﻿102 
  Messrs. 
  A. 
  Holt 
  and 
  E. 
  Hopkinson 
  on 
  Decomposition 
  

  

  Though 
  many 
  of 
  the 
  phenomena 
  attending 
  the 
  decom- 
  

   position 
  of 
  a 
  gas 
  by 
  electric 
  sparks 
  can 
  thus 
  be 
  readily 
  ex- 
  

   plained 
  by 
  diffusion, 
  this 
  explanation, 
  in 
  its 
  unmodified 
  

   form, 
  is 
  insufficient 
  to 
  account 
  both 
  for 
  the 
  important 
  fact 
  

   observed 
  by 
  Thomson 
  that 
  when 
  very 
  short 
  sparks 
  are 
  used 
  

   the 
  poles 
  at 
  which 
  the 
  excesses 
  of 
  oxygen 
  and 
  hydrogen 
  

   appear 
  are 
  reversed, 
  and 
  for 
  the 
  interesting 
  observation 
  made 
  

   by 
  Chapman 
  and 
  Lidbury, 
  that 
  the 
  total 
  amount 
  of 
  water- 
  

   vapour 
  decomposed 
  and 
  the 
  extent 
  of 
  the 
  separation 
  of 
  the 
  

   decomposition-products 
  is 
  very 
  much 
  greater 
  when 
  the 
  current 
  

   of 
  vapour 
  enters 
  the 
  spark-tube 
  in 
  the 
  neighbourhood 
  of 
  the 
  

   cathode 
  instead 
  of 
  near 
  the 
  anode. 
  

  

  When 
  sparks 
  are 
  passed 
  between 
  platinum 
  electrodes 
  in 
  

   water-vapour, 
  some 
  of 
  the 
  metal 
  is 
  shot 
  or 
  sprayed 
  off 
  the 
  

   electrodes 
  in 
  the 
  form 
  of 
  extremely 
  minute 
  particles 
  which 
  

   adhere 
  to 
  the 
  surface 
  of 
  the 
  glass, 
  and 
  gradually 
  produce 
  a 
  

   metallic 
  film 
  on 
  those 
  portions 
  of 
  the 
  spark-tube 
  which 
  imme- 
  

   diately 
  surround 
  the 
  anode 
  and 
  cathode. 
  The 
  amount 
  of 
  

   metal 
  deposited 
  from 
  the 
  anode 
  is 
  extremely 
  small 
  compared 
  

   with 
  that 
  from 
  the 
  cathode. 
  

  

  The 
  minute 
  particles 
  which 
  constitute 
  this 
  metallic 
  spray 
  

   have 
  in 
  all 
  probability 
  a 
  very 
  powerful 
  catalytic 
  action, 
  since 
  

   they 
  must 
  closely 
  resemble 
  platinum-black, 
  and 
  the 
  film 
  which 
  

   they 
  form 
  when 
  deposited 
  on 
  the 
  glass 
  would 
  also 
  have, 
  

   though 
  in 
  a 
  lesser 
  degree, 
  the 
  property 
  of 
  combining 
  electro- 
  

   lytic 
  gas. 
  

  

  It 
  appeared 
  to 
  us 
  that 
  this 
  metallic 
  spray 
  and 
  deposit, 
  with 
  

   its 
  catalytic 
  action, 
  must 
  influence 
  the 
  amount 
  of 
  water- 
  

   vapour 
  decomposed 
  by 
  the 
  sparks, 
  and 
  might 
  possibly 
  account 
  

   for 
  the 
  various 
  phenomena 
  not 
  wholly 
  explicable 
  by 
  the 
  

   hypothesis 
  of 
  gaseous 
  diffusion. 
  

  

  In 
  order 
  to 
  study 
  this 
  catalytic 
  action, 
  it 
  was 
  necessary 
  to 
  

   carry 
  out 
  parallel 
  experiments 
  with 
  platinum 
  and 
  some 
  metal 
  

   which 
  did 
  not 
  give 
  any 
  perceptible 
  spray 
  under 
  the 
  influence 
  

   of 
  the 
  spark, 
  and 
  which 
  did 
  not 
  react 
  chemically 
  with 
  water- 
  

   vapour 
  when 
  heated 
  to 
  a 
  moderate 
  temperature. 
  

  

  Aluminium 
  seemed 
  very 
  suitable 
  for 
  this 
  purpose 
  as 
  it 
  does 
  

   not 
  readily 
  oxidize 
  in 
  water-vapour, 
  ami 
  no 
  metallic 
  film 
  can 
  

   be 
  formed 
  from 
  it. 
  

  

  A 
  series 
  of 
  spark-tubes 
  of 
  the 
  shape 
  shown 
  in 
  fig. 
  5, 
  A 
  

   and 
  B, 
  were 
  used, 
  the 
  sole 
  difference 
  between 
  them 
  being 
  the 
  

   diameter 
  of 
  the 
  bulbs 
  surrounding 
  the 
  electrodes, 
  which 
  in 
  

   type 
  A 
  was 
  1 
  cm., 
  and 
  in 
  type 
  B 
  5 
  cms. 
  In 
  souk 
  1 
  tubes 
  the 
  

   electrodes 
  were 
  of 
  aluminium, 
  and 
  in 
  others 
  of 
  platinum. 
  

   With 
  this 
  latter 
  metal 
  various 
  thicknesses 
  of 
  electrodes 
  were 
  1 
  

  

  tried 
  in 
  order 
  to 
  see 
  whether 
  the 
  results 
  were 
  affected, 
  for 
  

   the 
  thinner 
  the 
  electrode 
  the 
  hotter 
  would 
  the 
  cathode 
  become 
  

  

  