'.S!mt(!fo)'tn
s, ~)H'i))~')'t.')'i!n'r'thy/<7:
t))urt.'t'(j)'cbcwrittcn

as \hc)t).h);

motion isuxit'pntiuntof
'i'))~p!U'ticuhn'MnhtLit))t ))'ay

w)u<'h)))u.st,l~<'t)c']':))ix<'(thy:),t.)-i}))os)t)n)t):tt)<)t),W)th respect to
:< intc~'r:)) v:)hn's uf'~ ;t))(.) )),;))) id.so \ith rL'.spccttoati U~
V!).h)('.sut/f,(.tL;t(.')')uiHL'(thyt))CL'')uati<))),

Th pun')y ftxi:d vibrations curt't.'spondto n. zcro v:).[ucof 7t",

nf)tinch)')cditit))ctabtc.
3~L Die contplutc intcgnd of thc cqun.tio)!

w))('nth<')'(!isno)i)nit:)<.i~))!t.s
to th .thscuccof:),.source.'t.tt!x'
))~)(',mvo)v(,'s!),sccotit1 fonction ()f?',whichm~y))o<)c!)ot('(1
by
T)'"s, omitting unuccL;ss:u'y con.stiUtt multipiicrs, we may
'(~')takG(2()0)

Whc'n
bntt))uH('c')ru)s(.'ric-s]'c()uii'c's)n<)dific:tt)0)),if))bc!i)ttc'gr:d.
/<=(), <hc hvo sries bccom ith-'utica), nnd tims thc imnic~ia.tu
Th~
)'~utt~t'.s))))p()si))g~=()in(2)t:Lch.sthncccss!U'ygcner:Uity.

270

C!HRAL SOLUTION.

['341.

rt'tjnit'cd.solution )nny,])nwt'V(')',hc'<'Lt:)i))(''t1Ly{hcnn1!)):t l'y )'utc

ap))]icah)'su<'hf':L~s.
Dt'))~tn)~t])t.: couOici~nts ot'~tfu~i
in (~j by .(/'), /'(- '<), wu ]):L\'c

:i.]]

HXt'HH.SSION BY DESCEND)~;

S)':I{')KS.

271

Tho fu)-m))):L<)f(]t'rivat.ir))t (.~tnnyhc

oLCuxcddirr'o~yfrom
()tcditrm'cnti:tt<([)t:)tiun()).
~r!t)))~))'/<')':Uhf;)Ht.ti)~

which is equivaicnt.to
mb~t)t<(~ta)i('n.s.

(.')), .sincc thc constant.s in

iu-earbitrary

Tin' scri:Lt (.'x)')'ssujns for

thns obiaincd arc cf'nvc)'<nt fnr
:d) v:duc.s of titc argument, Lut arc practic:J)y u.s<)css whcn th(.'
ar~mount i.sgruat. Jn suc]) cases wc tnxst. hve r~course to .suuuconvur~cut scncs cort'c.spondu)~ to <)):tt of (10) 200.
Equation (1) may he put into titc fonn

DIVHiKiHXT
WAVE.

f:~n.
L..

Whcn H is intc.L;-)'a),(~c.so sries !~m infimtcand

u!ti.natc]y
<).ve,ut,
hut(~~)().
;~)thi.scircu.nstaucc
dsnot interfre
Wtt])t)K'H-pr;t.ic;d)tti!i)y.
J.s toThc)))u.sti)np.j)t:u)t,)p)ic:ionnft])ccu!)tpl(.tcl))tp~ra]nf(~
rq~.nt.h.sturt,an
.!ivcr~n~frn.nthHp]c,
~prnh)c.
w'ch)n,.sbt.ntn.at.n,ySL.,).sin).i.s.n..moironth..(.on.municanon ~vth.-ati.s<.0!s.
T~.cu.~iti..nL).;L<,t).cdi.sturhan
n.)~s..ntc.)
by (1;~) s),)) Lu ..xdu.sivt.jy divergent is
.sin.piy
'~J'i~a~or
e
CiKnl
l Y yll 1
l)C 1'l',1
Ly.s..ppn.si.~?1I In5111~r
9' tOhevu.-y
'3 ~t;
,~t'c~ut;
m
tlu; 1)rilmil~;)1 1 1.
t).upnn<.ip~<Hfncnity<,j-'
fI'
l uf
t.c q.st.un
(.sist.s
H, .]i.s(.vcri.~w)~tr..)atiun
b.jtw~. th
f-o<fhc.~nts.-i't)~ as~jh~
.sc.riL-.sc.)rn.s)H,,).Lst.)t],is~n.].ti..n
~.r~nc).purp..scSt.sc.n,p)..y.st)n'.su!.t(i.muf'())i,.tL~m-.u
ofa. <.h.f,n,t. !nt~ra].
Wc.shaDattaiu
t!.c.sa.nuobjcct,purInLp.s
t~oro.S)).)p]y,1,ynsH)~t))cr~))]t.S.jf~0~.

-md thustho qucsti.,nn.dncc.s itsdft.uthc

dtermination ofthc
f.'nu of tlie r.~).t-ha,n) H.u.nber .,f
(1.~ wi.c.i is ~n~H. By (:)
;~and
(.) 200 wchuvc
~)

+~.(~)}

=~+~'7r+ higher terms in~(l.-j),

s~U~taHth~rcma~sistufmdthc~nnofthcddn~cintc~~t
ni (l- ~hen ..s s.na)i.
P~ti,~ ~~+~=~

{~~

341.] ]

DIVERGENT

WAVE.

273

whcre y is Euler's constant ('5772.); and, as we

m~y casily
tiatisiy ourscivcs by Intgration hy parts, th other Intgrais do not
contnbute anythmg to the Ic-a.dingtemM. TIius, \vltcnz is
very
small,

and compa.rmgwith the form assumcd by (4.),

Reptacing by
when ?'is sma! wc sec tliat in order to makc thc sones identicat
we must tako

so that a.sries of w;),vesdtvo-gmg from tlic ple, wLusc

expression1
in desccNdingsns is

In applying th formula of drivation (11) to tlie

dcsccndiug
series, the parts coutaining e- and e+'~ as fa.ctors will evidently
romain distinct, and tlic complete intgrt for th gnral value
of ?, subject to th condition that the part containing e+" sha.11
not appear, will be got by diferentiittion from tlie
complete
intgral for M= 0 subject to the same condition. Thus, sinco
by(5)~=~,

274

SOUNDINGBOARDS.

[341.

or, in tcrms of tlie asccnding series,

Thcso expressions arc apphed by rrof: Stukcs

tosbewbowfccbly
the vibrations of a string,
(corrcsponding to th term of order
onc), arc cotmnunicatcd to thc surt-onnding gas. For titis purposo
ite makcs a comparison bctwccu th actuiU sound, .md wh:~ wonid
h.ivc been onittc-d in thb ~mc direction, wcre thc latral Motion
of thc gas m tlie ncighbourhood of thc
stnng prcveutcd. For a.
piano string corrc.spondiog to tho middic C, thc radius of t)ie
wirc may be abont-02 u.ch, and is about 25
inchcs; and it
appcars that t)tc sonnd is ncarty 4(),()0()timcswcaker than it wou!d
h.ivc bcctt if t))o motion of thc partic)e.sofair )md taln
place m
p)anL-spassittg tbt-ough thc a.s of titu strittg. "Thi.s shews tho
vit:d nuportance uf .so)t))(]ing-horn-d.s
i)t strmged instruments.
AIthough thc amp)itudc of vibration of thc pin-tidus of th sound~
n)g-board is extrcnK.Jysma]I compa.rcd with tha.t of t))c partic)cs
of thc string, yet as it prGMcnts
a broad surface tu tho air it is able
to excite loud sonorous vibrations, whcreas wcrc the
string
in
an
supported
absotnteJy rigi(t manncr, the vibrations whici)
cou)d excite directiy in t)tc air w~u)d bc so .s.nal)as to be abnust
or
altogcther Inaudibic."
l''it!'!t.

"Thc incrcase of soun.! pro<!ucc<I thc

of tatcr~
hy
stoppao-c
motion ,~y Le prcttilycxLihitct]
hya vory .silnpie cxpen-.ncnt
i~ke a tuuing-fork, and holdmg it in thc
fit.gcrs aftcr it bas bccn

341.]

