﻿Canal 
  Rays 
  from 
  Hollow 
  Cathodes. 
  181 
  

  

  decreasing 
  pressure, 
  as 
  will 
  be 
  seen 
  from 
  the 
  figures 
  16 
  and 
  

   17, 
  corresponds 
  to 
  the 
  blue 
  beams, 
  starting 
  from 
  cylindrical 
  

   tubes, 
  indicated 
  in 
  the 
  figures 
  3 
  and 
  10. 
  We 
  have 
  seen 
  that 
  

   these 
  blue 
  beams 
  begin 
  to 
  disappear 
  while 
  the 
  rings 
  of 
  

   phosphorescence 
  shown 
  in 
  fig. 
  4 
  gradually 
  start 
  into 
  existence 
  

   with 
  decreasing 
  pressure. 
  Those 
  rings 
  are 
  due 
  to 
  cathode 
  rays 
  

   produced 
  by 
  the 
  impact 
  of 
  positive 
  ions 
  against 
  the 
  inner 
  

   surface 
  of 
  the 
  metal 
  tube 
  ; 
  the 
  positive 
  ions 
  themselves 
  are 
  

   due 
  to 
  the 
  ionization 
  of 
  the 
  gas 
  in 
  the 
  negative 
  glow 
  pro- 
  

   duced 
  by 
  the 
  corpuscles. 
  Just 
  as 
  in 
  the 
  case 
  of 
  a 
  cylindrical 
  

   cathode 
  the 
  rings 
  appear 
  gradually 
  while 
  the 
  blue 
  beams, 
  

   different 
  in 
  origin 
  from 
  the 
  primary 
  cathode 
  rays, 
  disappear, 
  

   so 
  we 
  are 
  to 
  expect 
  from 
  the 
  view 
  given 
  that 
  the 
  maximum 
  

   of 
  luminosity 
  of 
  the 
  blue 
  cathode 
  rays 
  will 
  appear 
  at 
  a 
  higher 
  

   pressure 
  than 
  the 
  maximum 
  of 
  brightness 
  of 
  the 
  positive 
  rays. 
  

   And 
  that 
  is 
  exactly 
  what 
  happens. 
  

  

  A 
  further 
  argument 
  in 
  favour 
  of 
  the 
  view 
  developed 
  is 
  the 
  

   appearance 
  of 
  the 
  narrow 
  strips 
  of 
  phosphorescence 
  repre- 
  

   sented 
  in 
  fig. 
  20, 
  which 
  accompany 
  the 
  central 
  band. 
  They 
  

   occur 
  at 
  the 
  same 
  stage 
  of 
  pressure 
  at 
  which 
  the 
  red 
  positive 
  

   rays 
  get 
  very 
  bright. 
  These 
  narrow 
  bands 
  are 
  obviously 
  of 
  

   the 
  same 
  origin 
  as 
  the 
  ring 
  a 
  in 
  fig. 
  4. 
  Positive 
  particles 
  

   travelling 
  inwards 
  from 
  the 
  negative 
  glow 
  DE 
  may 
  either 
  

   strike 
  against 
  the 
  metal 
  or 
  pass 
  through 
  between 
  the 
  two 
  

   plates. 
  The 
  former 
  ones 
  cause 
  cathode 
  rays 
  which 
  are 
  

   indicated 
  by 
  the 
  bright 
  bands 
  of 
  phosphorescence 
  ; 
  the 
  latter 
  

   constitute 
  the 
  red 
  beams 
  from 
  the 
  corners 
  of 
  the 
  triangle. 
  

  

  We 
  also 
  understand 
  that 
  the 
  three 
  bands 
  of 
  phosphorescence 
  

   along 
  the 
  meridian 
  of 
  the 
  bulb 
  are 
  widening 
  out 
  at 
  places 
  in 
  

   front 
  of 
  the 
  corners 
  and 
  contracting 
  before 
  the 
  sides 
  of 
  the 
  

   triangle. 
  If 
  we 
  consider 
  a 
  corpuscle 
  just 
  leaving 
  one 
  plate 
  

   at 
  a 
  corner 
  in 
  the 
  direction 
  towards 
  the 
  other 
  plate, 
  it 
  will 
  be 
  

   attracted 
  by 
  the 
  stream 
  of 
  positive 
  electricity, 
  and 
  its 
  deflexion 
  

   from 
  the 
  normal 
  or 
  polar 
  direction 
  will 
  be 
  less 
  in 
  the 
  

   beginning 
  of 
  its 
  career 
  than 
  when 
  it 
  starts 
  from 
  the 
  middle 
  

   of 
  one 
  side, 
  in 
  which 
  case 
  it 
  will 
  be 
  driven 
  away 
  from 
  the 
  

   normal 
  direction 
  by 
  the 
  stream 
  of 
  negative 
  electricity 
  between 
  

   the 
  two 
  plates. 
  Thus 
  corpuscles 
  rising 
  from 
  the 
  corners 
  will 
  

   move 
  in 
  a 
  path 
  further 
  apart 
  from 
  the 
  meridian 
  plane 
  lying 
  

   parallel 
  to 
  the 
  triangular 
  plates, 
  than 
  will 
  corpuscles 
  which 
  

   start 
  from 
  the 
  middle 
  of 
  a 
  side 
  of 
  the 
  triangles. 
  

  

  From 
  the 
  theory 
  illustrated 
  in 
  fig. 
  21 
  we 
  draw 
  the 
  follow- 
  

   ing 
  conclusions 
  : 
  — 
  

  

  (1) 
  If 
  we 
  compose 
  a 
  cathode 
  of 
  two 
  circular 
  plates 
  instead 
  

   of 
  two 
  triangular 
  ones, 
  we 
  should 
  expect 
  starting 
  from 
  any 
  

   point 
  of 
  the 
  circumference 
  of 
  the 
  free 
  space 
  between 
  the 
  

  

  