﻿Chemistry 
  and 
  Physics. 
  71 
  

  

  (such 
  as 
  the 
  solar 
  spectrum) 
  and 
  when 
  it 
  is 
  kept, 
  of 
  course, 
  

   invariable 
  in 
  the 
  two 
  cases. 
  As 
  Lord 
  Bayleigh 
  has 
  shown, 
  the 
  

   coefficient 
  of 
  diffusion 
  varies 
  inversely 
  as 
  the 
  fourth 
  power 
  of 
  

   the 
  wave-length, 
  for 
  scattering 
  by 
  gas 
  molecules. 
  In 
  this 
  case, 
  

   the 
  shorter 
  wave-lengths 
  will 
  predominate 
  in 
  the 
  diffused 
  light. 
  

   For 
  relatively 
  large 
  particles 
  the 
  composition 
  of 
  the 
  scattered 
  

   light 
  will 
  depend 
  upon 
  the 
  specific 
  properties 
  of 
  the 
  material 
  of 
  

   the 
  particles. 
  In 
  general, 
  the 
  violet 
  radiations 
  will 
  be 
  absorbed 
  

   to 
  a 
  greater 
  degree, 
  and 
  hence 
  less 
  diffused, 
  than 
  the 
  red 
  rays. 
  

   In 
  this 
  case 
  spectrophotometric 
  observations 
  on 
  the 
  scattered 
  

   light 
  would 
  furnish 
  valuable 
  information 
  concerning 
  the 
  nature 
  

   and 
  identification 
  of 
  the 
  diffusing 
  matter. 
  Considerations 
  of 
  

   the 
  kind 
  just 
  outlined 
  enabled 
  Fabry 
  to 
  derive 
  suitable 
  formulae 
  

   and 
  to 
  obtain, 
  with 
  their 
  aid, 
  the 
  following 
  interesting 
  theoret- 
  

   ical 
  results. 
  

  

  Consider 
  a 
  cubic 
  meter 
  of 
  air, 
  under 
  standard 
  conditions 
  of 
  

   pressure 
  and 
  temperature, 
  when 
  illuminated 
  by 
  the 
  sun 
  on 
  a 
  

   clear 
  day 
  in 
  summer. 
  In 
  a 
  direction 
  at 
  right 
  angles 
  to 
  the 
  inci- 
  

   dent 
  beam 
  the 
  intensity 
  of 
  the 
  diffused 
  light 
  is 
  found 
  by 
  compu- 
  

   tation 
  to 
  be 
  0-062 
  "candle." 
  A 
  volume 
  of 
  16 
  cubic 
  meters 
  of 
  

   air 
  would 
  give 
  one 
  candle. 
  The 
  mean 
  spherical 
  intensity 
  is 
  

   about 
  0-08 
  candle 
  per 
  cubic 
  meter. 
  If 
  the 
  air 
  in 
  a 
  hall, 
  having 
  

   a 
  volume 
  of 
  1000 
  cubic 
  meters, 
  could 
  conserve 
  the 
  luminosity 
  

   that 
  it 
  has 
  in 
  full 
  sunlight 
  it 
  would 
  give 
  an 
  intensity 
  of 
  80 
  

   candles. 
  

  

  Comets 
  exhibit 
  two 
  spectra, 
  one 
  of 
  bright 
  lines, 
  and 
  another 
  

   of 
  the 
  continuous 
  type 
  which 
  is 
  supposed 
  to 
  be 
  scattered 
  sunlight. 
  

   Assuming 
  the 
  latter 
  to 
  be 
  due 
  to 
  gaseous 
  molecules 
  it 
  is 
  shown 
  

   that 
  the 
  density 
  of 
  the 
  tail 
  of 
  the 
  comet 
  would 
  be 
  of 
  the 
  order 
  

   10 
  -11 
  , 
  even 
  when 
  the 
  tail 
  has 
  a 
  thickness 
  as 
  small 
  as 
  ten 
  times 
  

   the 
  diameter 
  of 
  the 
  earth. 
  The 
  density 
  just 
  given 
  is 
  about 
  the 
  

   same 
  as 
  that 
  of 
  the 
  residual 
  gas 
  in 
  the 
  highest 
  vacua 
  obtained 
  by 
  

   experimental 
  processes. 
  

  

  If 
  the 
  light 
  of 
  the 
  nocturnal 
  sky 
  were 
  due 
  to 
  diffusion 
  of 
  sun- 
  

   light 
  by 
  hydrogen 
  gas 
  then 
  the 
  density 
  would 
  have 
  to 
  be 
  of 
  the 
  

   order 
  10~ 
  14 
  . 
  This 
  is 
  an 
  extremely 
  high 
  degree 
  of 
  rarefaction 
  for, 
  

   under 
  the 
  circumstances, 
  one 
  gram 
  of 
  hydrogen 
  would 
  fill 
  a 
  cube 
  

   measuring 
  ten 
  kilometers 
  along 
  an 
  edge. 
  The 
  mean 
  free 
  path 
  

   of 
  a 
  molecule 
  would 
  be 
  1600 
  kilometers, 
  and 
  a 
  cubic 
  millimeter 
  

   would 
  contain 
  only 
  300 
  molecules. 
  

  

  The 
  density 
  of 
  hydrogen 
  in 
  the 
  solar 
  corona 
  would 
  decrease 
  

   very 
  rapidly 
  from 
  6-5 
  X 
  10~ 
  9 
  near 
  the 
  surface 
  to 
  10 
  _1 
  ° 
  at 
  a 
  dis- 
  

   tance 
  from 
  the 
  surface 
  equal 
  to 
  the 
  sun's 
  radius. 
  Accordingly, 
  

   it 
  is 
  not 
  difficult 
  to 
  see 
  how 
  comets 
  may 
  pass 
  through 
  this 
  attenu- 
  

   ated 
  atmosphere 
  without 
  suffering 
  appreciable 
  perturbations. 
  

  

  The 
  last 
  section 
  of 
  the 
  paper 
  deals 
  with 
  the 
  influence 
  of 
  the 
  

   Doppler-Fizeau 
  effect 
  called 
  into 
  play 
  by 
  the 
  rapid 
  motion 
  of 
  

   gaseous 
  molecules. 
  It 
  is 
  shown 
  that 
  fine 
  dark 
  lines 
  would 
  be 
  

  

  