﻿700 
  Mr. 
  J. 
  R. 
  Boon 
  on 
  the 
  Effect 
  of 
  Pressure 
  on 
  the 
  

  

  to 
  the 
  pressure, 
  and 
  consequently 
  the 
  wind-pressure 
  slopes 
  in 
  

   both 
  signs 
  of 
  discharge 
  are 
  altered 
  in 
  the 
  same 
  ratio. 
  

  

  It 
  would 
  seem 
  that 
  a 
  further 
  check 
  on 
  the 
  results 
  might 
  be 
  

  

  obtained 
  by 
  comparing 
  the 
  slopes 
  -J- 
  for 
  a 
  given 
  sign 
  at 
  

  

  different 
  atmospheric 
  pressures. 
  Neglecting 
  the 
  effect 
  of 
  

  

  any 
  back 
  discharge 
  which 
  may 
  be 
  present, 
  -j- 
  for 
  a 
  given 
  

  

  sign 
  should 
  be 
  proportional 
  to 
  II 
  since 
  it 
  is 
  inversely 
  pro- 
  

   portional 
  to 
  the 
  specific 
  velocity 
  of 
  the 
  ions. 
  

  

  Unfortunately, 
  the 
  pressure-gauge 
  and 
  its 
  calibrator, 
  in 
  the 
  

   particular 
  form 
  in 
  which 
  they 
  were 
  set 
  up, 
  were 
  not 
  designed 
  

   for 
  absolute 
  measurements 
  of 
  pressure-difference 
  at 
  atmo- 
  

   spheric 
  pressures 
  far 
  removed 
  from 
  normal 
  ; 
  and 
  in 
  the 
  

   sensitive 
  form 
  in 
  which 
  the 
  gauge 
  had 
  to 
  be 
  used, 
  strains 
  set 
  

   up 
  in 
  its 
  framework 
  at 
  lower 
  pressures 
  would 
  introduce 
  a 
  

   change 
  in 
  constant. 
  The 
  change 
  in 
  constant 
  could 
  not 
  be 
  

   estimated, 
  and 
  to 
  obtain 
  the 
  absolute 
  values 
  of 
  wind-pressure, 
  

   at 
  other 
  than 
  normal 
  atmospheric 
  pressure, 
  the 
  scale 
  deflexions 
  

   would 
  have 
  to 
  be 
  multiplied 
  by 
  some 
  unknown 
  constant 
  

   depending 
  on 
  the 
  value 
  of 
  II. 
  

  

  Conclusions. 
  

  

  Several 
  views 
  may 
  be 
  put 
  forward 
  to 
  explain 
  the 
  z 
  

   phenomenon. 
  

  

  (1) 
  The 
  ion 
  may 
  not 
  be 
  fully 
  formed 
  on 
  leaving 
  the 
  point 
  

   and 
  may 
  travel 
  a 
  small 
  distance 
  before 
  completely 
  clustering 
  ; 
  

   z 
  may 
  be 
  taken 
  as 
  a 
  measure 
  of 
  this 
  distance. 
  On 
  this 
  view 
  

   the 
  distance 
  travelled 
  by 
  an 
  unclustered 
  ion 
  would 
  be 
  probably 
  

   jointly 
  dependent 
  on 
  the 
  field 
  and 
  on 
  the 
  number 
  of 
  collisions 
  

   in 
  a 
  given 
  time. 
  Thus 
  if 
  F 
  = 
  field 
  at 
  the 
  point 
  where 
  com- 
  

   plete 
  clustering 
  occurs 
  and 
  11= 
  pressure 
  in 
  apparatus, 
  z 
  

   would 
  be 
  a 
  function 
  of 
  F/U. 
  Now 
  not 
  only 
  does 
  the 
  field 
  

   diverge 
  rapidly 
  from 
  the 
  point 
  but 
  it 
  also 
  varies 
  according 
  

   to 
  a 
  complex 
  law 
  with 
  pressure. 
  Owing 
  to 
  this 
  and 
  to 
  the 
  

   absence 
  of 
  experimental 
  data 
  or 
  of 
  any 
  satisfactory 
  theory 
  of 
  

   the 
  motion 
  of 
  the 
  ions 
  in 
  very 
  strong 
  fields, 
  it 
  is 
  impossible 
  to 
  

   say 
  what 
  form 
  the 
  function 
  would 
  take. 
  One 
  would 
  certainly 
  

   not, 
  however, 
  expect 
  such 
  a 
  simple 
  relationship 
  between 
  z 
  

   and 
  II 
  as 
  experiment 
  gives. 
  

  

  (2) 
  In 
  the 
  strong 
  fields 
  near 
  the 
  point 
  an 
  ion 
  may 
  not 
  

   move 
  viscously, 
  and 
  may 
  not 
  therefore 
  exert 
  its 
  full 
  drag 
  on 
  

   the 
  gas. 
  The 
  same 
  remarks 
  as 
  to 
  lack 
  of 
  data 
  or 
  theory 
  apply 
  

   also 
  in 
  this 
  case, 
  and 
  one 
  cannot, 
  therefore, 
  completely 
  dismiss 
  

   this 
  as 
  an 
  explanation. 
  The 
  simplicity 
  of 
  the 
  experimental 
  

   results 
  again 
  seems 
  to 
  render 
  this 
  explanation 
  decidedly 
  

   improbable. 
  

  

  