﻿218 
  Messrs. 
  H. 
  G. 
  J. 
  Moseley 
  and 
  C. 
  Gr. 
  Darwin 
  on 
  

  

  possible 
  eliminated 
  the 
  effect 
  of 
  selective 
  reflexion, 
  since 
  the 
  

   prominence 
  of 
  the 
  latter 
  in 
  the 
  general 
  intensity 
  curves 
  de- 
  

   pends 
  entirely 
  on 
  the 
  arrangement 
  of 
  the 
  slits. 
  In 
  the 
  curves 
  

   given 
  in 
  fig. 
  2 
  the 
  special 
  angles 
  are 
  marked 
  with 
  letters, 
  the 
  

   meaning 
  of 
  which 
  will 
  be 
  explained 
  later. 
  In 
  these 
  curves 
  

   the 
  intensity 
  is 
  plotted 
  against 
  the 
  glancing 
  angle. 
  It 
  was 
  

   found 
  that 
  the 
  shape 
  of 
  a 
  curve 
  was 
  greatly 
  influenced 
  by 
  the 
  

   state 
  of 
  the 
  tube. 
  As 
  an 
  example 
  of 
  this, 
  an 
  experiment 
  may 
  

   be 
  mentioned 
  in 
  which 
  rocksalt 
  was 
  used 
  as 
  the 
  reflector. 
  The 
  

   radiation 
  from 
  the 
  tube 
  was 
  allowed 
  to 
  become 
  gradually 
  softer. 
  

   The 
  maximum 
  of 
  the 
  curve 
  was 
  at 
  first 
  close 
  to 
  4°. 
  During 
  

   the 
  course 
  of 
  the 
  experiment 
  it 
  shifted 
  systematically 
  towards 
  

   the 
  broader 
  angles 
  and 
  finally 
  approached 
  6°. 
  Doubtless 
  if 
  a 
  

   great 
  part 
  of 
  the 
  soft 
  radiation 
  had 
  not 
  been 
  absorbed 
  in 
  its 
  

   passage 
  through 
  the 
  air, 
  the 
  maximum 
  would 
  have 
  occurred 
  

   at 
  a 
  still 
  larger 
  angle. 
  It 
  was 
  therefore 
  most 
  important 
  that 
  

   the 
  state 
  of 
  the 
  tube 
  should 
  remain 
  constant, 
  and 
  the 
  results 
  

   suffer 
  somewhat 
  from 
  the 
  impossibility 
  of 
  fulfilling 
  this 
  con- 
  

   dition 
  accurately. 
  It 
  was 
  found 
  that 
  the 
  tube 
  usually 
  remained 
  

   much 
  steadier 
  if 
  it 
  was 
  kept 
  warm 
  by 
  burning 
  an 
  electric 
  

   lamp 
  inside 
  the 
  lead 
  box. 
  The 
  tube 
  was 
  run 
  for 
  half 
  a 
  minute 
  

   at 
  regular 
  intervals 
  of 
  two 
  minutes. 
  As 
  soon 
  as 
  the 
  rays 
  

   were 
  cut 
  off, 
  the 
  electrometer 
  E 
  was 
  disconnected 
  from 
  the 
  

   detector, 
  in 
  order 
  to 
  avoid 
  the 
  natural 
  ionization 
  in 
  the 
  

   detector 
  during 
  the 
  time 
  taken 
  by 
  the 
  needle 
  to 
  come 
  to 
  

   rest. 
  The 
  ratio 
  of 
  the 
  readings 
  of 
  the 
  two 
  electrometers 
  E 
  

   and 
  F 
  was 
  taken 
  to 
  be 
  a 
  measure 
  of 
  the 
  intensity 
  of 
  reflexion, 
  

   after 
  allowing 
  for 
  the 
  natural 
  effect 
  in 
  the 
  detector. 
  In 
  the 
  

   more 
  accurate 
  work 
  the 
  current 
  through 
  the 
  detector 
  was 
  

   measured 
  by 
  a 
  balance 
  method. 
  It 
  was 
  found 
  in 
  practice 
  

   that 
  F 
  did 
  not 
  give 
  an 
  entirely 
  reliable 
  measure 
  of 
  the 
  

   intensity 
  of 
  the 
  primary 
  beam, 
  probably 
  because 
  the 
  focus 
  

   point 
  of 
  the 
  cathode 
  stream 
  was 
  liable 
  to 
  shift, 
  and 
  this 
  

   affected 
  F 
  very 
  little, 
  while 
  it 
  made 
  a 
  great 
  difference 
  in 
  the 
  

   amount 
  of 
  radiation 
  which 
  got 
  through 
  the 
  second 
  slit. 
  It 
  

   was 
  therefore 
  necessary 
  to 
  take 
  the 
  reflexion 
  at 
  one 
  angle 
  as 
  

   a 
  standard 
  and 
  return 
  to 
  it 
  after 
  every 
  few 
  readings. 
  In 
  

   fig. 
  2, 
  curve 
  I. 
  refers 
  to 
  reflexion 
  from 
  selenite. 
  The 
  state 
  

   of 
  the 
  tube 
  remained 
  very 
  nearly 
  constant 
  during 
  the 
  course 
  

   of 
  the 
  experiment, 
  which 
  took 
  several 
  days 
  to 
  complete. 
  

   Curve 
  II. 
  was 
  taken 
  with 
  rocksalt 
  and 
  curve 
  IV. 
  with 
  

   potassium 
  ferrocyanide. 
  The 
  state 
  of 
  the 
  tube 
  was 
  roughly 
  

   the 
  same 
  in 
  the 
  three 
  cases 
  and 
  1 
  mm. 
  of 
  aluminium 
  cut 
  

   down 
  the 
  primary 
  beam 
  by 
  50 
  or 
  60 
  per 
  cent. 
  Curve 
  III. 
  was 
  

   taken 
  with 
  the 
  ferrocyanide 
  crystal 
  after 
  hardening 
  the 
  

   primary 
  beam 
  by 
  passing 
  it 
  through 
  1 
  mm. 
  of 
  aluminium. 
  

  

  