﻿Energy 
  of 
  the 
  Ions 
  emitted 
  by 
  Hot 
  Bodies. 
  901 
  

  

  thus: 
  ©. 
  The 
  smooth 
  curve 
  drawn 
  is 
  the 
  graph 
  of 
  

  

  i=Ze~ 
  m2x 
  \ 
  It 
  will 
  be 
  observed 
  that 
  with 
  the 
  exception 
  o£ 
  

  

  the 
  last 
  four 
  points 
  on 
  the 
  left 
  all 
  the 
  rest 
  fall 
  on 
  the 
  curve 
  

  

  within 
  the 
  limits 
  of 
  experimental 
  error. 
  It 
  is 
  probable 
  that 
  

  

  the 
  high 
  value 
  of 
  i 
  for 
  these 
  points 
  was 
  due 
  to 
  some 
  part 
  of 
  

  

  the 
  strip 
  having 
  a 
  slight 
  twist 
  in 
  it, 
  causing 
  the 
  electric 
  field 
  

  

  near 
  that 
  portion 
  of 
  the 
  strip 
  to 
  throw 
  the 
  ions 
  emitted 
  from 
  

  

  it 
  over 
  to 
  the 
  left. 
  It 
  is 
  obvious 
  that 
  lack 
  of 
  symmetry 
  of 
  

  

  this 
  kind 
  is 
  due 
  to 
  instrumental 
  defects. 
  Making 
  allowance 
  

  

  tor 
  these 
  points 
  the 
  agreement 
  with 
  the 
  theoretical 
  type 
  of 
  

  

  curve 
  is 
  quite 
  satisfactory. 
  

  

  V 
  ne 
  

   Comparing 
  with 
  equation 
  (9), 
  we 
  see 
  that 
  -p^ 
  2 
  =632. 
  

  

  Putting 
  V 
  = 
  78-3 
  XlO 
  3 
  , 
  n* 
  = 
  -407, 
  6 
  = 
  1100° 
  abs. 
  and 
  ^='534, 
  

   we 
  find 
  11 
  = 
  4*04 
  x 
  10 
  3 
  . 
  Th^ 
  value 
  of 
  II 
  is 
  thus 
  even 
  nearer 
  

   the 
  standard 
  value 
  (3*7 
  x 
  10 
  3 
  ) 
  than 
  that 
  obtained 
  with 
  the 
  

   negative 
  ions. 
  

  

  A 
  number 
  of 
  series 
  of 
  observations 
  were 
  taken 
  with 
  the 
  

   positive 
  ionization 
  and 
  all 
  but 
  two 
  would 
  have 
  been 
  approxi- 
  

   mately 
  fitted 
  by 
  the 
  curve 
  shown 
  in 
  fig. 
  4. 
  They 
  would 
  

   therefore 
  have 
  led 
  to 
  the 
  same 
  value 
  of 
  H, 
  roughly, 
  of 
  course. 
  

   But 
  it 
  was 
  noticed 
  that, 
  with 
  two 
  curves 
  taken 
  after 
  long 
  

   continued 
  heating 
  of 
  the 
  strip, 
  the 
  two 
  halves 
  were 
  much 
  

   closer 
  together. 
  This 
  indicates 
  either 
  a 
  smaller 
  amount 
  of 
  

   sideways 
  energy 
  or 
  a 
  greater 
  charge 
  on 
  the 
  ions. 
  These 
  

   curves 
  were 
  fairly 
  well 
  represented 
  by 
  i 
  = 
  5'35s~ 
  l26ix2 
  * 
  

   Assuming 
  that 
  the 
  value 
  of 
  H0 
  had 
  not 
  changed 
  this 
  would 
  

   lead 
  to 
  ne 
  = 
  '814: 
  and 
  the 
  ions 
  divalent. 
  On 
  the 
  other 
  hand, 
  

   if 
  we 
  still 
  suppose 
  w£ 
  = 
  "407 
  we 
  should 
  get 
  about 
  half 
  the 
  

   normal 
  value 
  of 
  R. 
  There 
  is 
  nothing 
  in 
  the 
  present 
  experi- 
  

   ments 
  which 
  enables 
  us 
  to 
  decide 
  definitely 
  between 
  these 
  

   hypotheses. 
  A 
  priori, 
  the 
  balance 
  of 
  probability 
  would 
  seem 
  

   to 
  be 
  in 
  favour 
  of 
  the 
  kinetic 
  energy 
  being 
  constant 
  and 
  the 
  

   charge 
  doubled 
  ; 
  but 
  simultaneous 
  measurements 
  of 
  e/m 
  for 
  

   the 
  ions 
  indicated 
  that 
  this 
  also 
  had 
  changed, 
  and 
  was 
  re- 
  

   duced 
  to 
  nearly 
  one 
  half 
  its 
  previous 
  value. 
  So 
  that 
  if 
  the 
  

   charge 
  had 
  doubled 
  the 
  mass 
  must 
  have 
  increased 
  to 
  almost 
  

   four 
  times 
  its 
  earlier 
  value. 
  This 
  would 
  make 
  it 
  very 
  difficult 
  

   to 
  assign 
  these 
  ions 
  to 
  any 
  definite 
  chemical 
  substance. 
  The 
  

   matter 
  evidently 
  requires 
  further 
  investigation. 
  The 
  value 
  

   of 
  e/m 
  for 
  this 
  case 
  is 
  the 
  last 
  one 
  in 
  the 
  table 
  of 
  values 
  of 
  

   e/m 
  for 
  the 
  positive 
  ions 
  from 
  hot 
  platinum 
  in 
  the 
  author's 
  * 
  

   paper 
  on 
  that 
  subject. 
  

  

  * 
  Phil. 
  Mag. 
  [6] 
  vol. 
  xvi. 
  p."7o9 
  (Nov. 
  1908). 
  

  

  