,,cl]k ,,maxwell',s ,,electromagnetic ,,!ory ,! ,r$e ,lecture = #aibc ,,by ,h4 ,a4 ,,lor5tz ,,cambridge ,,at ,,! ,,univ]sity ,,press #aibc ,,cl]k ,,maxwell',s ,,electromagnetic ,,!ory ,,:5 ,i _h ! hon\r to be invit$ to deliv] a lecture 9 ! ,r$e ,f\ndation1 ,i ?\ght ,i might p]haps pres5t to y a brief review ( ! electromagnetic !ory ( y\r great physici/ ,james ,cl]k ,maxwell4 ,! *oice seem$ ! m appropriate z x is n[ exactly fifty ye>s ago t ! "w :i* rais$ him at once to ! v fir/ rank ( inve/igators ( all ages1 ! ,treatise on ,electricity & ,magnetism1 0 publi%$4 ,9 ?is "w x 0 prov$ beyond all d\bt t electric & magnetic actions c be conceiv$ z be+ transmitt$ ?r\gh a m$ium & ! !ory 0 cr[n$ by ! wond]ful revelation t light is an electromagnetic ph5om5on4 ,maxwell's !ory 0 also a great simplification4 ,inde$1 be=e 8 "t !re 0 mu* unc]ta9ty & confusion 9 ?is "p ( physics & _m cont5d+ !ories 7 9 ! field4 ,9 electrodynamics1 = in/ance1 we _h ! laws ( ,ampre & ,grassmann = ! actions betwe5 elem5ts ( curr5t1 &1 :5 we w5t fur!r1 we f\nd ! speculations ( ,web]1 ,riemann & ,clausius ab\t ! mutual actions ( p>ticles ( electricity4 ,9 connexion ) ^! !ories !re 0 a good deal ( discussion on ! ph5om5a t 7 to be expect$ 9 ! case ( clos$ & 9 t ( op5 circuits4 ,x 0 ?\ght 9 ^? days t ! curr5t 9 a wire by means ( :i* a metallic conductor is *>g$1 ends on t conductor1 & ev5 ! dis*>ge curr5t ( a cond5s] 0 consid]$ n to be clos$2 !re 0 a gap 9 ! circuit1 because we _h no idea t some?+ is go+ on 9 ! insulat+ lay] betwe5 ! coat+s4 ,9 optics we _h no less tr\ble4 ,x is true t ! g5]al pr9ciples ( ! undulatory !ory ( light _h be5 firmly e/abli%$ & physici/s 7 ju/ly pr\d ( ! success t _h be5 a*iev$ 9 ! explanation ( int]f]5ce & diffraction1 d\ble refraction & pol>ization4 ,yet1 :5 we tri$ to p5etrate some:at deep]1 we 7 confront$ ) s]i\s difficulties4 ,:5 we want$ to acc\nt = ! diff]5t optical prop]ties ( v>i\s sub/ances1 ( air & wat] = in/ance1 we _h ! *oice betwe5 two assumptions4 ,fresnel _h s\ght ! cause ( ! diff]5ce 9 an inequality ( ! d5sity ( ! e!r 9 ! two sub/ances1 ! ela/icity be+ ! same 9 bo?4 ,f4 ,e4 ,neumann1 on ! o!r h&1 _h suppos$ ! d5sities to be ! same1 b ! ela/icities to be diff]5t4 ,on ei!r ( ^! suppositions1 & 9 no o!r way1 x _h be5 f\nd possible to d$uce ! "r value = ! ratio betwe5 ! amplitude ( ! reflect$ & t ( ! incid5t light4 ,y "k t 9 ?is problem two pr9cipal cases mu/ be di/+ui%$1 ! vibrations be+ normal to ! plane ( incid5ce 9 ! "o case & p>allel to t plane 9 ! o!r4 ,! two values = ! ratio 9 "q are s9 7i -- r7 _/ s9 7i r7 & tg 7i - r7 _/ tg 7i r71 if i is ! angle ( incid5ce & r ! angle ( refraction2 & x is rem>kable t1 ( ! two rival !ories1 "o l$ to ! expression ) ! s9es :5 ! o!r requir$ t ) ! tang5ts1 & conv]sely4 ,9 connexion ) ?is ,fresnel suppos$ ! vibrations ( plane pol>iz$ light to be at "r angles to ! plane ( pol>ization1 :]eas ,neumann want$ !m to be p>allel to t plane4 ,"h 0 a problem t long baffl$ ! ef=ts ( physici/s1 & _m attempts 7 made to det]m9e exp]im5tally ! direction ( ! vibrations4 ,"o _c say t ! result has be5 v satisfactory & ! "q rema9$ op5 until ,maxwell's !ory settl$ x once = all4 ,i h enum]at$ "s only ( ! difficulties ) :i* we _h to /ruggle4 ,i c\ld h m5tion$ simil> problems t arose 9 ! !ory ( d\ble refraction & ,i may add t 9 "s cases longitud9al vibrations intrud$ !mselves & complicat$ ! !ory4 ,maxwell