﻿218 
  ANNUAL 
  EEPORT 
  SMITHSONIAN 
  INSTITUTION, 
  192 
  9 
  

  

  When 
  the 
  physicist 
  speaks 
  of 
  Hght, 
  he 
  refers 
  to 
  all 
  the 
  radiations 
  

   included 
  in 
  this 
  vast 
  range. 
  We 
  believe 
  that 
  they 
  are 
  all 
  the 
  same 
  

   kind 
  of 
  thing, 
  and 
  that 
  anything 
  which 
  may 
  be 
  said 
  about 
  the 
  nature 
  

   of 
  the 
  rays 
  in 
  one 
  part 
  of 
  this 
  region 
  is 
  equally 
  true 
  of 
  the 
  rest. 
  

  

  LIGHT 
  CONSISTS 
  OF 
  WAVES 
  

  

  There 
  are 
  many 
  ways 
  in 
  which 
  light 
  acts 
  like 
  a 
  wave 
  in 
  an 
  elastic 
  

   medium. 
  Such 
  elastic 
  waves 
  move 
  with 
  a 
  speed 
  which 
  is 
  the 
  same 
  for 
  

   all 
  wave 
  lengths 
  and 
  all 
  intensities, 
  just 
  as 
  does 
  light. 
  Waves, 
  like 
  

   Hght 
  rays, 
  can 
  be 
  reflected 
  and 
  refracted. 
  The 
  polarization 
  of 
  light 
  

   is 
  a 
  property 
  characteristic 
  of 
  the 
  transverse 
  waves 
  in 
  an 
  elastic 
  solid. 
  

   It 
  is 
  true 
  that 
  if 
  one 
  examines 
  the 
  constancy 
  of 
  the 
  speed 
  of 
  light 
  in 
  

   detail, 
  difficulties 
  arise; 
  for 
  it 
  is 
  found 
  that 
  its 
  speed 
  is 
  the 
  same 
  rela- 
  

   tive 
  to 
  an 
  observer 
  no 
  matter 
  how 
  fast 
  the 
  observer 
  is 
  going. 
  This 
  

   would 
  not 
  be 
  true 
  if 
  light 
  were 
  a 
  wave 
  in 
  an 
  ordinary 
  elastic 
  medium. 
  

   Maxwell's 
  identification 
  of 
  light 
  as 
  electromagnetic 
  waves, 
  however, 
  

   removes 
  this 
  difficulty. 
  

  

  The 
  crucial 
  test 
  for 
  the 
  existence 
  of 
  waves, 
  however, 
  has 
  always 
  been 
  

   that 
  of 
  diffraction 
  and 
  interference. 
  Imagine 
  a 
  row 
  of 
  pebbles 
  dropped 
  

   into 
  a 
  pond 
  at 
  the 
  same 
  instant. 
  The 
  effect 
  would 
  be 
  similar 
  to 
  that 
  

   showTi 
  in 
  Plate 
  1 
  , 
  Figure 
  1 
  . 
  In 
  this 
  figure 
  we 
  picture 
  a 
  series 
  of 
  waves 
  

   passing 
  through 
  a 
  succession 
  of 
  openings 
  in 
  a 
  grid. 
  After 
  passing 
  

   through, 
  the 
  crests 
  of 
  the 
  emerging 
  wavelets 
  recombine 
  to 
  form 
  a 
  new 
  

   wave 
  going 
  straight 
  ahead. 
  But 
  in 
  addition, 
  the 
  wavelet 
  just 
  emerg- 
  

   ing 
  from 
  one 
  opening 
  may 
  combine 
  with 
  the 
  first 
  wave 
  from 
  the 
  next 
  

   opening, 
  the 
  second 
  from 
  the 
  next, 
  and 
  so 
  on, 
  forming 
  a 
  new 
  wave 
  

   front 
  inclined 
  at 
  a 
  definite 
  angle 
  to 
  the 
  first. 
  The 
  angle 
  between 
  these 
  

   two 
  waves, 
  as 
  will 
  be 
  seen 
  from 
  this 
  diagram, 
  is 
  determined 
  by 
  the 
  dis- 
  

   tance 
  between 
  successive 
  waves, 
  i. 
  e., 
  the 
  wave 
  length, 
  and 
  by 
  the 
  

   distance 
  between 
  successive 
  openings 
  in 
  the 
  grid. 
  The 
  figure 
  at 
  the 
  

   right 
  shows 
  how 
  the 
  emergent 
  wave 
  may 
  combine 
  with 
  the 
  second 
  wave 
  

   from 
  the 
  adjacent 
  opening, 
  the 
  fourth 
  from 
  the 
  second 
  opening, 
  and 
  so 
  

   on, 
  and 
  form 
  a 
  wave 
  front 
  propagated 
  at 
  a 
  larger 
  angle. 
  

  

  That 
  such 
  a 
  variety 
  of 
  wave 
  formation 
  is 
  not 
  purely 
  imaginary 
  is 
  

   shown 
  in 
  Plate 
  1, 
  Figure 
  2, 
  which 
  is 
  a 
  photograph 
  of 
  ripples 
  on 
  the 
  sur- 
  

   face 
  of 
  mercury, 
  taken 
  after 
  they 
  have 
  passed 
  through 
  a 
  comblike 
  

   grid. 
  Notice 
  how 
  one 
  group 
  of 
  waves 
  combines 
  to 
  form 
  a 
  wave 
  front 
  

   going 
  straight 
  ahead. 
  But 
  in 
  addition, 
  on 
  either 
  side 
  of 
  the 
  central 
  

   beam, 
  we 
  find 
  two 
  beams 
  forming 
  where 
  the 
  paths 
  from 
  successive 
  

   openings 
  in 
  the 
  grid 
  differ 
  by 
  one 
  wave 
  length. 
  Out 
  at 
  a 
  large 
  angle 
  

   we 
  see 
  even 
  the 
  second 
  order 
  of 
  the 
  diffracted 
  beam. 
  

  

  If 
  we 
  were 
  unable 
  to 
  see 
  the 
  separate 
  waves, 
  but 
  knew 
  the 
  kind 
  of 
  

   grid 
  through 
  which 
  the 
  beam 
  of 
  ripples 
  had 
  passed, 
  not 
  only 
  could 
  we 
  

   say 
  that 
  this 
  is 
  the 
  way 
  the 
  beam 
  should 
  be 
  split 
  up 
  if 
  it 
  consists 
  of 
  

   waves, 
  but 
  we 
  could 
  even 
  tell 
  what 
  the 
  wave 
  length 
  of 
  the 
  ripples 
  must 
  

   be 
  in 
  order 
  to 
  give 
  these 
  particular 
  angles 
  between 
  the 
  diffracted 
  beams. 
  

  

  