﻿Dr. 
  C. 
  Chree's 
  Notes 
  on 
  Thermometry. 
  309 
  

  

  to 
  be 
  the 
  case 
  when 
  the 
  meniscus 
  changes 
  shape, 
  its 
  influence 
  

   cannot 
  be 
  neglected. 
  

  

  Regarding 
  the 
  experimental 
  determination 
  of 
  ft 
  "as 
  some- 
  

   what 
  uncertain, 
  Guillaume 
  * 
  recommends 
  that 
  it 
  be 
  deduced 
  

   from 
  the 
  observed 
  value 
  of 
  ft, 
  employing 
  the 
  theoretical 
  

   connexion 
  between 
  the 
  two 
  coefficients. 
  

  

  The 
  capillarity 
  difficulty 
  is 
  dwelt 
  on 
  in 
  the 
  Reichsanstalt's 
  

   publications 
  f 
  ; 
  but, 
  notwithstanding, 
  the 
  plan 
  recommended 
  

   is 
  to 
  attach 
  equal 
  weights 
  to 
  experimentally 
  determined 
  

   values 
  ft/, 
  ft' 
  of 
  the 
  two 
  coefficients, 
  and 
  to 
  calculate 
  values 
  

   ft 
  and 
  ft 
  by 
  means 
  of 
  these 
  and 
  the 
  theoretical 
  value 
  of 
  

   A 
  - 
  — 
  ft. 
  For 
  instance, 
  in 
  a 
  verre 
  dur 
  thermometer 
  we 
  should 
  

   have 
  

  

  ft=i(ft'+ft'--0000154), 
  

  

  ft 
  = 
  i(ft/ 
  + 
  ft' 
  + 
  -0000154). 
  

  

  A 
  table 
  X 
  of 
  observed 
  and 
  calculated 
  values 
  in 
  various 
  ther- 
  

   mometers 
  of 
  different 
  patterns 
  shows 
  a 
  pretty 
  close 
  agreement 
  

   between 
  the 
  mean 
  observed 
  difference 
  ft'— 
  ft/ 
  and 
  the 
  

   theoretical 
  ; 
  but 
  in 
  individual 
  cases 
  the 
  observed 
  difference 
  

   varies 
  from 
  45 
  X 
  10~ 
  7 
  to 
  250 
  x 
  10~ 
  7 
  . 
  The 
  explanation 
  

   favoured 
  is 
  apparently 
  the 
  uncertainty 
  in 
  the 
  capillary 
  effect. 
  

  

  Personally, 
  I 
  am 
  disposed 
  to 
  think 
  that 
  too 
  much 
  reliance 
  

   is 
  placed 
  on 
  the 
  supposed 
  constancy 
  of 
  ft— 
  ft 
  in 
  all 
  thermo- 
  

   meters 
  of 
  the 
  same 
  glass. 
  The 
  supposed 
  relation 
  is 
  exact 
  

   under 
  two 
  restrictions, 
  absolute 
  homogeneousness 
  in 
  the 
  

   glass, 
  and 
  uniformity 
  in 
  the 
  distribution 
  of 
  the 
  external 
  and 
  

   internal 
  pressures 
  over 
  their 
  respective 
  surfaces, 
  inclusive 
  

   of 
  bulb, 
  stem, 
  and 
  any 
  auxiliary 
  chambers. 
  Under 
  these 
  

   circumstances 
  a 
  simple 
  proof 
  of 
  the 
  relation 
  has 
  been 
  given 
  

   by 
  myself 
  §, 
  and 
  independently 
  by 
  Schuster 
  and 
  Gannon 
  ||. 
  

   The 
  supposed 
  uniformity 
  of 
  pressure 
  is, 
  however, 
  non-existent, 
  

   in 
  the 
  case 
  more 
  especially 
  of 
  the 
  internal 
  pressure, 
  and 
  the 
  

   means 
  of 
  allowing 
  for 
  this 
  in 
  the 
  theoretical 
  investigation 
  

   have 
  not 
  yet 
  been 
  devised. 
  

  

  In 
  applying 
  the 
  internal 
  pressure-correction 
  allowance 
  is 
  

   made 
  at 
  the 
  Reichsanstalt 
  ^[ 
  for 
  the 
  diminution 
  in 
  the 
  density 
  

   of 
  the 
  mercury-column 
  as 
  the 
  temperature 
  rises. 
  Guillaume** 
  

   omits 
  this, 
  on 
  the 
  ground 
  that 
  the 
  elasticity 
  of 
  the 
  glass 
  also 
  

  

  * 
  Loc. 
  cit. 
  p. 
  103. 
  

  

  f 
  Wiss. 
  Abhandl 
  vol. 
  i. 
  1894, 
  p. 
  80. 
  

   \ 
  L. 
  c. 
  p. 
  79. 
  

  

  § 
  Phil. 
  Mag. 
  Oct. 
  1894, 
  p. 
  371. 
  

   || 
  Phil. 
  Trans. 
  A., 
  1895, 
  p. 
  435. 
  

   % 
  Wiss. 
  Abhandl. 
  vol. 
  i. 
  1894, 
  p. 
  68. 
  

  

  ** 
  Thermometrie, 
  p. 
  107, 
  also 
  Tliermometres 
  Etalons, 
  p. 
  18 
  (Bureau's 
  

   Travanx 
  et 
  Memoires, 
  vol. 
  x. 
  1894). 
  

  

  