reliev$ u ( all ^! d\bts & unc]ta9ties4 ,by 8 bold assumption t 9 a non-conduct+ body1 9 a dielectric z he call$ x1 !re c exi/ :at is truly a motion ( electricity1 & t1 if ?is motion1 ! dielectric displacem5t1 is tak5 into acc\nt1 electricity c always be said to move z an incompressible fluid1 ! op5 curr5ts & ! longitud9al vibrations t 7 closely alli$ to !m 7 made to vani% f ! sc5e4 ,fur!r1 ! optical behavi\r ( non-conduct+ sub/ances 0 %[n to dep5d on two prop]ties1 ea* *>act]iz$ by a physical con/ant1 ! dielectric con/ant1 or ,f>aday's specific inductive capacity1 & ! magnetic p]meability4 ,x is true t ! way 9 :i* ^! con/ants are det]m9$ by ! con/itution ( matt]1 by ! /ructure ( molecules & atoms1 0 n consid]$ & t1 s f>1 !y 7 no less inaccessible ?an ! e!real d5sity & ela/icity ( ! old !ories1 b !re 0 ?is important diff]5ce t1 :]eas ^! latt] con/ants _h no connexion ) any o!r ph5om5a1 ! dielectric con/ant & ! magnetic p]meability c be measur$ by means ( /atical exp]im5ts1 s t1 at lea/ 9 c]ta9 simple cases1 we c d$uce ! optical prop]ties ( a sub/ance f :olly diff]5t data4 ,x 0 f\nd t 9 ! new !ory ! treatm5t ( ! reflexion problem 0 mu* l t 9 ! old "o2 "o is l$ to ! two =mulae :i* ,i recall$ to y1 if "o supposes ei!r ! dielectric con/ant or ! magnetic p]meability to be ! same 9 ! two sub/ances4 ,! *oice betwe5 ^! alt]native suppositions aga9 entail$ a decision conc]n+ ! direction ( ! vibrations ) respect to ! plane ( pol>ization1 b ! *oice 0 n d\btful n[1 z x _h be5 asc]ta9$ exp]im5tally t ! ratios betwe5 ! magnetic p]meabilities ( transp>5t sub/ances are little diff]5t f unity1 :]eas ! dielectric con/ants div]ge to a mu* great] ext5t4 ,x 0 !re=e at once e/abli%$ t ! electric vibrations are normal to ! plane ( pol>ization4 ,?is implies t ! magnetic vibrations are 9 t plane1 s t1 9 a s5se1 ! cont5d+ p>ties bo? _h _! w4 ,9 ! case ( cry/als x became c]ta9 t _! d\ble refraction is due to an inequality ( ! dielectric prop]ties 9 diff]5t directions4 ,s1 _m difficulties & out/&+ problems melt$ away z sn[ be=e ! sun4 ,inde$1 a review] 9 ,nature actually comp>$ ,maxwell's "w to ! sun1 8 only criticism be+ t !re are spots on ! sun itself1 :i*1 h[ev]1 are n visible save to ^? ^: eyes c be> ! full gl>e ( ! gl[+ orb4 ,my eyes c]ta9ly 7 n z /rong z t4 ,i c\ld n see ! spots1 b :at ,i c\ld see 0 t ! sun 0 n entirely uncl\d$2 :at sun always 08 ,x 0 n always easy to grasp ,maxwell's ideas1 & "o feels a want ( unity 9 8 book1 due to ! fact t x fai?fully reproduces 8 gradual transition f old to new ideas4 ,:5 we read :at ,maxwell says ( ,ampre & ,f>aday1 ( ! =m] hav+ remov$ all traces ( ! scaffold+ by :i* he _h built up a p]fect demon/ration ( 8 law1 :]eas ,f>aday %ews u 8 unsuccessful z well z 8 successful exp]im5ts1 & 8 crude ideas z well z 8 develop$ ones1 we feel t1 great ?\gh ! diff]5ce may be betwe5 ! ,exp]im5tal ,rese>*es & ,maxwell's l>gely ma!matical ,treatise1 yet ! two works 7 writt5 9 ! same _s4 ,9 fact1 ,maxwell repeat$ly expresses 8 indebt$ness to ,f>aday1 f :om he _h borr[$ "p ( 8 fundam5tal ideas1 s t1 :5 !re is "q ( ,maxwell's !ory1 we mu/ (t5 ?9k ( ,f>aday also4 ,maxwell's foll[]s1 ( :om !re 7 _m1 9 ?is c\ntry & else:]e1 h p]fect$ ! !ory 9 its =m & ext5d$ x by ! introduction ( new ideas4 ,?9k1 = in/ance1 ( ,poynt+'s beautiful & important !orem on ! fl[ ( en]gy1 det]m9$ at e po9t by ! electric & ! magnetic =ce exi/+ 9 ! field1 