﻿568 
  Prof. 
  E. 
  H. 
  Barton 
  and 
  Mr. 
  W. 
  B. 
  Kilby 
  on 
  

  

  mathematically 
  shown 
  to 
  follow 
  from 
  the 
  conductivity 
  pro- 
  

   duced 
  by 
  the 
  state 
  of 
  ionization 
  which 
  is 
  believed 
  to 
  exist 
  in 
  

   those 
  parts 
  of 
  the 
  atmosphere 
  under 
  consideration. 
  

  

  Other 
  speakers 
  showed 
  that 
  this 
  curvature 
  of 
  the 
  path 
  of 
  

   radiation 
  could 
  not 
  be 
  attributed 
  to 
  diffraction 
  and 
  spoke 
  

   highly 
  of 
  the 
  theory 
  propounded 
  by 
  Dr. 
  Eccles. 
  

  

  But 
  although 
  this 
  theory 
  seemed 
  so 
  preeminently 
  satis- 
  

   factory, 
  it 
  appeared 
  desirable 
  to 
  obtain, 
  if 
  possible, 
  some 
  

   experimental 
  evidence 
  along 
  the 
  same 
  lines. 
  Indeed, 
  

   Prof. 
  Fleming 
  expressed 
  to 
  one 
  of 
  us 
  his 
  wish 
  for 
  experi- 
  

   mental 
  research 
  as 
  to 
  the 
  effects 
  of 
  ionization 
  on 
  the 
  dielectric 
  

   Constant 
  and 
  on 
  the 
  conductivity. 
  Dr. 
  Eccles 
  similarly 
  

   expressed 
  his 
  interest 
  in 
  any 
  experimental 
  evidence 
  as 
  to 
  

   the 
  effect 
  of 
  ionization 
  on 
  the 
  velocity 
  of 
  electric 
  waves. 
  

  

  Other 
  kindred 
  topics 
  also 
  suggested 
  themselves 
  as 
  needing 
  

   investigation, 
  and 
  various 
  methods 
  of 
  attacking 
  them 
  were 
  

   considered. 
  It 
  seemed 
  however 
  preferable, 
  at 
  the 
  outset, 
  

   to 
  examine 
  the 
  relation 
  between 
  the 
  ionization 
  of 
  the 
  air 
  and 
  

   its 
  conductivity 
  for 
  the 
  alternating 
  currents 
  occurring 
  in 
  

   oscillations 
  whose 
  frequency 
  is 
  of 
  the 
  order 
  used 
  in 
  wireless 
  

   telegraphy. 
  For 
  obviously, 
  if 
  the 
  increased 
  conductivity 
  

   to 
  such 
  alternations 
  proved 
  detectable, 
  it 
  would 
  favour 
  the 
  

   theory 
  of 
  increased 
  speed 
  of 
  propagation 
  of 
  which 
  this 
  

   conductivity 
  is 
  an 
  integral 
  part. 
  

  

  It 
  appeared 
  to 
  us 
  that 
  the 
  change 
  of 
  conductivity 
  due 
  to 
  

   ionization 
  might 
  be 
  inferred 
  from 
  the 
  change 
  of 
  frequency 
  

   of 
  a 
  circuit 
  as 
  detected 
  by 
  the 
  Fleming 
  cymometer 
  or 
  some 
  

   appropriate 
  modification 
  of 
  it. 
  On 
  trying 
  the 
  method, 
  the 
  

   need 
  for 
  unusual 
  sensitiveness 
  was 
  soon 
  apparent. 
  This 
  was 
  

   attained 
  by 
  modifications 
  of 
  the 
  capacity 
  and 
  inductance 
  in 
  

   the 
  responding 
  circuit, 
  and 
  by 
  the 
  substitution 
  of 
  a 
  special 
  

   electrometer 
  as 
  a 
  quantitative 
  detector 
  of 
  best 
  resonance 
  

   instead 
  of 
  the 
  neon 
  tube. 
  Frequencies 
  of 
  the 
  order 
  two- 
  

   million 
  per 
  second 
  were 
  used, 
  the 
  wave-lengths 
  being 
  about 
  

   150 
  metres. 
  

  

  A 
  full 
  and 
  rigorous 
  mathematical 
  theory 
  of 
  the 
  phenomena 
  

   occurring 
  in 
  the 
  primary 
  and 
  responding 
  circuits 
  as 
  used,, 
  

   if 
  developed, 
  might 
  prove 
  intractable 
  or 
  too 
  complicated 
  for 
  

   real 
  usefulness 
  in 
  the 
  absence 
  of 
  precise 
  knowledge 
  as 
  to 
  the 
  

   values 
  of 
  the 
  constants 
  involved. 
  But 
  early 
  in 
  the 
  course 
  

   of 
  the 
  work 
  a 
  simple 
  theory 
  was 
  derived 
  for 
  the 
  discharge 
  of 
  

   a 
  leaky 
  condenser 
  through 
  an 
  inductive 
  resistance. 
  And 
  it 
  

   is 
  hoped 
  that 
  this 
  represents 
  with 
  sufficient 
  approximation 
  

   the 
  salient 
  features 
  of 
  the 
  case. 
  This 
  elementary 
  theory 
  

   shows 
  that 
  the 
  growth 
  from 
  zero 
  of 
  a 
  leak 
  in 
  the 
  condenser 
  

   at 
  first 
  increases 
  the 
  frequency 
  natural 
  to 
  the 
  circuit. 
  With 
  

  

  