﻿946 
  Extension 
  of 
  Principal 
  Series 
  of 
  Sodium 
  Spectrum, 
  

  

  a 
  metro-radius 
  ruled 
  with 
  15,000 
  Lines 
  to 
  the 
  inch. 
  By 
  

   employing 
  an 
  exceedingly 
  fine 
  slit 
  and 
  baking 
  especial 
  pains 
  

   about 
  the 
  Eocns, 
  Q 
  the 
  wave-lengths 
  can 
  be 
  determined 
  to 
  within 
  

   about 
  0"2 
  of 
  an 
  Angstrom 
  unit. 
  The 
  number 
  of 
  lines 
  which 
  

   appear 
  in 
  the 
  absorption-spectrum 
  increases 
  with 
  the 
  vapour- 
  

   density, 
  and 
  I 
  feel 
  sure 
  that 
  by 
  employing 
  a 
  longer 
  tube, 
  and 
  

   giving 
  a 
  longer 
  exposure, 
  many 
  more 
  linos 
  would 
  come 
  into 
  

   view. 
  Oneor 
  the 
  photographs 
  is 
  reproduced 
  on 
  PL 
  XXYI1I., 
  

   greatly 
  enlarged 
  of 
  course, 
  with 
  the 
  spark-spectrum 
  of 
  iron 
  

   as 
  a 
  comparison 
  spectrum. 
  On 
  the 
  original 
  negative 
  the 
  

   absorption-lines 
  can 
  be 
  seen 
  a< 
  Ear 
  down 
  as 
  the 
  cadmium 
  line 
  

   2418, 
  while 
  the 
  theoretical 
  end 
  <){' 
  the 
  series 
  is 
  at 
  the 
  iron 
  

   line 
  2411, 
  Both 
  of 
  these 
  points 
  1 
  have 
  indicated 
  on 
  the 
  

   print. 
  

  

  The 
  wave-lengths 
  are 
  as 
  follows 
  : 
  — 
  

  

  ■ 
  

  

  5898 
  (mean) 
  

  

  ;;. 
  03 
  

  

  2852-9 
  

  

  2680-5 
  

  

  2593-8 
  

  

  25448 
  

  

  2512-2 
  

  

  2491-5 
  

  

  2476-4 
  

  

  2464-9 
  

  

  24568 
  

  

  2450-0 
  

  

  2444-7 
  

  

  2440-4 
  

  

  2486-9 
  

  

  2484-2 
  

  

  2432 
  < 
  I 
  

  

  24300 
  

  

  2428-2 
  

  

  2426-7 
  

  

  L'i_'.' 
  - 
  

  

  24250 
  

  

  n 
  = 
  30 
  2418-1 
  cal. 
  

  

  n 
  = 
  2409-8 
  

  

  Under 
  the 
  microscope 
  of 
  the 
  dividing-machine, 
  on 
  which 
  

   the 
  plates 
  were 
  measured, 
  five 
  or 
  six 
  more 
  lines 
  could 
  be 
  

   seen 
  crowded 
  together 
  between 
  the 
  last 
  line 
  measured 
  and 
  

   the 
  cadmium 
  line 
  2418, 
  which 
  is 
  in 
  agreement 
  with 
  the 
  

   calculated 
  position 
  for 
  the 
  line 
  for 
  which 
  ?z 
  = 
  30. 
  

  

  I 
  hope 
  to 
  have 
  an 
  opportunity 
  in 
  the 
  near 
  future 
  of 
  

   photographing 
  the 
  spectrum 
  with 
  a 
  large 
  quartz 
  spectro- 
  

   graph, 
  which 
  will 
  make 
  a 
  more 
  accurate 
  determination 
  of 
  

   the 
  wave-lengths 
  possible. 
  I 
  feel 
  sure 
  that 
  those 
  in 
  the 
  table 
  

  

  n 
  

  

  = 
  

  

  8 
  

  

  a 
  

  

  — 
  

  

  4 
  

  

  n 
  

  

  = 
  

  

  5 
  

  

  v 
  

  

  = 
  

  

  <*» 
  

  

  ti 
  

  

  -- 
  

  

  7 
  

  

  n 
  

  

  := 
  

  

  - 
  

  

  n 
  

  

  — 
  - 
  

  

  9 
  

  

  n 
  

  

  = 
  

  

  hi 
  

  

  H 
  

  

  — 
  

  

  n 
  

  

  n 
  

  

  — 
  

  

  12 
  

  

  n 
  

  

  = 
  

  

  13 
  

  

  a 
  

  

  — 
  

  

  14 
  

  

  a 
  

  

  = 
  

  

  L5 
  

  

  

  = 
  

  

  16 
  

  

  a 
  

  

  = 
  

  

  17 
  

  

  V 
  

  

  ss 
  

  

  18 
  

  

  It 
  

  

  — 
  

  

  L9 
  

  

  n 
  

  

  = 
  

  

  20 
  

  

  

  = 
  

  

  21 
  

  

  ii 
  

  

  = 
  

  

  22 
  

  

  n 
  

  

  = 
  

  

  23 
  

  

  

  = 
  

  

  24 
  

  

  