﻿Polarization 
  of 
  tjie 
  Light 
  of 
  Resonance 
  Spectra. 
  847 
  

  

  excitation 
  was 
  by 
  the 
  green 
  line 
  alone. 
  This 
  was 
  to 
  avoid 
  

   the 
  complication 
  resulting 
  from 
  simultaneous 
  excitation 
  at 
  

   three 
  points 
  of 
  the 
  spectrum, 
  as 
  I 
  wished 
  particularly 
  to 
  

   ascertain 
  whether 
  the 
  polarization 
  was 
  as 
  strong 
  for 
  the 
  lines 
  

   far 
  removed 
  from 
  the 
  exciting 
  line, 
  as 
  for 
  those 
  which 
  were 
  

   close 
  to 
  it. 
  

  

  Every 
  precaution 
  was 
  taken 
  to 
  avoid 
  scattered 
  light 
  and 
  

   to 
  have 
  the 
  luminous 
  iodine 
  vapour 
  seen 
  against 
  an 
  absolutely 
  

   black 
  background. 
  

  

  In 
  front 
  of 
  the 
  tube 
  a 
  screen 
  was 
  mounted 
  perforated 
  with 
  

   a 
  vertical 
  slit 
  about 
  1 
  mm. 
  in 
  width. 
  Close 
  against 
  this 
  slit 
  

   the 
  Fresnel 
  double 
  prism 
  was 
  placed 
  so 
  oriented 
  that 
  the 
  

   fringes, 
  seen 
  by 
  means 
  of 
  a 
  Nicol 
  prism, 
  ran 
  in 
  a 
  horizontal 
  

   direction. 
  These 
  fringes 
  are 
  due 
  to 
  the 
  circumstance 
  that 
  

   the 
  plane 
  of 
  polarization 
  at 
  each 
  point 
  of 
  the 
  slit 
  is 
  rotated 
  

   by 
  an 
  amount 
  depending 
  upon 
  the 
  difference 
  between 
  the 
  

   thicknesses 
  of 
  the 
  right- 
  and 
  ] 
  eft-handed 
  quartz 
  elements 
  at 
  

   the 
  point 
  in 
  question. 
  If 
  no 
  polarization 
  is 
  present, 
  no 
  bands 
  

   are 
  seen, 
  and 
  by 
  compensating 
  the 
  polarization 
  with 
  one 
  

   or 
  more 
  glass 
  plates 
  turned 
  through 
  a 
  known 
  angle, 
  the 
  

   percentage 
  of 
  polarization 
  can 
  be 
  determined 
  by 
  the 
  usual 
  

   formula 
  (Physical 
  Optics, 
  2nd 
  ed. 
  p. 
  299). 
  A 
  long-focus 
  

   lens 
  and 
  a 
  direct-vision 
  prism 
  spread 
  out 
  the 
  image 
  of 
  the 
  

   slit 
  into 
  a 
  spectrum, 
  and 
  the 
  polarization 
  fringes 
  were 
  found 
  

   to 
  cut 
  across 
  all 
  of 
  the 
  lines 
  of 
  the 
  resonance 
  spectrum 
  thus 
  

   formed. 
  By 
  means 
  of 
  the 
  compensating 
  plates 
  it 
  was 
  found 
  

   that 
  the 
  percentage 
  of 
  polarization 
  had 
  practically 
  the 
  same 
  

   value 
  in 
  all 
  of 
  the 
  lines 
  including 
  the 
  exciting 
  or 
  R.R. 
  line 
  

   (see 
  previous 
  paper). 
  

  

  If 
  the 
  exciting 
  light 
  is 
  polarized 
  with 
  its 
  electric 
  vector 
  

   vertical, 
  before 
  it 
  enters 
  the 
  iodine 
  vapour, 
  the 
  fringes 
  

   become 
  more 
  distinct; 
  if 
  polarized 
  with 
  its 
  electric 
  vector 
  

   horizontal, 
  the 
  fringes 
  disappear, 
  as 
  would 
  be 
  expected 
  on 
  any 
  

   theory. 
  Two 
  photographs 
  of 
  the 
  phenomenon 
  are 
  reproduced 
  

   in 
  PL 
  XVII. 
  They 
  have 
  not 
  been 
  retouched 
  or 
  intensified 
  in 
  

   any 
  way, 
  and 
  may 
  be 
  taken 
  as 
  furnishing 
  an 
  exact 
  record 
  of 
  

   the 
  distinctness 
  of 
  the 
  polarization 
  fringes. 
  The 
  percentage 
  

   of 
  polarization 
  is 
  not 
  nearly 
  as 
  high 
  as 
  is 
  the 
  case 
  with 
  the 
  

   vapours 
  of 
  sodium 
  and 
  potassium, 
  for 
  which 
  I 
  have 
  observed 
  

   30 
  per 
  cent, 
  of 
  polarization 
  with 
  the 
  exciting 
  light 
  polarized. 
  

   L. 
  Dunoyer 
  has 
  observed 
  48 
  per 
  cent, 
  under 
  similar 
  con- 
  

   ditions, 
  working 
  however 
  with 
  the 
  vapour 
  in 
  glass 
  bulbs. 
  In 
  

   my 
  experiments 
  I 
  used 
  a 
  steel 
  tube 
  provided 
  with 
  a 
  lateral 
  

   branch 
  for 
  observation. 
  

  

  In 
  the 
  case 
  of 
  iodine 
  excited 
  by 
  the 
  green 
  mercury 
  line, 
  

  

  