a !orem t has produc$ m cle>ness p]haps ?an any o!r & :i* is n[ s ess5tial t we c h>dly recall ! /ate 9 :i* physics 0 :5 we did n "k x4 ,yet1 not)/&+ all innovations ( ?is k9d1 we always speak1 & ) full ju/ice1 ( ,maxwell's ,!ory4 ,we cont9ue to d s n[ t we h be5 l$ to introduce electric *>ges suppos$ to exi/ 9 ! int]ior ( molecules & atoms1 by :i* we h come to ! !ory ( electrons4 ,& :5 we ref] to ^? wond]fully simple equations 9 :i* ! fundam5tal laws ( electromagnetism are embodi$ ) a concis5ess t c\ld nev] h be5 dream$ ( be=e1 we call !m ,maxwell's ,equations4 ,surely1 ?\gh ,maxwell did n use !m 9 _! mod]n =m1 no "n c\ld be m appropriate1 = ! g5]al relations :i* !y express are ^? t 7 con/antly 9 8 m9d4 ,"t does n p]mit me to dwell at l5g? on ! v]ifications ( ,maxwell's !ory1 b ,i %\ld l to make an exception = two ( !m4 ,all[ me1 9 ! fir/ place1 to say "s words on ! optical prop]ties ( metals4 ,if ! m$ium1 s we read 9 ,maxwell1 in/ead ( be+ a p]fect insulator1 is a conductor1 ! di/urbance 7viz4 t :i* is produc$ by an incid5t beam ( light7 w consi/ n only ( electric displacem5ts b ( curr5ts ( conduction1 9 :i* electric en]gy is trans=m$ into heat1 s t ! undulation is absorb$ by ! m$ium4 ,aft] hav+ /at$ 9 ^! words "o ( ! mo/ important consequ5ces drawn f 8 !ory1 ,maxwell goes on to calculate ! coeffici5t ( absorption z a function ( ! conductivity1 & he proce$s3 ,gold1 silv] & plat9um are good conductors1 & yet1 :5 =m$ into v ?9 plates1 !y all[ light to pass ?r\gh !m4 ,f exp]im5ts :i* ,i h made on a piece ( gold leaf1 x appe>s t its transp>5cy is v mu* great] ?an is consi/5t ) our !ory1 unless we suppose t !re is less loss ( en]gy :5 ! electromotive =ces are rev]s$ = e semi-vibration ( light ?an :5 !y act = s5sible times1 z 9 our ord9>y exp]im5ts4 ,lat] rese>*es h amply confirm$ :at ,maxwell says "h2 obvi\sly1 bodies1 bo? conductors & dielectrics1 behave 9 g5]al diff]5tly t[>ds rapidly alt]nat+ electric =ces & t[>ds /ation>y ones4 ,yet1 ,hag5 & ,rub5s h be5 able to %[ t :51 in/ead ( work+ ) visible light1 "o uses infra-r$ rays ( suffici5tly great wave-l5g?1 ! prop]ties ( metals w1 9 ! limit1 exactly con=m to ! !ory1 if we reckon ) ! ord9>y conductivity4 ,hag5 & ,rub5s did n measure ! am\nt ( radiation t is transmitt$ ?r\gh a ?9 plate b ! coeffici5t ( reflexion ( a ?ick mirror4 ,= ! case ( normal incid5ce1 ?is coeffici5t & !re=e also ! loss ( en]gy1 i4e4 ! quantity t is absorb$ by ! mirror1 c easily be calculat$ z a function ( ! conductivity4 ,= ! residual rays ( sylv91 ^: wavel5g? is #ab 1 & = silv]1 copp]1 gold & plat9um1 ! absorb$ en]gy 0 f\nd to be respectively #aae1 #af1 #ba & #ce p] c5t1 ( ! incid5t en]gy1 :]eas x ought to h be5 #ac1 #ad1 #af1 & #ce p] c5t1 accord+ to ! !oretical =mula4 ,! agreem5t became /ill bett] :5 ! residual rays ( fluorite ) a wave-l5g? ( #bee 7 us$4 ,s9ce1 h[ev]1 = waves ( ?is l5g? ! reflexion becomes ne>ly complete1 x 0 n possible to det]m9e ! loss ( en]gy ) suffici5t precision4 ,hag5 & ,rub5s ov]came ?is difficulty by measur+ ! emissivity ( ! diff]5t metals1 or r ! ratio betwe5 ?is emissivity & t ( a p]fectly black body at ! same temp]ature1 a ratio :i*1 by ,kir*h(f's law1 is equal to t betwe5 ! absorb$ & ! incid5t en]gy = a beam fall+ on ! metal f ! outside1 s t x c be calculat$ by ! same =mula z ?is latt] ratio4 ,= ! f\r metals j m5tion$ 7at a temp]ature ( #agj ,c47 ! ratio 9 "q 0 f\nd to be 7aft] multiplication by #ajj7 #aac1 #aag1 #aef & #bhb4 ,! !oretical values 7 #aae1 #abi1 #aci & #bif4 ,^! numb]s %[ conclusively t1 h[ev] complicat$ ?