SYMMETRICAL DIVERGENT WAVES.

275

ma.dc to vibrato, place a. sbpct of paper, or th bladc of a. broad

knifn, witb its cd~e par~tiu) L~Lhu :t,xis of {.hc fork, and as ncfn' to
tho fork ft.sconvc'DKjnUymnybcwithout toucjnng. If thc plane; of
th bstac)c coincide witit eit.))Ht'of thc p)ancs of symmetry of thc
furk, a-s t'cpt'csL'titcd in section at ~1 or j~, no effect is produced
but if it bc p)accd in au into'mcdia.tc position, such as (7, ttic
Sound becomos mucb strougor
~42. Tbo rcfd expression for the velocity-potential of symmctricn.1 wavu.s divo'ging in two dimensions is obt~incd from (1M)
~-H aftcr introduction of tbc time factor e' by rujccting th
imaginary pa.rt it is

in which, as usu;d, twn :n'hit)'aryconstants may hc InscTtcd,one as

a. mu!tip)ier of thu w)))c:cxprcs.siouand th othcr as an addttion
to thc thnc.
TitG prnDcm of a. lincfu'source of uniforni Intcnsity may f).1so
hc tt'e:itGdby the g'cncnd )nut!)')d~ppliciL~tcIn thrcc dl;ncnsions.
Thus by (3) 277, if p bu thc distance oi'any ctcmott ~.<;from 0,
thc point at which t)te potoitin.! is to hc cstimatcd, and r hc thc
smaHest value ofp, so thut =)'' +?' we ]nay tal<e

from which thc vfirious cxjx'cssions fu!!ow a.s in (14) 341. When
~r is grcat, an nppn'ximit.t~ ViUueof t!ic mtu~r:).l may ue obtaiucd
sinco on [(.ccount of the
by nc~ecttng titc vm'iatioa cr \/(2r+y),
wu uecd attend
rapid Hnctuntiou uf sign caused by th factur e'
r/(t!. yra'

18C8.

183

276

LINEAR SOURCE.

[342.

as tlie value of thc vclocity-potcntial at a grcat distance. A

.shnilar argument is appHcabIe to shcw that (1) is aiso the
expression fur tlie velocity-potential on one sidc of an Innnite
p!anc
( 273) due to thc unifurm normal motion of an Inmutcshnal strip
boundcd by paraUel Unes.
In Ii!~ mn.nncr wn )n:ty regard thc tcrm of t)ic first or<!er
(20) :~1 as the expression of the vclocity-potentinl due to double
sources uniform)y distributcd aloug an infinit strai~ht linc.
FrMn tlie point of view of tlie prsent section we sec the
winch f~ppcars in (1) and in
si~ificancs of thc rctardation of
the results of thc foUowingsection (l(i), (17). In tlie
ordinary
intgration for .surface distributions by Huyghcns' zones ( 283)
th who)c effuct is tlie lmlf of tliat of tlie nr.st zone,and the phase
of thc cnect of the first zone is midway between the ph~ea due
to its extreme parts, i.c.
behind thc phase duc to thc central
point. In thc prsent case tho retardation of th rsultant rcla.tivcly to the central clement is less, on account of thc preponderance of the central parts.
343. In illustration of tlie formut of 341 wc may take
the prohicm of tho disturbancc of plane waves of sound by a
cyhndrical obstacle, whose radius is small in comparison with
the Icngth of the wavcs, and whose axis is paraUel to tlicir
plane. (Compare 335.)
Let tlie plane waves be represented by
Thc gnral expansion of in Fonncr's sries may be readily
uSeetod, the coefficients of thc various tcrms being, as might

343.]

CYLINDRrCAL

OBSTACLE.

277

bc antieipa.ted, simply thc Bcs.scI'Hfmictions of corresponding

ordcrs; but, as wc confine onrscivcs ho-e to the case wLere c
t])Cmdms f thc cyliudcr Is sma.li, Avewill at once cxpa.ud 1)1
powursof 1'.
Thus, when )'=c, if e" be omitted,

Th amount and cvoi thc la.wof th disturbance dpends upon

t]tc c])ara.ctet' of thc obstacle. We will bc~m by supposing th
mn.tcrial of thc cylinder to bu a. gas of dcnsity o-'and comprcssibility ?/6' thu solution of tbc probtem for a, rigtd obstacle may
finaUybe durivod by suitu.btc suppositions with respect to o- ?/t'.
If K' bc tli mtern~I vduc of K, w have inside th cyliuder by
thc condition that thc axis is not a source ( 3~0),

278

CYLINDRICA.L
OBSTACLE.

[343.

Th tact that
varius invcr.sc)y .as \"S might Ijave been
anticipatcd by thc motitod uf duncn.sionsas in thu con-espondmg
proDum fur the sphcrc (33.-)). As in that case, thc synunctric:3
pfn-t of tlie divcr~nt w:L\-cdopcnd.s upon t)tc van~tiuu of comp.ssibiJity,and wouht di.s:tppc;n-in thu application to an actua.1
gas, and tlie turm of the first ordet- deponis npon tlic variation of
dL'nsity.
By snpposin~o-' and 7~' to becomc innnitc, in sud) a manner
that their ratio rcmaius fiuite, we obtain tlie solution corresponding to a rigid and iM)novcab]cobstacle,

Thc aualysis of this section is appUcaDc to th

matlicm~ticaDy
anaingous problum of H)tding t!ic ctt'ect of a cytindrica! obstacle

343.]

PASSACiH0F SOUNDTUROLTGIIFABRICS.

279

on plane wavcs of transvo'sc vibration ni an clastic solid, thc

directioti of vibration bcing p:u'aHcl to tlie :txis of th cyHndcr.
If tlie dcusities be o-, o-' and thc ri~iditics be M, and Ydnote
thc tmnsvcrs disptacctncnt, thc bound.u'y conditions arc

Fur an application to the thcory of light thc roader is rcferrcd

to a papcr by thc author, 'On tl)c manufacture aud theory of
diffracLiungra.ting~
T)tc cxcceding smaUness of th obstrnction nncrcd by fine
wh'cs or nbl'es to th passage of sonud i.s stt'ikin~Iy iHustra.tcd
i)i some of Tyndatl's cxpcrnnonts. A pice of stUF fcit hait an
iuch in thickness allows mucit more sou)id to pass than a we~erZ
pockct-handkcrchicf, which in consquence of tho ctosinf of
its porcs behaves rather as a thin lamina. For the same rcason
fogs, aud even raiti aud snow, interfre but little with th freo
propagation of sounds of moderato wave-Iength. In tlie case
of a hiss, or other very acute sound, the cilcct would perbaps
be apparent.
P~t<1/y.Vul.xn-n. 187.1.

C1IAPTER XIX.
FLUID FRICTION. PRINCIPLE 0F DYNAMICAL8IMFLARITY.
344. TiE quations of Chapter XI.aud the consquencesthat
wehavcdcducett fromthctn arc b~scdupon thc assumption (230),
that the mututd action betwccn anytwo portions of fluid separatcd
by au imagina.rysurface is normal to that surface. Actua.1Sui<tn
howcvcr do not corne np to thi.s idal iu many phcnoinona thc
dcfcct of Huu)it.y,usually callcd viscosityor ftuid friction, plays an
important a,)idevoi a prcponderating part. It will therefore Lo
proper to inquire whctijcr the laws ofacri:d vibrations are sensibly
iunuenced by tlie viscosity of air, and if so in what mauncr.
In order to understand clearly the nature of viscosity, let us
conceivea nuid dividcd into parallel strata. in such a manner that
wItHecach stratum moves in its own plane with uniform velocity,
a change of velocityoccurs in passing irom one stratmn to anothcr.
TIie simplest supposition which we ca)i makc is that thc vclocitics
ofa)l thc strata are in tho sanjc direction, but incrcasc uniformly
ht magnitude as wo pass along a linc perpendicular to tlie planes
of stratification. Under thse ch'cumstances a tangential force
betwccn contiguo~tsstrata is caHnd into play, in the direction of
the relative motion,and of magnitude proportional to th rate at
which the velocitycitangcs, and to a coeflicicut of viscosity,commonly dcnotcd by the lutter
Thus, if the strata bo paraUcI to
a'~ and t)ie directionof thch' motion bc pamUulto titc tangcntia!
force, reckoned (iikc a pressure) pcr nuit of area, is

Thc dimensions of~Marc [J7Z''7~'j.

