﻿410 
  Anomalous 
  Dispersion 
  hy 
  Metallic 
  Vapours. 
  

  

  case, 
  however, 
  there 
  is, 
  under 
  the 
  conditions 
  of 
  the 
  experi" 
  

   ment, 
  a 
  considerable 
  amount 
  of 
  hydrogen 
  present 
  even 
  

   when 
  the 
  tube 
  is 
  exhausted 
  initially. 
  A 
  certain 
  amount 
  of 
  

   hydroxide 
  is 
  inevitably 
  introduced 
  with 
  the 
  metal, 
  and 
  this 
  

   on 
  heating 
  with 
  the 
  free 
  metal 
  evolves 
  hydrogen 
  ; 
  so 
  that 
  

   presumably 
  a 
  steady 
  condition 
  is 
  reached 
  when 
  the 
  pressure 
  

   of 
  the 
  hydrogen 
  reaches 
  the 
  value 
  corresponding 
  to 
  the 
  

   equilibrium 
  pressure 
  of 
  the 
  metallic 
  hydride. 
  

  

  In 
  the 
  case 
  of 
  sodium 
  vapour 
  Wood 
  has 
  shown 
  that 
  

   anomalous 
  dispersion 
  takes 
  place 
  at 
  the 
  first 
  three 
  pairs 
  of 
  

   the 
  principal 
  series 
  *, 
  and 
  further 
  that 
  the 
  absorption 
  spec- 
  

   trum 
  consists 
  of 
  lines 
  o£ 
  the 
  principal 
  series 
  t? 
  together 
  

   with 
  a 
  channelled 
  spectrum. 
  If 
  the 
  absorption 
  spectrum 
  be 
  

   observed 
  as 
  the 
  temperature 
  of 
  the 
  sodium 
  vapour 
  is 
  raised 
  

   and 
  its 
  density 
  increased, 
  absorption 
  is 
  observable 
  of 
  gradu- 
  

   ally 
  increasing 
  amount 
  near 
  the 
  D 
  lines 
  and 
  soon 
  the 
  whole 
  

   of 
  the 
  green 
  disappears, 
  but 
  there 
  is 
  no 
  apparent 
  connexion 
  

   between 
  this 
  absorption 
  and 
  the 
  position 
  of 
  the 
  green 
  lines 
  

   of 
  sodium 
  ; 
  the 
  maximum 
  absorption 
  in 
  the 
  bands 
  which 
  

   appear 
  does 
  not 
  coincide 
  even 
  approximately 
  with 
  the 
  green 
  

   line. 
  The 
  behaviours 
  of 
  sodium 
  and 
  potassium 
  are 
  thus 
  very 
  

   much 
  alike. 
  In 
  the 
  case 
  of 
  lithium, 
  no 
  absorption 
  except 
  in 
  

   the 
  region 
  of 
  the 
  red 
  line 
  has 
  as 
  yet 
  been 
  detected. 
  This 
  

   may 
  simply 
  be 
  due 
  to 
  the 
  fact 
  that 
  a 
  small 
  quantity 
  of 
  

   lithium 
  was 
  used 
  and 
  a 
  much 
  higher 
  temperature 
  may 
  be 
  

   required 
  to 
  show 
  effects 
  analogous 
  to 
  those 
  observed 
  in 
  the 
  

   cases 
  of 
  sodium 
  and 
  potassium. 
  The 
  absolute 
  amount 
  of 
  

   metallic 
  vapour 
  present 
  to 
  produce 
  the 
  dispersions 
  observed 
  

   must 
  be 
  very 
  small, 
  as 
  a 
  potassium 
  tube 
  can 
  be 
  used 
  over 
  

   and 
  over 
  again 
  without 
  apparent 
  change 
  in 
  the 
  amount 
  of 
  

   potassium 
  remaining 
  at 
  the 
  bottom 
  of 
  the 
  tube. 
  The 
  potas- 
  

   sium 
  does 
  slowly 
  distil 
  to 
  the 
  cooler 
  parts 
  of 
  the 
  tube, 
  but 
  

   the 
  distillation 
  is 
  extremely 
  slow 
  and 
  the 
  tube 
  may 
  be 
  used 
  

   for 
  many 
  hours 
  without 
  any 
  diminution 
  of 
  effect. 
  

  

  It 
  can 
  be 
  seen 
  that 
  the 
  amount 
  of 
  dispersion 
  due 
  to 
  the 
  

   potassium 
  vapour 
  is 
  different 
  in 
  the 
  neighbourhood 
  of 
  the 
  two 
  

   red 
  lines, 
  as 
  we 
  should 
  expect, 
  seeing 
  that 
  one 
  of 
  the 
  lines 
  is 
  

   more 
  intense 
  than 
  the 
  other 
  in 
  the 
  emission 
  spectrum. 
  The 
  

   dispersion 
  spectrum 
  is 
  thus 
  unsymmetrical 
  about 
  the 
  mean 
  

   position 
  of 
  the 
  two 
  lines. 
  In 
  the 
  case 
  of 
  sodium 
  the 
  same 
  

   phenomenon 
  occurs, 
  and 
  apparently 
  the 
  same 
  is 
  true 
  of 
  the 
  

   dispersion 
  due 
  to 
  lithium. 
  The 
  question 
  is 
  one 
  not 
  easy 
  to 
  

   decide 
  definitely 
  as 
  the 
  dispersion 
  is 
  small 
  and 
  the 
  apparent 
  

   effect 
  is 
  masked 
  by 
  the 
  differences 
  in 
  absorption 
  that 
  occur 
  ; 
  

  

  * 
  Phil. 
  Mag. 
  1904, 
  viii. 
  p. 
  296. 
  

   t 
  Phil. 
  Mag. 
  Dec. 
  1908. 
  

  

  