﻿Interferometer 
  for 
  the 
  Study 
  of 
  Band 
  Spectra. 
  179 
  

  

  at 
  5165. 
  Lines 
  which 
  do 
  not 
  properly 
  belong 
  to 
  the 
  series 
  

   forming 
  the 
  band 
  produce 
  kinks 
  in 
  the 
  patterns, 
  a 
  con- 
  

   spicuous 
  one 
  being 
  indicated 
  by 
  an 
  arrow 
  on 
  fig. 
  1. 
  A 
  

   careful 
  inspection 
  of 
  the 
  figure 
  shows 
  also 
  that 
  the 
  band 
  is 
  

   made 
  up 
  of 
  two 
  series 
  of 
  lines 
  superposed; 
  for 
  we 
  have 
  a 
  

   dark 
  series 
  of 
  hyperbolae, 
  with 
  a 
  fainter 
  series 
  almost 
  exactly 
  

   out-of-step 
  with 
  the 
  first. 
  That 
  these 
  patterns 
  show 
  the 
  

   presence 
  of 
  structure 
  beyond 
  the 
  resolving 
  power 
  of 
  tbe 
  

   spectroscope 
  employed 
  to 
  photograph 
  them 
  is 
  clearly 
  shown 
  

   by 
  fig. 
  2, 
  for 
  no 
  trace 
  of 
  the 
  lines 
  forming 
  the 
  bands 
  is 
  

   shown 
  in 
  the 
  vicinity 
  of 
  the*' 
  heads" 
  in 
  the 
  narrow 
  photo- 
  

   graph 
  of 
  the 
  aluminium 
  spectrum 
  taken 
  with 
  the 
  same 
  

   slit-width 
  after 
  the 
  interferometer 
  had 
  been 
  removed 
  from 
  

   the 
  path 
  of 
  the 
  light. 
  

  

  To 
  ascertain 
  the 
  effect 
  which 
  the 
  law 
  governing 
  the 
  spacing 
  

   of 
  the 
  lines 
  had 
  upon 
  the 
  appearance 
  of 
  the 
  bands 
  in 
  the 
  

   spectrum, 
  I 
  made 
  a 
  number 
  of 
  imitation 
  spectra 
  by 
  ruling 
  

   lines 
  with 
  a 
  dividing-engine 
  on 
  the 
  film 
  of 
  a 
  photographic 
  

   plate 
  which 
  had 
  been 
  fixed 
  without 
  exposure, 
  the 
  spacing 
  of 
  

   the 
  lines 
  being- 
  made 
  to 
  follow 
  some 
  definite 
  law. 
  These 
  

   plates 
  were 
  then 
  superposed 
  on 
  the 
  photograph 
  of 
  the 
  

   spectrum 
  of 
  a 
  white 
  source 
  containing 
  the 
  dark 
  arcs 
  

   previously 
  alluded 
  to. 
  The 
  symmetrical 
  bands 
  at 
  once 
  

   appeared. 
  

  

  Two 
  photographs 
  of 
  the 
  superposed 
  plates 
  are 
  reproduced 
  

   on 
  Plate 
  IV. 
  figs. 
  3 
  and 
  4, 
  the 
  former 
  taken 
  with 
  the 
  head 
  of 
  

   the 
  series 
  of 
  lines 
  towards 
  the 
  long 
  wave-length 
  end 
  of 
  the 
  

   spectrum, 
  the 
  hitter 
  with 
  the 
  head 
  turned 
  the 
  other 
  way. 
  

   In 
  one 
  case 
  we 
  have 
  the 
  hyperbola?, 
  and 
  in 
  the 
  other 
  the 
  

   ellipses. 
  The 
  artificial 
  series 
  of 
  lines 
  appear 
  alone 
  at 
  

   the 
  top 
  of 
  each 
  picture, 
  the 
  bands 
  appearing 
  only 
  where 
  

   they 
  cut 
  across 
  the 
  dark 
  arcs 
  formed 
  in 
  the 
  continuous 
  

   spectrum 
  by 
  the 
  interference 
  caused 
  by 
  the 
  half-silvered 
  

   plates. 
  

  

  A 
  much 
  more 
  extended 
  study 
  of 
  the 
  subject 
  will 
  be 
  

   necessary 
  before 
  the 
  method 
  can 
  be 
  of 
  much 
  use 
  in 
  the 
  study 
  

   of 
  band 
  and 
  line 
  spectra; 
  and 
  the 
  object 
  of 
  the 
  present 
  paper 
  

   is 
  merely 
  to 
  draw 
  attention 
  to 
  the 
  possibilities 
  of 
  the 
  method. 
  

   I 
  do 
  not 
  expect 
  to 
  continue 
  the 
  study 
  of 
  the 
  subject 
  myself, 
  

   at 
  least 
  for 
  the 
  present; 
  and 
  it 
  is 
  my 
  hope 
  that 
  other 
  

   investigators 
  will 
  be 
  able 
  to 
  develop 
  the 
  idea. 
  

  

  N2 
  

  

  