﻿Ionization 
  in 
  various 
  Gases. 
  881 
  

  

  found 
  to 
  be 
  suitable, 
  e. 
  g. 
  about 
  150 
  mm. 
  divisions 
  per 
  volt 
  

   at 
  1 
  metre 
  from 
  the 
  mirror. 
  

  

  The 
  chamber 
  communicated 
  through 
  the 
  glass 
  side 
  tube, 
  

   already 
  referred 
  to, 
  with 
  a 
  manometer, 
  a 
  Toepler 
  pump, 
  and 
  

   the 
  gas-generating 
  apparatus. 
  All 
  the 
  gases 
  dealt 
  with 
  were 
  

   prepared 
  by 
  the 
  writer 
  from 
  the 
  purest 
  materials 
  obtainable 
  

   (usually 
  from 
  Kahlbaum), 
  and 
  were 
  carefully 
  purified 
  before 
  

   use. 
  

  

  The 
  manometer 
  was 
  of 
  the 
  U-type, 
  one 
  limb 
  communi- 
  

   cating 
  with 
  the 
  chamber, 
  while 
  by 
  a 
  special 
  arrangement 
  the 
  

   other 
  could 
  be 
  periodically 
  evacuated 
  with 
  the 
  pump 
  and 
  

   then 
  isolated. 
  At 
  first 
  mercury 
  was 
  used 
  in 
  the 
  manometer, 
  

   but 
  it 
  was 
  found 
  that 
  the 
  ionization 
  current 
  in 
  the 
  case 
  of 
  air 
  

   was 
  unmanageably 
  large 
  except 
  for 
  pressures 
  less 
  than 
  about 
  

   1 
  cm. 
  of 
  mercury. 
  This 
  did 
  not 
  render 
  accurate 
  pressure- 
  

   readings 
  easy; 
  so 
  the 
  mercury 
  was 
  replaced 
  by 
  sulphuric 
  acid. 
  

   This 
  liquid 
  had 
  the 
  additional 
  advantage 
  of 
  being 
  without 
  

   action 
  on 
  any 
  of 
  the 
  gases 
  used. 
  

  

  In 
  nearly 
  every 
  measurement 
  the 
  procedure 
  was 
  as 
  

   follows 
  : 
  — 
  The 
  chamber 
  was 
  evacuated 
  to 
  about 
  ^qq 
  w^n^. 
  of 
  

   mercury, 
  and 
  tests 
  were 
  made 
  both 
  of 
  the 
  tightness 
  of 
  the 
  

   vessel 
  and 
  of 
  the 
  perfection 
  of 
  the 
  insulation. 
  The 
  sensitiveness 
  

   of 
  the 
  electrometer 
  was 
  ascertained 
  by 
  connecting 
  to 
  a 
  Clark 
  

   cell, 
  and 
  two 
  indexes 
  were 
  placed 
  on 
  the 
  semitransparent 
  

   scale 
  at 
  a 
  distance 
  apart 
  corresponding 
  to 
  the 
  E.M.F. 
  of 
  the 
  

   Clark 
  cell. 
  A 
  little 
  dry 
  air 
  was 
  introduced 
  into 
  the 
  chamber, 
  

   w^hich 
  after 
  a 
  minute 
  or 
  so 
  was 
  re-evacuated. 
  More 
  dry 
  air 
  

   was 
  slosvly 
  run 
  in 
  until 
  the 
  pressure 
  rose 
  to 
  about 
  10 
  cms. 
  of 
  

   HoSO^. 
  The 
  earth 
  connexion 
  to 
  the 
  electrode 
  was 
  then 
  

   broken 
  and 
  the 
  time 
  taken 
  for 
  the 
  spot 
  of 
  light 
  to 
  travel 
  

   between 
  the 
  indexes 
  on 
  the 
  electrometer-scale 
  was 
  measured 
  

   with 
  a 
  chronograph 
  watch. 
  Several 
  successive 
  time-readings 
  

   were 
  taken 
  for 
  each 
  pressure-reading 
  ; 
  and 
  when 
  conditions 
  

   were 
  favourable 
  a 
  succession 
  of 
  three 
  or 
  four 
  difiering 
  by 
  

   not 
  more 
  than 
  ? 
  second 
  was 
  deemed 
  sufficient 
  for 
  the 
  

   accuracy 
  desired. 
  Such 
  sets 
  of 
  measurements 
  corresponding 
  

   to 
  each 
  pressure 
  were 
  made 
  for 
  three 
  or 
  four 
  pressures. 
  The 
  

   air 
  was 
  removed 
  by 
  the 
  pump, 
  any 
  residual 
  air 
  being 
  " 
  washed 
  

   out" 
  with 
  a 
  little 
  of 
  the 
  pure 
  gas 
  under 
  experiment. 
  The 
  

   gas 
  was 
  then 
  introduced 
  and 
  its 
  ionization 
  currents 
  and 
  cor- 
  

   responding 
  pressures 
  w^ere 
  measured 
  in 
  the 
  same 
  way: 
  and 
  

   in 
  its 
  turn 
  it 
  was 
  removed 
  and 
  " 
  washed 
  out 
  " 
  with 
  dry 
  air, 
  

   and 
  a 
  second 
  set 
  of 
  air-readings 
  was 
  taken. 
  Frequent 
  obser- 
  

   vations 
  of 
  the 
  temperature 
  of 
  the 
  ionization-vessel 
  were 
  made 
  

   by 
  means 
  of 
  a 
  thermometer 
  placed 
  in 
  contact 
  with 
  its 
  outer 
  

   surface. 
  

  

  