Thc exa.minatiotiof t!)c origit) of th tangenti~l force
Lclon~s
to motecutar science. It )ias bee]i exp!:uncd by Maxwell in ac-

345.1
~j

FLUID FRICTION.

281

cordance
with tlie kinctic theory of gascs M resulting from mterof
change of molcules between thc strata, giviug riso to diffusion
mon~ntum. Both by theory and experiment th rcmarka-ble
conclusion haa been esta.bliahed that within widc limits th forco
is mdepcndent of tlie density of th gas. For air at Centigrade
Maxwell' found
tlie centimtre, gramme, and second being units.
345. Th investiga-tion of th equations of nuid motion in
which regard is pfdd to viscous forces c~u sca.rccly be considcrcd
to belong to thc subject of tins work, but it may bc of service
to some readers to point out its close conneetion with th more
geMera.llyknown tlieory of solid cla~ticity.
Thc potential encrgy of unit of volume of uniformly stra.incd
2
bc
mattcr
exprcssed"
may
isotropic

in which 8(= e +/+~) is tlie dihitn.tion,e,~ < a, c arc tlie six

Mtua,!displaccmeuts a,AY
componcnts of stmin, couuceted with tlic
by tlie quations

of which M mcasurcs thc n'yi'~y, or i-Gnist~ccto &7~(t?- a.nd K

mcapurcs th rsistance to change of ~o~t?~. T!)C componeuts of
a, &,C,
stress P,
r,
corresponduig rcspectively to e,~
are f'jund from by simple ditiercutiation with respect to those
<l)innh)t.)f'H
thtIS

1 On tho Viscosityor Interne Friction of Air nnd other Gases. Phil. 2'rM.
18CC.
T'oxo~/ty. AppondixC.
Thomsou and Tait's ~<f<xr<~

S 82

EQUATIONS0F MOTION.

f345.

If ~Y,F, Z bc thc component.sof tlic applied force reckoned

pcr
unit of voknnc, tlie equ~ti~ns ofequilibrium arc ot' thc form

from \v1)ichthc cqu~ion.s of motion arc

immedia-telyobtfuna.bio
hymens oi'D'Aionburt.sprincipIu.
In tenus uf tlie di.sphtcctncut.-ja,
thcsc c<{U!tti'jus
bccutnc

1)1th ordin.n-ythcory of ihtid friction no forces of restitution

!n'C!lnch)ded,b)ttont)motiiCt-!]a.)idwe)ta.vetoconsiderviscous
forces w))oscrc~tion to thu vu)ucities(,M) of'thc nuid cicments
is of prucisctythu s~mu c)mracteras t)t:).tof th forces of restitution
to thc <)iHp)accincnts
(a, ~3,~)of :m isotropic sulid. TiiUHif S' bc
tlie vetocity of dittt.ta.tion,Hothat

Sofaj- x~)d?!f).rcn.)-bitr:uy constants; Lut

Ititasbccnar~ncd
wit)i ~-C!ttforce byPruf. Stukus, th.-tt there is no rc!t.son
wt~ya,
motion ot'(ti):).tatioau)ntunn ni al! (.)i)-cetio)i.ssLou)<t
givcriscto
vi.scunsfor, 0- c:u)su tho prc.ssut-ctu diffcr from thc s~ticat
pressure-correspond)!)~to t!tc actu~
density. In a.ccordnncewit)i this
iu'gmncntwc :u-cto put /t=0; and.as ~ppcars from (C),~ eoincides
witit th ()u:uitityprcviou.slydenoted by
Th frictions! terms
a.re thcreforc

345.~

rr.ANn WAVES.

283

or, if thcrc bc no n.pplicd forces :ui.d tlie squa.rc of thc motion bo

nc~ectud.,

cotisidered
Wcmn.y observe ttm.tthcdis.sip~tivcfurccsiturc
correspond tu :). dissipation fnnctiuri, whosc furnt is thc .santCwit)i
with rospect to a, /3, y, i)i ttie ttiuut'y
u,?u vs tlt:).t of
rL'spcct tu
ui'i:jotropic soiids. Thus puttin~ A:= 0, wc Itavc frotu (1)

in forcement with Prof. St.okcn' ca,!cu!a,tion'. T!n.: theoryuf friction

ib)' thc c~sc of n compi'cssibio Huid was first given by Fuissur~.
Wc will )iow app)y tlie diiei'e;ntl!U quations to thc inin'e
vcsUgft.tiun of pt~nc wa.VL'sof sonnJ. Suppo.sin~ t.h:t.tv a)id
&c. arc functions of ou)y, wc obtit.in from
xo'u a.nd tha-t n,
~13) ~
:~fi.

which is thc cquatlo~~gtvcn by Stokc.s~.

Lct us now inquu'e how a, traiu of harmonie waves of wavcwhich are inanitcuncd at tlie ongni (a;= 0), Me a.way
Jcugth
18:')1.g~9.
Com~rf'f7'M;);!ffc<tnH.<,
JoNntff/ </e <'A'co/<'~/)/f<'c'7t)tt'~)', t. xni. cah. 20, p. 139.
C<tMi')')'f~<'2'<')t.i<fc<<('x.184:

284

EFFECTS

OF FRICTION.

[346.

as .c iucrcascs. Assuming tliat M varies M e' we find as ia

14.8.

lu tho application to air at ordinary pressures ma.y bc conHidcred to bo a vury smaM qu:mtity and its square may Le
ue~lected. Thus

It appca-rs th~t to tilis ordcr of a.pproxima.tiontlie vclocity of

sound is unnH'cctcdLy Huid friction. If we rcptuce M by 27ra\
thc expression fur the cocfHcicntof d(jc:t.ybccomcs

s)icwM)gthat tue inimcncc of viscosity is greatest on the wavcs of

short wavc-)L'j)gth. Tlie :unplitudc is ditnhus!tud iu thu ratio
C 1, wlicu x =f" In c. O.S. mca.surc wu may take

Thus the amplitude of wavcs of one centimtre wavc-Icngth is

diminishod in the ratio e 1 after travc)IIng a, distance of 88
jnctres. A wave-lcngth of 10 centimtres would correspond ncarly
to
for this case a; = 8800 mtres. It a.ppe:u'stherefore thu.t at
atmospheric pressures t))e influence of fricLionis not Hkdy to bu
sensible to ordiuary observation, cxcept nc:).rtite upperII)nit of the
musical sca)e. 'Die mellowingof soonds by distance,as obscrved Iti
mountainous countrics, is pcrhaps to bc attribnted to friction, by
th opration of which the higher and Iuu's))cr componcnts arc
gradually climinated. It must oftcn have bccn noticecl that the
suund s is scareciy, if at al], rctnrncd by echos, and I hve fuund~
that at a,distance of 200 nictrcs a powcrfui hiss losesits charactcr,
even whcnUicrc is no refiection. Proba.b!y Uns enect aiso is duc
to viscosity.
AcofitictU

Observations,

P/t~. ~/<t.'7., Junc,

1877.

34G.]

TRANSVERSE

VIBRATIONS.

