﻿the 
  Absolute 
  Scale 
  of 
  Temperature. 
  

  

  371 
  

  

  sures 
  as 
  we 
  are 
  here 
  considering), 
  but 
  to 
  vary 
  considerably 
  

   with 
  the 
  temperature. 
  Thus 
  at 
  0° 
  

  

  0-1870 
  caL, 
  or 
  7,839,000 
  ergs 
  (Regnault) 
  I 
  ? 
  994 
  Q0Q 
  

  

  0-1948 
  „ 
  8,149,000 
  „ 
  (Wiedemann) 
  / 
  mean 
  ' 
  jJy4 
  ' 
  UUU 
  > 
  

  

  and 
  at 
  100° 
  

  

  0-2145 
  cal.,or 
  8,991,000 
  ergs 
  (Regnault) 
  | 
  q 
  Q48 
  00Q 
  

  

  0-2177 
  „ 
  9,106,000 
  „ 
  (Wiedemann) 
  j 
  mean 
  y,U40 
  ' 
  UUU 
  ' 
  

  

  the 
  average 
  over 
  the 
  range 
  0° 
  to 
  100° 
  being 
  8,520.000 
  ergs. 
  

  

  Specific 
  Volume. 
  

  

  

  Density 
  at 
  0° 
  

  

  Specific 
  volume 
  

  

  

  and 
  760 
  mm. 
  

  

  at 
  0° 
  and 
  760 
  mm. 
  

  

  at 
  0° 
  and 
  1000 
  mm. 
  

  

  Hydrogen 
  

  

  0-0000 
  8987 
  

   00012 
  93 
  

   00012 
  57 
  

   0-0019 
  65 
  

  

  11 
  1270 
  

   773-3 
  

   795-5 
  

   508-9 
  

  

  8457-5 
  

   587-5 
  

   604-7 
  

   385-9 
  

  

  Air 
  

  

  Carbon 
  dioxide 
  ... 
  

  

  i 
  

  

  Cooling 
  on 
  Free 
  Expansion. 
  — 
  This 
  quantity 
  also 
  enters 
  

   only 
  in 
  the 
  correction 
  term, 
  but 
  it 
  has 
  not 
  been 
  determined 
  

   with 
  sufficient 
  accuracy 
  even 
  for 
  that. 
  If 
  we 
  take 
  the 
  pro- 
  

   bable 
  error 
  of 
  /3 
  as 
  measured 
  by 
  Chappuis 
  at 
  2 
  x 
  10 
  -7 
  (or 
  

   1/18000 
  part 
  of 
  its 
  whole 
  value) 
  — 
  an 
  opinion 
  which 
  seems 
  

   justified 
  by 
  the 
  close 
  agreement 
  of 
  his 
  observations 
  among 
  

   themselves 
  — 
  then 
  the 
  cooling 
  e 
  resulting 
  from 
  a 
  fall 
  of 
  1 
  

   atmosphere 
  in 
  the 
  pressure 
  should 
  be 
  measured 
  to 
  about 
  

   0°*004 
  to 
  influence 
  the 
  result 
  equally. 
  This 
  is 
  about 
  1/50 
  of 
  

   the 
  value 
  of 
  e 
  for 
  air 
  ; 
  that 
  degree 
  of 
  accuracy 
  has 
  certainly 
  

   not 
  been 
  attained, 
  even 
  for 
  air, 
  much 
  less 
  for 
  the 
  other 
  gases 
  

   which 
  Joule 
  and 
  Thomson 
  studied. 
  

  

  Hydrogen. 
  — 
  With 
  this 
  gas 
  twelve 
  experiments 
  were 
  made 
  

   at 
  about 
  7° 
  and 
  five 
  at 
  about 
  90°. 
  The 
  mean 
  of 
  the 
  twelve 
  

   at 
  the 
  low 
  temperature 
  gave 
  e 
  = 
  o 
  *301 
  per 
  100 
  inches 
  of 
  

   mercury 
  pressure 
  for 
  pure 
  hydrogen 
  ; 
  but 
  in 
  only 
  five 
  cases 
  

   did 
  the 
  analysis 
  of 
  the 
  gas 
  used 
  show 
  as 
  much 
  as 
  90 
  per 
  cent, 
  

   of 
  hydrogen 
  — 
  the 
  rest 
  being 
  air 
  — 
  and 
  these 
  five 
  experiments 
  

   o-ave 
  e=0°* 
  108, 
  while 
  the 
  one 
  case 
  in 
  which 
  the 
  purity 
  rose 
  

   to 
  98*2 
  per 
  cent, 
  gave 
  e=0°-075 
  for 
  the 
  pure 
  gas. 
  It 
  is 
  

  

  