﻿Ions 
  in 
  Gases 
  under 
  Different 
  Pressures. 
  301 
  

  

  entirely 
  different 
  method. 
  The 
  method 
  which 
  he 
  used 
  was 
  

   that 
  of 
  sending 
  a 
  current 
  of 
  ionized 
  air 
  along 
  a 
  tube 
  con- 
  

   taining 
  an 
  insulated 
  electrode 
  connected 
  to 
  an 
  electrometer. 
  

   By 
  taking 
  the 
  deflexion 
  for 
  the 
  electrode 
  at 
  different 
  positions 
  

   along 
  the 
  tube, 
  he 
  was 
  able 
  to 
  determine 
  the 
  rate 
  at 
  which 
  the 
  

   number 
  of 
  ions 
  diminished 
  in 
  consequence 
  of 
  recombination. 
  

   In 
  that 
  paper 
  he 
  gives 
  the 
  value 
  as 
  3420 
  e. 
  The 
  agreement 
  

   between 
  the 
  two 
  determinations 
  is 
  remarkably 
  close 
  con- 
  

   sidering 
  the 
  diversity 
  of 
  methods 
  and 
  apparatus 
  employed 
  

   by 
  Prof. 
  Townsend 
  and 
  myself. 
  

  

  The 
  value 
  of 
  a. 
  which 
  has 
  been 
  obtained 
  by 
  this 
  determina- 
  

   tion 
  for 
  air 
  at 
  atmospheric 
  pressure 
  may 
  be 
  considered 
  as 
  the 
  

   value 
  of 
  the 
  coefficient 
  of 
  recombination 
  for 
  air 
  at 
  all 
  other 
  

   pressures, 
  at 
  least 
  between 
  the 
  limits 
  of 
  *125 
  and 
  3 
  atmo- 
  

   spheres, 
  since 
  it 
  has 
  been 
  shown 
  in 
  a 
  former 
  part 
  of 
  this 
  paper 
  

   that 
  a. 
  is 
  independent 
  of 
  the 
  pressure 
  of 
  the 
  air. 
  

  

  It 
  might 
  be 
  worth 
  while 
  to 
  mention 
  briefly 
  the 
  method 
  

   which 
  was 
  employed 
  to 
  determine 
  the 
  electric 
  capacities 
  in 
  

   the 
  circuit, 
  as 
  it 
  is 
  rather 
  an 
  unusual 
  one. 
  

  

  To 
  determine 
  the 
  value 
  of 
  1ST, 
  the 
  capacity 
  in 
  the 
  circuit 
  

   was 
  simply 
  that 
  of 
  the 
  plate 
  electrodes, 
  the 
  connexions, 
  and 
  

   electrometer. 
  To 
  determine 
  Q, 
  however, 
  it 
  was 
  necessary 
  to 
  

   introduce 
  a 
  considerably 
  larger 
  capacity 
  into 
  the 
  circuit, 
  as 
  

   the 
  rate 
  of 
  movement 
  of 
  the 
  electrometer-needle 
  was 
  too 
  rapid 
  

   without 
  it. 
  For 
  this 
  purpose, 
  an 
  air-condenser 
  was 
  connected 
  

   in 
  parallel, 
  as 
  shown 
  at 
  M 
  in 
  fig. 
  1. 
  It 
  was 
  necessary 
  there- 
  

   fore 
  to 
  determine 
  the 
  capacity 
  both 
  of 
  this 
  condenser 
  and 
  of 
  

   the 
  electrometer, 
  together 
  with 
  the 
  connexions. 
  To 
  do 
  this, 
  

   the 
  following 
  method 
  was 
  used. 
  

  

  If 
  the 
  air 
  between 
  two 
  insulated 
  parallel 
  plates 
  be 
  ionized 
  

   by 
  means 
  of 
  uranium 
  oxide, 
  and 
  if 
  one 
  of 
  the 
  plates 
  be 
  

   charged 
  while 
  the 
  other 
  is 
  connected 
  to 
  the 
  electrometer 
  ; 
  

   then 
  the 
  deflexion 
  of 
  the 
  electrometer-needle 
  per 
  second 
  is 
  

   inversely 
  proportional 
  to 
  the 
  capacity 
  in 
  the 
  circuit. 
  

  

  Two 
  parallel 
  zinc 
  plates 
  were 
  therefore 
  arranged 
  horizon- 
  

   tally 
  in 
  a 
  closed 
  metal 
  box. 
  The 
  inclosing 
  box 
  was 
  connected 
  

   to 
  earth, 
  while 
  the 
  two 
  plates 
  were 
  carefully 
  insulated. 
  On 
  

   the 
  lower 
  plate 
  a 
  thin 
  layer 
  of 
  uranium 
  oxide 
  was 
  placed. 
  The 
  

   lower 
  plate 
  was 
  connected 
  to 
  the 
  battery, 
  while 
  the 
  upper 
  one 
  

   was 
  connected 
  in 
  parallel 
  with 
  the 
  electrometer 
  circuit. 
  This 
  

   of 
  course 
  introduced 
  the 
  additional 
  capacity 
  of 
  the 
  two 
  parallel 
  

   plates. 
  The 
  ratio 
  of 
  the 
  capacity 
  of 
  these 
  plates 
  to 
  that 
  of 
  

   the 
  rest 
  of 
  the 
  electrometer 
  circuit 
  was, 
  however, 
  determined 
  

   by 
  the 
  method 
  of 
  mixtures. 
  

  

  The 
  rate 
  of 
  leak 
  was 
  taken 
  with 
  a 
  condenser 
  of 
  known 
  

   capacity 
  in 
  the 
  place 
  of 
  M, 
  fig. 
  1. 
  Then 
  the 
  known 
  capacity 
  

  

  