﻿Special 
  Case 
  of 
  Gaseous 
  Conduction, 
  919 
  

  

  6. 
  The 
  figures 
  in 
  the 
  tables 
  clearly 
  justify 
  the 
  assertion 
  

   that 
  over 
  a 
  considerable 
  range 
  of 
  pressure, 
  i/i 
  is 
  independent 
  

   of 
  the 
  pressure 
  for 
  small 
  values 
  of 
  V. 
  It 
  will 
  be 
  observed 
  that 
  

   within 
  this 
  range 
  the 
  values 
  of 
  i/1 
  are 
  maxima, 
  decreasing 
  

   either 
  for 
  higher 
  or 
  lower 
  pressures. 
  But 
  the 
  figures 
  may 
  

   not 
  seem 
  to 
  bear 
  out 
  the 
  statement 
  that 
  these 
  maximum 
  values 
  

   are 
  independent 
  of 
  I 
  and 
  of 
  the 
  nature 
  of 
  the 
  gas. 
  However, 
  

   several 
  times 
  in 
  the 
  course 
  of 
  the 
  work 
  it 
  was 
  found 
  that 
  the 
  

   value 
  of 
  i/1 
  changed 
  suddenly 
  and 
  abnormal 
  values 
  (not 
  

   recorded 
  in 
  the 
  tables) 
  were 
  found 
  ; 
  after 
  a 
  time, 
  especially 
  

   if 
  the 
  apparatus 
  was 
  taken 
  down 
  and 
  put 
  up 
  again, 
  i/1 
  would 
  

   return 
  nearly 
  but 
  not 
  quite 
  to 
  its 
  previous 
  value. 
  Further 
  

   investigation 
  of 
  these 
  changes 
  showed 
  that 
  they 
  took 
  place 
  

   at 
  the 
  surface 
  of 
  the 
  plates 
  of 
  the 
  condenser 
  ; 
  it 
  appears 
  that 
  

   under 
  certain 
  conditions 
  the 
  electrodes 
  may 
  refuse 
  to 
  absorb 
  

   all 
  the 
  ions 
  falling 
  on 
  them 
  in 
  small 
  electric 
  fields. 
  The 
  

   exact 
  circumstances 
  determining 
  the 
  change 
  have 
  not 
  been 
  

   investigated, 
  but 
  it 
  was 
  found 
  that 
  aluminium, 
  especially 
  if 
  

   carefully 
  cleaned, 
  was 
  peculiarly 
  subject 
  to 
  the 
  trouble 
  ; 
  and 
  

   that 
  it 
  could 
  be 
  cured 
  completely 
  by 
  covering 
  the 
  surface 
  of 
  

   the 
  plates 
  with 
  soot. 
  This 
  procedure 
  was 
  adopted 
  in 
  all 
  the 
  

   later 
  experiments, 
  which 
  showed 
  that 
  the 
  values 
  of 
  i/1 
  

   obtained 
  previously 
  were 
  all 
  slightly 
  too 
  low 
  and 
  that 
  the 
  

   difference 
  between 
  them 
  was 
  only 
  due 
  to 
  differences 
  in 
  the 
  

   condition 
  of 
  the 
  electrodes 
  ; 
  when 
  sooted 
  electrodes 
  were 
  

   used, 
  no 
  difference 
  could 
  be 
  detected 
  between 
  the 
  maximum 
  

   values 
  of 
  i/1 
  in 
  the 
  different 
  tables. 
  

  

  7. 
  Observations 
  were 
  then 
  made 
  with 
  greater 
  values 
  of 
  V. 
  

   Variations 
  in 
  the 
  value 
  of 
  i/1 
  were 
  now 
  more 
  difficult 
  to 
  

   detect 
  because 
  the 
  values 
  were 
  in 
  all 
  cases 
  so 
  nearly 
  1 
  ; 
  but 
  

   it 
  appeared 
  again 
  that 
  there 
  is 
  a 
  range 
  of 
  pressure 
  in 
  each 
  

   gas 
  and 
  for 
  each 
  value 
  of 
  I 
  within 
  which 
  the 
  values 
  of 
  i/1 
  

   are 
  a 
  maximum, 
  independent 
  of 
  the 
  nature 
  of 
  the 
  gas 
  and 
  

   of 
  /. 
  This 
  range 
  was 
  not 
  quite 
  the 
  same 
  as 
  in 
  the 
  case 
  of 
  

   the 
  smaller 
  values 
  of 
  V, 
  being 
  displaced 
  towards 
  higher 
  

   pressures. 
  The 
  following 
  table 
  gives 
  a 
  few 
  typical 
  figures: 
  — 
  

  

  

  

  Table 
  VII. 
  

  

  

  

  

  

  Air; 
  21 
  = 
  2-6 
  ram. 
  

  

  

  

  P 
  •• 
  

  

  ... 
  60 
  

  

  104 
  212 
  353 
  

  

  468 
  

  

  762 
  

  

  V=4. 
  i/1 
  .. 
  

  

  .... 
  0-971 
  

  

  0-982 
  0-988 
  0-987 
  

  

  0-963 
  

  

  0-921 
  

  

  Y=2. 
  i/1 
  .. 
  

  

  .... 
  0-958 
  

  

  0-967 
  0-976 
  0-977 
  

   Hydrogen 
  ; 
  b0'9 
  mm. 
  

  

  0-950 
  

  

  0-855 
  

  

  P 
  •• 
  

  

  ... 
  Ill 
  

  

  291 
  392 
  504 
  

  

  741 
  

  

  

  V=4. 
  i/1 
  .. 
  

  

  ... 
  0-987 
  

  

  0-988 
  - 
  0-985 
  0-987 
  

  

  0-984 
  

  

  

  V=2. 
  i/1 
  .. 
  

  

  ... 
  0-962 
  

  

  0-969 
  0975 
  0976 
  

  

  0-972 
  

  

  

  