﻿On 
  the 
  Velocity 
  of 
  Light. 
  331 
  

  

  of 
  uncertainty 
  of 
  the 
  velocity 
  of 
  light 
  concluded 
  above 
  

   (299,920 
  kilometres) 
  may 
  be 
  fairly 
  taken 
  at 
  70 
  kilometres. 
  

  

  " 
  (2) 
  Uncertainty 
  in 
  the 
  coefficient 
  of 
  the 
  light 
  equation 
  

   from 
  observations 
  of 
  the 
  satellites 
  of 
  Jupiter. 
  . 
  . 
  . 
  The 
  amount 
  

   of 
  uncertainty 
  is 
  probably 
  not 
  far 
  from 
  one 
  second 
  of 
  time*. 
  

  

  " 
  (3) 
  Uncertainty 
  in 
  the 
  constant 
  of 
  sidereal 
  aberration. 
  

   I 
  conceive 
  that 
  a 
  variation 
  of 
  0*025 
  in 
  this 
  well-determined 
  

   constant 
  t 
  will 
  not 
  be 
  regarded 
  far 
  from 
  the 
  limit 
  of 
  un- 
  

   certainty. 
  . 
  . 
  . 
  

  

  " 
  (4) 
  Uncertainty 
  in 
  the 
  relation 
  of 
  the 
  absolute 
  terres- 
  

   trial 
  velocity 
  to 
  the 
  velocity 
  in 
  space. 
  . 
  . 
  . 
  The 
  impossibility 
  

   oi 
  an 
  experimental 
  determination 
  of 
  this 
  relation 
  renders 
  the 
  

   assumption 
  of 
  identity 
  necessary. 
  

  

  " 
  In 
  conclusion, 
  then, 
  all 
  the 
  experimental 
  determinations 
  

   of 
  the 
  velocity 
  of 
  light 
  hitherto 
  made 
  give, 
  when 
  combined 
  

   with 
  astronomical 
  constants, 
  the 
  mean 
  equatorial 
  horizontal 
  

   parallax 
  of 
  the 
  sun, 
  

  

  8"-808 
  + 
  0-006. 
  

  

  The 
  corresponding 
  mean 
  radius 
  of 
  the 
  terrestrial 
  orbit 
  is 
  

   149,345,000 
  kilometres." 
  

  

  Roughly 
  speaking, 
  it 
  appears 
  that 
  the 
  velocity-of-light 
  

   method 
  of 
  obtaining 
  the 
  solar 
  parallax 
  is 
  about 
  ten 
  times 
  as 
  

   accurate 
  as 
  the 
  astronomical 
  method. 
  

  

  It 
  also 
  appears 
  that 
  in 
  the 
  last-named 
  method 
  the 
  order 
  of 
  

   accuracy 
  in 
  the 
  determination 
  of 
  the 
  astronomical 
  factors 
  is 
  

   about 
  one 
  one-thousandth 
  part. 
  

  

  The 
  undulatory 
  theory 
  does 
  not 
  specify 
  the 
  nature 
  of 
  the 
  

   undulations, 
  and 
  it 
  is 
  no 
  argument 
  against 
  it 
  that 
  the 
  special 
  

   mode 
  of 
  motion 
  assumed 
  by 
  Fresnel 
  (vibrations 
  of 
  an 
  elastic 
  

   solid, 
  which 
  so 
  elegantly 
  explains 
  all 
  the 
  known 
  phenomena 
  

   of 
  light) 
  has 
  met 
  with 
  some 
  serious 
  objections. 
  If, 
  according 
  

   to 
  Maxwell, 
  these 
  undulations 
  are 
  electromagnetic 
  changes, 
  

   these 
  objections 
  no 
  longer' 
  apply, 
  and 
  this 
  electromagnetic 
  

   theory 
  of 
  light 
  is 
  now 
  universally 
  accepted 
  J. 
  

  

  In 
  accordance 
  with 
  Maxwell's 
  theory, 
  the 
  ratio 
  of 
  any 
  

   electric 
  or 
  magnetic 
  unit 
  measured 
  electrostatically 
  to 
  the 
  

  

  * 
  This 
  is 
  taken 
  from 
  the 
  observations 
  of 
  Delambre, 
  493'2, 
  weight 
  1, 
  

   and 
  of 
  Glasenapp, 
  500-84+1-02, 
  weight 
  2. 
  - 
  : 
  

  

  + 
  Professor 
  Todd 
  accepts 
  Struve's 
  value 
  2" 
  - 
  044o. 
  

  

  X 
  While 
  the 
  electromagnetic 
  theory 
  gives 
  a 
  satisfactory 
  explanation 
  

   •of 
  light 
  and 
  goes 
  a 
  long 
  way 
  toward 
  the 
  explanation 
  of 
  the 
  mechanism 
  

   of 
  radiation, 
  yet 
  it 
  seems 
  not 
  at 
  all 
  unlikely 
  that 
  a 
  return 
  to 
  something 
  

   analogous 
  to 
  the 
  now 
  discarded 
  elastic 
  solid 
  theory 
  may 
  be 
  necessary 
  to 
  

   •explain 
  electromagnetic 
  phenomena. 
  

  

  