﻿Constitution 
  of 
  the 
  Mercury 
  Green 
  Line. 
  369 
  

  

  Results. 
  — 
  When 
  the 
  small 
  arc 
  lamp 
  was 
  used, 
  visual 
  

   observations 
  of 
  the 
  line 
  showed 
  the 
  characteristic 
  pattern 
  — 
  

   one 
  bright 
  centre 
  line, 
  two 
  not 
  far 
  apart 
  on 
  the 
  left 
  (the 
  

   side 
  of 
  greater 
  wave-length), 
  a 
  closer 
  doublet 
  on 
  the 
  right, 
  

   and 
  an 
  isolated 
  line 
  further 
  off 
  to 
  the 
  right, 
  almost 
  centrally 
  

   placed 
  between 
  the 
  two 
  successive 
  spectra. 
  These 
  lines 
  

   have 
  the 
  same 
  positions 
  relatively 
  to 
  the 
  centre 
  line 
  as 
  the 
  

   satellites 
  numbered 
  respectively 
  (5), 
  (6), 
  (4), 
  (3), 
  (1), 
  in 
  the 
  

   -appended 
  table. 
  As 
  no 
  variation 
  of 
  current 
  strength 
  or 
  

   lamp 
  temperature 
  seemed 
  to 
  affect 
  this 
  visible 
  pattern, 
  

   photographic 
  methods 
  were 
  afterwards 
  adopted. 
  

  

  In 
  the 
  small 
  lamp 
  a 
  current 
  of 
  somewhat 
  less 
  than 
  one 
  

   ampere 
  was 
  used, 
  and 
  the 
  vapour 
  pressure 
  could 
  not 
  exceed 
  

   50 
  mm. 
  of 
  mercury 
  and 
  was 
  usually 
  about 
  40 
  mm. 
  The 
  

   temperature 
  of 
  the 
  lamp 
  was 
  about 
  230° 
  0. 
  With 
  the 
  

   Cooper 
  Hewitt 
  lamp 
  currents 
  varying 
  from 
  2 
  amps, 
  to 
  3*7 
  

   amps, 
  were 
  used, 
  giving 
  a 
  variation 
  of 
  temperature 
  from 
  

   130° 
  C. 
  to 
  400° 
  C, 
  and 
  consequently 
  of 
  vapour 
  pressure 
  from 
  

   1*2 
  mm. 
  of 
  mercury 
  to 
  about 
  1500 
  mm. 
  Photographs 
  were 
  

   taken 
  at 
  various 
  points 
  along 
  the 
  length 
  of 
  the 
  tube 
  

   from 
  about 
  5 
  cm. 
  from 
  the 
  anode 
  to 
  about 
  5 
  cm. 
  from 
  the 
  

   cathode. 
  With 
  all 
  the 
  lamps 
  only 
  " 
  side 
  on 
  " 
  exposures 
  

   were 
  obtainable. 
  The 
  self-induction 
  of 
  the 
  circuit 
  was 
  not 
  

   varied. 
  

  

  Notwithstanding 
  these 
  variations 
  in 
  the 
  conditions 
  of 
  the 
  

   lamps 
  and 
  in 
  the 
  type 
  and 
  size 
  of. 
  lamp, 
  there 
  was 
  absolutely 
  

   no 
  measurable 
  change 
  in 
  the 
  line 
  pattern. 
  Photography 
  

   revealed 
  three 
  fainter 
  satellites 
  not 
  visible 
  to 
  the 
  eye, 
  and 
  

   obtained 
  on 
  about 
  70 
  per 
  cent, 
  of 
  the 
  plates 
  measured. 
  

   These 
  faint 
  satellites 
  are 
  nos. 
  (2), 
  (7), 
  and 
  (8) 
  in 
  the 
  following 
  

   list. 
  The 
  results 
  of 
  measurements 
  give 
  the 
  values 
  as 
  shown 
  

   in 
  the 
  second 
  last 
  column 
  of 
  Table 
  L, 
  for 
  the 
  positions 
  of 
  

   the 
  satellites. 
  These 
  are 
  average 
  results 
  ; 
  but 
  the 
  agreement 
  

   of 
  the 
  measurements 
  from 
  different 
  photographs 
  was 
  very 
  

   close, 
  the 
  " 
  probable 
  error 
  " 
  of 
  the 
  measurements 
  being 
  

   *001 
  A.U., 
  i. 
  e. 
  fifty 
  per 
  cent, 
  of 
  the 
  measurements 
  do 
  not 
  

  

  differ 
  from 
  these 
  means 
  by 
  more 
  than 
  *001 
  A. 
  U. 
  Our 
  

   plates 
  show 
  four 
  satellites 
  on 
  each 
  side 
  of 
  the 
  main 
  line. 
  

   We 
  could 
  detect 
  no 
  evidence 
  of 
  the 
  centre 
  line 
  itself 
  being- 
  

   complex. 
  All 
  our 
  measurements 
  are 
  taken 
  from 
  the 
  middle 
  

   of 
  this 
  main 
  line. 
  

  

  Comparison 
  icitli 
  Previous 
  Results. 
  — 
  In 
  Table 
  I. 
  we 
  give 
  also 
  

   the 
  results 
  obtained 
  by 
  other 
  observers 
  for 
  the 
  positions 
  and 
  

   relative 
  intensities 
  of 
  the 
  satellites. 
  Fig. 
  1 
  shows 
  dia- 
  

   grammatically 
  the 
  various 
  patterns 
  observed, 
  the 
  lengths 
  of 
  

   the 
  lines 
  being 
  drawn 
  proportional 
  to 
  the 
  intensities. 
  

  

  