285

lu highiy rarcned air th value of a as givcn in (8) is much

incrcased, being constant. Sounds even of grave pitch may thon
bc affected withiu niodcrate distances.
From the observations of CoHadon in thc Iakc of Gcncva,it
would appcar that in water grave sonnds are more rapidiy da.mped
than acute sounds. At a moderato distance from a bcH, struck
undcr water, he found the sound short and sharp, without musical
charactcr.
347. Th effect of viscosity in modifying thc motion of air in
contact with vlbratingsolldswill be best uudcrstood from th solution of tho probicm for a verysimple case givcn by Stokes. Lct us
suppose thn.tan innnito plane (~) excutes harmonie vibrations In
a direction (y) parallel to itsclf. TI~c motion bcing in parallel
strata, u and M vanish, and thc variable quantitics are funetions of a; otdy. Tho nrst of quations (13) 345 shews that the
pressure is constant; th correspondiug quation in v takes tho
furm

sun)]n,r to thc quation for tho tincn.r conduction of hcat. If wc

now suppose tlm.t v is proportional to e' tlie resulting quation
in a; Is

If th gas bc on the positive si.doof tlie vlbrating plane the motion

Hence J9=0, and thc value of v
is to vanish when a:=+cc.
bccomca on rejection of the imaginary part

a.t a;==0. Th velocity of thc fnnd in contact with thc plane is

usually assu)aed to be tlie sn.mea~ tha.t of tlie plane itscif on the

28G

PROPAGATION

0F

SOUND

[347.

would imply nn
nppa,rent!y snfHcicnt g)'f)nnd thiittLcc'mtnn'y
itif-mitc'y ~-ren.tersjnc'ithofss of <n (htid w;th "(..sp~f-t t~ t.hi! soiit'
ti):m wit.)t respect tu itscU'. On t.)ns
supposition (5) expresses t!)u
nn'tion uf t)ic finid on tlic positive sidc due to a motion of tliu
p!:uic biven by (G).
Thc tangcnti:d force pet- unit a)'c:i
n.cting on th plane is

'f~t.=l.
Thc fir.st tcrm t'cp)-(jscnts a dissipativo forcci holding to
stop thc motion thu second rGprcscnt.s a f")'cn G<~)iv:Jc-ntto an
ino-caHOin thc ino-ti.i of t.])c vibrating body. Thc m~mtudc uf
both .t'urccs(h;pcmds upon thc fn-f~tuncy ofthc vibration.
Wc wi)i app!y th)src.sulttoca!cu!atcappt-ox!)n:t,t(j!y th VL-iocity
of sound in tubes so Ufu'row tbat t)K' vi.sco.sitynf air (.'x~rciscsa.
scnsiDu mfiucncc. As in 2(!), !ct JV dnote thc total transfur of
<i))Ki across t)tc section of t)m tube at thc point .f. T))c fo~'c,
duo to hydrostatic pi'cssurt. actih~ on th sticc bctwcun a; and
.e +
i.s, as usua),

Thc force uuu tu vi.sco.sitymay bc infcn-cd fro)n t.hc

invcstigatinn
H)r :Lvibi-ftting' p):u)u, pt-ovidL-dUi~ thu t))iduiuss of thc
fayur of
:ur atUt(j)-i)~ to thu W!i)).sof the tulxj bc sn~]! in
co]np!m.s'))i with
t))u (tliDn~t~-r. Thus, if 7~ bc t)tc puntm.'t.ur cf thc tube, :n)<) rbc
thc vutocity of thc auTent at a distance in'm 0~; w:i.!Lsxf thu
tube, thc tangcntiid force on t.hc s!ico, w))usc volume is
is
Ly (7)

347.]

1~ NARROW TUBES.

287

Thc rc.sult cxpresscd in (12) wn.s first oht.aincd by Hchnhoitz.

An <'):dx'mtu invustig:)tiu)t of this pruUon ])as :)!.sob(.-cngivcn by
Kn'chhoff, who inchKicd iti las c~cuht.t.ion aot only tlic dU'ct, of
i'rictiou but atsn th:)i ufthe couduction ofin.-at. Kit'chhofr.s n.'su]h
is d' thc samc form as (12), but ~/(~.p'') i.si'cplucud hy t](c q)):uitity
(c~Hcd ~)

whore /<is Newtons va)ue of thc vcioeity of sound, and t/ is a cocfHciunt of conductiun, equal according to th kinctic thuory of
gasusto~p" `.
Thu dintinntiou of thc vcloclty of sound in nn.n'ow tubes, aa
iudiciitcd by t.hc w:t.VL!-t(.'ngt)Kjfst:Lticna.ry
vibrations, was cbscrvcJ
by Kundt ( 2CO),and bas becu speci:d)y invcstigatcd by
Scimeebuli' and A. Sucbeck~ It appcars thn.t thc ditninution of
vclocity -variesas?' in nccordancewitb (12), but, wbcn n varies,it
is proportionat rat]ter to K'~ than to
Since is indcpcnduut
of thc dunsity (~)),t]ic effoctwou!d bc incrcasodin rarcficd air,
34'8. In the course of this work wc Lave )iad frquent occasion
to notice the importance of the conclusionsthat may be arrived at
by the mcthod of dimensions. Now that we are in :t. position to
draw Hhtstrations from a grcatcr varicty of acoustical phcnomcn:),
re!a.t.ing-to th vibrations of both so)ids and iiuids, it will be convenicnt to rcsume th subjcet, and to dcvefopcsonewhat in dtail
the principes upon -\v]tic]ithe mcthod rcsts.
In th case of Systems,such as heHs or tt)ning-fo)\ks,formed of
uniform isotropic mntcna!, and vibmting in virtuc of cJasticity, th
'7'n~)t).t.cxxx!7.177. 1RM.
~7'f~t)f.t.cxxx!X.l(tL 1870.

''ro~in);.t.cxxxvi.29(!. 1SC9.

288

DYNAMICAL

8IMILARITY.

[348.

acoustical clments are th shape, th Hnear dimension c, thc

constants of clastieity q and ( 149), and th density p. Hcnec-,
by thc method of dimensions, tho periodic time varies cfe<e?'M
as th lincar dimension, at lea.st if th amplitude of vibra~rM'
tion be in thc same proportion; aud, if thc ia\v of Isochronism
be assamcd, tI)G!a.st-namedrestriction may be dispcnsed witL. Jn
fact, since thc dimensions o.f q and p arc respectively [Jt7'Z''2')
and [~Z'"], wliile
is a mre number, the only combination
capable ofrepresenting a thue is y'~ ./3~ c.
Th argument which undcrUes this mathcmatical shorthand is
ofthc followingnature. Conoivetwo gnomctricallysimitar bodics,
whose mccitanicai constitotion at corrcsponding points is the
same, to excute similar muvcmc'nts in such a manncr that t))o
corrcspondingctiangcsoccnpytimcs' which are proportional to tho
linear dimensionsin thc ratio, say, of 1 ?;. Then, if the ono
movement be possible as a consquence of the elastic forces,th
other will bc also. For th nasses to bc moveclarc as 1 ?", th acclerations as 1
and tlicrefore the necessary forces arc as 1 M';
and, sincc the strains are thc same, tins is in fact th ratio of tlic
clastic forces duc to them when rcferrcd to corrcsponding a-reas.
If th elastic forces are competent to producc the supposcdmotion
in the first case, they arc atso competent to produce thc
supposcd
motion in the second case.
Thc dynamical similarity is disturbed by th opration of a
force like gravity, proportions! tothc cubes, and not to thc squares,
of eorresponding lines; but in cases whcre
gravity is th sole
motive power, dynamical similarity may bc sccured by a different
relation between con-csponding spaccs and corrcsponding times.
Titus if thc ratio of corresponding spaces bc 1 ?), and that of
corresponding times bc 1 )r, th acclrations are in both cases
thc same, and may bc thc effectsof forces in thc ratio 1 m"acting
on masses whieh are in th same ratio. As examples comingunder
t!us head may be mentionc<)the common pendufum, sca-waves,
whose velocity varies as th s()uarc root of the wave-]ength,and th
whole theory of th comparison of sliips and thcir models by
which Mr Froudo prcdicts th behaviour of ships from experiments made on models of moderate dimensions.
1 Thconception
of an altrationof sealoin spacchas~ccnmadofamiliarby
th nuiverfifd
useof mapoand n)0(tc)s,
buttho con-ospondiu~
fortimo
conception
isoftenlessdistinct. Rfrence
tothocaoonf musieai composition
at
performed
difleroutspcodsmaynssistthimnginntion
oftbostudent.

DYNAMICAL

348.]

SIMILARITE.