+s may be = %ort] waves1 we c calculate ! optical prop]ties ( metals 9 ! extreme infra-r$ by means ( ,maxwell's equations1 simply sub/itut+ = ! conductivity ! value t has be5 d$uc$ f exp]im5ts ) con/ant or sl[ly alt]nat+ curr5ts4 ,?is is c]ta9ly a mo/ spl5did confirmation1 ! c\nt]p>t to ! v]ification1 :i* = gase\s bodies at lea/ has be5 v satisfactory1 ( ,maxwell's relation ( ! dielectric con/ant to ! index ( refraction4 ,! ph5om5on ( ! pressure ( radiation may s]ve z a second example ( v]ification4 ,t a beam ( light fall+1 say 9 ! normal direction1 on a mirror ex]ts on x a pressure proportional to ! int5sity ( ! beam 0 d$uc$ by ,maxwell f 8 =mulae1 & he calculat$ ! =ce t may be expect$ 9 ! case ( sunlight4 ,x la/$ a qu>t] ( a c5tury be=e ,leb$ew succe$$ 9 obs]v+ ?is small =ce1 :i*1 = sunlight1 am\nts to no m ?an ab\t a t5 million? "p ( a gramme weight p] cm4#b1 & :i* x is !re=e difficult to dis5tangle f o!r =ces t are caus$ by ! surr\nd+ gas1 ev5 :5 ?is is highly r>efi$4 ,"s ye>s lat] ,e4 ,f4 ,ni*ols & ,hull repeat$ ! exp]im5t ) ! utmo/ c>e & 7 able to measure ! pressure & to prove t its int5sity agrees ) ,maxwell's calculation4 ,we are n[ q sure ( ?is ph5om5on :i* has come to play a great "p 9 /ell> physics4 ,:5 we are conc]n$ ) v small p>ticles ne> or 9 a />1 ! radiation pressure may v well become great] ?an ! =ce ( gravitation1 & x is tak5 into acc\nt by _m a/ronom]s 9 _! speculations ab\t ! /ate ( heav5ly bodies4 ,! =ces ex]t$ by rays ( light or heat are a special case ( :at we call pond]omotive =ces1 i4e4 ( ! =ces ) :i* ! electromagnetic field acts on mat]ial bodies4 ,maxwell %[$ h[1 9 g5]al1 ^! =ces c be d$uc$ f ! values ( ! electromagnetic en]gy correspond+ to diff]5t positions ( a sy/em ( bodies1 or f a consid]ation ( c]ta9 /resses :i* exi/ 9 ! electromagnetic field & ( :i* he taught u to det]m9e ! direction & ! int5sity4 ,e /ud5t1 ev5 ( r elem5t>y physics1 n[ kn[s t ! mutual attraction ( two conduct+ plates betwe5 :i* !re is a diff]5ce ( pot5tial1 e4g4 ( ! plates ( an absolute electromet]1 may be consid]$ z due to /resses along ! l9es ( =ce1 t ! same may be said ( ! attraction betwe5 a magnetic pole & a piece ( iron1 & t 9 ! case ( an electromagnetic motor we are conc]n$ ) ! tang5tial /resses act+ at ! surface ( ! revolv+ sy/em4 ,"h aga9 !re has be5 a great deal ( lat] developm5t1 b we cont9ue to speak ( ,maxwell's /resses4 ,! notion ( ! electromagnetic mom5tum1 :i* ,maxwell seems n to h _h1 ?\gh he 0 q ne> x1 has also prov$ v fruitful4 ,a beam ( light has a def9ite mom5tum1 mu* l a mov+ ball1 & :5 ! beam is normally reflect$ by a mirror1 s t ! mom5tum is inv]t$1 we c d$uce ! =ce act+ on ! mirror f ! *ange ( ! mom5tum1 exactly z we c d 9 ! case ( ! ball or ( a /ream ( mat]ial p>ticles4 ,9 mod]n !ory--,i allude to ! !ory ( relativity--"o has f\nd good reasons = comb9+ into "o unity1 :i* we call ! /ress-en]gy-t5sor1 all ^! quantities ( :i* ,i h spok51 viz4 ! en]gy1 ! fl[ ( en]gy1 ,maxwell's /resses & ! electromagnetic mom5tum4 ,9 ,e9/e9's !ory ( gravitation ?is t5sor det]m9es ! gravitation field t is produc$ by an electromagnetic sy/em & 9 virtue ( :i* su* a sy/em has an influ5ce on ! motion ( mat]ial p>ticles1 un=tunately mu* too small to be obs]v$4 ,i %\ld be l$ too