﻿Constitution 
  of 
  the 
  Mercury 
  Green 
  Line. 
  373 
  

  

  We 
  add 
  a 
  diagram 
  (fig. 
  2) 
  o£ 
  the 
  spectrum 
  of 
  the 
  line- 
  

   showing 
  the 
  relative 
  widths 
  of 
  the 
  components. 
  

  

  Fig. 
  2. 
  

  

  JUu_nJLJ 
  UU 
  

  

  With 
  the 
  lamps 
  used 
  we 
  found 
  no 
  change 
  in 
  the 
  pattern 
  

   after 
  the 
  lamps 
  had 
  been 
  running 
  for 
  even 
  six 
  hours 
  on 
  end. 
  

  

  Various 
  kinds 
  of 
  photographic 
  plates 
  were 
  tried. 
  The 
  

   best 
  results 
  were 
  obtained 
  from 
  " 
  Imperial 
  Orthochrome. 
  

   Special 
  Sensitive.'" 
  The 
  exposures 
  varied 
  from 
  twenty 
  to 
  

   eighty 
  minutes, 
  and 
  the 
  plates 
  were 
  all 
  developed 
  under 
  the 
  

   same 
  conditions. 
  

  

  Conclusions. 
  — 
  It 
  would 
  appear 
  from 
  these 
  investigations 
  

   that 
  the 
  discrepancies 
  in 
  the 
  positions 
  of 
  these 
  satellites 
  as 
  

   determined 
  by 
  different 
  observers 
  are 
  for 
  the 
  most 
  part 
  

   apparent, 
  not 
  real, 
  and 
  are 
  dependent 
  upon 
  the 
  difficulty 
  of 
  

   judging 
  the 
  centre 
  of 
  the 
  main 
  wide 
  line. 
  See 
  Table 
  II. 
  

  

  Janicki 
  and 
  Galitzin 
  and 
  Wilip 
  observed 
  that 
  a 
  change 
  of 
  

   pattern 
  occurred 
  when 
  the 
  lamp 
  reached 
  a 
  high 
  temperature. 
  

   Accounts 
  of 
  this 
  phenomenon 
  have 
  also 
  been 
  given 
  by 
  

   Stansfield 
  and 
  by 
  McLennan, 
  who 
  do 
  not 
  state 
  what 
  tem- 
  

   perature 
  was 
  attained. 
  Within 
  the 
  range 
  of 
  temperature 
  

   employed 
  by 
  us 
  no 
  such 
  effect 
  was 
  observed. 
  

  

  The 
  discrepancies 
  in 
  the 
  relative 
  brightnesses 
  of 
  corre- 
  

   sponding 
  satellites 
  in 
  the 
  different 
  cases 
  appear 
  not 
  to 
  be 
  due 
  

   to 
  any 
  variations 
  of 
  temperature 
  or 
  of 
  vapour 
  pressure, 
  and 
  

   as 
  they 
  cannot 
  well 
  be 
  attributed 
  to 
  absorption 
  they 
  remain 
  

   at 
  present 
  unexplained. 
  

  

  We 
  desire 
  to 
  thank 
  Professor 
  Gray 
  for 
  the 
  advice 
  he 
  

   has 
  given 
  us, 
  and 
  for 
  the 
  interest 
  he 
  has 
  taken 
  in 
  these 
  

   experiments. 
  

  

  Note. 
  

  

  Since 
  writing 
  the 
  above 
  we 
  have 
  made 
  further 
  experi- 
  

   ments 
  with 
  a 
  view 
  to 
  studying 
  the 
  change 
  of 
  pattern 
  ob- 
  

   served 
  by 
  Janicki 
  and 
  by 
  Galitzin 
  and 
  Wilip. 
  In 
  order 
  to 
  

   observe 
  the 
  effect 
  under 
  conditions 
  similar 
  to 
  those 
  under 
  

   which 
  it 
  was 
  first 
  obtained, 
  we 
  employed 
  vacuum-tubes 
  

   filled 
  and 
  exhausted 
  by 
  ourselves 
  and 
  driven 
  by 
  a 
  large 
  

  

  