﻿Energy 
  of 
  the 
  Ions 
  emitted 
  by 
  Hot 
  Bodies. 
  913 
  

  

  Q>) 
  Positive 
  Ions. 
  — 
  In 
  applying 
  this 
  method 
  to 
  the 
  positive 
  

   ions, 
  the 
  experimental 
  points 
  in 
  fig. 
  4 
  and 
  some 
  additional 
  

   ones 
  from 
  the 
  same 
  set 
  of 
  observations 
  which 
  fell 
  outside 
  the 
  

   limits 
  of 
  that 
  diagram 
  were 
  treated 
  exactly 
  like 
  those 
  in 
  

   fig. 
  3 
  were 
  treated 
  in 
  the 
  case 
  of 
  the 
  negative 
  ions. 
  The 
  

   resulting 
  values 
  of 
  ix 
  1 
  have 
  been 
  plotted 
  against 
  

  

  Fio-. 
  8. 
  

  

  x 
  in 
  fig. 
  8. 
  

  

  

  

  : 
  

  

  

  

  

  

  

  

  

  

  

  

  X 
  

  

  

  X 
  

  

  

  

  

  4- 
  

  

  

  

  

  

  ; 
  

  

  c 
  

  

  

  

  

  

  n 
  

  

  3 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  NJ 
  

  

  

  

  

  

  

  

  

  

  

  

  

  £ 
  

  

  2 
  

  

  

  

  

  

  

  

  

  

  

  

  1 
  X 
  

  

  

  X 
  

  

  

  

  

  

  

  

  

  

  

  

  

  X 
  

  

  X 
  

  

  654-32 
  10 
  1234- 
  

  

  Scale 
  of 
  oc 
  ( 
  1= 
  -03/8 
  cm). 
  

  

  It 
  will 
  be 
  observed 
  that 
  this 
  treatment 
  greatly 
  emphasises 
  

   the 
  lack 
  of 
  symmetry 
  of 
  the 
  original 
  figure. 
  The 
  area 
  

   embraced 
  by 
  the 
  left-hand 
  portion 
  of 
  the 
  curve 
  is 
  now 
  very 
  

   much 
  greater 
  than 
  the 
  right. 
  It 
  is 
  also 
  fairly 
  clear 
  that 
  the 
  

   points 
  at 
  a 
  distance 
  from 
  the 
  central 
  axis 
  do 
  not 
  fall 
  on 
  

   the 
  same 
  analytical 
  curve 
  as 
  the 
  rest. 
  Since 
  these 
  corre- 
  

   spond 
  to 
  very 
  small 
  currents 
  they 
  are 
  affected 
  by 
  a 
  very 
  

   large 
  ^ 
  experimental 
  error 
  in 
  any 
  event; 
  but 
  we 
  have 
  seen 
  

   that 
  it 
  is 
  quite 
  probable 
  that 
  the 
  greater 
  part 
  of 
  the 
  current 
  

   through 
  the 
  slit 
  at 
  these 
  large 
  distances 
  arises 
  from 
  ions 
  

   scattered 
  by 
  collisions 
  with 
  molecules 
  of 
  gas 
  in 
  the 
  tube 
  and 
  

   not 
  from 
  ions 
  having 
  originally 
  the 
  very 
  large 
  amounts 
  of 
  

   -sideways 
  energy 
  indicated. 
  Any 
  errors 
  of 
  this 
  kind 
  are 
  

   greatly 
  emphasised 
  by 
  the 
  present 
  method 
  of 
  treatment. 
  

   For 
  these 
  reasons, 
  in 
  calculating 
  the 
  mean 
  value 
  of 
  the 
  

   energy, 
  the 
  points 
  away 
  from 
  the 
  centre 
  have 
  not 
  been 
  

   taken 
  into 
  consideration, 
  and 
  the 
  curves 
  have 
  been 
  com- 
  

   pleted, 
  as 
  shown, 
  with_soine 
  regard 
  to 
  reasonable 
  probability. 
  

   The 
  value 
  of 
  ±nmu 
  2 
  has 
  been 
  determined 
  from 
  fig. 
  8 
  in 
  

   the 
  manner 
  already 
  described. 
  Separate 
  values 
  have 
  been 
  

   deduced 
  from 
  (1) 
  the 
  part 
  to 
  the 
  right 
  of 
  # 
  = 
  0, 
  (2) 
  the 
  part 
  

   to 
  the 
  left 
  of 
  a? 
  = 
  0, 
  and 
  (3) 
  the 
  whole 
  curve. 
  In 
  each 
  case 
  

   the 
  corresponding 
  regions 
  of 
  the 
  ix 
  diagram 
  were 
  used. 
  

   The 
  values 
  obtained 
  were 
  : 
  — 
  

  

  (1) 
  Right-hand 
  half 
  

  

  (2) 
  Left-hand 
  half 
  

  

  ±nmu 
  2 
  = 
  2-35 
  x 
  10 
  6 
  ergs. 
  

   5-9 
  xlO 
  6 
  

  

  (3) 
  Whole 
  curve 
  ^mw 
  2 
  = 
  4'45x 
  10 
  6 
  „ 
  

  

  Phil. 
  Mag. 
  S. 
  6. 
  Vol. 
  16. 
  No. 
  96. 
  Uec. 
  1908. 
  3 
  

  

  