搜索
A.Einstein,Ann.Phys.17,1321905ConcerninganHeuristicPointofViewTowardtheEmissionandTransformationofLightA.EinsteinBern,17March1905(ReceivedMarch18,1905)TranslationintoEnglishAmericanJournalofPhysics,v.33,n.5,May1965——¦♦¦——AprofoundformaldistinctionexistsbetweenthetheoreticalconceptswhichphysicistshaveformedregardinggasesandotherponderablebodiesandtheMaxwelliantheoryofelectromagneticprocessesinso–calledemptyspace.Whileweconsiderthestateofabodytobecompletelydeterminedbythepositionsandvelocitiesofaverylarge,yetfinite,numberofatomsandelectrons,wemakeuseofcontinuousspatialfunctionstodescribetheelectromagneticstateofagivenvolume,andafinitenumberofparameterscannotberegardedassufficientforthecompletedeterminationofsuchastate.AccordingtotheMaxwelliantheory,energyistobeconsideredacon-tinuousspatialfunctioninthecaseofallpurelyelectromagneticphenomenaincludinglight,whiletheenergyofaponderableobjectshould,accordingtothepresentconceptionsofphysicists,berepresentedasasumcarriedovertheatomsandelectrons.Theenergyofaponderablebodycannotbesubdividedintoarbitrarilymanyorarbitrarilysmallparts,whiletheenergyofabeamoflightfromapointsource(accordingtotheMaxwelliantheoryoflightor,moregenerally,accordingtoanywavetheory)iscontinuouslyspreadaneverincreasingvolume.Thewavetheoryoflight,whichoperateswithcontinuousspatialfunc-tions,hasworkedwellintherepresentationofpurelyopticalphenomena1andwillprobablyneverbereplacedbyanothertheory.Itshouldbekeptinmind,however,thattheopticalobservationsrefertotimeaveragesratherthaninstantaneousvalues.Inspiteofthecompleteexperimentalconfirma-tionofthetheoryasappliedtodiffraction,reflection,refraction,dispersion,etc.,itisstillconceivablethatthetheoryoflightwhichoperateswithcon-tinuousspatialfunctionsmayleadtocontradictionswithexperiencewhenitisappliedtothephenomenaofemissionandtransformationoflight.Itseemstomethattheobservationsassociatedwithblackbodyradia-tion,fluorescence,theproductionofcathoderaysbyultravioletlight,andotherrelatedphenomenaconnectedwiththeemissionortransformationoflightaremorereadilyunderstoodifoneassumesthattheenergyoflightisdiscontinuouslydistributedinspace.Inaccordancewiththeassumptiontobeconsideredhere,theenergyofalightrayspreadingoutfromapointsourceisnotcontinuouslydistributedoveranincreasingspacebutconsistsofafinitenumberofenergyquantawhicharelocalizedatpointsinspace,whichmovewithoutdividing,andwhichcanonlybeproducedandabsorbedascompleteunits.InthefollowingIwishtopresentthelineofthoughtandthefactswhichhaveledmetothispointofview,hopingthatthisapproachmaybeusefultosomeinvestigatorsintheirresearch.1.ConcerningaDifficultywithRegardtotheThe-oryofBlackbodyRadiationWestartfirstwiththepointofviewtakenintheMaxwellianandtheelectrontheoriesandconsiderthefollowingcase.Inaspaceenclosedbycompletelyreflectingwalls,lettherebeanumberofgasmoleculesandelectronswhicharefreetomoveandwhichexertconservativeforcesoneachotheroncloseapproach:i.e.theycancollidewitheachotherlikemoleculesinthekinetictheoryofgases.1Furthermore,lettherebeanumberofelectronswhichareboundtowidelyseparatedpointsbyforcesproportionaltotheirdistancesfromthesepoints.Theboundelectronsarealsotoparticipateinconserva-tiveinteractionswiththefreemoleculesandelectronswhenthelattercome1Thisassumptionisequivalenttothesuppositionthattheaveragekineticenergiesofgasmoleculesandelectronsareequaltoeachotheratthermalequilibrium.Itiswellknownthat,withthehelpofthisassumption,HerrDrudederivedatheoreticalexpressionfortheratioofthermalandelectricalconductivitiesofmetals.2veryclose.Wecalltheboundelectrons“oscillators”:theyemitandabsorbelectromagneticwavesofdefiniteperiods.Accordingtothepresentviewregardingtheoriginoflight,theradiationinthespaceweareconsidering(radiationwhichisfoundforthecaseofdynamicequilibriuminaccordancewiththeMaxwelliantheory)mustbeidenticalwiththeblackbodyradiation—atleastifoscillatorsofalltherelevantfrequenciesareconsideredtobepresent.Forthetimebeing,wedisregardtheradiationemittedandabsorbedbytheoscillatorsandinquireintotheconditionofdynamicalequilibriumassociatedwiththeinteraction(orcollision)ofmoleculesandelectrons.Thekinetictheoryofgasesassertsthattheaveragekineticenergyofanoscillatorelectronmustbeequaltotheaveragekineticenergyofatranslatinggasmolecule.Ifweseparatethemotionofanoscillatorelectronintothreecomponentsatanglestoeachother,wefindfortheaverageenergyEofoneoftheselinearcomponentstheexpressionE=(R/N)T,whereRdenotestheuniversalgasconstant.Ndenotesthenumberof“realmolecules”inagramequivalent,andTtheabsolutetemperature.TheenergyEisequaltotwo-thirdsthekineticenergyofafreemonatomicgasparticlebecauseoftheequalitythetimeaveragevaluesofthekineticandpotentialenergiesoftheoscillator.Ifthroughanycause—inourcasethroughradiationprocesses—itshouldoccurthattheenergyofanoscillatortakesonatime-averagevaluegreaterorlessthanE,thenthecollisionswiththefreeelectronsandmoleculeswouldleadtoagainorlossofenergybythegas,differentontheaveragefromzero.Therefore,inthecaseweareconsidering,dynamicequilibriumispossibleonlywheneachoscillatorhastheaverageenergyE.Weshallnowproceed