f> a/ray if ,i dwelt on ^! que/ions1 b :at ,i want to po9t out is ?is1 t we c\ld nev] h gone s f> if we _h cont5t$ ourselves ) ! actions at a di/ance1 if we _h n fix$ our att5tion on ! int]v5+ m$ium1 localiz+ ! en]gy 9 x & consid]+ x z ! seat ( mom5ta & /resses :i* manife/ !mselves 9 ! obs]v$ motions ( bodies4 ,all ^! mod]n ideas h _! orig9 9 ,maxwell's "w4 ,we are also conc]n$ ) a /ress-en]gy-t5sor1 simil> to ! electromagnetic "o1 :5 we consid] a sy/em ( mat]ial p>ticles1 :e!r unconnect$ l ! molecules ( a gas1 or held toge!r by int]nal =ces z 9 an ela/ic body or a fluid4 ,! "q naturally arises3 are ^! /ress-en]gy-t5sors1 ! electromagnetic &1 let me say1 ! mat]ial "o1 :olly indep5d5t or c "o be r$uc$ to ! o!r8 ,"o has (t5 tri$ to d s &1 m p>ticul>ly1 to imag9e electromagnetic ph5om5a z produc$ by "s invisible me*anism mov+ accord+ to ! laws ( dynamics4 ,?is 0 a fav\rite idea ( ,maxwell's & "o ( 8 mo/ brilliant *apt]s is devot$ to ! dynamical !ory ( electromagnetism4 ,x is ! m important because x %[s t su* a !ory c be develop$ on v g5]al l9es1 x n be+ necess>y to make def9ite assumptions reg>d+ ! und]ly+ me*anism4 ,maxwell %[$ t 9 ! case ( l9e> circuits c>ry+ electric curr5ts we c acc\nt = ! pond]omotive =ces & = ! ph5om5a ( self & mutual induction by ,lagrange's or ,hamilton's equations ( motion1 provid$ t we introduce1 besides ! coord9ates :i* det]m9e ! positions ( ! mat]ial circuits1 a c]ta9 numb] ( new coord9ates1 "o = ea* circuit1 ! velocities belong+ to ^! coord9ates = ! sev]al circuits be+ proportional to ! curr5t int5sities4 ,9 fact1 ! new or int]nal coord9ate = ea* circuit repres5ts ! total quantity ( electricity t has trav]s$ "s section s9ce a fix$ in/ant t is *os5 z ! orig9 ( "t4 ,:5 ! int]nal coord9ates are giv5 ?is mean+1 ! magnetic en]gy becomes ! k9etic en]gy ( ! sy/em1 :]eas ! electric en]gy has to be id5tifi$ ) ! pot5tial en]gy & is comp>able to ! en]gy ( de=mation ( an ela/ic body4 ,:ile he 0 apply+ ! laws ( dynamics 9 ?is v g5]al way1 ,maxwell 0 l$ to discuss c]ta9 ph5om5a t might p]haps be expect$ to exi/ & "s ( :i* h be5 actually obs]v$ 9 our days ?\gh ,maxwell 0 n able to detect !m ) ! means at 8 disposal4 ,9 ! !ory ( dynamical sy/ems !re are z _m velocities z !re are coord9ates & ! k9etic en]gy is a homog5e\s quadratic function ( ^! velocities1 9 :i* 9 g5]al n only ! squ>es b also ! products ( ! velocities appe>4 ,:5 we h "o or m circuits c>ry+ electric curr5ts1 we c di/+ui% 9 ! k9etic en]gy "o "p t dep5ds on ! mat]ial velocities only1 & ?is is ! k9etic en]gy ( ord9>y me*anics1 & a second "p conta9+ only ! velocities correspond+ to ! int]nal coord9ates2 ?is is ! magnetic en]gy t manife/s itself 9 s _m ways4 ,n[1 is ?is all8 ,!re w\ld c]ta9ly be a ?ird "p ( ! k9etic en]gy if an electric curr5t consi/$ 9 a real motion ( "s sub/ance along ! conduct+ wire1 = if ! wire 7 mov+1 say 9 ! direction ( its l5g?1 ) ! velocity v & if v' 7 ! int]nal velocity proportional to ! curr5t1 ! total velocity ( ! mov+ sub/ance w\ld be v v' & 9 its squ>e we %\ld h ! t]m #bvv'4 ,"o is l$ to a simil> conclusion on o!r less simple assumptions & s1 indep5d5tly ( an special conception1 ! "q arises :e!r any "p ( ! k9etic en]gy consi/s ( products ( ord9>y velocities & /r5g?s ( electric curr5ts4 ,maxwell ?9ks ?is "q to be ( great importance & deems x desirable t exp]im5ts %\ld be made on ! subject ) great c>e4 ,he !n proce$s to exam9e