﻿of 
  Pressure 
  which 
  accompany 
  Point 
  Discharge. 
  27 
  

  

  useful 
  in 
  absorbing 
  the 
  water 
  -vapour 
  formed 
  during 
  

   discharge 
  ; 
  the 
  gas 
  having 
  to 
  pass 
  through 
  them 
  in 
  entering 
  

   and 
  leaving 
  B. 
  

  

  The 
  usual 
  method 
  of 
  experimenting 
  was 
  to 
  observe 
  the 
  gauge 
  

   every 
  two 
  minutes 
  for 
  8 
  minutes 
  before 
  and 
  for 
  about 
  a 
  

   quarter 
  of 
  an 
  hour 
  after 
  the 
  discharge. 
  In 
  Curve 
  la 
  is 
  given 
  

   the 
  record 
  of 
  a 
  complete 
  experiment 
  of 
  this 
  kind 
  for 
  the 
  purest 
  

   hydrogen 
  we 
  were 
  able 
  to 
  obtain. 
  A 
  is 
  the 
  beginning 
  and 
  

   B 
  the 
  end 
  of 
  the 
  discharge. 
  The 
  greater 
  part 
  of 
  the 
  rise 
  of 
  

   pressure 
  between 
  A 
  and 
  B 
  is 
  seen 
  to 
  be 
  temporary, 
  and 
  is 
  no 
  

   doubt 
  mainly 
  due 
  to 
  heat. 
  There 
  is 
  also 
  a 
  noticeable 
  drift 
  

   of 
  the 
  gauge 
  zero, 
  which 
  owing 
  to 
  the 
  extreme 
  sensitiveness 
  

  

  © 
  © 
  / 
  © 
  

  

  of 
  the 
  gauge 
  we 
  were 
  unable 
  to 
  avoid, 
  and 
  which 
  altered 
  in 
  

   a 
  more 
  or 
  less 
  regular 
  manner 
  with 
  the 
  temperature 
  of 
  the 
  

   room. 
  We 
  found 
  that 
  within 
  the 
  limits 
  of 
  about 
  twenty 
  

   minutes 
  to 
  half 
  an 
  hour 
  we 
  could 
  depend 
  on 
  the 
  change 
  in 
  

   this 
  drift 
  being 
  small 
  and 
  uniform, 
  but 
  not 
  for 
  much 
  longer 
  

   intervals. 
  Curve 
  1 
  b 
  represents 
  Curve 
  1 
  a 
  with 
  the 
  drift 
  

   eliminated. 
  It 
  shows 
  that 
  the 
  discharge 
  has 
  brought 
  about 
  a 
  

   small 
  permanent 
  lowering 
  of 
  the 
  gas 
  pressure 
  by 
  the 
  

   amount 
  p. 
  Curve 
  lc 
  represents 
  a 
  similar 
  set 
  of 
  readings 
  for 
  

   hydrogen 
  containing 
  a 
  trace 
  of 
  oxygen, 
  the 
  drift 
  having 
  been 
  

   again 
  eliminated. 
  

  

  Discharge 
  in 
  Pure 
  Hydrogen. 
  

  

  When 
  the 
  amount 
  of 
  oxygen 
  present 
  is 
  above 
  about 
  one 
  

   part 
  by 
  volume 
  in 
  a 
  million 
  of 
  hydrogen, 
  negative 
  discharge 
  

   gives 
  higher 
  values 
  of 
  p 
  than 
  positive. 
  But 
  as 
  the 
  hydrogen 
  

   is 
  purified 
  below 
  this 
  point 
  by 
  continued 
  discharge 
  and 
  

   absorption 
  of 
  the 
  water- 
  vapour 
  formed, 
  the 
  values 
  of 
  p 
  for 
  

   positive 
  and 
  negative 
  discharge, 
  in 
  addition 
  to 
  becoming 
  very 
  

   small, 
  seem 
  to 
  approach 
  equality, 
  in 
  the 
  case 
  of 
  discharge 
  

   against 
  a 
  copper 
  plate. 
  

  

  To 
  illustrate 
  this 
  we 
  have 
  plotted 
  as 
  ordinates 
  in 
  Curves 
  II. 
  

   the 
  values 
  ofp 
  for 
  a 
  series 
  of 
  two-minute 
  discharges 
  of 
  2*16 
  

   microamperes 
  against 
  the 
  copper 
  plate 
  used 
  in 
  most 
  of 
  our 
  

   work. 
  

  

  The 
  exact 
  initial 
  amount 
  of 
  oxygen 
  present 
  in 
  this 
  parti- 
  

   cular 
  case 
  was 
  unknown, 
  but 
  in 
  the 
  light 
  of 
  what 
  follows 
  we 
  

   may 
  put 
  it 
  at 
  about 
  O'OOl 
  per 
  cent. 
  The 
  observations 
  are 
  

   arranged 
  horizontally 
  in 
  the 
  order 
  in 
  which 
  they 
  were 
  made. 
  

   For 
  convenience 
  of 
  identification 
  the 
  points 
  for 
  positive 
  dis- 
  

   charge 
  (circles) 
  and 
  for 
  negative 
  (dots) 
  are 
  connected, 
  each 
  

   set 
  by 
  a 
  separate 
  line 
  ; 
  but 
  it 
  will 
  be 
  understood 
  that 
  they 
  

   were 
  obtained, 
  positive 
  and 
  negative, 
  as 
  they 
  occur 
  along 
  the 
  

  

  