289

Thc same comparison that we bave c'mpJoycdabovc for clastic

sobds app)ics a)soto acrial vibrations. T)<L-pressurer,in th casesto
bc compared arc th same, aud thercforc wlicn acth)~ ovcr areasin.
thc ratio 1 ?r, givc forces in tbc sarnc ratio. Thse forces operato
on masses in th ratio 1 ?~,and threforu pro(h)ccacclrationsin
thc ratio 1 /f, winch is th ratio of th(; acLuataccek'rationswJten
both spaccs and times arc as 1 :7:. Accordix~lythe pcr!odicti)ncs
of simiJar rosonantcavitius, H))t.'d~'ith thc sa.rnci~as,arc dirc'ctiyas
thc lincar dimcnsiuna Yc'ryimportant )aw fh'st f<n'niu)atcdhy
t-i:LYart.
Sincc thc s:m)c mut.hodof conpaD.son :)pp1i<sboth to clastic
.so)id.sand to cta.stic nuids, an cxtcnsiua mayhc nmdc to Systems
into winch both ~ind.sof vibration c'ntcr. For cx:nuph?,tho sc:do
of a systcm compoundcd ofa hining-fork and of nn nir rcsonnfor
maybe supposed to bc altcrcd wit))out change in tbc motion ot.hc'r
~han that i:)vo!vcdin takit~ thc timcs in thc .samcratio as th
hncar dl~ncnsions.
JIitItcrto thc altration of sca!c bas b<;unsnpposcd to bc
nnifonn m iiU dimensions, but Hierc arc cnscs, not c'oniiog undcr
this hcad, to whieh thc principe of dynannMdsimihn'ity maybc
most usefuDyapplicd. Let us considcr, for c'xamptc,titc i!cxural
vibrations of a sy.st.emconposed oi' a tbin clastic hunin~, phtncor
cnrved. Dy 2L-i-,2L5wu sec that thc thicknc.ssof thu Iann<t:t
and thc mechanica) const.'u~tsy and p, will occur onty in tbe cotnbinations
and
and thns a contp:n'ison may bc made cvcn
although tbc attcradon of tinckncss bc not in. t]tc Hamoproportion
as for thc ot-!)ordimensions. If c bc thu ]Inear ditncnsion who~
thc tilickncss is disrcgardcd, t)ic timc.s must vary c~o'/' ~)fu't'~<s
as ~3~.c\
yor a givcn uiatcrial, ttnckncss, and shapc, thc
titnes arc thcrcforc as tbe.s~~' of tbc )inoar dnnc'nsion. Jt must
t)ot bc forgottcn, Itowcvci')t)tat rc'snits such as thcsc, which invo)vc
a )aw whosc truth is ody approximatc, stand on a dinercnt Jcvct
from thc more Immdiate consquences of t))u principic of shniIm'ity.

THE END.

R. !.

1<)

APPENDIX

A.

( 307).

probtcm of dctcrmming tho correction for th opcu cud of a

Innnitc
tubei8oneofconaIdct'a.b)cdif)icuKy,cvcnwhenHu-rci~au
HtLnge. Itisproved in t]<ctext ( 307) that th co)')-uction a in grcnter t]mu
Q
Tho latter vahio is obtuuied l'y c:duuhiting
~7r7?, n.ud tcsa tha: 7~.
Tho

th encrgy of tho !not!o)t on tho suppositiou t)mt thc vn]oc!ty p:u'a))c)

to thc axis is constant ovcr tho plane of thc mouth, and con)])arit)g this
encrgy with th square of tho total ctu't'ent. TJtC actnat volocity, no
doubt, mercasca froni tho ccntro ontw:u'ds, bnconnng i))<inito at thc sl)~)'p
tidge a.ud tho assumption of tLconstant value ia a sotucwhat violent onc.
NovcrthcJoss tho value of a so Cidculatcd turns eut to bc not grfatly in
Tt M vident t]tat wo should hf: justifier in cxcxeess of tho truth.
pecting i).Ycry good rcsult, if we assunic n.n axdal velocity of tito for]u

?' dcnotmg thc distance of tlio point cojisido'ed ft'ovntho centre of tho
Mul so as to tunko t)tc whole eucrgy tt
mouth, and thcn dctcnuino
minuuuin. Ttte cncrgy so ca-lcult~cd,tho~tghucccssurily in cxccss, must
bo a very goodupproximn.tionto t!tu tt'uth.
In can'yingout this phui.'wohft.votwo distinct pro~lems to de;d wit.)),
l,
tho dotenninatioli of tho motion (I) ontsidc, tuid (2) it~ido t))o eylindcr.
Th former, bcing t!io e~Iei', wo will takc (Irst.
TIte conditions are tha.t ~) Y{mis!ia.t uiduty, n.ndthat wlien = 0,
MtC
whcre
van.iBh,except over tho area,of thc circle r=

Uadcr thcso circumstanccs we know ( 27'8) that

l')2

2S2

CORRECTION

FOR OPEN ENDS.

of t)( pumt wlicro

wLero p dcnoies tho (]iu<.ti!ico
ft'mnLhec'Ioncntofn.rca.f/u-. 'Nuw

Is to bo estiinated

Thn vahin hf 7' is tu l'f c~cuhtt.f~ hy t.hc ~K'Lhod onpioyud in Lho t,(.t
(~ 307) for t).unifortu dt'n.sity. At, thu edgc uf tho dise, witeji eut du\)t
tu radius f<,wo)t!L\'(:L)t(;puLL'ntiai

on c<ct,Ii)~ thc Integmtinn. This qu)Utt,ity<U\'idcdLy

kitit,Iccncrgy ofthe uiotiou JcfmcJ Ly (1).

givcs twice ijto

Th tut:d cm'n'ut,

Woha.vcnoxtt.ocons!<!c)' thc pi-oLIt'jnofdctcrnunit]~ thc motion of nn

incotupt'pssiLie ~uidwithin )-i~i([ cy)im)frundcr thc coudit.ions t])at thc
!txi:d vt.']ucity shitH bc unifut'nt w)n'~.'<;
M and w)t(;n a: 0 sItaU ho of
t))Cfurm

if for thc sft!:e of brcvify wc })ut 7~ 1,

'Thp~f')))i)tyofthcnunlisH))pposcdtf))r')n))ty.

CORRECTION FOR OPEN ENDS.

293

Now<~ ma.y bc cxpauduj in thc ncr!cs

E)t.ch tenu ofUtis

sries sitLi.sfic.sUic conditmn ofgivin~

no l'tutitd
innlno!))<)tio)ioE
]tt
vu!octt.y,v))('u)'-lj
!myki)n!,w][cun;co.
rctninns to dtitcrnuao tlio coc(!iclcnt.n
so as to saLiijfy (G), Avhun-O.
Fron )'- 0 to r- I, wo nmst hin'o

thc sunnrtiLttoncxtcndu~ to :dt Hic admissible values of 1;. Wc ]m.\c

uow to iin'l t))0 cnorgy of jnotioti of so nmch of tlie flilid as Is mcitutcd
bctwccn x = 0, and = l, witoru is so grca.t that tho volocity la tliere
sensibly constant.
By Grccu's tlicoroni

Thc uumcrieal vn.Iucacf thc roots aro appi'OMmatc)y

~t= 3-83170~
~=13-321,

~= 7-015,
p,i=l(W71,

~=10'17~.
~=19-C1C.

294

CORRECTION FOR OFE~ ENDS.

KothiLtt.hoscccmdtot'mis 7!7(1+~+~')".
Iti catculating tho ~rst tcnn, wc must rcnx'mhcr thitt if
two difTuL'cnt
values of~,

nnd~ tx:

To t.)us must hc :ult)<;dtho cno~y of ttic rnoUon on thc poHitIvcHido

nf~~O. Ou(.))owho~

FOROFENENDS.
CORRECTION

2955

)Utd our uljcct is to detei-tninc its mnxhnuta Ytdno. la geneDt] if:

<S':uid Le two qu~dmtic functionH, th mnxmuun and mminnnn vaincs
of x= <S'S" tu'o gh'cn by tbo cubic cqn~tioti

and 0', A', arc dm'tvcd from 0 and A by nu.et'changing thc acccntcd and
unacccntcd Ictturs.
lu t)io prsent case, si)ice<S"is a product of liucar factors, A'=0,
A 0 sunpiy.
0' =- 0, so that
iLud aincc th two factors arc thc MiLUtt-,
tho tmuu'ricul v:dm.s, !md <-f)'uctingLhu c:).!c(dado!is, wc
Substituting
(iud = = '~!898G~,whiu)i M tho nm.xiuumLY.duo uf Lho fracdi). uonsistcuL
wibh i'< values of !(.ttd
Tho con-cHpoudmg Y:duo of a is -821:227. LhtUi winch Lhc tnn:
corrcctiti ciumob bc ~ruatL'r.
If wo assumu
-0, tliu ~rua(,(.'st Ytduc of.: titcn possible is '021~63,
whn:!i fi\us
whh:h
givus
a .'8281.iG7~.
On th othor Iiand if wc put
=0, tho maximum value of s conics
out '027G53, winjucc

a-82

535 3 A'.