diff]5t ways 9 :i* ! t]ms 9 "q might be made to reveal !mselves1 ! fir/ ( :i* he expla9s z foll[s3 ,if any "p ( ! k9etic en]gy dep5ds on ! product ( an ord9>y velocity & ! /r5g? ( a curr5t1 x w probably be mo/ easily obs]v$ :5 ! velocity & ! curr5t are 9 ! same or 9 opposite directions4 ,we !re=e take a circul> coil ( a great _m w9d+s1 & susp5d x by a f9e v]tical wire1 s t its w9d+s are horizontal1 & ! coil is capable ( rotat+ ab\t a v]tical axis1 ei!r 9 ! same direction z ! curr5t 9 ! coil1 or 9 ! opposite direction4 ,we %all suppose ! curr5t to be convey$ into ! coil by means ( ! susp5d+ wire1 &1 aft] pass+ r\nd ! w9d+s1 to complete its circuit by pass+ d[nw>ds ?r\gh a wire 9 ! same l9e ) ! susp5d+ wire & dipp+ into a cup ( m]cury4 ,a v]tical mirror is atta*$ to ! coil to detect any motion 9 azimu?4 ,n[ let a curr5t be made to pass ?r\gh ! coil 9 ! direction ,n4,e4,s4,w4 ,if electricity 7 a fluid l wat]1 fl[+ along ! wire1 !n1 at ! mom5t ( />t+ ! curr5t1 & z long z its velocity is increas+1 a =ce w\ld require to be suppli$ to produce ! angul> mom5tum ( ! fluid 9 pass+ r\nd ! coil1 & z ?is mu/ be suppli$ by ! ela/icity ( ! susp5d+ wire1 ! coil w\ld at fir/ rotate 9 ! opposite direction or ,w4,s4,e4,n41 & ?is w\ld be detect$ by means ( ! mirror4 ,on /opp+ ! curr5t !re w\ld be ano!r movem5t ( ! mirror1 ?is "t 9 ! same direction z t ( ! curr5t4 ,x does n appe> t ,maxwell actually tri$ ! exp]im5t2 he only says3 no ph5om5on ( ?is k9d has yet be5 obs]v$4 ,n[1 if = ,maxwell's coil we sub/itute a rod ( iron1 ! magnetization & demagnetization ( :i* are comp>able to ! />t+ & /opp+ ( a curr5t 9 ! coil1 we h exactly ! ,ri*>dson-,e9/e9-de ,haas effect t 0 really obs]v$ by ,e9/e9 & de ,haas & by "s o!r physici/s4 ,y "k t x am\nts to ?is1 t a cyl9drical rod ( iron susp5d$ 9 a v]tical direction is set rotat+1 ) a sudd5 j]k1 :5 x is rapidly magnetiz$ or demagnetiz$4 ,:5 ! magnetization is p]iodically rev]s$1 ! rod is made to oscillate & ! amplitude ( ?is motion may be increas$ by adju/+ ! frequ5cy ( ! rev]sals to t ( ! free oscillations ( ! rod4 ,:]eas ,maxwell seems n to h tri$ ! above exp]im5t ) ! coil1 he tri$ to obs]ve ano!r effect4 ,a coil1 ?r\gh :i* a curr5t c\ld be pass$ & :i* c\ld be provid$ ) an iron core1 0 plac$ 9 a sy/em rapidly revolv+ ab\t a v]tical axis1 ! arrangem5t be+ su* t ! coil 0 free to rotate 9 ! revolv+ sy/em1 s t ! axis ( ! coil c\ld be incl9$ to diff]5t degrees ) respect to ! v]tical axis ( rotation4 ,if 9 ! =mula = ! k9etic en]gy !re 7 t]ms ( ! k9d :i* ,maxwell want$ to detect1 !re w\ld be a t5d5cy = ! coil to place itself ) its axis p>allel to ! axis ( rotation2 x w\ld behave z a gyroscope & might be call$ an electromagnetic gyroscope4 ,no trace ( a ph5om5on ( ?is k9d c\ld1 h[ev]1 be obs]v$4 ,i may ins]t ! rem>k t1 if ! t5d5cy ( :i* ,i spoke exi/$ to an appreciable ext5t1 a magnetic ne$le w\ld1 ev5 9 ! abs5ce ( ! t]re/rial magnetic field1 /ill take a def9ite position t w\ld be det]m9$ by ! rotation ( ! e>?2 9 fact1 9 virtue ( its int]nal motions ! magnetic ne$le w\ld be comp>able to a gyroscopic compass1 su* z has lately come into use 9 navigation4 ,n[ t we "k ! int5sity ( ! ,e9/e9 effect we c say ) c]ta9ty t1 if only 8 in/rum5tal means _h be5 m ref9$4 ,maxwell's exp]im5t w\ld h _h a positive result4 ,we c evaluate ! magnitude ( ! effect & we c also calculate t ! =ce :i* a magnetic ne$le exp]i5ces on acc\nt ( its int]nal motions & ( ! rotation ( ! e>? is ?