﻿210 
  Mr. 
  E. 
  Haigh 
  on 
  Orthobaric 
  Volumes 
  in 
  

  

  small 
  that 
  the 
  general 
  results 
  contained 
  in 
  Tables 
  I.-VI. 
  would 
  

   hardly 
  be 
  affected. 
  With 
  the 
  necessary 
  alteration 
  the 
  first 
  value 
  

   of 
  2 
  for 
  benzene, 
  Table 
  III., 
  becomes 
  15*857, 
  and 
  in 
  other 
  cases 
  

   the 
  correction 
  would 
  be 
  still 
  smaller.] 
  

  

  It 
  is 
  evident 
  from 
  these 
  tables 
  that 
  the 
  initial 
  proposition 
  

   of 
  this 
  article 
  is 
  substantially 
  true, 
  but 
  there 
  are, 
  in 
  some 
  

   cases, 
  outstanding 
  deviations 
  of 
  considerable 
  magnitude 
  which 
  

   remain 
  unaccounted 
  for. 
  It 
  is 
  difficult 
  to 
  say 
  in 
  what 
  pro- 
  

   portions 
  such 
  deviations 
  must 
  be 
  ascribed 
  to 
  the 
  inadequacy 
  

   of 
  the 
  equation, 
  to 
  experimental 
  error, 
  or 
  to 
  circumstances 
  

   of 
  the 
  problem 
  which 
  have 
  not 
  been 
  taken 
  into 
  account. 
  For 
  

   example, 
  no 
  allowance 
  has 
  been 
  made 
  for 
  any 
  degree 
  of 
  poly- 
  

   merization 
  whether 
  occurring 
  in 
  the 
  vapour 
  or 
  in 
  the 
  liquid 
  

   phase 
  of 
  the 
  substance. 
  

  

  (Stress 
  may 
  rightly 
  be 
  laid 
  upon 
  the 
  closeness 
  of 
  the 
  agree- 
  

   ment 
  in 
  the 
  case 
  of 
  isopentane, 
  for 
  if 
  the 
  dual 
  equation 
  were 
  

   radically 
  defective, 
  its 
  inadequacy 
  would 
  at 
  once 
  be 
  made 
  

   evident 
  by 
  employing 
  it 
  to 
  represent 
  the 
  properties 
  of 
  a 
  

   compound 
  for 
  which 
  the 
  experimental 
  investigation 
  has 
  been 
  

   of 
  the 
  most 
  exact 
  character. 
  The 
  contrary 
  is, 
  however, 
  tin- 
  

   case 
  and 
  in 
  some 
  instances, 
  where 
  the 
  agreement 
  is 
  not 
  

   so 
  satisfactory, 
  there 
  is 
  reason 
  to 
  believe 
  that 
  the 
  determi- 
  

   nation 
  of 
  some 
  of 
  the 
  magnitudes 
  involved 
  has 
  been 
  less 
  

   precise 
  *. 
  Stannic 
  chloride 
  gives 
  values 
  of 
  2 
  deviating 
  from 
  

   the 
  norm 
  to 
  a 
  much 
  greater 
  extent 
  than 
  those 
  of 
  carbon 
  

   tetrachloride 
  ; 
  but 
  reference 
  to 
  Young's 
  paper 
  on 
  the 
  former 
  

   compound 
  will 
  show 
  that 
  there 
  is 
  some 
  doubt 
  as 
  to 
  the 
  exact 
  

   value 
  of 
  the 
  critical 
  temperature, 
  and 
  consequently 
  of 
  the 
  

   critical 
  volume, 
  in 
  this 
  case. 
  

  

  Where 
  the 
  law 
  of 
  Cailletet 
  and 
  Mathias 
  does 
  not 
  strictly 
  

   hold 
  good 
  a 
  slight 
  uncertainty 
  attaches 
  to 
  the 
  value 
  of 
  the 
  

   critical 
  volume, 
  if 
  determined 
  by 
  this 
  method, 
  and 
  small 
  

   errors 
  in 
  the 
  value 
  of 
  this 
  constant 
  have 
  a 
  marked 
  influence 
  

   on 
  that 
  of 
  the 
  invariant 
  function 
  when 
  calculated 
  for 
  low 
  

   temperatures. 
  The 
  factor 
  3^r— 
  1, 
  or 
  t/t 
  — 
  A 
  (which 
  in 
  the 
  

   reduced 
  equation 
  is 
  the 
  analogue 
  of 
  van 
  der 
  Waals's 
  factor. 
  

   r 
  — 
  //), 
  is 
  very 
  sensitive 
  to 
  small 
  variations 
  in 
  the 
  values 
  of 
  its 
  

   terms 
  when 
  these 
  approach 
  equality. 
  Such 
  variations 
  may 
  

   be 
  duo 
  to 
  small 
  errors 
  in 
  the 
  measurement 
  of 
  the 
  volume 
  o£ 
  

   the 
  liquid, 
  in 
  the 
  determination 
  of 
  the 
  critical 
  volume, 
  or 
  may 
  

   be 
  caused 
  by 
  variations 
  in 
  the 
  value 
  of 
  "A." 
  In 
  van 
  der 
  

   Waals's 
  original 
  investigation 
  it 
  is 
  pointed 
  out 
  that 
  the 
  intro- 
  

   duction 
  of 
  the 
  factor 
  /- 
  — 
  A 
  is 
  subjecl 
  to 
  the 
  condition 
  that 
  u 
  l>" 
  

   is 
  relatively 
  small 
  compared 
  with 
  u 
  v 
  93 
  and 
  that 
  below 
  the 
  

   limit 
  r='2/> 
  the 
  formula 
  Bor 
  gases 
  cannot 
  be 
  expected 
  to 
  hold 
  

  

  good 
  owing 
  to 
  the 
  intersection 
  of 
  the 
  spheres 
  of 
  action 
  of 
  the 
  

   : 
  See 
  note 
  bi 
  1 
  >r. 
  S. 
  Young. 
  

  

  