﻿524 
  Mr. 
  F. 
  Soddy 
  on 
  the 
  Production 
  of 
  

  

  the 
  tube 
  attached 
  to 
  the 
  limb 
  h 
  is 
  exhausted 
  and 
  sealed. 
  In 
  

   reopening 
  it, 
  it 
  is 
  advisable 
  to 
  break 
  the 
  tip 
  under 
  a 
  rubber 
  

   cap 
  and 
  to 
  admit 
  the 
  air 
  slowly, 
  although 
  each 
  mercury 
  tap, 
  

   before 
  being 
  used, 
  is 
  tested 
  under 
  working 
  conditions 
  by 
  a 
  

   sudden 
  break 
  to 
  ensure 
  that 
  it 
  is 
  correctly 
  dimensioned. 
  

   Recently 
  such 
  a 
  sudden 
  break 
  occurred 
  with 
  the 
  thorium 
  

   flask, 
  doing 
  no 
  harm. 
  

  

  Spectroscopic 
  Test 
  for 
  the 
  presence 
  of 
  Helium. 
  

  

  Before 
  discussing 
  the 
  results 
  so 
  far 
  obtained, 
  it 
  is 
  necessary 
  

   to 
  state 
  exactly 
  what 
  is 
  and 
  what 
  is 
  not 
  considered 
  sufficient 
  

   evidence 
  of 
  the 
  presence 
  of 
  these 
  infinitesimal 
  quantities 
  of 
  

   helium. 
  The 
  D 
  3 
  line 
  is 
  generally 
  the 
  only 
  line 
  visible, 
  and 
  

   fortunately 
  its 
  position 
  is 
  quite 
  easy 
  to 
  verify 
  with 
  accuracy 
  

   owing 
  to 
  its 
  proximity 
  to 
  the 
  sodium 
  lines. 
  The 
  D 
  3 
  line 
  becomes 
  

   quite 
  brilliant 
  under 
  all 
  circumstances 
  before 
  any 
  of 
  the 
  other 
  

   lines 
  in 
  the 
  helium 
  spectrum 
  appear. 
  Now 
  the 
  hydrogen 
  

   secondary 
  spectrum, 
  which 
  is 
  composed 
  of 
  so 
  many 
  lines 
  in 
  

   the 
  yellow 
  and 
  green 
  region 
  that 
  it 
  appears 
  more 
  or 
  less 
  con- 
  

   tinuous, 
  has 
  a 
  line 
  so 
  near 
  D 
  3 
  that 
  it 
  cannot 
  be 
  distinguished 
  

   by 
  visual 
  observation 
  even 
  with 
  a 
  good 
  spectroscope. 
  It 
  is 
  

   always 
  less 
  brilliant 
  than 
  other 
  lines 
  on 
  the 
  more 
  refrangible 
  

   side. 
  An 
  ordinary 
  hydrogen 
  tube, 
  filled 
  at 
  not 
  too 
  low 
  

   pressure, 
  shows 
  this 
  line; 
  and 
  it 
  is 
  useless 
  to 
  attempt 
  to 
  detect 
  

   the 
  minimum 
  amount 
  of 
  helium 
  in 
  presence 
  of 
  the 
  hydrogen 
  

   secondary 
  spectrum. 
  It 
  is 
  also 
  useless 
  when 
  argon 
  is 
  at 
  all 
  

   strong, 
  as 
  there 
  are 
  lines 
  in 
  the 
  spectrum 
  very 
  close 
  to 
  D 
  3 
  , 
  

   although 
  there 
  is 
  always 
  more 
  than 
  one, 
  relatively 
  weak, 
  in 
  

   the 
  small 
  tubes 
  used, 
  compared 
  to 
  the 
  other 
  lines. 
  In 
  the 
  

   cases 
  mentioned 
  the 
  test 
  would 
  be 
  rejected. 
  What 
  is 
  required 
  

   is 
  that 
  D 
  3 
  should 
  be 
  seen 
  on 
  a 
  black 
  background 
  with 
  no 
  

   lines 
  in 
  its 
  immediate 
  neighbourhood, 
  except 
  D 
  1 
  and 
  D 
  2 
  if 
  a 
  

   soda-glass 
  spectrum 
  is 
  used, 
  and 
  the 
  position 
  of 
  the 
  line 
  must 
  

   be 
  exactly 
  found 
  both 
  with 
  reference 
  to 
  the 
  sodium 
  line, 
  and 
  

   by 
  comparison 
  with 
  a 
  helium-tube. 
  

  

  A 
  considerable 
  quantity 
  of 
  helium, 
  say 
  10 
  or 
  20 
  times 
  the 
  

   minimum 
  amount, 
  can 
  be 
  detected 
  in 
  presence 
  of 
  a 
  strong 
  

   argon 
  spectrum, 
  as 
  D 
  3 
  under 
  these 
  circumstances 
  is 
  far 
  

   brighter 
  than 
  the 
  neighbouring 
  argon 
  lines. 
  Very 
  small 
  

   quantities 
  of 
  argon 
  do 
  not 
  interfere 
  greatly 
  with 
  the 
  obser- 
  

   vation 
  of 
  D 
  3 
  ; 
  for 
  under 
  these 
  circumstances 
  only 
  the 
  blue 
  

   or 
  spark 
  spectrum 
  of 
  argon 
  is 
  observed, 
  which 
  has 
  practically 
  

   no 
  lines 
  in 
  the 
  less 
  refrangible 
  part 
  of 
  the 
  spectrum. 
  

  

  With 
  the 
  infinitesimal 
  quantities 
  of 
  gas 
  dealt 
  with 
  in 
  

   these 
  experiments, 
  the 
  passage 
  of 
  even 
  a 
  weak 
  discharge 
  

   causes 
  the 
  spectrum-tube 
  to 
  become 
  non-conducting 
  in 
  a 
  few 
  

  

  