It wouhl nppca)' from this n'unit th~t Un) variable parL f'f thu
norinid vclocity at thc nnjuth is buttai' t-~pruscatud t'y a tct-m vm'yio~ as
tliau by une vnryihg as
Jf thu
TI~ value M '82 12~ is probably p)-(.y clo.sc to thf! truth.
ifofLitnji-m 1
normal vclocity bu assmncd constant, a-'8~S2GA',
aud wi~'u Uh;
isHuitabtydctcrntined;
w)f'n
1+u.?'~ a-'828157~,
iuiutbcr arbitrary constant, is !na<h:
forni l+/jL!~+~
e'jntami)~
tho fomulation of thc ca!cn)atiou, w" gt;t a --82i2A'.
Tlic truc Ya]uc ofa is pt-ubab)y about '82/t*.
In thc case of

thu nnnhnmn uno-gy con'csponds to

l'U.):

so tliat

On thia snpposition thc nornud yclucity of thu cdgf; (<' -/<') wou!d 1~~
about double of that ncar t)tc (.'t-htrc'.
'X~tcsfnUc~t.'l'fifunctious.

J'/tf/<Xov.187:

29G

NOTE

TC)~ 27;

A nK'Lhnd nf o))t..unm~ L'uisson's soluLio!! (8) ~ivoi l'y Mouvitlu'

wru'U)yfjfn()<iu(.\

in

!f)'L('thr'p()]~]')'at!t)ts\T'ct,()r]nMsur('(tf)'o)na!)ypnintO,n.n<lt))<'
g<')u')':t!d)f)i')'(')tLiid<qu!).t,iutt))('htt(~)'at('()()V(;)'t)n!V(~))un(iin(:I)utud
wciitul on tnu)ni'<j)')n!).Lut.w('t'nH~)t(-rica)H)n-f)te('i()i'ra(1ii)'aud
!'+<
ii')itoft)t(!.s(;(-'o))dint.('~)'!d~y(h'c(')t'.st.]n'or<'m

mw)nch<\ ~\`
~~<~r,thaLi.st.os!)yi.sj!n)pnrt.!u))aItothH]aca)tva!m'
of </)rcckotu'd ovcr t])c sphf.'t'ioUsurr~cf;of nuiiu.s )'. Eqnatiun (n) nuiy
Lo i'(~)H-(tcdas a)i cx~'asion of ()) 27iJ; it may a).so)'o pruved from
thc cxto'f.ssioti(.5) 241 fur ~<~ I)i teru)s of Lhc ordmary polar co-ordinatcs ?', M.
TJic gcm'nd soluttou of (fjt)iH

wp
H!)7

jVo~ <))?7~)'<.s'tM !)~(u~(~<t

JA(~eMf<<t'<<s'()f7y, )~

7'oc<'c~ty.'i
/~r. ~'u.ll').

</; Zo;t(~

Jt hiY.softcn ~ceu rcnuu'~cd thaf, whcn .~ro~pof Win'os a~vanccs

iuLo sti!) ~'att')', Oa: Y<(;iLy of tlio ~t'oup is Ic.ss thati th~t of t!)ti individuitt w~vcsoi'whichitis
conpum'd; t.]tCwaY(.'snpp('!ti'toa'h'anu(.'
')j'")~ !Lwny!'s t!t('yn]~)t-o:(c))it.s :)n)r!)'ior!i)ntt..
t)))'ou~))t)!u~ruuj',
t ))t'.)[u'<iu'Htcxp)!L)m!d))yHt())<cs,w]torc'l'his jtjn'nottu'nunwu.s,
~ardcd tho ~r~up ns onncd t'y t-))o sopo'pusition ot' two infinittttmins
of wttvcs, of C(~]!t)tUttrittudus imd of ])c:D')y cqun) waYG-tc'ngUiS, adMyat,t,c))tion\as(:'a))c<ltoth(!sn~uct
vancingintit(i.S)U~odi)'ccti.on.
ahout two yc'ars MinceLy ~n' Fronde, !U)d (fx; siuuo cxplanatiou t]ten
occun-cd tu me i)LdcpcndcntIy'.
]u inyboo];; on t)io "Thuoryof
Suund" ( 191), 1 La\'c eonsidcrcd tho question more gaiendty, and
lu~'H shcwn Hott, if r Lo tito 'ctocity ci' pt'opng~tiou of tmy kind of
and K 27r\
tl)cn
w:LVCs\viiose w.LVc-L'ngt]).i.s
thc vclocity of
!). gronp compo.scd of n. gi'cat tnnnbcr cf wn.Ycs, atldmo'ving into (m undisturbed p.n't of Utc ntedium, ia cxpresscd )'y

In fttct, if tlio two itiiinito tmius bo rcprc.seutc.d !jy eon()"):)

and cos ' ( t~-x), tlicu' resulta.nt is roprcseutcd Ly
cos K ( F< ) + cos x' (

a:),

1 Another phouomcuou, aJ.-iOmcutiotictl to lue hy ~[[ Froujc!, ndmits of a

similar txpittun.tiou. A steam In.uueti moving quickty thron};!)tho water is nccomp&uiedby a pcculinr Hy.stculof Jiver~i])~ Wftvcs,of which tho most stri]:i))g
fon.turoia tlt0 obliquity of tito Iiuo coutn.iuiHgtho grcatcst devn.tious of suecossivo
waycs to tho wttvo-fronts. Thia ~mvo tmtturii rnny bo oxpiaincd by tho snporof slightiy diiering wft.Ye-Icngths,
positiou of two (or more)infinito trains of wn,Yes,
whoso du'cctionHand velocities of propngn.tiouaro so related in ench eas th~t thero
is no eLfiago of position rcla.tivcly to tlio boat. Tho modoof composition will bo
bcst, uudcrstootl by Jrawiog on papcr two sots of parnUtiland cquidistaut lincs,
eubjeet to tho abovc condition, to rcprcaeut tho crcsts of tho componcut trains. lu
tho cnso of two trains of slightiy diiercut wavc-Ioneths, it may Lo pro~cd that tho
tangent cf tho angle bctwcou tlio liuo of maxima and tho wavo-fronts ia hait tho
tangent of tho angle betwccutho waye-fronta and tbe boat's courso.

2:~8

PROGRESSIVE

WAVHS.

whichiHuquidt-u

If K'-K,
])o Hinft)), wc htivc n. tmin of W!).vcswhosc amptitudu
vanus slowly from ono point to Miotlier hctwcen tho limits 0 tmd 2,
formiug <).scrics of groups separatcd from ono anoUtor hy l'cgiona cotnpM-attVc!y -ec froni di~turhanco. TLo position at tuno t of tho middiu
of Umt group, which w~s imti:L])y at th origm, iHgivcu by

winch. shews that

UmveiocityoftItC!

group

is(/<r)-(K').
InthcHu)it,wi)('uthc]U))nLfrufw!LVusnieaehgroupisindoTinItt']y
grent, titia rcsult eoincidus witit (t).
T))0 fo!)owu)g particutar ca.scs arc worth notice, a.ud :u'o hcro tabulatcd fur convctiicuco of comparison
Fcc,
fee~,
ree.
)' M .

6'

r =c ",

6'

~-0,
~)
r,
r,
2 r,

]!('yno!ds'di.sconncetedpcndut)))ns.
Dt'<'p-wat<jr~ravit,yw:Lvcs.
A(.')'ia)-a\-('M,Ac.
(.')t])iHary watur wavcs.
Figura) waY.;s.

T]t0 ci~fiihu-y watcrwavcs arc tLcH whosc \)n'c-]t.-ngUt i.s su sinidi

thiLt tho furcc of restitution duc to Ctq'mfU-it.y I'n-~c)y oxeceds t)tat duu
to gn~-ity.
TJtcir theo-y JtaH hccu givu Ly TitOtnson
(/<
J/f~
Thc j)cxund wav(.'s, for w])ic)t ~=.2 F, arc t)toso corNov. 1871).
ruapohding to thu Pouding uf an clastiu rud or jdatc ("TJtcory of
Sound,"

191).

