﻿176 
  ANNUAL 
  REPORT 
  SMITHSONIAN 
  INSTITUTION, 
  1927 
  

  

  A 
  very 
  similar 
  method 
  may 
  be 
  used 
  to 
  determine 
  the 
  atomic 
  weight 
  

   of 
  the 
  atoms 
  of 
  which 
  the 
  stars 
  are 
  composed. 
  A 
  star 
  is 
  in 
  effect 
  

   nothing 
  but 
  a 
  huge 
  X-ray 
  apparatus. 
  We 
  know 
  the 
  total 
  mass 
  of 
  

   many 
  stars, 
  and 
  we 
  can 
  readily 
  calculate 
  the 
  rate 
  at 
  which 
  they 
  

   are 
  generating 
  X 
  rays 
  — 
  it 
  is 
  merely 
  the 
  rate 
  at 
  which 
  they 
  are 
  

   radiating 
  energy 
  away 
  into 
  space. 
  If 
  we 
  could 
  shut 
  our 
  Maxwell 
  

   demon 
  inside 
  a 
  star 
  and 
  make 
  him 
  cut 
  each 
  atomic 
  nucleus 
  in 
  half, 
  

   keeping 
  the 
  star's 
  mass 
  and 
  rate 
  of 
  generation 
  unaltered, 
  we 
  should 
  

   halve 
  the 
  coefficient 
  of 
  opacity 
  of 
  the 
  star. 
  This 
  would 
  necessitate 
  a 
  

   change 
  in 
  the 
  star's 
  build; 
  in 
  actual 
  fact 
  its 
  radius 
  would 
  increase 
  

   fourfold 
  while 
  its 
  surface 
  temperature 
  would 
  be 
  halved. 
  We 
  could 
  

   follow 
  the 
  progress 
  of 
  the 
  demon's 
  work 
  by 
  watching 
  the 
  changes 
  in 
  

   the 
  surface 
  temperature 
  of 
  the 
  star. 
  Hence 
  from 
  the 
  observed 
  surface 
  

   temperature 
  of 
  any 
  star 
  the 
  mass 
  and 
  luminosity 
  of 
  which 
  are 
  known, 
  

   it 
  must 
  be 
  possible 
  to 
  estimate 
  the 
  atomic 
  weight 
  of 
  the 
  atoms 
  of 
  

   which 
  the 
  star 
  is 
  composed. 
  The 
  formula 
  given 
  above 
  provides 
  

   the 
  means. 
  

  

  I 
  ought 
  perhaps 
  to 
  mention 
  in 
  passing 
  that 
  Eddington 
  and 
  others 
  

   have 
  approached 
  this 
  question 
  from 
  the 
  other 
  end, 
  assigning 
  con- 
  

   jectural 
  values 
  to 
  N'/A 
  from 
  our 
  knowledge 
  of 
  the 
  elements 
  which 
  

   occur 
  in 
  the 
  atmospheres 
  of 
  the 
  sun 
  and 
  stars. 
  Such 
  a 
  course 
  appears 
  

   to 
  be 
  very 
  risky. 
  A 
  star's 
  spectrum 
  gives 
  no 
  indication 
  of 
  the 
  selec- 
  

   tion 
  of 
  elements 
  which 
  occur 
  in 
  its 
  interior; 
  and 
  there 
  is 
  at 
  least 
  

   an 
  a 
  priori 
  possibility 
  that 
  the 
  elements 
  occurring 
  there 
  may 
  be 
  

   entirely 
  different 
  from 
  those 
  which 
  appear 
  in 
  its 
  surface; 
  consider 
  

   into 
  what 
  errors 
  an 
  extra-terrestrial 
  observer 
  might 
  be 
  led 
  if 
  he 
  

   assumed 
  that 
  the 
  earth 
  contained 
  no 
  chemical 
  elements 
  beyond 
  

   those 
  appearing 
  in 
  its 
  atmosphere. 
  

  

  When, 
  however, 
  the 
  risk 
  has 
  been 
  taken, 
  and 
  such 
  values 
  assigned 
  

   to 
  N-/A, 
  all 
  the 
  quantities 
  which 
  occur 
  in 
  the 
  luminosity 
  formula 
  

   are 
  known, 
  and 
  the 
  only 
  question 
  which 
  remains 
  is 
  whether 
  the 
  

   values 
  calculated 
  for 
  the 
  luminosity 
  agree 
  with 
  those 
  observed 
  

   through 
  the 
  telescope. 
  They 
  do 
  not. 
  

  

  It 
  is 
  clear 
  that 
  the 
  value 
  of 
  N-/A 
  must 
  be 
  adjusted 
  until 
  agree- 
  

   ment 
  is 
  obtained, 
  and 
  this 
  amounts 
  to 
  precisely 
  the 
  same 
  thing 
  as 
  

   determining 
  N-/A, 
  directly 
  and 
  at 
  once, 
  from 
  the 
  luminosity 
  formula. 
  

   On 
  doing 
  this 
  for 
  a 
  series 
  of 
  stars, 
  I 
  have 
  found 
  (Mon. 
  Not. 
  K. 
  A. 
  S., 
  

   June, 
  1926) 
  that 
  two 
  very 
  significant 
  facts 
  emerge. 
  First, 
  most 
  of 
  

   the 
  values 
  so 
  determined 
  prove 
  to 
  be 
  higher 
  than 
  the 
  value 
  for 
  

   uranium, 
  the 
  heaviest 
  element 
  known 
  on 
  earth. 
  Second, 
  the 
  different 
  

   values 
  of 
  N^/A 
  show 
  an 
  ordered 
  arrangement, 
  the 
  youngest 
  stars 
  

   generally 
  giving 
  the 
  highest 
  values 
  for 
  N''/A, 
  and 
  this 
  value 
  falling 
  

   as 
  we 
  pass 
  to 
  older 
  stars. 
  

  

  The 
  second 
  of 
  these 
  results 
  has 
  far-reaching 
  implications. 
  Con- 
  

   trary 
  to 
  the 
  views 
  of 
  the 
  early 
  spectroscopists, 
  and 
  contrary 
  to 
  what 
  

  

  