﻿the 
  Absolute 
  Scale 
  of 
  Temperature. 
  369 
  

  

  modynamic 
  scale 
  both 
  take 
  Regnault's 
  0*003665 
  as 
  the 
  co- 
  

   efficient 
  of 
  pressure 
  at 
  the 
  ordinary 
  density 
  ; 
  it 
  seems 
  to 
  me 
  

   there 
  can 
  be 
  little 
  doubt 
  that 
  is 
  too 
  low. 
  Another 
  series 
  of 
  

   observations 
  by 
  Regnault 
  himself 
  gives 
  0*0036679, 
  while 
  

   Magnus 
  finds 
  0*0036678 
  (reduced 
  to 
  the 
  boiling-point 
  under 
  

   76 
  cm. 
  pressure), 
  and 
  Jolly 
  00036695, 
  all 
  higher 
  than 
  the 
  

   first 
  quoted 
  number. 
  Further, 
  Mendeleef 
  *, 
  on 
  recalculating 
  

   these 
  three 
  observations 
  and 
  applying 
  certain 
  corrections 
  

   neglected 
  by 
  the 
  experimenters, 
  concluded 
  that 
  they 
  show 
  a 
  

   much 
  closer 
  agreement 
  than 
  when 
  uncorrected, 
  thus 
  : 
  — 
  

  

  Regnault 
  0*0036694 
  

  

  Magnus 
  700 
  

  

  Jolly 
  702 
  

  

  Mendeleef 
  s 
  reduction 
  is 
  regarded 
  probably 
  with 
  justice 
  as 
  

   the 
  most 
  exact, 
  and 
  we 
  will 
  take, 
  therefore, 
  the 
  mean 
  result 
  

   as 
  being 
  true 
  for 
  the 
  normal 
  density 
  (corresponding 
  to 
  a 
  

   pressure 
  of 
  1 
  atmo 
  at 
  0°). 
  

  

  The 
  variation 
  with 
  the 
  density 
  of 
  the 
  air 
  was 
  taken 
  by 
  

   Jochmann 
  as 
  0*0000115, 
  the 
  unit 
  of 
  density 
  being, 
  as 
  before, 
  

   the 
  density 
  at 
  normal 
  temperature 
  and 
  pressure. 
  This 
  re- 
  

   presents 
  the 
  set 
  of 
  measurements 
  taken 
  by 
  Regnault 
  for 
  the 
  

   express 
  purpose 
  ; 
  it 
  also 
  agrees 
  moderately 
  well 
  with 
  some 
  

   observations 
  by 
  Melander, 
  quoted 
  in 
  Landolt 
  and 
  Bornstein's 
  

   tables, 
  the 
  only 
  others 
  I 
  have 
  been 
  able 
  to 
  find 
  bearing 
  on 
  the 
  

   point. 
  Vie 
  may 
  then 
  put 
  

  

  {3= 
  0*0036700 
  + 
  0*0000115( 
  /3 
  -l). 
  

  

  Nitrogen 
  (atmospheric, 
  containing 
  argon). 
  — 
  Regnault 
  

   found 
  0*0036682 
  for 
  one 
  atmo 
  pressure 
  at 
  the 
  freezing- 
  

   point, 
  Chappuis 
  0*00367466 
  for 
  995 
  mm. 
  pressure. 
  We 
  

   adopt 
  the 
  latter 
  value, 
  and 
  in 
  the 
  absence 
  of 
  any 
  more 
  definite 
  

   information 
  take 
  the 
  same 
  density 
  variation 
  as 
  for 
  air. 
  

  

  ^ 
  = 
  0*0036710 
  + 
  0*0000115(^-1) 
  

  

  where 
  p 
  is 
  the 
  density 
  expressed 
  in 
  terms 
  of 
  that 
  at 
  N.T.P. 
  

   as 
  unit. 
  

  

  Carbon 
  Dioxide. 
  — 
  This 
  gas 
  has 
  been 
  very 
  much 
  studied, 
  but 
  

   the 
  results 
  are 
  more 
  discordant 
  than 
  for 
  air. 
  Regnault 
  gives 
  

   0*0036871 
  and 
  0*0036856 
  as 
  the 
  result 
  of 
  two 
  experiments, 
  

   Magnus 
  0*0036937, 
  Jolly 
  0*0037060, 
  all 
  at 
  normal 
  density, 
  

   while 
  Regnault's 
  observations 
  on 
  the 
  density 
  variation 
  are 
  

   well 
  represented 
  by 
  0*00004668, 
  the 
  number 
  adopted 
  by 
  

   Jochmann. 
  The 
  values 
  given 
  here 
  are 
  not 
  corrected 
  in 
  the 
  

  

  * 
  Mendeleef, 
  Ber. 
  d. 
  Deutsck. 
  Chem. 
  Ges. 
  x. 
  p. 
  81. 
  

  

  