In '<-paper rcml at <.itorfynumt.)) )ncct.ing of t))(! British Association

.Nature," ~t)g. ~:), ]877), Prof. Osbornc
(a.ftc'rwardM printcd in
l!.(.'yn<j]ds gi'.vu n dyuiunicid L'xp).t!i)ttio)t of th fact that ft group of
of titoindidccp-waturwavcs
ad\U)cuswit)ion)yIta]f
therapidity
vulual v'avcs. ]t aj'pt-fu'ri t))!tt thc cnc'rgy p)-opagat';d in-ross any
point.
whun a tral)i of wavcMi.s pa.s.sing, is oniy onc-h:df of t))o ouurgy nccesnary t~ supp)y tii w.n-c.s w)m.t pass in thc samc ti)m', so t)tat, if thu
train of wavt's hL- limitL'd, IL i.s i)))j'ossi)j]c th:it its front can Lo
prop:t~atcd wit.L ttic fuH vctocity of tin! wavcs, Lccat~sc this -ou)d imply tue
ac~uisitmn of morf ('ttf'rgy thnti (-:mm faet hc su))))!icd. Prof. Rcyno)ds
did not contonpiatc t)i(i cases whurc ?/!e'rt; t'])c'rgy is propagatcd thau
correspondu tu th(; wavcs passing m tho saine tinu' but Lis argument,
applied converst')y to tho rt'suits a)ruady givcn, s)icws tftat suu)t cnso3
must cxjst. Tito ratio of thc cuo-~y pro]'ag!ited to titat of tho passin'~
wavcsis
in tho unit thncis
r; t))ust))n<'ncrgypr()pagatcd
F

PROGRESSIVE WAVES.

299

of Ut)t.tcxL~ing in n. Icngth F, or U timcs tlmt cxis(,ing in tim UltiL

Icngtb. Accorttmgly
Energy pt'opagittcd in unit timo Etx'rgy contn.hicd (on au ~vcragc')
~y(~).
=~(!"):(/'<,
inunihiungth
As tili cx)unp!c, 1 wH)tiJ tho c:)scof s!n:dl in'otjttional wnvcs i)i
If x t'c ncasut'cd downwai'ds from the Hurfacr,
Wittct'of fhuto dcpt!i
!md thc cluvtt.Lion(A)of tho waye bo douotcd hy

Tins vainc of

tho goicral dKrcrctitlid Cf~mtmn for m'ot~

fiu.t!afLcs

tionid motion (\7~=0),

and-wf'cn
J<

makcs tho vertical

Yclci~y-~zuro~-hen

~==0. Thc vclocity of propngniion is giveJt t'y

Wc inay now cftieuhtto t,!)Cenor~y contiuuRd in n. hngth :c, which is

:<.Jiumber of wayes t!mt fmctio:ml ptu-b
HUpxoscd to includu no grcut
nin.y bc lui't ont of account.
For tho potentud cncrgy wc I~c

thf ar~mncnt advn.nccd ut U)n

hy (1) funi (G). If, m nccordanco wiUt
und of thiM p~rM', t!)C cqua!ity of F, nnd 7' bo asanmed, tho Ytduc of
Thc
tho vclocity ci' propngattou f'ottow.s ft'oxt t)ic prfsoit exprossions.
is thp!'cfn)-c (fur cncit
)i Ibn~Lh
wholo encrgy itt tlio w.n'csoeeuj'yin~
umtofbrcadLh)

r.+~(~),
7/ dcnoting thc !)iax!i~mn c!cv:Lt.i(jn.
rruf. HcyMida eo!)HukMUtGtroeh.~M Wftvoof n~nkhtc nud Frotulc, which
mvoh'es tnakcuittr rottitifin.

~00

PROGRESSIVE

WAVJ.:S.

Wu I):n-c ucxL to cai'uh~t' U)e enci~y p)-o]<:)~att'(t i~ thue < ncross :t

pLuMfo- which~i.s
constant, or, i <)t)tcrwo)-(!.s,thcw(M-k(!~)t.h:)t t
mustL('<1nniuo:'tot'tuHU.st!ti)tt!)(;]))<)tinaofthopLm(-((;o!t!ji()t~'cd
an!Lf!t'ib!c))nni)ta)i)it!)('f)tccoi't!mf]ttidpresHU)'c.s
acLtn~unouDit'
i')-o)tLof it. T))H Y:u'iaMo pitrt of Hic prc.ssuro (~<), at
in
dr)!t,h
~i\'(.t~y

As nn cxfunph' of t!ic direct crdcuLttionof

we !nay tllke titc ca.sc
of waves tnovjng und<;rth joint mHnenccof ~'n.vity and cohsion.
Itis{)roved)'yTi)ontsunt))at

Whcn is snmU,t]t0 surface tension i.snngligi)')c, and th~n ~'= r;

butwhcjt, on tbo contrary, is Ifu'go, ~r,
iLSlias alt-eady hccn
s~tcd. Witcn ?"y,
~'=r.
This con-cspoiuls to tho nunhmntt
vclocity of propngitttou i)ivcst!gatcd ~y Thomson.
Althongh t]io nrgtimcnt from mtcrfo-cncc groupa scons s~tisfactory,
fui mdcpoidunt invcstigatio]i H dcsit-a)')c of thc ruititi.ou Letwcen
cncrgy cxisting :md enct-~ypropitgatcd. Fu)' sotnc tmio 1 Anmat ft !os8
fur Mcthod applicable to ail khul.s of wavcs, not socing in pfu'ticular
why tho comparison of nergies should hitroducc Uie consLderat,Io!iof

PROGRESSIVE WAYES.

301

a variation ofwa.vc-lcngth. Tho fullowlng investigation, in. whichtho

inci'cincjttiof wave-IcngHi is ~~Ky<f<r~, inay porhaps Le considcred to
]nect thc want,
Lct us suppose that tht! motion of cvcry part of tho jncdium Is
re.sistcdt'ya.forcoofYcrysmaUmaguituduproportionaltothcmaHS
and to thc yc)ocity of' tho part, thc cn'cct f winch will bo t))at wavcs
~cnct'atcd at DiCorigin graduaUy die :Lwayas a; inovascs. Ti)c motion,
whieti m thca~sonof frictujn would )j('rcp)'(;HC!tt(.'d
by co8(~),
under ttto iniincuco of friction is rcprescjttcd hy e'~eos (;<<-.'<'),
who'c is a StuaH posiLn-ococicicnt:1~ stnctncss t!)(; vainc of K is
aLsoaltcrcd l'y Dte friction; but tho adoration is of titc second ontcr as
r'~ards th ft'ictional forccH,and ]nay Lu omittcd ~tidor thf circumstitnccs hci'c suppo.s<;d. Thc cucr~y of tho waA'cspcr unit IcngHt nt
anysta~oofdc~radatiott in proportiotiaitothf! square of thc a]np)itudc,
and thus thc w)t(j]cL'nprgyon titc positive sidc nf tho ori~in is to tho
cncrgy of' HOjnuch of tho wa.vcs at their ~reatest Yiduo, i.e., at thc
],
origin, as wooid hecoutainod lu tho unit of lun~H), ns~e'~f/.E
or as (~)'
Thocncr~y trnnstnittcd throug)it!tc engin in th
unit time is t!tC satnc as thc cncrgy dissipatcd and, if t))C frictional
force ncting on thc cloncnt of mass Mt bo/t7~M,whurc'uistitcvclucity
of thc c)cmcnt and is constant, th onnrgy dissijtatcd in unit thno is
or 2/
/t~?HU"
Luing tho kmctic oiergy. Thus, on tho assmnptio~
t)mt tho kinctic cncr~y Is ])a]f tho wholo cucrgy, wc .rntd that tho
cncrgy transuntted in tho unit tuao is to tho grcatcst encrgy cxisting
in tho unit lungth ns A 3~. It rcmains to tiud thc conncction Lotwccn /t and
For this purposc it will bf conYcniunt to regard cos (M<-M') as tho
rca~ part ofc"~c')
!t.ndtoinqnirohowisa.fruct('d,w)x'n~isgi\'('n,
hy thc mtroductiou of friction. New t)tn c'n'octof friction is rcju'L'scnt.cd
in thc thft'crentiai quations of motiojt hy tho substitution
of j+/tin place of
.j,

or, mnco titc wl)o!o motion is proportional to e'

Ly

snhstituting -?~+t7~ for ?~. Hcncf thc introduction of friction

corresponds to an a!tt;ruLionof M from M to M-Ai/t (t~c Bqnam of
hcing ung)cctcd), and n.ccordingjy is aht-rcd from tu K-

<

./<
T))C

sohttion

t!m3

bccojucs
,.ff)fdl<

part ia rcjcctcd, e
A:

e
,1'

'eo8(~<-<.T),

(;'0'<r),

or,

wl(cn

80 t]<at

T))(''mtioofthf'('ncr~vtra))sn)ittf'dinth<'

thc

~<

int.~gina.ry
<

and
unirtinK't"

J'KOCKES.StVH
WAVES.

