﻿Dispersion, 
  and 
  Surface- 
  Colour 
  of 
  Selenium. 
  615 
  

  

  the 
  second 
  lithium 
  band. 
  The 
  first 
  dark 
  band 
  moved 
  nearer 
  to 
  

   the 
  thin 
  edge 
  with 
  decreasing 
  wave-length. 
  A 
  screen 
  consisting 
  

   of 
  a 
  glass 
  cell 
  filled 
  with 
  a 
  solution 
  of 
  cuprammonium 
  and 
  a 
  

   sheet 
  of 
  dense 
  cobalt 
  glass 
  was 
  now 
  placed 
  in 
  front 
  of 
  a 
  

   Welsbach 
  light. 
  This 
  gave 
  a 
  deep 
  blue 
  light 
  free 
  from 
  red. 
  

   In 
  this 
  case 
  the 
  first 
  dark 
  band 
  was 
  the 
  only 
  one 
  that 
  

   appeared, 
  but 
  it 
  was 
  distinctly 
  nearer 
  the 
  edge 
  than 
  in 
  the 
  

   case 
  of 
  the 
  thallium 
  light, 
  which 
  shows 
  that 
  the 
  wave-length 
  

   of 
  the 
  blue 
  light 
  in 
  the 
  selenium 
  is 
  considerably 
  less 
  than 
  the 
  

   wave-length 
  of 
  the 
  thallium 
  light, 
  which 
  would 
  not 
  be 
  the 
  

   case 
  if 
  the 
  refractive 
  index 
  for 
  the 
  blue 
  was 
  considerably 
  less 
  

   than 
  for 
  the 
  green. 
  In 
  this 
  last 
  case 
  the 
  band 
  was 
  too 
  ill- 
  

   defined 
  to 
  admit 
  of 
  an 
  accurate 
  measurement 
  of 
  its 
  position. 
  

  

  A 
  preliminary 
  examination 
  of 
  the 
  light 
  transmitted 
  by 
  a 
  

   very 
  thin 
  film 
  squeezed 
  out 
  between 
  two 
  plates 
  of 
  quartz, 
  by 
  

   means 
  of 
  a 
  quartz 
  spectrograph, 
  showed 
  that 
  there 
  was 
  no 
  

   return 
  of 
  transparency 
  in 
  the 
  ultra-violet, 
  at 
  least 
  up 
  to 
  

   wave-length 
  '00028 
  or 
  thereabouts. 
  To 
  determine 
  if 
  possible 
  

   the 
  position 
  of 
  the 
  centre 
  of 
  the 
  band, 
  which 
  gives 
  us 
  \ 
  m 
  in 
  

   the 
  dispersion 
  formula, 
  a 
  series 
  of 
  photometric 
  measurements 
  

   were 
  made 
  of 
  the 
  visible 
  and 
  ultra-violet 
  light 
  transmitted 
  

   by 
  a 
  thin 
  film 
  deposited 
  in 
  vacuo 
  by 
  means 
  of 
  the 
  cathode 
  

   discharge. 
  In 
  the 
  visible 
  spectrum 
  measurements 
  were 
  made 
  

   by 
  placing 
  the 
  glass 
  plate, 
  partly 
  covered 
  by 
  the 
  film, 
  in 
  

   contact 
  with 
  the 
  double 
  slit 
  of 
  a 
  spectrometer, 
  the 
  line 
  ter- 
  

   minating- 
  the 
  film 
  being 
  brought 
  into 
  coincidence 
  with 
  the 
  

   junction 
  of 
  the 
  slits. 
  The 
  thickness 
  of 
  the 
  film 
  was 
  determined 
  

   by 
  means 
  of 
  fringe 
  displacements, 
  making 
  use 
  of 
  the 
  data 
  

   obtained 
  with 
  the 
  selenium 
  prisms. 
  

  

  Owing 
  to 
  the 
  difficulty 
  of 
  obtaining 
  accurate 
  data 
  in 
  the 
  

   blue 
  and 
  violet 
  by 
  the 
  visual 
  method, 
  and 
  to 
  enable 
  us 
  to 
  

   carry 
  on 
  a 
  study 
  of 
  the 
  absorption 
  in 
  the 
  ultra-violet, 
  a 
  photo- 
  

   graphic 
  method 
  was 
  adopted, 
  the 
  details 
  of 
  which 
  were 
  worked 
  

   out 
  largely 
  by 
  Mr. 
  Pfund. 
  

  

  The 
  double 
  slit 
  was 
  removed 
  and 
  a 
  single 
  slit 
  put 
  in 
  its 
  

   place, 
  which 
  was 
  opened 
  to 
  a 
  width 
  of 
  0*5 
  mm. 
  The 
  ocular 
  

   slit 
  of 
  the 
  spectrophotometer 
  was 
  also 
  opened 
  to 
  the 
  same 
  

   width. 
  A 
  short 
  distance 
  behind 
  the 
  ocular 
  slit, 
  the 
  photo- 
  

   graphic 
  plate 
  was 
  mounted 
  with 
  the 
  selenium 
  film 
  covering 
  

   the 
  upper 
  portion. 
  A 
  part 
  of 
  the 
  plate 
  was 
  therefore 
  exposed 
  

   to 
  light 
  coming 
  through 
  clear 
  glass, 
  and 
  part 
  to 
  light 
  which 
  

   had 
  passed 
  through 
  the 
  film 
  of 
  selenium. 
  This 
  method 
  was 
  

   found 
  to 
  give 
  better 
  results 
  than 
  a 
  previous 
  method 
  which 
  we 
  

   tried, 
  which 
  consisted 
  in 
  substituting 
  a 
  photographic 
  plate 
  

   for 
  the 
  eyepiece 
  of 
  the 
  instrument, 
  and 
  using 
  the 
  Vierordt 
  

   slit 
  as 
  in 
  the 
  visual 
  method. 
  To 
  determine 
  the 
  ratio 
  of 
  the 
  

  

  