﻿to 
  the 
  Measurement 
  of 
  Short 
  Electric 
  Waves. 
  257 
  

  

  since 
  only 
  one 
  of 
  the 
  terminals 
  wears 
  away, 
  a 
  pair 
  consisting 
  of 
  

   one 
  with 
  a 
  silver 
  tip 
  and 
  one 
  with 
  an 
  iridium 
  tip 
  is 
  about 
  as 
  

   good 
  as 
  two 
  of 
  iridium. 
  Also, 
  the 
  decomposition 
  of 
  the 
  oil 
  

   deposits 
  carbon 
  on 
  the 
  terminals, 
  changing 
  the 
  oscillations 
  by 
  

   temporary 
  short-circuits, 
  making 
  the 
  readings 
  irregular. 
  I 
  

   have 
  not 
  attempted 
  to 
  avoid 
  this 
  except 
  by 
  frequent 
  renewal 
  

   of 
  the 
  oil. 
  It 
  might 
  perhaps 
  better 
  be 
  done 
  by 
  forcing 
  a 
  con- 
  

   tinuous 
  jet 
  of 
  oil 
  across 
  the 
  spark. 
  As 
  a 
  liquid 
  dielectric 
  for 
  

   the 
  spark, 
  a, 
  I 
  have 
  used 
  vaseline 
  oil 
  ; 
  in 
  this 
  following 
  the 
  

   example 
  of 
  Mr. 
  Righi. 
  

  

  The 
  Break 
  of 
  the 
  Induction 
  Coil. 
  — 
  To 
  get 
  a 
  strong, 
  con- 
  

   stant 
  source 
  of 
  the 
  waves, 
  it 
  is 
  important 
  to 
  have 
  an 
  interrupter 
  

   for 
  the 
  primary 
  circuit 
  of 
  the 
  induction 
  coil 
  that 
  is 
  rapid 
  and 
  

   uniform. 
  To 
  this 
  end 
  I 
  have 
  employed 
  a 
  motor-break 
  of 
  the 
  

   Foucault 
  type. 
  A 
  platinum 
  rod, 
  attached 
  directly 
  by 
  an 
  eccen- 
  

   tric 
  to 
  the 
  shaft 
  of 
  an 
  alternating-current 
  motor, 
  is 
  plunged 
  

   up 
  and 
  down 
  in 
  a 
  cup 
  of 
  mercury. 
  The 
  motor 
  is 
  of 
  the 
  type 
  

   supplied 
  by 
  the 
  General 
  Electric 
  Company 
  to 
  run 
  their 
  small 
  

   portable 
  fans, 
  and 
  makes 
  about 
  1,500 
  revolutions 
  per 
  minute. 
  

   The 
  surface 
  of 
  the 
  mercury 
  is 
  kept 
  clean 
  by 
  a 
  stream 
  of 
  water 
  

   that 
  flows 
  continually 
  over 
  it 
  — 
  a 
  device 
  previously 
  employed 
  

   by 
  Mr. 
  Maclean* 
  in 
  this 
  laboratory. 
  In 
  spite 
  of 
  these 
  precau- 
  

   tions 
  I 
  have 
  not 
  been 
  able 
  to 
  obtain 
  a 
  constant 
  source 
  of 
  waves 
  ; 
  

   the 
  oscillator 
  slowly 
  deteriorates, 
  and 
  the 
  readings 
  of 
  the 
  

   instrument 
  get 
  gradually 
  smaller. 
  This 
  source 
  of 
  error 
  is, 
  

   however, 
  quite 
  regular, 
  and 
  in 
  most 
  of 
  the 
  work 
  of 
  comparing 
  

   quantities 
  can 
  be 
  fairly 
  well 
  eliminated 
  by 
  taking 
  alternate 
  

   readings 
  and 
  averaging. 
  

  

  For 
  good 
  adjustment 
  of 
  the 
  oil 
  spark-gap 
  a 
  deflection 
  of 
  

   8 
  cm 
  can 
  be 
  obtained 
  with 
  the 
  oscillator 
  at 
  50 
  cm 
  from 
  the 
  resona- 
  

   tor. 
  For 
  better 
  regularity 
  of 
  the 
  measurements, 
  the 
  spark 
  was 
  

   ordinarily 
  adjusted 
  so 
  as 
  to 
  give 
  between 
  1 
  and 
  3 
  cm 
  deflections. 
  

  

  Quant 
  itativeness 
  of 
  the 
  readings. 
  — 
  At 
  the 
  outset 
  the 
  ques- 
  

   tion 
  arises 
  as 
  to 
  what 
  relation 
  exists 
  between 
  the 
  deflections 
  of 
  

   the 
  instrument 
  and 
  the 
  intensity 
  of 
  the 
  radiations 
  that 
  produce 
  

   them. 
  We 
  should 
  suppose 
  beforehand 
  that 
  these 
  two 
  quanti- 
  

   ties 
  are 
  directly 
  proportional, 
  for 
  the 
  deflections 
  are 
  propor- 
  

   tional 
  to 
  the 
  electromotive 
  force 
  at 
  the 
  thermal 
  knot. 
  This 
  

   e.m.f. 
  should 
  be 
  very 
  approximately 
  proportional 
  to 
  the 
  heat 
  

   developed 
  there, 
  which 
  being 
  of 
  the 
  dimensions 
  of 
  energy 
  

   should 
  be 
  proportional 
  to 
  the 
  intensity 
  of 
  the 
  waves. 
  

  

  Attempt 
  to 
  test 
  this 
  point 
  by 
  depriving 
  the 
  oscillator 
  of 
  its 
  

   parabolic 
  reflector 
  and 
  measuring 
  the 
  deflections 
  produced 
  in 
  

   the 
  resonator 
  at 
  various 
  distances 
  from 
  the 
  oscillator 
  presents 
  

   difficulty 
  ; 
  for, 
  in 
  the 
  first 
  place, 
  if 
  we 
  start 
  with 
  the 
  conclu- 
  

   sions 
  of 
  Hertz's 
  mathematical 
  theory 
  of 
  the 
  oscillator 
  that 
  the 
  

  

  *G-. 
  V. 
  Maclean, 
  this 
  Journal, 
  vol. 
  viii, 
  July, 
  1899. 
  

  

  