302

U)C cnc'rgy cxiKting in t))R unit h'ngth is tl)('cfoi'(- c'xpro'.sr'd hy

or
'/ <f

11\!pI'OY(,(1.
was to hc
113wns
as
provc-d.

It )iaa ofton h<'cu noticed, in particuh~r cases of progressive wavcs,

t))at tho pobential nnd kiupLicnergies arc cqual Lut 1 do not cfdl to
]nind any gcncra! trcatmcnt of Dm question. Tito tileorem ia not
usutdly truc for Lhc jndividuat j'arts of t)'c tnndhnn', Lut must Le
imdcrstood to rcfct' cithcr to au intcgral tnm~K'r of wave-tcugths, or to
:t spacc so considcrftLIcthat t)t0 outstandin~ fra(;t,ion:dparts of waves
may hc k'ft out of nccouut. As an cxtunp)e wt'Il adaptcd to givo inHr'ht into thc question, 1 will tn.k<ithc case of a uniform. stretched
f'ireular jncjn~rano ("Tiieory of Sound," 200) viLmting with a. g)ve;i
tmmLcr of nodal eirck'a nnd diaumtcrs. Thc fundamcnta! jnodcH arc
not quitc dt'tcrminatc in couijcqucucc of thc synunctry, fur any dianttcr may bo mado nodaL In ordcr io get rid of this in()cternunatcncss, wc tn~y Hupposo tito mcinLrauc to carry n,sniaUJoad attachcd to
it anywhcrn exempt on a Jioda! circic. Thcre arc t]~'n two dt'nnito
fundamontal modt's, in oue of which titc Jo~d lies ou a uodal. ditunctcr,
thus ]~roduumgno cfrcct, n.ad in thc othcr jnidway bctwccn nodal dianictcrs, wh~)'c it ])roducoa n. maximum eftoct ("Tix;o)'y of Sound,"
208). If vibrations ofboth modes arc going on. siuuntaneousiy, thc
ne calculatcd
notcuti~ and kitietic cnorgi.cs of tho vhotc motion nmy
by St'y~e <t~to~ of thosc of tho componcnts. Lcb us now, aupposing
thc load to di)ninisl) wIDtout htnit., haaginc that tlic vibrations arc of
cqual ampHtudo aud diu't'i' in phase hy a quartor of a pcriod. T!ie
rcstdt is a y~'o~ressti'cwavc, whos~ potcutml nn.d kinctic cnergics arc
tho suma of tliosc of thc stationary wavcs of whictt it is composcd.
For tho rat componcnt wo Iiavo r,==~'cos*?~, ~==J?sin"~<; a!id
so that
~=~'eos'~j
for tho second eomponf-nt, ~siu'?~,
+ P' ==7' + 7~ 2', or t)tc potcnti~t and kinctic nergies of tite
as t)io w!io!o encrgy of
pro'n'cssivo '\vavo aro equa), hcing tho samo
cither of th compojicnts. TIio nifthod of proof hcro entployed appcara
to bo suSIeIentIy gnera!, tiLOUgIiit is rathcr diilicult to cxprcaa it in
Innguago which is a-ppropriate to aU kindHof wa.vcs.
AMal warea Mo an important exception.

CAMftIUDO!

t'K)!)TKD

i)Y C. J. CLAY,

H.A.

AT THE

UStVKtta)TT

MESS

TUE

THEORY OF SOUND.
VOL. I.

8\-o. dotl), pricc 12$. G~.

Thc Authnr wi)l mo'it in thc highest dcgrcc tho thimk.s of ;dl w)to atudy
thc Munc tf):n)nor in
physics and inttUtMn~tics if ho c~htinuus Uto wot-k ia
Tho Aut))('r bas routtcred it
w]tiuh l)c has ]~nn it in thc first Yhunc.
ofthc whulc, fur tho
po.sHi))!o,)'y th yery cnnvcuicot .sy.stutn~Licarrangutncut
)nost diMeuit pn'hions of acou.stics to 'jo now ntudiudwiHtfi.n' grcator eMe
IIchtiho~tx in Nature.'
Umn hitIicrto.Pt'of.
Wolook forward wit.ti tttc grcatc.sb inhcrest tu t!io [tj'peamnco of tt)C subof
squent volumes, for whiuh t)[is proparcs th way. Thc highor study
acoustics will bc n diit'crent titiag attoguthcr wilcu LhoyM'c in onr tutnds.
J ('t~)y.
IttACMILLAN AN!)

CO. LONDO~.

In. Crowii

8vo. pricc

S~. G~.

SOUND AND i~rusrc.

A Non-Mathcmaticfd Trcn.tiric Oti t))0 Physica) C~OHtituLh'n of Mn.sK'd
S<)H)n).'jatHUIann<)))y,i))L')ndh~t])cchiufA(.'nu.st,ic:dDi.sc~YcnL!snf]'roHi.ssor[Idn)h.)it/. LiySi'JIJijf'
TAVLOU,
~I.A., latc ycUuw of Trinity
CoHcgo,(Ja)uhrittgc.
"In no prcviou~ .sciuntiftctreatitiodn wc remcinhcr so cxhimst-ivc ami
Hol'ichly i)tu.stt'atcd a. <Iuscript!(Utof form.suf vibration iUtduf w:n'cttlut.imtianuidtj.J/tM/(.'<:<S'~</n~<n<
ON SOUND AND ATMOSPHUHIC
VIBRATIONS.
Wit!. thc
Mat)tctnatic:d Eturnuntu of ~fn.sic. Dy Sir G. K AIHY, Astronon~r
l{nyfd. Sccund cdit~n], ruvi.scd .md cnt.n'gcd. Ct'own Hvo. Oif.
AN ELEMENTARY
TREATtSE
ON MUSICAL
INTERVALS
AND TE~t'EXA~fHXT.
With an accouxt of an Enhfu-innmc Uarntoniutn (.-xttit'itcdin tho Lo.~HCn))t;ct.i~n<~Scic-nti~c nstnunotts, Soutit
Kunsit~ton, Ls7(!; i alsodf:ut Hnimnnonic 0)-K:m<x))ihitcd tn Lhc ~tn~ical
Associ.ttit'n of' Lando)), May, 1S7.'). ]!y R. 11. M. BOSANQUET, FcHuw
<)fStJ(jltt)'~C'un(;gc,Oxfnrd. Hvo. (if).
SOUND

AND

MUSIC.

Two Lectures
By Dr W. n. SON).
dc!tVcrcdat.Snut)tK(;)t.si))~tun. muhtt'~tcd. Cruw)tM\-o. (!<

LECTURES
ON SO~rE
RECENT
ADVANCES
JN PHYM!CAL SC'rKXC!
~y I'r.,fc.s.~)rP. C:. TAfT, M.A. H)n.stratud. Sucund
Edition, ('nhu~ud. L'ruwn.iv~ !).<.
THE

APPLTCATrON.S

OF PHYSICAL
FORC'ES.
Trans~trd hy Mt-.s L.~yt'r, aodcditcdwith
1iyA.(:Unj,EM)X.
Additions a))dXut.('().yJ.I~.ckyu)',F.J;.S.
\VithCut<)n)\.))P!a)<-sat)d
inuori'~u.sinustr.ditU~.
Hny.d~v~. :}).<.(!
~).\('fLLA~

AX)) ~a

3
L~Xno\.