﻿904 
  Prof. 
  0. 
  W. 
  Richardson 
  on 
  the 
  Kinetic 
  

  

  It 
  seems 
  practically 
  certain 
  that 
  this 
  arises 
  from 
  ions 
  

   which 
  are 
  deviated 
  by 
  collisions 
  with 
  molecules 
  of 
  gas 
  in 
  the 
  

   tube. 
  Although 
  the 
  pressure 
  of 
  the 
  residual 
  gas, 
  as 
  registered 
  

   by 
  the 
  McLeod 
  gauge, 
  was 
  always 
  very 
  small 
  (not 
  more 
  

   than 
  5 
  x 
  10~ 
  4 
  mm.), 
  effects 
  of 
  this 
  kind 
  will 
  be 
  important 
  

   on 
  account 
  of 
  the 
  very 
  rapid 
  way 
  in 
  which 
  the 
  theoretical 
  

   current 
  through 
  the 
  slit 
  diminishes 
  as 
  we 
  recede 
  from 
  the 
  

   centre. 
  The 
  fact 
  that 
  this 
  effect 
  was 
  much 
  smaller 
  in 
  the 
  

   experiments 
  on 
  the 
  negative 
  ionization 
  is 
  strongly 
  in 
  favour 
  

   of 
  this 
  explanation. 
  In 
  the 
  first 
  place 
  the 
  potential-difference 
  

   was 
  only 
  about 
  one-eighth 
  as 
  great, 
  so 
  that 
  the 
  ions 
  spread 
  

   out 
  much 
  more 
  by 
  virtue 
  of 
  their 
  initial 
  energy, 
  and 
  scatter- 
  

   ing 
  by 
  collisions 
  would 
  be 
  relatively 
  less 
  important 
  ; 
  and 
  in 
  

   tlw 
  second 
  place 
  the 
  mean 
  free 
  path 
  of 
  a 
  negative 
  1 
  ion 
  is 
  

   Bome 
  four 
  times 
  as 
  great 
  as 
  that 
  of 
  a 
  positive 
  ion 
  in 
  gas 
  at 
  

   the 
  same 
  pressure, 
  bo 
  thai 
  there 
  would 
  be 
  less 
  scattering 
  of 
  

   the 
  negative 
  than 
  of 
  the 
  positive 
  ions 
  by 
  collisions 
  even 
  under 
  

   conditions 
  which 
  were 
  otherwise 
  equal. 
  

  

  On 
  the 
  above 
  grounds 
  it 
  seemed 
  reasonable 
  to 
  conclude 
  

   that 
  the 
  kinetic 
  energy 
  effect 
  is 
  that 
  which 
  determines 
  the 
  

   -lope 
  of 
  the 
  graph 
  of 
  log 
  }=/(.r- 
  ) 
  in 
  the 
  neighbourhood 
  of 
  

   ./• 
  = 
  (), 
  and 
  that 
  the 
  behaviour 
  of 
  this 
  graph 
  when 
  x 
  2 
  is 
  large 
  

   can 
  be 
  rejected 
  as 
  being 
  due 
  to 
  causes 
  which 
  lie 
  outside 
  the 
  

   scope 
  of 
  the 
  present 
  investigation. 
  Substituting 
  the 
  values 
  

   of 
  .'y 
  — 
  .''■/ 
  and 
  of 
  Iogi 
  « 
  2 
  — 
  log 
  10 
  «i 
  obtained 
  from 
  the 
  line 
  

   passing 
  through 
  the 
  points 
  lying 
  between 
  x 
  2 
  = 
  and 
  

   .r 
  = 
  2 
  x 
  10 
  _3 
  cm.- 
  in 
  equation 
  (10), 
  together 
  with 
  the 
  values 
  

   of 
  6, 
  ne, 
  V, 
  and 
  : 
  already 
  given 
  (p. 
  ( 
  J01), 
  we 
  find 
  for 
  the 
  

   value 
  of 
  R, 
  R 
  = 
  3'3xl0 
  3 
  / 
  

  

  This 
  agre.- 
  with 
  the 
  value 
  previously 
  found 
  and 
  with 
  the 
  

   standard 
  value 
  within 
  the 
  limits 
  of 
  experimental 
  error, 
  and 
  

   the 
  results 
  which 
  have 
  just 
  been 
  discussed 
  show 
  that 
  the 
  

   sideways 
  kinetic 
  energy 
  of 
  the 
  positive 
  ions 
  obeys 
  the 
  same 
  

   laws 
  as 
  that 
  of 
  the 
  negative. 
  

  

  We 
  have 
  seen 
  already 
  that 
  a 
  series 
  of 
  observations 
  taken 
  

   after 
  the 
  strip 
  had 
  been 
  heated 
  for 
  a 
  long 
  time, 
  when 
  

   analysed 
  by 
  the 
  first 
  method 
  gave 
  a 
  value 
  of 
  R 
  almost 
  one- 
  

   half 
  of 
  the 
  above, 
  for 
  the 
  same 
  assumed 
  value 
  of 
  ne. 
  This 
  

   curve 
  was 
  also 
  examined 
  by 
  the 
  present 
  method 
  and, 
  

   although 
  it 
  was 
  somewhat 
  unsymmetrical, 
  both 
  sides 
  of 
  the 
  

   curve 
  were 
  made 
  use 
  of 
  as 
  there 
  w 
  r 
  as 
  no 
  evidence 
  tending 
  

   to 
  favour 
  the 
  rejection 
  of 
  one 
  side 
  rather 
  than 
  the 
  other. 
  

   With 
  this 
  exception 
  it 
  was 
  treated 
  in 
  the 
  same 
  way 
  as 
  

   the 
  curve 
  in 
  fig. 
  4, 
  and 
  led 
  to 
  the 
  numbers 
  in 
  the 
  next 
  

   table. 
  

  

  