﻿70 
  L. 
  DeForest 
  — 
  Reflection 
  of 
  Hertzian 
  Waves. 
  

  

  between 
  the 
  two 
  bridges 
  decreases 
  the 
  discrepancies 
  between 
  

   throws 
  are 
  seen 
  to 
  diminish, 
  approaching 
  zero. 
  The 
  effective 
  

   length 
  of 
  the 
  wave-train 
  is 
  about 
  8 
  meters. 
  

  

  I^rom 
  this 
  curve 
  the 
  exponential 
  decrement 
  of 
  amplitudes 
  is 
  

   found 
  to 
  be 
  a:? 
  = 
  -00095 
  per 
  cm. 
  

  

  Experiments 
  vnth 
  the 
  Coherer. 
  

  

  A 
  coherer 
  was 
  used 
  in 
  looking 
  for 
  polarization 
  in 
  the 
  field 
  

   near 
  the 
  wires. 
  A 
  glass 
  tube 
  half 
  filled 
  with 
  coarse 
  copper 
  

   filings 
  was 
  placed 
  in 
  the 
  nose 
  of 
  a 
  large 
  zinc 
  funnel, 
  soldered 
  

   into 
  the 
  side 
  of 
  a 
  tin 
  box. 
  The 
  galvanometer, 
  dry 
  cell, 
  and 
  a 
  

   graphite 
  resistance 
  of 
  some 
  150,000 
  ohms, 
  were 
  shielded 
  in 
  this, 
  

   leaving 
  only 
  a 
  small 
  aperture 
  through 
  which 
  to 
  read 
  the 
  

   mirror. 
  Thus 
  protected 
  the 
  coherer 
  responded 
  only 
  to 
  the 
  

   waves 
  from 
  the 
  parallel 
  wires 
  placed 
  near. 
  

  

  Three 
  gratings 
  were 
  tried 
  : 
  one 
  made 
  of 
  half-inch 
  strips 
  of 
  

   zinc, 
  alternating 
  with 
  paper 
  ; 
  the 
  second, 
  of 
  lengths 
  of 
  insulated 
  

   copper 
  wire 
  3™"" 
  in 
  diameter, 
  placed 
  as 
  closely 
  as 
  possible 
  ; 
  the 
  

   third, 
  of 
  wires 
  0*6"°' 
  diameter, 
  strung 
  5""°^ 
  apart. 
  

  

  Along 
  the 
  length 
  of 
  the 
  Lecher 
  wires 
  absolutely 
  no 
  polar- 
  

   ized 
  displacements 
  were 
  found, 
  whether 
  the 
  coherer 
  lay 
  in 
  the 
  

   plane 
  of 
  the 
  two 
  parallel 
  wires, 
  or 
  in 
  one 
  perpendicular 
  to 
  

   this 
  and 
  equally 
  distant 
  from 
  both 
  wires 
  — 
  and 
  whether 
  the 
  

   coherer 
  were 
  close 
  to 
  or 
  far 
  from 
  these. 
  The 
  waves 
  are 
  pro- 
  

   pagated 
  in 
  approximately 
  circular 
  forms 
  from 
  the 
  wires, 
  with 
  

   displacements 
  tangentially 
  perpendicular 
  to 
  them 
  ; 
  but 
  there 
  

   are 
  also 
  displacements 
  parallel 
  to 
  the 
  wires 
  from 
  the 
  flow 
  of 
  

   the 
  currents 
  in 
  them. 
  One 
  at 
  least 
  of 
  these 
  two 
  components 
  

   will 
  always 
  penetrate 
  the 
  grating. 
  

  

  That 
  the 
  displacements 
  are 
  not 
  radial 
  here 
  is 
  shown 
  by 
  the 
  

   complete 
  shielding 
  from 
  them 
  afforded 
  by 
  a 
  tin 
  pipe, 
  as 
  large 
  

   as 
  3 
  inches 
  in 
  diameter, 
  fitted 
  on 
  to 
  the 
  end 
  of 
  the 
  funnel. 
  If 
  

   this 
  extended 
  more 
  than 
  three 
  inches 
  the 
  action 
  was 
  entirely 
  

   cut 
  off. 
  This 
  amount 
  of 
  protection 
  was 
  necessary, 
  however, 
  to 
  

   shield 
  the 
  coherer 
  from 
  diffuse 
  radiation. 
  

  

  The 
  coherer 
  was 
  next 
  placed 
  beyond 
  and 
  facing 
  the 
  ends 
  of 
  

   the 
  parallel 
  wires 
  (S'^'" 
  apart). 
  Here 
  also 
  at 
  a 
  distance 
  greater 
  

   than 
  15*^'" 
  from 
  the 
  ends 
  the 
  waves 
  were 
  not 
  polarized, 
  and 
  

   were 
  cut 
  oft' 
  as 
  before 
  by 
  a 
  long 
  ])rojecting 
  tin 
  cap. 
  When, 
  

   however, 
  the 
  ends 
  were 
  brought 
  up 
  close 
  to 
  the 
  grating, 
  or 
  about 
  

   gem 
  fj.Qjyj 
  x\\Q 
  coherer, 
  a 
  very 
  decided 
  polarization 
  was 
  found. 
  

   The 
  plane 
  of 
  the 
  two 
  wires 
  being 
  horizontal, 
  the 
  galvanometer 
  

   gave 
  deflections 
  averaging 
  17 
  for 
  the 
  grating 
  strips 
  vertical, 
  

   and 
  none 
  whatever 
  for 
  them 
  horizontal. 
  This 
  was 
  with 
  the 
  

   first-mentioned 
  analyzer. 
  The 
  second 
  gave 
  average 
  deflections 
  

   corresponding 
  to 
  16 
  and 
  2 
  respectively 
  for 
  the 
  two 
  positions 
  ; 
  

  

  m 
  

  

  