﻿Minimum 
  Spark 
  Potentials. 
  

  

  457 
  

  

  potentials 
  near 
  350 
  volts 
  a 
  rapid 
  series 
  of 
  discharges 
  or 
  

   short-circuits 
  -would 
  occur, 
  accompanied 
  by 
  a 
  humming 
  note, 
  

   very 
  like 
  that 
  o£ 
  the 
  Trevelyan 
  rocker 
  ; 
  evidently 
  the 
  elec- 
  

   trodes 
  were 
  set 
  in 
  rapid 
  vibration 
  by 
  the 
  electrostatic 
  force 
  

   between 
  the 
  charged 
  electrodes. 
  Clearly 
  the 
  supports 
  were 
  

   not 
  sufficiently 
  rigid. 
  

  

  This 
  phenomenon 
  suggests 
  the 
  possibility 
  that 
  what 
  have 
  

   been 
  considered 
  spark-potentials 
  for 
  very 
  short 
  sparks 
  may 
  

   not 
  have 
  been 
  at 
  all 
  the 
  potentials 
  required 
  to 
  give 
  a 
  disruptive 
  

   discharge 
  through 
  the 
  gas 
  between 
  the 
  electrodes, 
  but 
  were 
  

   rather 
  the 
  potentials 
  necessary 
  to 
  give 
  sufficient 
  electrostatic 
  

   attraction 
  between 
  the 
  electrodes 
  to 
  cause 
  a 
  displacement 
  

   equal 
  to 
  the 
  spark-gap, 
  thus 
  bringing 
  the 
  electrodes 
  in 
  con- 
  

   tact. 
  If 
  this 
  be 
  true, 
  the 
  potential 
  required 
  to 
  effect 
  a 
  

   discharge, 
  with 
  these 
  short 
  distances, 
  would 
  be 
  a 
  function 
  of 
  

   the 
  size 
  of 
  the 
  electrode, 
  especially 
  if 
  that 
  size 
  were 
  com- 
  

   parable 
  in 
  magnitude 
  to 
  the 
  spark-gap. 
  

  

  In 
  order 
  to 
  investigate 
  this 
  question 
  the 
  following 
  apparatus 
  

   w 
  T 
  as 
  devised 
  : 
  Spherical 
  electrodes 
  of 
  minute 
  dimension 
  were 
  

   obtained 
  by 
  fusing 
  in 
  the 
  oxy-hydrogen 
  flame 
  the 
  end 
  of 
  fine 
  

   platinum 
  wire, 
  "0057 
  cm. 
  in 
  diameter 
  ; 
  the 
  smallest 
  spheres 
  

   thus 
  formed 
  had 
  a 
  radius 
  very 
  approximately 
  '0035 
  cm. 
  and 
  

   were 
  essentially 
  true 
  spheres. 
  Two 
  such 
  electrodes 
  were 
  

   mounted 
  on 
  the 
  interferometer, 
  as 
  shown 
  in 
  fig. 
  1 
  (horizontal 
  

  

  Fis-. 
  1. 
  

  

  section). 
  The 
  electrodes 
  e 
  e 
  are 
  carried 
  on 
  screws 
  that 
  thread 
  

   very 
  tightly 
  through 
  the 
  steel 
  supports 
  ; 
  one 
  support 
  is 
  

   rigidly 
  clamped 
  to 
  the 
  moving 
  carriage 
  of 
  the 
  interferometer, 
  

  

  