﻿ARCHITECTURE 
  OF 
  ATOMS 
  — 
  ABBOT. 
  163 
  

  

  The 
  great 
  question 
  which 
  this 
  table 
  raises 
  is 
  this 
  : 
  If 
  a 
  star 
  actually 
  

   passes 
  through 
  the 
  states 
  enumerated, 
  why 
  do 
  the 
  spectrum 
  lines 
  of 
  

   the 
  metals 
  disappear 
  as 
  the 
  star 
  approaches 
  its 
  culmination 
  as 
  a 
  

   blue-white 
  giant 
  and 
  then 
  reappear 
  at 
  a 
  later 
  stage 
  as 
  a 
  yellowish 
  

   dwarf? 
  Can 
  we 
  believe 
  that 
  the 
  mere 
  exaltation 
  of 
  temperature 
  

   destroys 
  the 
  metals, 
  resolving 
  them 
  into 
  the 
  simpler 
  atoms 
  of 
  hydro- 
  

   gen 
  and 
  helium? 
  Such 
  a 
  view 
  at 
  first 
  sight 
  seems 
  reasonable, 
  in 
  

   view 
  of 
  the 
  now 
  well-known 
  fact 
  that 
  some 
  of 
  the 
  heavy 
  metals 
  — 
  

   uranium, 
  thorium, 
  and 
  radium 
  — 
  continually 
  break 
  down 
  sponta- 
  

   neously 
  with 
  formation 
  of 
  helium, 
  and 
  that 
  Kutherford 
  and 
  Chad- 
  

   wick 
  have 
  indeed 
  disintegrated 
  the 
  lighter 
  elements 
  — 
  nitrogen, 
  boron, 
  

   fluorine, 
  sodium, 
  and 
  phosphorus 
  — 
  with 
  expulsion 
  of 
  hydrogen. 
  

   But 
  the 
  proportion 
  of 
  the 
  atoms 
  affected 
  was 
  only 
  trifling 
  and 
  the 
  

   means 
  employed 
  for 
  knocking 
  out 
  the 
  protons 
  from 
  the 
  atomic 
  

   nuclei 
  were 
  the 
  alpha 
  rays 
  of 
  radium, 
  means 
  as 
  powerful 
  as 
  would 
  

   be 
  the 
  maintenance 
  of 
  a 
  temperature 
  of 
  millions 
  of 
  degrees, 
  perhaps 
  

   a 
  thousand 
  times 
  as 
  powerful 
  as 
  the 
  effects 
  of 
  the 
  actual 
  temperatures 
  

   at 
  the 
  surface 
  of 
  the 
  blue 
  stars. 
  Accordingly, 
  we 
  must 
  regretfully 
  give 
  

   up 
  this 
  simple 
  temperature 
  explanation 
  of 
  the 
  disappearance 
  of 
  the 
  

   metals 
  from 
  the 
  spectra 
  of 
  the 
  hotter 
  stars. 
  It 
  demands 
  temperatures 
  

   higher 
  than 
  the 
  stars 
  afford, 
  for 
  measurements 
  already 
  made 
  limit 
  

   us 
  there 
  to 
  surface 
  temperatures 
  not 
  exceeding 
  20,000° 
  C. 
  

  

  What 
  appears 
  to 
  be 
  the 
  main 
  factor 
  in 
  the 
  true 
  explanation 
  is 
  

   simpler 
  still 
  in 
  essence, 
  though 
  involving 
  such 
  intricate 
  and 
  obscure 
  

   properties 
  of 
  the 
  structure 
  of 
  the 
  atoms 
  as 
  not 
  yet 
  to 
  be 
  fully 
  worked 
  

   out 
  by 
  physicists, 
  at 
  least 
  in 
  the 
  cases 
  of 
  any 
  but 
  the 
  simplest 
  of 
  

   atoms. 
  In 
  short, 
  rise 
  of 
  temperature, 
  or, 
  speaking 
  more 
  broadly, 
  

   increased 
  intensity 
  of 
  excitation, 
  alters 
  the 
  atoms 
  by 
  setting 
  free 
  

   electrons 
  in 
  such 
  a 
  way 
  that 
  the 
  spectrum 
  lines 
  given 
  out 
  shift 
  over 
  

   from 
  the 
  visible 
  and 
  photographic 
  regions 
  of 
  the 
  spectrum 
  to 
  those 
  

   of 
  shorter 
  wave 
  length 
  beyond 
  the 
  violet, 
  where 
  the 
  earth's 
  atmos- 
  

   phere 
  is 
  not 
  transparent 
  to 
  them. 
  Thus 
  the 
  metallic 
  spectra 
  are 
  still 
  

   present 
  even 
  in 
  the 
  hottest 
  stars, 
  but 
  are 
  merely 
  removed 
  from 
  the 
  

   familiar 
  spectrum 
  region, 
  where 
  they 
  can 
  be 
  observed, 
  to 
  another 
  

   where 
  they 
  can 
  not 
  penetrate 
  our 
  atmosphere. 
  1 
  

  

  This 
  behavior 
  of 
  the 
  spectrum 
  lines 
  of 
  the 
  elements 
  is 
  illustrated 
  

   by 
  laboratory 
  work 
  and 
  explained 
  by 
  theoretical 
  researches 
  of 
  Bohr, 
  

   Saha, 
  and 
  others 
  on 
  the 
  relations 
  of 
  atomic 
  structure 
  to 
  spectra. 
  

   Laboratory 
  spectra 
  are 
  of 
  three 
  principal 
  kinds, 
  called 
  flame, 
  arc, 
  

   and 
  spark 
  spectra, 
  according 
  as 
  they 
  are 
  excited 
  by 
  sources 
  like 
  low 
  

   temperature 
  flames, 
  the 
  higher 
  excitation 
  of 
  the 
  electric 
  arc, 
  or 
  the 
  

   still 
  more 
  intense 
  action 
  of 
  the 
  electric 
  spark 
  forced 
  by 
  high 
  elec- 
  

   trical 
  potential 
  differences 
  and 
  often 
  reenforced 
  by 
  condenser 
  dis- 
  

  

  1 
  There 
  is 
  another 
  factor 
  which 
  we 
  shall 
  touch 
  upon 
  later. 
  

  

  