﻿relation 
  to 
  Pressure 
  and 
  Temperature. 
  209 
  

  

  Table 
  II. 
  — 
  The 
  maximum 
  deviation 
  for 
  isopentane 
  in 
  no 
  ease 
  

   exceeds 
  one-third 
  per 
  cent., 
  and 
  the 
  mean 
  deviation, 
  — 
  '014 
  in 
  16, 
  

   is 
  less 
  than 
  one 
  part 
  in 
  one 
  thousand. 
  These 
  magnitudes 
  are 
  not 
  

   far 
  removed 
  from 
  the 
  limits 
  of 
  experimental 
  error. 
  In 
  view 
  of 
  the 
  

   fact 
  that 
  this 
  substance 
  has 
  been 
  investigated 
  in 
  great 
  detail, 
  and 
  

   with 
  especial 
  care 
  to 
  make 
  due 
  allowance 
  for 
  all 
  known 
  sources 
  of 
  

   error, 
  it 
  is 
  highly 
  satisfactory 
  to 
  find 
  so 
  close 
  an 
  agreement 
  of 
  the 
  

   equation 
  with 
  experimental 
  results. 
  

  

  Table 
  III. 
  — 
  Benzene, 
  di-isopropyl, 
  and 
  hexameth}'lene 
  closely 
  

   resemble 
  each 
  other 
  in 
  the 
  magnitude 
  and 
  character 
  of 
  their 
  

   deviations, 
  which 
  are 
  small 
  and 
  mainly 
  positive. 
  In 
  four 
  other 
  

   cases, 
  viz., 
  n-pentane*, 
  ^-hexane, 
  di-isobutyl, 
  and 
  n- 
  octane, 
  the 
  

   deviations 
  are 
  negative 
  in 
  sign 
  and 
  numerically 
  much 
  greater, 
  being- 
  

   highest 
  at 
  the 
  lowest 
  temperature 
  and 
  decreasing 
  numerically 
  as 
  

   the 
  temperature 
  increases. 
  

  

  Table 
  IV. 
  — 
  Carbon 
  tetrachloride 
  closely 
  resembles 
  n-hexane 
  in 
  

   the 
  negative 
  sign 
  and 
  numerical 
  magnitude 
  of 
  its 
  deviations, 
  and 
  

   consequently 
  in 
  the 
  value 
  of 
  the 
  mean. 
  

  

  Stannic 
  chloride 
  f 
  similarly 
  resembles 
  n-octane, 
  the 
  deviations 
  

   being 
  numerically 
  much 
  greater. 
  

  

  Table 
  Y. 
  — 
  This 
  table 
  has 
  been 
  calculated 
  at 
  intervals 
  of 
  

   10° 
  C. 
  from 
  40° 
  C. 
  to 
  190° 
  C. 
  (0=-67O2 
  to 
  0=-9914). 
  As 
  a 
  dis- 
  

   tinctly 
  normal 
  substance 
  the 
  deviations 
  for 
  ether 
  are 
  greater 
  than 
  

   one 
  would 
  expect 
  to 
  find, 
  being 
  not 
  much 
  less 
  than 
  for 
  propyl 
  

   alcohol. 
  

  

  Leduc 
  (/Nouvelles 
  recherches 
  sur 
  les 
  Gaz,' 
  Paris, 
  1899) 
  has 
  

   similarly 
  found 
  that 
  his 
  calculated 
  results 
  for 
  the 
  isothermals 
  of 
  

   ether 
  are 
  in 
  defect. 
  He 
  attributes 
  the 
  deficiency 
  to 
  a 
  possible 
  

   experimental 
  error. 
  [" 
  II 
  faut 
  remarquer 
  que 
  le 
  volume 
  occupe 
  par 
  

   la 
  vapeur 
  dans 
  le 
  piezometre 
  n'est 
  que 
  d'un 
  centimetre 
  cube 
  environ. 
  

   L'ecart 
  .... 
  doit 
  etre 
  attribu.ee 
  sans 
  hesitation 
  a 
  une 
  erreur 
  de 
  

   0,0 
  m 
  s 
  r 
  sur 
  la 
  pesee, 
  ce 
  que 
  correspond 
  a 
  5 
  mmc 
  ou 
  6 
  mmc 
  sur 
  la 
  volume 
  

   observce."] 
  

  

  Table 
  YI. 
  — 
  Results 
  for 
  two 
  alcohols 
  have 
  been 
  added 
  as 
  examples 
  

   of 
  substances 
  for 
  which 
  there 
  is 
  evidence 
  of 
  marked 
  polymerization. 
  

   It 
  will 
  be 
  seen 
  that 
  even 
  in 
  the 
  case 
  of 
  a 
  highly 
  polymerized 
  sub- 
  

   stance 
  like 
  methyl 
  alcohol 
  the 
  equation 
  is 
  still 
  approximately 
  true. 
  

  

  Table 
  YII. 
  — 
  Many 
  of 
  the 
  numerical 
  calculations 
  required 
  for 
  

   this 
  investigation 
  were 
  made 
  at 
  a 
  relatively 
  early 
  date, 
  and 
  this 
  

   table 
  gives 
  only 
  the 
  critical 
  constants 
  actually 
  employed 
  in 
  

   obtaining 
  " 
  reduced 
  " 
  values 
  of 
  the 
  variables. 
  Some 
  of 
  the 
  critical 
  

   volumes 
  have 
  since 
  been 
  slightly 
  altered 
  by 
  later 
  determinations 
  of 
  

   the 
  critical 
  densities. 
  [See 
  Young, 
  " 
  jNote 
  on 
  the 
  law 
  of 
  Cailletet 
  

   and 
  Mathias," 
  Phil. 
  Mag. 
  vol. 
  1. 
  p. 
  291 
  (1900). 
  The 
  differences 
  

   between 
  "old" 
  and 
  "corrected" 
  critical 
  densities 
  usually 
  are 
  so 
  

  

  * 
  See 
  note 
  by 
  Dr. 
  S. 
  Young, 
  

   t 
  See 
  note 
  bv 
  Dr. 
  S. 
  Young". 
  

  

  