﻿Minute 
  Metal 
  Particles 
  for 
  Ligld-ivaves. 
  397 
  

  

  metallic 
  deposits 
  on 
  glass 
  which 
  the 
  microscope 
  shows 
  to 
  be 
  

   made 
  up 
  of 
  particles 
  smaller 
  than 
  the 
  wave-lengths 
  of 
  light, 
  

   which 
  by 
  transmitted 
  light 
  exhibit 
  colours 
  quite 
  as 
  brilliant 
  as 
  

   those 
  produced 
  by 
  aniline 
  dyes. 
  I 
  have 
  sought 
  in 
  every 
  way 
  

   to 
  explain 
  these 
  colours, 
  by 
  all 
  the 
  well-known 
  principles 
  of 
  

   interference 
  and 
  diffraction, 
  and 
  at 
  the 
  present 
  am 
  forced 
  to 
  

   accept 
  the 
  hypothesis 
  suggested 
  in 
  the 
  title 
  of 
  this 
  paper, 
  

   which 
  I 
  have 
  chosen 
  not 
  without 
  some 
  misgivings. 
  

  

  In 
  the 
  present 
  paper 
  I 
  shall 
  content 
  myself 
  with 
  giving 
  a 
  

   full 
  account 
  of 
  the 
  experimental 
  data 
  which 
  I 
  have 
  thus 
  far 
  

   obtained, 
  postponing 
  for 
  the 
  present 
  anything 
  like 
  a 
  definite 
  

   conclusion. 
  

  

  The 
  metallic 
  deposits 
  are 
  obtained 
  by 
  heating 
  small 
  frag- 
  

   ments 
  of 
  the 
  alkali 
  metals 
  in 
  glass 
  bulbs, 
  thoroughly 
  exhausted 
  

   and 
  hermetically 
  sealed. 
  Only 
  the 
  small 
  portion 
  of 
  the 
  bulb 
  

   on 
  which 
  the 
  mstal 
  particle 
  lies 
  * 
  is 
  heated, 
  leaving 
  the 
  

   remainder, 
  where 
  the 
  condensation 
  is 
  to 
  take 
  place, 
  quite 
  cold. 
  

   The 
  metallic 
  film 
  which 
  condenses 
  on 
  the 
  wall, 
  when 
  viewed 
  

   by 
  transmitted 
  light, 
  shows 
  colours 
  of 
  excessive 
  brilliancy, 
  as 
  

   brilliant 
  in 
  fact 
  as 
  films 
  strongly 
  stained 
  with 
  aniline 
  dyes. 
  

   On 
  first 
  obtaining 
  these 
  films, 
  I 
  was 
  of 
  the 
  opinion 
  that 
  these 
  

   colours 
  could 
  be 
  attributed 
  in 
  some 
  way 
  to 
  interference, 
  but 
  

   more 
  careful 
  consideration 
  showed 
  that 
  many 
  difficulties 
  were 
  

   present. 
  In 
  the 
  first 
  place, 
  the 
  transmitted 
  colours, 
  in 
  the 
  

   case 
  of 
  thin 
  plates, 
  are 
  never 
  of 
  any 
  considerable 
  intensity, 
  

   being 
  diluted 
  with 
  a 
  large 
  excess 
  of 
  white 
  light. 
  Moreover, 
  

   the 
  extraordinarily 
  high 
  absorption 
  -coefficient 
  of 
  metallic 
  

   sodium 
  makes 
  it 
  impossible 
  for 
  us 
  to 
  apply 
  the 
  theory 
  of 
  the 
  

   colours 
  of 
  thin 
  plates, 
  except 
  on 
  the 
  assumption 
  that 
  the 
  

   thickness 
  of 
  the 
  film 
  is 
  exceedingly 
  small 
  in 
  comparison 
  with 
  

   the 
  wave-length, 
  and 
  that 
  there 
  is 
  a 
  relative 
  change 
  of 
  phase 
  

   at 
  the 
  two 
  surfaces, 
  which 
  varies 
  with 
  the 
  wave-length. 
  This 
  

   seemed 
  on 
  the 
  whole 
  not 
  very 
  probable, 
  for, 
  as 
  far 
  as 
  experi- 
  

   ments 
  go, 
  indications 
  are 
  that 
  the 
  phase-change 
  in 
  metallic 
  

   reflexion 
  is 
  nearly 
  independent 
  of 
  the 
  wave-length. 
  Moreover, 
  

   metallic 
  films 
  obtained 
  in 
  other 
  ways, 
  as 
  by 
  chemical 
  or 
  

   cathodic 
  deposition, 
  do 
  not 
  show 
  these 
  colours. 
  Thin 
  films 
  of 
  

   metal 
  show 
  a 
  more 
  or 
  less 
  marked 
  colour 
  by 
  transmitted 
  light, 
  

   but 
  the 
  colour 
  is 
  fixed 
  for 
  any 
  definite 
  metal, 
  and, 
  except 
  in 
  

   the 
  case 
  of 
  gold 
  and 
  silver, 
  not 
  very 
  pronounced. 
  The 
  sodium 
  

   films, 
  on 
  the 
  contrary, 
  may 
  be 
  deep 
  purple, 
  blue, 
  apple-green, 
  

   or 
  red 
  according 
  to 
  conditions. 
  I 
  noticed 
  at 
  the 
  outset 
  that 
  

   where 
  the 
  deposit 
  was 
  somewhat 
  dense 
  in 
  the 
  immediate 
  

   vicinity 
  of 
  the 
  sodium 
  drop, 
  the 
  outside 
  of 
  the 
  bulb, 
  seen 
  by 
  

   reflected 
  light, 
  exhibited 
  halo 
  effects, 
  similar 
  to 
  those 
  shown 
  

   by 
  certain 
  minerals. 
  This 
  pearly 
  or 
  silky 
  lustre, 
  which 
  is 
  

  

  