﻿Anomalous 
  Conduction 
  in 
  a 
  /Solid 
  Dielectric. 
  687 
  

  

  ratio 
  1*03 
  to 
  unity. 
  When 
  allowance 
  is 
  made 
  for 
  the 
  slight 
  

  

  difference 
  between 
  these 
  quantities, 
  the 
  curves 
  for 
  the 
  

  

  two 
  plates 
  become 
  even 
  more 
  exactly 
  the 
  images 
  of 
  each 
  

  

  other. 
  

  

  Another 
  factor 
  which 
  tends 
  to 
  destroy 
  the 
  exact 
  equality 
  

  

  8TT 
  £17 
  

   between 
  -~ 
  and 
  ~ 
  arises 
  from 
  the 
  quantity 
  p 
  referred 
  

  

  to 
  above. 
  This 
  quantity 
  arises 
  mainly 
  from 
  the 
  appreciable 
  

   edge 
  effect 
  of 
  the 
  plate 
  C. 
  It 
  results 
  from 
  a 
  positive 
  charge 
  

   coining 
  to 
  the 
  centre 
  portion 
  of 
  the 
  plate 
  from 
  the 
  edges, 
  

   owing 
  to 
  the 
  fall 
  in 
  potential 
  experienced 
  by 
  the 
  plate, 
  and 
  

   caused 
  by 
  the 
  electricity 
  leaving 
  it. 
  The 
  effect 
  of 
  the 
  

   quantity 
  p 
  on 
  the 
  rise 
  of 
  potential 
  of 
  the 
  electrometer 
  is 
  

   greater 
  when 
  the 
  electrometer 
  is 
  connected 
  to 
  A 
  than 
  when 
  

   it 
  is 
  connected 
  to 
  B, 
  owing 
  to 
  the 
  nearer 
  proximity 
  of 
  the 
  

   plate 
  in 
  the 
  first 
  case, 
  but 
  the 
  effect 
  is 
  a 
  positive 
  one 
  in 
  

   both 
  cases, 
  so 
  that 
  it 
  acts 
  in 
  such 
  a 
  way 
  as 
  to 
  increase 
  the 
  

   movement 
  of 
  the 
  needle 
  when 
  the 
  electrometer 
  is 
  connected 
  

   to 
  A, 
  and 
  to 
  decrease 
  the 
  movement 
  when 
  the 
  electrometer 
  

   is 
  connected 
  to 
  B. 
  This 
  provides 
  another 
  reason 
  for 
  the 
  

   slight 
  difference 
  in 
  the 
  heights 
  of 
  the 
  ordinates 
  of 
  the 
  

   curves. 
  The 
  ratio 
  of 
  the 
  corresponding 
  ordinates 
  of 
  the 
  two 
  

   curves 
  is 
  practically 
  constant. 
  

  

  The 
  necessity 
  for 
  the 
  equality 
  in 
  the 
  alteration 
  of 
  potential 
  

   of 
  the 
  two 
  plates 
  rests 
  on 
  the 
  assumption 
  that 
  in 
  virtue 
  

   of 
  the 
  guard-rings, 
  the 
  plates 
  act 
  like 
  infinite 
  plates, 
  so 
  

   that 
  the 
  experiments 
  justify 
  this 
  assumption 
  very 
  well. 
  The 
  

   point 
  was 
  well 
  illustrated 
  by 
  some 
  preliminary 
  experiments, 
  

   taken 
  with 
  a 
  piece 
  of 
  apparatus 
  in 
  which 
  the 
  centre 
  plate 
  

   C 
  was 
  only 
  equal 
  in 
  size 
  to 
  the 
  plates 
  A 
  and 
  B 
  so 
  

   that 
  the 
  guard-ring 
  arrangement 
  was 
  imperfect. 
  Here 
  

   the 
  stream 
  lines 
  were 
  more 
  nearly 
  perpendicular 
  to 
  the 
  

   plates 
  between 
  C 
  and 
  A, 
  than 
  between 
  C 
  and 
  B. 
  The 
  effect 
  

   was 
  to 
  increase 
  the 
  rate 
  of 
  rise 
  of 
  potential 
  of 
  the 
  plate 
  A 
  

   owing 
  to 
  the 
  diminution 
  in 
  the 
  counteracting 
  effect 
  which 
  

   was 
  caused 
  by 
  the 
  electricity 
  going 
  from 
  to 
  B. 
  The 
  same 
  

   effect 
  caused 
  a 
  diminution 
  in 
  the 
  fall 
  of 
  potential 
  of 
  the 
  

   plate 
  B, 
  so 
  that 
  the 
  ordinates 
  of 
  the 
  curve 
  for 
  the 
  plate 
  B 
  

   were 
  less 
  than 
  the 
  corresponding 
  ordinates 
  for 
  plate 
  A 
  by 
  

   30 
  per 
  cent. 
  

  

  The 
  fact 
  that 
  in 
  the 
  experiments 
  made 
  with 
  the 
  final 
  

   apparatus 
  the 
  curve 
  for 
  the 
  plate 
  A 
  is 
  so 
  nearly 
  an 
  image 
  of 
  

   the 
  curve 
  for 
  the 
  plate 
  B 
  constitutes 
  an 
  important 
  fact 
  in 
  

   favour 
  of 
  the 
  experiments, 
  because 
  it 
  shows 
  that 
  the 
  alteration 
  

   in 
  the 
  potentials 
  of 
  the 
  plates, 
  which 
  should 
  be 
  zero 
  if 
  Ohm's 
  

   law 
  were 
  obeyed, 
  is 
  to 
  be 
  attributed 
  to 
  a 
  true 
  departure 
  

  

  