﻿of 
  Water-Vapour 
  by 
  Electric 
  Sparks. 
  95 
  

  

  the 
  spark. 
  The 
  hydrogen, 
  on 
  account 
  of 
  its 
  great 
  power 
  of 
  

   diffusion, 
  tends 
  to 
  become 
  uniformly 
  distributed 
  throughout 
  

   the 
  apparatus, 
  even 
  if 
  a 
  rapid 
  current 
  of 
  vapour 
  is 
  passing 
  

   through 
  it. 
  

  

  The 
  distribution 
  of 
  the 
  oxygen, 
  however, 
  depends 
  far 
  more 
  

   on 
  the 
  water-vapour 
  current 
  ; 
  it 
  will 
  be 
  divided 
  according 
  to 
  

   the 
  position 
  of 
  the 
  entrance-tube. 
  When 
  this 
  tube 
  is 
  in 
  the 
  

   neighbourhood 
  of 
  the 
  cathode, 
  the 
  greater 
  portion 
  of 
  the 
  

   oxygen 
  is 
  driven 
  out 
  at 
  the 
  anode, 
  the 
  result 
  being 
  that 
  an 
  

   excess 
  of 
  oxygen 
  is 
  found 
  at 
  the 
  anode 
  and 
  an 
  equivalent 
  

   excess 
  of 
  hydrogen 
  at 
  the 
  cathode. 
  

  

  If 
  this 
  entrance-tube 
  is 
  brought 
  to 
  the 
  anode, 
  the 
  hydrogen, 
  

   on 
  account 
  of 
  its 
  power 
  of 
  diffusion, 
  again 
  tends 
  to 
  become 
  

   equally 
  distributed 
  through 
  all 
  parts 
  of 
  the 
  apparatus. 
  The 
  

   greater 
  portion 
  of 
  the 
  oxygen, 
  however, 
  is 
  now 
  swept 
  out 
  at 
  

   the 
  cathode, 
  so 
  that 
  it 
  appears 
  in 
  excess 
  at 
  that 
  pole 
  while 
  

   hydrogen 
  is 
  found 
  at 
  the 
  anode. 
  

  

  If 
  equal 
  quantities 
  of 
  oxygen 
  are 
  swept 
  towards 
  both 
  poles, 
  

   pure 
  electrolytic 
  gas 
  should 
  be 
  collected 
  from 
  both 
  anode 
  and 
  

   cathode, 
  and 
  Chapman 
  and 
  Lidbury 
  found 
  that 
  this 
  condition 
  

   was 
  reached 
  when 
  the 
  entrance-tube 
  was 
  situated 
  near 
  the 
  

   centre 
  of 
  the 
  spark-gap. 
  

  

  The 
  faster 
  the 
  stream 
  of 
  vapour 
  passes 
  through 
  the 
  appa- 
  

   ratus 
  the 
  less 
  easily 
  does 
  the 
  oxygen 
  diffuse 
  against 
  it, 
  whereas 
  

   the 
  distribution 
  of 
  the 
  hydrogen 
  is 
  almost 
  unaffected, 
  so 
  the 
  

   amount 
  of 
  separation 
  of 
  the 
  decomposition 
  products 
  should 
  

   increase 
  as 
  the 
  stream 
  of 
  vapour 
  becomes 
  more 
  and 
  more 
  

   rapid. 
  

  

  In 
  order 
  to 
  investigate 
  the 
  truth 
  of 
  this 
  view 
  of 
  the 
  part 
  

   played 
  by 
  diffusion, 
  we 
  have 
  carried 
  out 
  a 
  number 
  of 
  experi- 
  

   ments 
  with 
  both 
  water-vapour 
  and 
  carbon 
  dioxide, 
  for 
  it 
  

   should 
  be 
  impossible 
  to 
  obtain 
  any 
  separation 
  of 
  the 
  oxygen 
  

   and 
  carbon 
  monoxide 
  produced 
  by 
  the 
  passage 
  of 
  sparks 
  

   through 
  this 
  latter 
  gas, 
  since 
  both 
  carbon 
  monoxide 
  and 
  

   oxygen 
  diffuse 
  at 
  almost 
  similar 
  rates, 
  and 
  would, 
  conse- 
  

   quently, 
  become 
  uniformly 
  distributed 
  in 
  all 
  parts 
  of 
  the 
  

   apparatus. 
  

  

  In 
  the 
  first 
  series 
  of 
  experiments, 
  the 
  discharge 
  from 
  an 
  

   induction-coil 
  with 
  a 
  large 
  condenser 
  was 
  passed 
  through 
  a 
  

   spark-tube 
  of 
  the 
  shape 
  shown 
  in 
  fig. 
  2 
  (p. 
  96). 
  A 
  rapid 
  

   current 
  of 
  water- 
  vapour 
  entered 
  the 
  bulb 
  D, 
  which 
  was 
  blown 
  

   in 
  the 
  centre 
  of 
  the 
  spark-gap. 
  The 
  stream 
  divided, 
  any 
  gas 
  

   which 
  might 
  diffuse 
  out 
  of 
  the 
  spark 
  was 
  driven 
  away 
  through 
  

   the 
  side-tube 
  E 
  to 
  a 
  condenser 
  which 
  was 
  surrounded 
  by 
  a 
  

   freezing 
  - 
  mixture, 
  while 
  the 
  decomposition 
  - 
  products 
  re- 
  

   maining 
  in 
  the 
  path 
  of 
  the 
  spark 
  were 
  swept 
  by 
  the 
  stream 
  

  

  