﻿520 
  Mr. 
  F. 
  Soddy 
  on 
  the 
  Production 
  of 
  

  

  on 
  account 
  of 
  the 
  presence 
  of 
  argon, 
  from 
  air 
  in 
  the 
  solutions. 
  

   The 
  last 
  test 
  with 
  thorium 
  was 
  completely 
  successful, 
  but 
  it 
  

   was 
  in 
  the 
  nature 
  of 
  a 
  blank 
  experiment, 
  the 
  period 
  of 
  

   ^^Ticcumulation 
  of 
  gas_ 
  in 
  the 
  thorium 
  being 
  only 
  about 
  

   22*5 
  hours. 
  No 
  sign 
  of 
  helium 
  was 
  seen. 
  The 
  minimum 
  

   quantity 
  detectable 
  under 
  the 
  circumstances 
  — 
  about 
  2 
  X 
  10 
  ~ 
  9 
  

   gram 
  — 
  would 
  have 
  been 
  observed, 
  as 
  the 
  experiment 
  was 
  

   a 
  very 
  perfect 
  one. 
  

  

  In 
  the 
  second 
  series 
  of 
  experiments, 
  vertical 
  all-glass 
  

   reflux 
  condensers 
  were 
  sealed 
  to 
  the 
  flasks, 
  and 
  the 
  principle 
  

   of 
  the 
  mercury 
  seal 
  was 
  pushed 
  to 
  its 
  logical 
  conclusion 
  by 
  

   making 
  the 
  two 
  upper 
  limbs 
  of 
  the 
  Y-tube 
  the 
  barometric 
  

   height, 
  so 
  that 
  there 
  could 
  be 
  a 
  vacuum 
  on 
  the 
  one 
  side 
  

   and 
  the 
  pressure 
  of 
  an 
  atmosphere 
  on 
  the 
  other. 
  The 
  

   principle 
  was 
  correct, 
  and 
  is 
  being 
  used 
  in 
  the 
  experiments 
  

   now 
  in 
  progress 
  ; 
  but 
  the 
  design 
  was 
  faulty, 
  and 
  the 
  arrange- 
  

   ment 
  need 
  not 
  therefore 
  be 
  described 
  in 
  more 
  detail. 
  Four 
  

   flasks 
  were 
  set 
  up 
  in 
  this 
  series, 
  but 
  only 
  two, 
  with 
  the 
  

   uranium 
  and 
  thorium 
  preparations 
  before 
  employed, 
  are 
  

   concerned 
  in 
  this 
  investigation. 
  The 
  first 
  and 
  the 
  only 
  test 
  

   with 
  the 
  thorium 
  solution 
  failed 
  on 
  account 
  of 
  argon 
  being- 
  

   present. 
  Carbon 
  dioxide 
  was 
  noticed 
  in 
  the 
  gas 
  as 
  it 
  was 
  

   admitted 
  to 
  the 
  furnace, 
  and, 
  after 
  the 
  absorption, 
  the 
  

   spectrum 
  was 
  that 
  of 
  pure 
  argon. 
  The 
  first 
  two 
  tests 
  with 
  

   uranium 
  failed 
  also 
  from 
  the 
  same 
  cause, 
  but 
  the 
  last 
  was 
  a 
  

   perfect 
  experiment. 
  The 
  solution 
  had 
  been 
  left 
  to 
  accumulate 
  

   for 
  six 
  days 
  and 
  four 
  hours, 
  and 
  D 
  3 
  was 
  present 
  faint 
  and 
  

   clear 
  on 
  the 
  spark-spectrum 
  of 
  argon. 
  The 
  spectrum-tube 
  

   was 
  of 
  soda- 
  glass 
  with 
  a 
  volume 
  of 
  0*3 
  c.c, 
  a 
  kind 
  that 
  has 
  

   been 
  subsequently 
  abandoned. 
  With 
  it, 
  a 
  quantity 
  of 
  about 
  

   2 
  x 
  10~ 
  9 
  gram 
  helium 
  could 
  be 
  detected. 
  

  

  During 
  the 
  experiment, 
  however, 
  a 
  violent 
  bumping 
  of 
  the 
  

   uranium 
  solution 
  during 
  boiling 
  projected 
  some 
  of 
  it 
  over 
  

   on 
  to 
  the 
  top 
  of 
  the 
  mercury 
  in 
  the 
  tap, 
  and 
  the 
  strongly 
  

   acid 
  liquid 
  at 
  once 
  began 
  to 
  attack 
  the 
  mercury 
  with 
  the 
  

   evolution 
  of 
  oxides 
  of 
  nitrogen. 
  These, 
  of 
  course, 
  did 
  no 
  

   harm, 
  as 
  the 
  calcium 
  absorbs 
  them, 
  but 
  the 
  flask 
  had 
  to 
  be 
  

   cut 
  down. 
  In 
  this 
  operation, 
  owing 
  to 
  faulty 
  design 
  of 
  the 
  

   mercury 
  taps, 
  the 
  whole 
  of 
  the 
  flasks 
  were 
  spoiled, 
  and 
  this 
  

   series 
  of 
  experiments 
  came 
  to 
  an 
  end. 
  With 
  regard 
  to 
  the 
  

   single 
  result, 
  it 
  must 
  be 
  borne 
  in 
  mind 
  that 
  the 
  experiment 
  

   was 
  performed 
  three 
  years 
  ago, 
  before 
  the 
  possible 
  sources 
  

   of 
  error 
  were 
  as 
  well 
  known 
  as 
  now, 
  and 
  no 
  significance 
  is 
  to 
  

   be 
  attached 
  to 
  it. 
  In 
  fact, 
  more 
  recent 
  results 
  have 
  failed 
  to 
  

   confirm 
  it. 
  

  

  In 
  the 
  first 
  experiment 
  of 
  the 
  third 
  series 
  with 
  thorium 
  now 
  

  

  