﻿on 
  Electric 
  Waves 
  in 
  Short 
  Wire 
  Systems. 
  603 
  

  

  in 
  the 
  primary 
  system, 
  C 
  the 
  capacity 
  of 
  the 
  secondary 
  con- 
  

   denser, 
  and 
  c 
  the 
  capacity 
  per 
  unit 
  length 
  of 
  the 
  secondary 
  

   wires, 
  the 
  capacities 
  being 
  in 
  electrostatic 
  units. 
  The 
  

   secondary 
  wires 
  may 
  he 
  considered 
  parallel, 
  so 
  Lord 
  

   Rayleigh's 
  value 
  for 
  c 
  may 
  be 
  adopted 
  (Phil. 
  Mag. 
  xliv. 
  

   p. 
  199, 
  1897). 
  Differentiating 
  the 
  equations 
  and 
  substituting 
  

   values 
  appropriate 
  to 
  the 
  apparatus 
  used 
  in 
  the 
  present 
  re- 
  

   research, 
  the 
  change 
  o£ 
  C 
  necessary 
  to 
  maintain 
  resonance 
  

   for 
  a 
  given 
  change 
  in 
  the 
  position 
  of 
  the 
  bridge 
  may 
  be 
  

   readily 
  found. 
  An 
  error 
  of 
  1 
  cm. 
  in 
  the 
  position 
  of 
  the 
  

   bridge 
  for 
  resonance 
  between 
  the 
  two 
  circuits 
  with 
  the 
  glass 
  

   between 
  the 
  condenser-plates, 
  would 
  produce 
  a 
  considerable 
  

   difference 
  in 
  the 
  calculated 
  value 
  of 
  K. 
  The 
  method 
  is 
  not 
  

   so 
  sensitive 
  as 
  the 
  one 
  which 
  has 
  been 
  used 
  in 
  the 
  present 
  

   investigation. 
  

  

  Many 
  observers, 
  in 
  determining 
  specific 
  inductive 
  capacities, 
  

   have 
  used 
  slabs 
  of 
  dielectrics 
  in 
  fields 
  not 
  rigorously 
  uniform 
  

   without 
  taking 
  that 
  fact 
  into 
  consideration, 
  and 
  have 
  neglected 
  

   altogether 
  the 
  edge 
  correction 
  to 
  the 
  capacity 
  of 
  the 
  con- 
  

   densers. 
  The 
  values 
  of 
  K 
  given 
  by 
  such 
  experimenters 
  may 
  

   be 
  very 
  much 
  in 
  error. 
  

  

  Northrup 
  (Phil. 
  Mag. 
  xxxix. 
  p. 
  78, 
  1895) 
  describes 
  a 
  

   method 
  for 
  comparing 
  the 
  values 
  of 
  the 
  specific 
  inductive 
  

   capacities 
  of 
  a 
  substance 
  under 
  slowly 
  and 
  rapidly 
  changing 
  

   fields. 
  His 
  apparatus, 
  a 
  modified 
  form 
  of 
  Gordon's 
  five- 
  

   plate 
  balance 
  (Phil. 
  Trans, 
  p. 
  427, 
  1879, 
  and 
  Hopkinson, 
  Phil. 
  

   Trans, 
  clxxii. 
  pt. 
  ii., 
  1881), 
  may 
  be 
  capable 
  of 
  showing 
  that 
  

   the 
  specific 
  inductive 
  capacity 
  of 
  a 
  material 
  is 
  not 
  greatly 
  

   different 
  in 
  slowly 
  and 
  in 
  rapidly 
  varying 
  fields; 
  but 
  his 
  

   method 
  is 
  quite 
  inadequate 
  for 
  determining 
  the 
  actual 
  value 
  

   of 
  the 
  constant. 
  He 
  neglects 
  altogether 
  the 
  edge 
  correction 
  

   to 
  the 
  capacity 
  of 
  the 
  condensers 
  formed 
  by 
  the 
  small 
  plates 
  

   and 
  the 
  middle 
  one. 
  As 
  the 
  glass 
  is 
  much 
  larger 
  than 
  the 
  

   small 
  plates, 
  and 
  as 
  one 
  of 
  these 
  is 
  not 
  always 
  in 
  contact 
  with 
  

   the 
  glass 
  plate, 
  the 
  capacity 
  of 
  the 
  condenser 
  of 
  which 
  the 
  

   glass 
  forms 
  part 
  of 
  the 
  dielectric 
  would 
  be 
  very 
  difficult 
  to 
  

   estimate. 
  A 
  balance 
  similar 
  to 
  that 
  described 
  by 
  Northrup 
  

   has 
  been 
  used 
  in 
  the 
  present 
  investigation, 
  a 
  Rutherford 
  

   detector 
  being 
  adopted 
  instead 
  of 
  the 
  spark 
  method 
  for 
  

   determining 
  when 
  the 
  small 
  plates 
  were 
  at 
  the 
  same 
  potential. 
  

   Following 
  Northrup's 
  method 
  exactly, 
  the 
  specific 
  inductive 
  

   capacity 
  of 
  a 
  plate 
  of 
  glass, 
  cut 
  at 
  the 
  same 
  time 
  and 
  

   from 
  the 
  same 
  sheet 
  as 
  the 
  disk 
  used 
  in 
  the 
  experiments 
  

   described 
  previously, 
  comes 
  out 
  10, 
  being 
  greater 
  than 
  the 
  

   true 
  value, 
  as 
  one 
  would 
  expect 
  from 
  the 
  incorrect 
  formula 
  

   used. 
  An 
  improved 
  method 
  at 
  once 
  suggests 
  itself 
  — 
  to 
  have 
  

  

  