﻿220 
  Messrs. 
  H. 
  Gr. 
  J. 
  Moseley 
  and 
  C. 
  G. 
  Darwin 
  on 
  

  

  Much 
  information 
  on 
  the 
  way 
  in 
  which 
  constituents 
  of 
  

   differing 
  penetrating 
  power 
  are 
  reflected 
  at 
  different 
  angles 
  

   can 
  be 
  obtained 
  from 
  a 
  comparison 
  of 
  curves 
  III. 
  and 
  IV. 
  

   A 
  direct 
  determination 
  was 
  also 
  made 
  of 
  the 
  extent 
  to 
  which 
  

   the 
  rays, 
  reflected 
  at 
  different 
  angles 
  from 
  rocksalt, 
  were 
  

   absorbed 
  by 
  1 
  mm. 
  of 
  aluminium. 
  The 
  state 
  of 
  the 
  tube 
  was 
  

   the 
  same 
  as 
  for 
  curve 
  II. 
  The 
  absorption 
  was 
  found 
  to 
  be 
  

   23 
  per 
  cent, 
  at 
  3°, 
  33 
  per 
  cent, 
  at 
  4°, 
  53 
  per 
  cent, 
  at 
  6°, 
  and 
  

   63 
  per 
  cent, 
  at 
  8°, 
  while 
  for 
  the 
  primary 
  beam 
  it 
  was 
  59 
  per 
  cent. 
  

   The 
  measurements 
  at 
  very 
  small 
  angles 
  were 
  not 
  easy 
  to 
  

   make, 
  since, 
  if 
  the 
  adjustment 
  either 
  of 
  the 
  slits 
  or 
  of 
  the 
  

   crystal 
  was 
  slightly 
  in 
  error, 
  part 
  of 
  the 
  primary 
  beam 
  failed 
  

   to 
  strike 
  the 
  surface 
  of 
  the 
  crystal. 
  This 
  point 
  was 
  carefully 
  

   tested 
  optically, 
  and 
  the 
  diminution 
  of 
  the 
  efficiency 
  of 
  

   reflexion 
  from 
  ferrocyanide 
  and 
  rocksalt 
  at 
  very 
  small 
  angles 
  

   Is 
  certainly 
  a 
  genuine 
  effect. 
  A 
  similar 
  result 
  was 
  obtained 
  

   with 
  selenite 
  but 
  was 
  not 
  accurately 
  measured. 
  The 
  upper 
  

   parts 
  of 
  curves 
  III. 
  and 
  IV. 
  have 
  been 
  drawn 
  from 
  the 
  results 
  

   of 
  a 
  number 
  of 
  experiments. 
  All 
  showed 
  the 
  general 
  

   characters 
  depicted, 
  though 
  the 
  results 
  obtained 
  on 
  different 
  

   days 
  were 
  not 
  in 
  perfect 
  agreement. 
  These 
  curves 
  will 
  be 
  

   discussed 
  later, 
  after 
  the 
  simpler 
  case 
  of 
  reflexion 
  of 
  mono- 
  

   chromatic 
  radiation 
  has 
  been 
  considered. 
  

  

  Even 
  at 
  the 
  most 
  favourable 
  angle 
  a 
  crystal 
  reflects 
  only 
  a 
  

   small 
  fraction 
  of 
  the 
  general 
  incident 
  radiation. 
  The 
  actual 
  

   reflexion 
  coefficient 
  depends 
  to 
  a 
  certain 
  extent 
  on 
  the 
  quality 
  

   of 
  the 
  primary 
  radiation. 
  Under 
  the 
  usual 
  working 
  con- 
  

   ditions 
  it 
  was 
  found 
  that 
  the 
  selenite 
  crystal 
  gave 
  a 
  maximum 
  

   reflexion 
  coefficient 
  *0037 
  at 
  1^°, 
  while 
  the 
  ferrocyanide 
  

   gave 
  a 
  maximum 
  of 
  *0035 
  at 
  I|°. 
  In 
  the 
  earliest 
  experi- 
  

   ments 
  * 
  a 
  reflexion 
  coefficient 
  of 
  '01 
  to 
  '02 
  was 
  found 
  for 
  mica, 
  

   but 
  it 
  is 
  probable 
  that 
  in 
  this 
  case 
  the 
  surface 
  was 
  slightly 
  

   curved 
  and 
  chanced 
  to 
  focus 
  the 
  very 
  broad 
  incident 
  beam 
  

   into 
  the 
  narrow 
  opening 
  of 
  the 
  detector. 
  Since 
  then 
  mica 
  

   has 
  been 
  little 
  used, 
  owing 
  to 
  the 
  difficulty 
  in 
  obtaining 
  a 
  

   flat 
  surface, 
  but 
  in 
  the 
  later 
  work 
  the 
  reflexion 
  coefficient 
  

   found 
  for 
  this 
  crystal 
  has 
  certainly 
  not 
  been 
  greater 
  than 
  

   about 
  -005. 
  

  

  The 
  Selective 
  Reflexion. 
  

  

  It 
  has 
  already 
  been 
  mentioned 
  that, 
  while 
  these 
  experiments 
  

   were 
  in 
  progress, 
  Prof. 
  W. 
  H. 
  and 
  W. 
  L. 
  Bragg 
  f 
  discovered 
  

   that 
  certain 
  crystals 
  reflected 
  exceptionally 
  well 
  at 
  three 
  

  

  * 
  Moseley 
  and 
  Darwin, 
  loc. 
  cit. 
  

  

  + 
  W. 
  H. 
  and 
  W. 
  L. 
  Bragg, 
  Proc. 
  Roy. 
  Soc. 
  A.lxxxviii. 
  p. 
  428 
  (1913). 
  

  

  