\s&s ( millions ( times small] ?an ! =ce due to ! e>?'s magnetic field1 s t y ne$ n fe> f ?is cause any error 9 measurem5ts 9 t]re/rial magnetism4 ,9 ! exp]im5ts j discuss$ we 7 conc]n$ ) =ces act+ on ! mat]ial bodies4 ,maxwell next consid]s cases 9 :i*1 always on acc\nt ( ! t]ms 9 "q1 n ! mat]ial sy/em b ! electricity conta9$ 9 x is set 9 motion4 ,"h he rev]ts aga9 to ! susp5d$ circul> coil1 & he po9ts out t :5 a rotation is sudd5ly imp>t$ to x !re c\ld be produc$ a transi5t electric curr5t4 ,simil>ly1 !re w\ld be a curr5t1 b n[ 9 ! opposite direction1 :5 ! motion ( ! coil is /opp$4 ,a v simple exp]im5t may s]ve to give y an idea ( ^! ph5om5a4 ,we take a cyl9drical tumbl]1 p>tly fill$ ) wat]1 & set x sudd5ly 9 rotation ab\t its v]tical axis4 ,!n ! friction betwe5 ! glass & ! wat] w make ! fluid rotate likewise1 b ?is w take "s "t2 dur+ a c]ta9 p]iod ! wat] w lag beh9d4 ,let u next suppose t ! motion ( ! vessel has be5 kept con/ant = a suffici5t l5g? ( "t1 s t ! wat] has acquir$ ! full angul> velocity & !n let ! vessel be sudd5ly br\ght to re/4 ,x is cle> t ! circulation ( ! wat] w cont9ue = a c]ta9 "t1 until x is exhau/$ by ! friction4 ,simil> ph5om5a c\ld be obs]v$ ) a clos$ circul> tube capable ( rotat+ ab\t its axis1 & !re mu/ be a correspond+ electromagnetic ph5om5on if a metallic wire conta9s some?+ l movable electricity1 z we are n[ 9 a position to ass]t1 because we h good reasons = believ+ t an electric curr5t consi/s 9 a motion ( negative electrons4 ,?\gh ! exp]im5t is mu* m delicate ) electricity ?an ) wat]1 ,tolman & ,/ew>t h p]=m$ x ) v satisfactory results4 ,us+ a coil ) a great numb] ( w9d+s1 rapidly rotat+ ab\t its axis1 !y 7 able to obs]ve ) a s5sitive galvanomet] ! transi5t electric curr5t t 0 produc$ on /opp+ ! motion by means ( a brake4 ,! electrons cont9u$ to move ov] "s di/ance j z ! wat] did 9 ! exp]im5t ) ! circul> tube4 ,! direction ( ! curr5t %[$ t ! movable p>ticles really h negative *>ges & ! obs]v$ deflections agre$ ) :at c be inf]r$ f ! ratio betwe5 ! *>ge & ! mass ( ! electrons1 a ratio t 0 f\nd = ! fir/ "t by ,zeeman & ,sir ,j4 ,j4 ,?omson some:at m ?an tw5ty-five ye>s ago & has be5 repeat$ly det]m9$ 9 lat] ye>s4 ,no less rem>kable ?an ,tolman & ,/ew>t's exp]im5ts are ^? made by ,mr & ,mrs ,b>nett4 ,!y f\nd t a cyl9drical rod ( iron1 rotat+ ab\t its geometrical axis1 becomes !reby magnetiz$ 9 ! direction ( its l5g?4 ,l ! ,e9/e9 effect1 to :i* x =ms a c\nt]p>t1 ?is new ph5om5on c be pr$icat$ on ! assumption t magnetization consi/s 9 a motion ( electrons 9 ! molecules ( ! metal4 ,?is all[s u to assign ! direction ( ! effects1 b = ! sake ( tru? ,i mu/ add t bo? ! rotation produc$ by ! magnetization ( rods & ! magnetization caus$ by a rotation are1 = "s reason :i* we d n yet und]/&1 only ab\t half :at ! !ory ( electrons _h l$ u to expect4 ,9 ,maxwell's "t ! electron 0 unkn[n & ! me*anism ( conduction 0 ev5 m my/]i\s ?an x is n[4 ,x 0 precisely = ?is reason t he atta*$ s mu* importance to ! ph5om5a to :i* ,i h drawn y\r att5tion4 ,he says 9 ?is connexion3 ,x appe>s to me t :ile we d]ive great advantage f ! recognition ( ! _m analogies betwe5 ! electric curr5t & a curr5t ( a mat]ial fluid1 we mu/ c>efully avoid mak+ any assumption n w>rant$ by exp]im5tal evid5ce1 & t !re is1 z yet1 no exp]im5tal evid5ce to %ew :e!r ! electric curr5t is really a curr5t ( a mat]ial sub/ance1 or a d\ble curr5t1 or :e!r its velocity is great or small z measur$ 9 feet p] second4 ,a k ( ^! ?