﻿Anomalous 
  Conduction 
  in 
  a 
  Solid 
  Dielectric. 
  691 
  

  

  Quite 
  apart 
  from 
  this, 
  however, 
  if 
  the 
  fact 
  that 
  the 
  potential 
  

   of 
  A 
  starts 
  to 
  rise 
  when 
  A 
  is 
  released 
  from 
  earth 
  is 
  to 
  be 
  

   explained 
  in 
  any 
  way 
  as 
  due 
  to 
  recovery 
  from 
  strain, 
  such 
  

   a 
  rise 
  during 
  part 
  of 
  the 
  experiment 
  can 
  only 
  be 
  accounted 
  

   for 
  by 
  allowing 
  that 
  during 
  another 
  period 
  of 
  the 
  expe- 
  

   riment 
  there 
  would 
  be 
  a 
  corresponding 
  fall 
  ; 
  at 
  least, 
  this 
  

   is 
  so 
  unless 
  we 
  make 
  the 
  very 
  unlikely 
  assumption 
  that 
  the 
  

   ratio 
  of 
  the 
  steady 
  displacements 
  between 
  C 
  and 
  A 
  to 
  that 
  

   between 
  C 
  and 
  B 
  is 
  less 
  than 
  the 
  ratio 
  of 
  the 
  corresponding- 
  

   fields. 
  A 
  condition 
  in 
  which 
  A 
  fell 
  in 
  potential 
  when 
  C 
  was 
  

   positively 
  charged 
  was 
  never 
  found 
  to 
  occur, 
  and 
  we 
  can 
  

   only 
  imagine 
  the 
  possibility 
  of 
  its 
  occurrence 
  by 
  relegating 
  

   it, 
  together 
  with 
  the 
  period 
  of 
  transition, 
  to 
  the 
  few 
  minutes 
  

   which 
  elapsed 
  between 
  charging 
  the 
  plate 
  C 
  and 
  taking 
  the 
  

   first 
  observations. 
  Somewhat 
  similar 
  arouments 
  to 
  the 
  

   above 
  may 
  be 
  applied 
  to 
  show 
  that 
  it 
  is 
  very 
  unlikely 
  that 
  

   the 
  present 
  results 
  are 
  due 
  to 
  the 
  fact 
  that 
  ions 
  do 
  not 
  

   remain 
  uniformly 
  distributed 
  between 
  the 
  plates, 
  and 
  that 
  

   an 
  appreciable 
  time 
  is 
  taken 
  by 
  them 
  in 
  attaining 
  a 
  steady 
  

   state. 
  

  

  The 
  first 
  idea 
  which 
  these 
  experiments 
  suggest, 
  is 
  that 
  of 
  

   an 
  ionization 
  increasing 
  with 
  the 
  field, 
  and 
  it 
  is 
  at 
  any 
  rate 
  

   worth 
  inquiring 
  as 
  to 
  how 
  far 
  this 
  phenomenon 
  is 
  capable 
  

   of 
  providing 
  an 
  explanation 
  of 
  the 
  observed 
  effects. 
  It 
  turns 
  

   out, 
  however, 
  that 
  since 
  the 
  mean 
  free 
  path 
  of 
  an 
  electron 
  in 
  

   a 
  solid 
  is 
  probably 
  not 
  greater 
  than 
  10" 
  7 
  cm., 
  the 
  energy 
  

   given 
  by 
  the 
  field 
  to 
  an 
  electron 
  while 
  it 
  is 
  travelling 
  over 
  its 
  

   mean 
  free 
  path, 
  is, 
  even 
  in 
  the 
  case 
  of 
  a 
  field 
  of 
  50,000 
  volts 
  

   per 
  cm., 
  only 
  equivalent 
  to 
  the 
  energy 
  obtained 
  in 
  a 
  fall 
  

   through 
  0*005 
  volt, 
  and 
  this 
  is 
  far 
  too 
  small 
  to 
  enable 
  the 
  

   electron 
  to 
  ionize. 
  

  

  A 
  simple 
  estimation 
  will 
  show 
  that 
  any 
  fields 
  which 
  we 
  

   can 
  produce 
  would 
  be 
  far 
  too 
  small 
  to 
  tear 
  an 
  atom 
  to 
  pieces, 
  

   or 
  in 
  fact 
  to 
  be 
  of 
  any 
  appreciable 
  importance 
  in 
  this 
  

   respect 
  compared 
  with 
  the 
  effects 
  due 
  to 
  the 
  atoms 
  on 
  each 
  

   other 
  ; 
  and 
  we 
  are 
  led 
  to 
  look 
  for 
  an 
  explanation 
  of 
  the 
  

   phenomenon 
  as 
  resulting 
  from 
  a 
  development 
  for 
  strong- 
  

   fields 
  of 
  that 
  particular 
  theory 
  of 
  corpuscular 
  conduction 
  

   which 
  attributes 
  the 
  conductivity 
  of 
  metals 
  as 
  due 
  to 
  a 
  

   polarization 
  of 
  the 
  medium,, 
  resulting 
  in 
  corpuscles 
  being- 
  

   emitted 
  more 
  copiously 
  in 
  a 
  direction 
  opposite 
  to 
  that 
  of 
  the 
  

   field 
  than 
  in 
  any 
  other 
  direction, 
  the 
  corpuscle 
  after 
  emission 
  

   being 
  immediately 
  drawn 
  into 
  the 
  atom 
  which 
  has 
  been 
  

   responsible 
  for 
  pulling 
  it 
  out 
  (see 
  ' 
  Corpuscular 
  Theory 
  of 
  

   Matter,' 
  J. 
  J. 
  Thomson, 
  p. 
  86). 
  If, 
  following 
  Thomson, 
  we 
  

   assume 
  that 
  the 
  number 
  of 
  corpuscles 
  whose 
  axes 
  make 
  an 
  

  

  