﻿Distribution 
  of 
  Energy 
  in 
  the 
  Spectra 
  of 
  Gases. 
  805 
  

  

  experiments 
  by 
  Vegard 
  (Ann. 
  der 
  Pliys. 
  vol. 
  xxxix. 
  pp. 
  111- 
  

   169, 
  1912). 
  In 
  agreement 
  with 
  Pascheti 
  and 
  Lunkenheimer, 
  

   Vegard 
  finds 
  that 
  the 
  distribution 
  of 
  energy 
  in 
  the 
  spectrnm 
  

   of 
  the 
  light 
  from 
  Kanalstrahlen 
  is 
  independent 
  of 
  the 
  

   velocity 
  of 
  the 
  latter, 
  but 
  that 
  the 
  distribution 
  in 
  the 
  

   spectrum 
  of 
  the 
  light 
  excited 
  in 
  the 
  stationary 
  atoms 
  by 
  

   the 
  Kanalstrahlen, 
  though 
  independent 
  also 
  of 
  the 
  voltage, 
  

   differs 
  markedly 
  from 
  the 
  distribution 
  in 
  the 
  moving 
  atoms. 
  

   This 
  explains 
  the 
  apparent 
  shift 
  of 
  energy 
  with 
  changing 
  

   voltage, 
  observed 
  in 
  the 
  transverse 
  light, 
  for 
  the 
  difference 
  

   between 
  the 
  distribution 
  in 
  the 
  "stationary" 
  and 
  "moving" 
  

   intensity 
  changes 
  with 
  the 
  voltage, 
  and 
  also 
  becomes 
  less 
  

   marked 
  as 
  the 
  pressure 
  is 
  lowered. 
  

  

  An 
  observation 
  of 
  Sir 
  J. 
  J. 
  Thomson's 
  (loc. 
  cit.) 
  to 
  the 
  

  

  effect 
  that 
  the 
  ratio 
  in 
  intensity 
  tj^ 
  is 
  greater 
  in 
  the 
  

  

  positive 
  column 
  than 
  in 
  the 
  negative 
  glow 
  is 
  also 
  confirmed. 
  

   These 
  results 
  throw 
  much 
  light 
  on 
  the 
  subject 
  of 
  the 
  

   radiation 
  from 
  Geissler 
  tubes, 
  but 
  owing 
  to 
  the 
  widely 
  

   differing 
  ranges 
  of 
  pressure 
  in 
  the 
  two 
  papers 
  mentioned 
  

   and 
  the 
  present 
  paper, 
  a 
  direct 
  comparison 
  of 
  results 
  is 
  

   impossible. 
  In 
  the 
  present 
  case 
  such 
  high 
  pressures 
  were 
  

   necessary 
  in 
  order 
  to 
  obtain 
  sufficient 
  energy 
  that 
  it 
  is 
  

   doubtful 
  whether 
  the 
  pressure-effects 
  found 
  by 
  other 
  ob- 
  

   servers 
  at 
  pressures 
  below 
  1 
  mm. 
  of 
  mercury 
  have 
  much 
  

   significance. 
  

  

  3. 
  Description 
  of 
  Preliminary 
  Experiments. 
  

  

  The 
  experiments 
  were 
  commenced 
  in 
  the 
  summer 
  of 
  1909. 
  

   At 
  first 
  an 
  attempt 
  was 
  made 
  to 
  obtain 
  measurable 
  amounts 
  

   of 
  energy 
  by 
  using 
  Pliicker 
  tubes 
  in 
  conjunction 
  with 
  a 
  

   Rubens 
  thermopile 
  and 
  a 
  d'Arsonval 
  galvanometer. 
  Several 
  

   forms 
  of 
  discharge-tube 
  were 
  tried, 
  but 
  all 
  failed. 
  A 
  Broca 
  

   galvanometer 
  was 
  next 
  tried. 
  The 
  sensitiveness 
  of 
  this 
  

   instrument 
  was 
  3x 
  10~ 
  9 
  amp. 
  for 
  1 
  mm. 
  deflexion 
  when 
  the 
  

   period 
  of 
  oscillation 
  was 
  17 
  sec. 
  This 
  again 
  proved 
  in- 
  

   sufficient 
  when 
  working 
  under 
  its 
  ordinary 
  conditions, 
  as 
  the 
  

   deflexions 
  for 
  the 
  total 
  energy 
  in 
  the 
  visible 
  spectrum 
  were 
  

   never 
  more 
  than 
  2 
  or 
  3 
  mm. 
  on 
  the 
  scale. 
  

  

  The 
  Broca 
  is, 
  incidentally, 
  scarcely 
  suited 
  for 
  this 
  kind 
  of 
  

   measurement 
  for 
  another 
  reason 
  : 
  — 
  The 
  suspended 
  system 
  is 
  

   a 
  heavy 
  one, 
  and 
  the 
  time 
  of 
  oscillation 
  large, 
  so 
  that 
  

   measurements 
  of 
  first 
  swings 
  cannot 
  be 
  made 
  with 
  that 
  

   rapidity 
  which 
  is 
  so 
  desirable 
  when 
  the 
  source 
  is 
  liable 
  to 
  

   fluctuations, 
  and 
  when, 
  owing 
  to 
  the 
  smallness 
  of 
  the 
  effect 
  

  

  