﻿WHAT 
  IS 
  LIGHT? 
  COMPTON" 
  227 
  

  

  respect 
  the 
  solution 
  is 
  indetermiiiate; 
  but 
  its 
  indeterminateness 
  cor- 
  

   responds 
  to 
  an 
  indeterminateness 
  in 
  the 
  experiment 
  itself. 
  There 
  is 
  

   no 
  way 
  of 
  performing 
  the 
  experiment 
  so 
  as 
  to 
  make 
  the 
  electron 
  recoil 
  

   in 
  a 
  definite 
  direction 
  as 
  a 
  result 
  of 
  an 
  encounter 
  with 
  a 
  photon. 
  It 
  is 
  

   a 
  beauty 
  of 
  the 
  theory 
  that 
  it 
  is 
  determinate 
  only 
  where 
  the 
  experi- 
  

   ment 
  itself 
  is 
  determinate, 
  and 
  leaves 
  arbitrary 
  those 
  parameters 
  

   which 
  the 
  experiment 
  is 
  incapable 
  of 
  defining. 
  

  

  It 
  is 
  not 
  usually 
  possible 
  to 
  describe 
  the 
  motion 
  of 
  either 
  a 
  beam 
  of 
  

   light 
  or 
  a 
  beam 
  of 
  electrons 
  without 
  introducing 
  both 
  the 
  concepts 
  of 
  

   particles 
  and 
  waves. 
  There 
  are 
  certain 
  localized 
  regions 
  in 
  which 
  at 
  a 
  

   certain 
  moment 
  energy 
  exists, 
  and 
  this 
  may 
  be 
  taken 
  as 
  a 
  definition 
  

   of 
  what 
  we 
  mean 
  by 
  a 
  particle. 
  But 
  in 
  predicting 
  where 
  these 
  localized 
  

   positions 
  are 
  to 
  be 
  at 
  a 
  later 
  instant, 
  a 
  consideration 
  of 
  the 
  propaga- 
  

   tion 
  of 
  the 
  corresponding 
  waves 
  is 
  usually 
  our 
  most 
  satisfactory 
  mode 
  

   of 
  attack. 
  

  

  Attention 
  should 
  be 
  called 
  to 
  the 
  fact 
  that 
  the 
  electromagnetic 
  

   waves 
  and 
  the 
  De 
  Broglie 
  waves 
  are 
  according 
  to 
  this 
  theory 
  waves 
  of 
  

   probability. 
  Consider 
  as 
  an 
  example 
  the 
  difi^raction 
  pattern 
  of 
  a 
  beam 
  

   of 
  light 
  or 
  of 
  electrons, 
  reflected 
  from 
  a 
  ruled 
  grating, 
  and 
  falling 
  on 
  

   a 
  photographic 
  plate. 
  In 
  the 
  intense 
  portion 
  of 
  the 
  diffraction 
  pattern 
  

   there 
  is 
  a 
  high 
  probability 
  that 
  a 
  grain 
  of 
  the 
  photographic 
  plate 
  will 
  be 
  

   affected. 
  In 
  corpuscular 
  language, 
  there 
  is 
  a 
  high 
  probability 
  that 
  a 
  

   photon 
  or 
  electron, 
  as 
  the 
  case 
  may 
  be, 
  will 
  strike 
  this 
  portion 
  of 
  the 
  

   plate. 
  Where 
  the 
  diffraction 
  pattern 
  is 
  of 
  zero 
  intensity, 
  the 
  proba- 
  

   bility 
  of 
  a 
  particle 
  striking 
  is 
  zero, 
  and 
  the 
  plate 
  is 
  unaffected. 
  Thus 
  

   there 
  is 
  a 
  high 
  probability 
  that 
  a 
  photon 
  will 
  be 
  present 
  where 
  the 
  

   "intensity" 
  of 
  an 
  electromagnetic 
  wave 
  is 
  great, 
  and 
  a 
  lesser 
  prob- 
  

   ability 
  where 
  this 
  "intensity" 
  is 
  smaller. 
  

  

  It 
  is 
  a 
  corollary 
  that 
  the 
  energy 
  of 
  the 
  radiation 
  lies 
  in 
  the 
  photons, 
  

   and 
  not 
  in 
  the 
  waves. 
  For 
  we 
  mean 
  by 
  energy 
  the 
  ability 
  to 
  do 
  work, 
  

   and 
  we 
  find 
  that 
  when 
  radiation 
  does 
  anything 
  it 
  acts 
  in 
  particles. 
  

  

  In 
  this 
  connection 
  it 
  may 
  be 
  noted 
  that 
  this 
  wave-mechanics 
  

   theory 
  does 
  not 
  enable 
  us 
  to 
  locate 
  a 
  photon 
  or 
  an 
  electron 
  definitely 
  

   except 
  at 
  the 
  instant 
  at 
  which 
  it 
  does 
  something. 
  When 
  it 
  activates 
  

   a 
  grain 
  on 
  a 
  photographic 
  plate, 
  or 
  ionizes 
  an 
  atom 
  which 
  may 
  be 
  

   observed 
  in 
  a 
  cloud 
  expansion 
  chamber, 
  we 
  can 
  say 
  that 
  the 
  particle 
  

   was 
  at 
  that 
  point 
  at 
  the 
  instant 
  of 
  the 
  event. 
  But 
  in 
  between 
  such 
  

   events 
  the 
  particle 
  can 
  not 
  be 
  definitely 
  located. 
  Some 
  positions 
  

   are 
  more 
  probable 
  than 
  others, 
  in 
  proportion 
  as 
  the 
  corresponding 
  

   wave 
  is 
  more 
  intense 
  in 
  these 
  positions. 
  But 
  there 
  is 
  no 
  definite 
  

   position 
  that 
  can 
  be 
  assigned 
  to 
  the 
  particle 
  in 
  between 
  its 
  actions 
  

   on 
  other 
  particles. 
  Thus 
  it 
  becomes 
  meaningless 
  to 
  attempt 
  to 
  

   assign 
  any 
  definite 
  path 
  to 
  a 
  particle. 
  It 
  is 
  like 
  assigning 
  a 
  definite 
  

   path 
  to 
  a 
  ray 
  of 
  light: 
  The 
  more 
  sharply 
  we 
  try 
  to 
  define 
  it 
  by 
  narrow 
  

   slits, 
  the 
  more 
  widely 
  the 
  ray 
  is 
  spread 
  by 
  diffraction. 
  

  

  82322—30 
  16 
  

  

  