﻿454 
  F. 
  A. 
  Gooch 
  and 
  T. 
  S. 
  Hart 
  — 
  Detection 
  and 
  

  

  retrace 
  a 
  step 
  without 
  trouble 
  in 
  case 
  a 
  portion 
  of 
  the 
  solution 
  

   has 
  been 
  unwittingly 
  diluted 
  too 
  much 
  ; 
  such 
  a 
  mode 
  of 
  pro- 
  

   ceeding 
  is, 
  of 
  course, 
  necessary 
  when 
  the 
  final 
  dilution 
  must 
  

   exceed 
  100 
  cm 
  3 
  , 
  unless 
  large 
  graduates 
  are 
  called 
  into 
  use. 
  

   Excepting 
  the 
  cases 
  of 
  very 
  concentrated 
  solutions, 
  no 
  signifi- 
  

   cant 
  loss 
  of 
  material 
  is 
  occasioned 
  by 
  the 
  filling 
  of 
  the 
  coils, 
  

   the 
  error 
  thus 
  introduced 
  being 
  trivial 
  in 
  comparison 
  with 
  that 
  

   inherent 
  in 
  all 
  photometric 
  processes. 
  The 
  following 
  is 
  the 
  

   record 
  of 
  our 
  experience 
  in 
  the 
  comparison 
  of 
  solutions 
  of 
  

   pure 
  potassium 
  chloride, 
  the 
  strength 
  of 
  the 
  test 
  solution 
  being 
  

   unknown 
  to 
  the 
  observer. 
  

  

  Experiment 
  T. 
  

  

  Experiment 
  II. 
  

  

  

  Characteristic 
  of 
  

  

  

  Characteristic 
  of 
  

  

  Volume 
  of 
  

  

  line 
  compared 
  

  

  Volume 
  of 
  

  

  line 
  compared 
  

  

  test-solution. 
  

  

  with 
  standard. 
  

  

  test-solation. 
  

  

  with 
  standard. 
  

  

  20 
  cm 
  3 
  

  

  stronger 
  

  

  30 
  cm 
  3 
  

  

  stronger 
  

  

  50 
  " 
  

  

  stronger 
  

  

  60 
  " 
  

  

  stronger 
  

  

  100 
  " 
  

  

  stronger 
  

  

  82 
  " 
  

  

  weaker 
  

  

  110 
  " 
  

  

  stronger 
  

  

  70 
  " 
  

  

  stronger 
  

  

  120 
  " 
  

  

  stronger 
  

  

  76 
  " 
  

  

  stronger 
  

  

  150 
  " 
  

  

  like 
  

  

  78 
  " 
  

  

  stronger 
  

  

  200 
  " 
  

  

  weaker 
  

  

  80 
  " 
  

  

  like 
  

  

  160 
  " 
  

  

  weaker 
  

  

  

  

  150 
  " 
  

  

  like 
  

  

  

  

  (150 
  X 
  0*0001 
  = 
  0-0150) 
  

  

  Potassium 
  found 
  0-0150 
  grin. 
  

  

  taken 
  0-0150 
  " 
  

  

  .0-0120 
  " 
  

   0-0160 
  " 
  

   Error 
  .0*0000 
  " 
  

  

  Limits 
  on 
  either 
  side 
  

  

  (80 
  X 
  0-0001 
  = 
  0-0080) 
  

  

  Potassium 
  found 
  0*0080 
  

  

  " 
  taken 
  .0*0080 
  

  

  J 
  0-0078 
  

   ' 
  \ 
  0-0082 
  

   Error 
  0-0000 
  

  

  Limits 
  on 
  either 
  side 
  - 
  

  

  These 
  results 
  show 
  a 
  degree 
  of 
  accuracy 
  in 
  the 
  process 
  quite 
  

   unexpected. 
  In 
  the 
  former 
  no 
  attempt 
  was 
  made 
  to 
  approxi- 
  

   mate 
  as 
  closely 
  as 
  possible 
  to 
  the 
  limits 
  of 
  dilution 
  on 
  both 
  

   sides 
  of 
  the 
  condition 
  of 
  equal 
  brightness 
  in 
  test 
  and 
  standard, 
  

   but 
  in 
  the 
  latter 
  great 
  care 
  w 
  T 
  as 
  taken 
  in 
  this 
  respect 
  and 
  the 
  

   possible 
  error 
  cannot 
  exceed 
  two 
  and 
  a 
  half 
  per 
  cent 
  of 
  the 
  

   entire 
  amount 
  of 
  potassium 
  involved. 
  

  

  Experiment 
  III 
  was 
  made 
  to 
  discover 
  the 
  effect 
  of 
  the 
  

   presence 
  of 
  a 
  reasonable 
  amount 
  of 
  sodium 
  chloride 
  upon 
  the 
  

   determination 
  of 
  the 
  potassium. 
  To 
  a 
  portion 
  of 
  the 
  solution 
  

   if 
  pure 
  potassium 
  chloride 
  containing 
  0*01 
  grm. 
  of 
  the 
  element 
  

   was 
  added 
  0-1 
  grm. 
  of 
  sodium 
  chloride 
  taken 
  from 
  the 
  salt 
  

   purified 
  as 
  previously 
  described. 
  This 
  solution 
  was 
  diluted 
  

   and 
  the 
  comparison 
  made 
  with 
  the 
  standard 
  according 
  to 
  the 
  

   accompanying 
  account. 
  

  

  