+s w\ld am\nt to at lea/ ! beg9n+ ( a complete dynamical !ory ( electricity1 9 :i* we %\ld reg>d electrical action1 n1 z 9 ?is treatise1 z a ph5om5on due to an unkn[n cause1 subject only to ! g5]al laws ( dynamics1 b z ! result ( kn[n motions ( kn[n portions ( matt]1 9 :i* n only ! total effects & f9al results1 b ! :ole int]m$iate me*anism & details ( ! motion1 are tak5 z ! object ( /udy4 ,maxwell did n always express himself s cauti\sly3 at o!r times he did n %r9k f imag9+ an elaborate me*anical model4 ,all physici/s "k its pr9cipal features4 ,! magnetic en]gy is consid]$ z a true vis viva1 ! magnetic field be+ ! seat ( invisible motions1 rotations ( small p>ticles ab\t ! l9es ( =ce4 ,! sy/em ( ^! p>ticles may be comp>$ to a :eelwork1 & ,maxwell has to expla9 h[ x c be t all ! :eels 9 an elem5t ( volume are rotat+ 9 ! same direction4 ,?is %[s t ! motion is n transmitt$ directly f "o :eel to ! next4 ,s ,maxwell is l$ to assume t betwe5 ^! :eels ( ! magnetic field & 9 contact ) !m1 !re are small] ones :i* transmit ! motion 9 ! mann] ( friction :eels4 ,:at ,i call$ :eels might 9 reality be sph]es capable ( turn+ ab\t axes 9 any direction &1 z ,maxwell %[$1 a sy/em ( ?is k9d is am5able to ma!matical analysis4 ,9 8 image ! friction :eels repres5t :at we call electricity2 9 a conductor we mu/ conceive !m to be freely movable1 :]eas ! c5tres ( ! l>g] :eels or balls h fix$ positions4 ,x is easily se5 h[ a motion ( translation imp>t$ to ! friction :eels c give rise to rotations ( ! l>g] balls2 ?is is h[ a curr5t produces a magnetic field4 ,o!r ph5om5a c be expla9$ on ! same l9es2 x is suffici5t = ?is t ,maxwell's equations c be d$uc$ by means ( ! model4 ,simil> models h be5 inv5t$1 ) m or less success1 by o!r physici/s1 9 su* quantities1 ev51 t ne>ly all t c\ld be done has be5 tri$4 ,! outcome ( ^! v>i\s attempts is1 9 my op9ion1 ?is3 t we mu/ admit ! possibility ( me*anical repres5tations1 a possibility t is already %[n by ! fact t ! =mulae ( ! electromagnetic field c be giv5 ! =m ( ! g5]al equations ( dynamics4 ,on ! o!r h& x _c be d5i$ t1 :5 we desire !m to be applicable to a comp>atively wide range ( ph5om5a1 ! !ories t h be5 propos$ become s complicat$ t !y c give u b little satisfaction4 ,s1 ,i ?9k1 ! majority ( physici/s w agree to atta* less importance to me*anical models & ev5 to ! g5]al dynamical analogies1 h[ev] helpful ^! may be to u1 ?an to ,maxwell's equations t summ>ize s admirably all t is ess5tial4 ,w x be possible to ma9ta9 ^! equations8 ,i am n ?9k+ "h ( ! comp>atively slight modifications t h be5 f\nd necess>y 9 ! !ory ( relativity2 & by :i* we expla91 = in/ance1 ! curvature ( a ray ( light 9 a gravitation field4 ,a great] & really s]i\s dang] is ?reat5+ f ! side ( ! quantum !ory1 = ! exi/5ce ( am\nts ( en]gy t rema9 conc5trat$ 9 small spaces dur+ _! propagation1 to :i* sev]al ph5om5a seem to po9t1 is 9 absolute contradiction to ,maxwell's equations4 ,h[ev] ?is may be1 ev5 if fur!r developm5t %\ld require pr(\nd alt]ations1 ,maxwell's !ory w always rema9 a /ep ( ! highe/ importance 9 ! progress ( physics4 ,b ,i mu/ n trespass any long] on y\r pati5ce4 ,let me r ?ank y = ! att5tion ) :i* y h li/5$ to :at ,i _h to say1 :i* 0 ma9ly int5d$ z a tribute to ! memory ( "o ( ! greate/ ,ma/]s ( ,sci5ce4