﻿Charge 
  of 
  tlie 
  Ions 
  emitted 
  by 
  Hot 
  Bodies 
  

  

  763 
  

  

  platinum 
  are 
  shown 
  in 
  fig. 
  5. 
  The 
  central 
  curve 
  is 
  for 
  a 
  

   potential-difference 
  of 
  10*6 
  volts. 
  For 
  the 
  curves 
  with 
  the 
  

   magnetic 
  field 
  on 
  the 
  potential-difference 
  was 
  41 
  volts. 
  The 
  

   ions 
  are, 
  of 
  course, 
  more 
  spread 
  out 
  the 
  lower 
  the 
  potential- 
  

   difference. 
  

  

  The 
  numbers 
  which 
  were 
  obtained 
  in 
  the 
  two 
  determinations 
  

   of 
  e/m 
  for 
  the 
  negative 
  ions 
  from 
  hot 
  platinum 
  are 
  given 
  in 
  

   the 
  next 
  table. 
  

  

  V. 
  

  

  H. 
  

  

  z. 
  

  

  *. 
  

  

  e/m. 
  

  

  40-8 
  

   41 
  

  

  13-5 
  

   13-5 
  

  

  •534 
  

   •534 
  

  

  1-7 
  

   1-69 
  

  

  1-46x107 
  

   l-44 
  3 
  xl0? 
  

  

  Mean 
  e/m 
  = 
  1*45 
  xlO 
  7 
  . 
  

  

  The 
  mean 
  value 
  of 
  e/m 
  for 
  the 
  negative 
  ions 
  is 
  thus 
  almost 
  

   the 
  same 
  for 
  both 
  carbon 
  and 
  platinum. 
  It 
  is 
  significant 
  

   that 
  not 
  only 
  the 
  mean 
  but 
  each 
  individual 
  value 
  of 
  e/m 
  is 
  

   distinctly 
  smaller 
  than 
  the 
  best 
  value, 
  which 
  is 
  1*88 
  x 
  10 
  7 
  , 
  

   with 
  some 
  uncertainty 
  as 
  to 
  the 
  last 
  figure. 
  

  

  A 
  Possible 
  Correction. 
  

  

  It 
  seems 
  likely 
  that 
  some 
  of 
  this 
  difference 
  is 
  due 
  to 
  the 
  

   electric 
  field, 
  in 
  the 
  immediate 
  neighbourhood 
  of 
  the 
  strip 
  of 
  

   filament, 
  not 
  being 
  the 
  same 
  as 
  if 
  the 
  hot 
  body 
  were 
  a 
  portion 
  

   of 
  an 
  infinite 
  plane. 
  Both 
  strip 
  and 
  filament 
  are 
  narrow 
  

   conductors 
  stretched 
  across 
  a 
  much 
  wider 
  gap. 
  It 
  is 
  difficult 
  

   to 
  know 
  exactly 
  what 
  difference 
  this 
  will 
  make 
  in 
  the 
  electric 
  

   field. 
  In 
  the 
  case 
  of 
  the 
  filament 
  it 
  would 
  seem 
  reasonable 
  

   to 
  suppose 
  that 
  the 
  deviation 
  of 
  the 
  field 
  from 
  uniformity 
  

   would 
  be 
  smaller 
  than, 
  but 
  comparable 
  with, 
  what 
  would 
  

   occur 
  if 
  the 
  filament 
  were 
  a 
  cylinder 
  of 
  equal 
  radius 
  a, 
  and 
  

   the 
  slit 
  plates 
  formed 
  part 
  of 
  a 
  concentric 
  cylinder 
  of 
  radius 
  b, 
  

   where 
  the 
  shortest 
  distance 
  between 
  the 
  filament 
  and 
  slit 
  

   plates 
  is 
  equal 
  to 
  b— 
  a. 
  

  

  To 
  illustrate 
  the 
  question 
  consider 
  the 
  motion 
  of 
  an 
  ion 
  

   projected 
  perpendicularly 
  from 
  the 
  surface 
  of 
  the 
  inner 
  

   cylinder 
  whose 
  radius 
  is 
  a. 
  Let 
  the 
  direction 
  of 
  projection 
  

   be 
  the 
  axis 
  of 
  £, 
  and 
  let 
  a 
  difference 
  of 
  potential 
  V 
  be 
  

   maintained 
  between 
  the 
  two 
  cylinders. 
  Since 
  the 
  electric 
  

   force 
  is 
  radial 
  the 
  ion 
  will 
  always 
  remain 
  on 
  the 
  axis 
  of 
  z. 
  

   Now 
  suppose 
  there 
  is 
  a 
  uniform 
  magnetic 
  force 
  H 
  parallel 
  

  

  