﻿310 
  Dr. 
  0. 
  Chree's 
  Notes 
  on 
  Thermometry. 
  

  

  changes 
  with 
  the 
  temperature, 
  and 
  that, 
  at 
  least 
  in 
  the 
  case 
  

   of 
  verve 
  dur, 
  the 
  two 
  influences 
  nearly 
  neutralize 
  one 
  another. 
  

   The 
  limits 
  of 
  temperature 
  within 
  which 
  this 
  happy 
  arrange- 
  

   ment 
  prevails 
  are, 
  I 
  suspect, 
  somewhat 
  uncertain. 
  At 
  

   temperatures 
  above 
  100° 
  C, 
  1 
  suspect 
  the 
  uncertainties 
  

   proceeding 
  from 
  the 
  above 
  sources 
  and 
  from 
  the 
  capillarity 
  

   are 
  appreciable. 
  Internal 
  pressure-corrections 
  of 
  the 
  order 
  

   o, 
  l 
  C. 
  are 
  by 
  no 
  means 
  unprecedented, 
  so 
  that 
  when 
  accuracy 
  

   of 
  the 
  order 
  0°*001 
  C. 
  is 
  aimed 
  at, 
  an 
  error 
  of 
  even 
  2 
  or 
  3 
  per 
  

   cent, 
  in 
  the 
  value 
  of 
  /3; 
  begins 
  to 
  tell. 
  

  

  § 
  28. 
  By 
  adopting 
  the 
  vertical 
  as 
  the 
  standard 
  position, 
  the 
  

   practice 
  followed 
  at 
  Kew 
  Observatory, 
  the 
  necessity 
  for 
  an 
  

   internal 
  pressure-correction 
  can 
  usually 
  be 
  avoided, 
  at 
  least 
  

   for 
  accuracy 
  of 
  the 
  order 
  o, 
  01 
  C. 
  As 
  already 
  pointed 
  out, 
  

   when 
  the 
  stem 
  is 
  vertical 
  the 
  reduction 
  of 
  the 
  reading, 
  like 
  

   its 
  two 
  contributory 
  causes 
  expansion 
  of 
  the 
  bulb 
  and 
  com- 
  

   pression 
  of 
  the 
  mercury, 
  is 
  proportional, 
  at 
  least 
  as 
  a 
  first 
  

   approximation, 
  to 
  the 
  length 
  of 
  the 
  mercury-column 
  measured 
  

   from 
  the 
  middle 
  of 
  the 
  bulb 
  ; 
  but 
  in 
  an 
  ordinary 
  thermometer, 
  

   increment 
  of 
  stem-length 
  is 
  sufficiently 
  nearly 
  proportional 
  

   to 
  increment 
  of 
  reading. 
  Hence 
  at 
  any 
  temperature 
  t 
  the 
  

   depression 
  due 
  to 
  internal 
  pressure 
  is, 
  in 
  stem-divisions, 
  

  

  g 
  + 
  t/p, 
  

  

  where 
  p 
  and 
  q 
  are 
  constants 
  for 
  the 
  thermometer. 
  

  

  Thus 
  to 
  make 
  a 
  Bureau 
  International 
  thermometer 
  register 
  

   correctly 
  in 
  the 
  vertical 
  position 
  we 
  need 
  only 
  lower 
  the 
  

   freezing-point 
  mark 
  q 
  divisions 
  below 
  the 
  position 
  answering 
  

   to 
  a 
  horizontal 
  reading, 
  and 
  shorten 
  each 
  degree-division 
  by 
  \jp 
  

   of 
  itself. 
  This 
  obviously 
  comes 
  to 
  the 
  same 
  thing 
  as 
  marking 
  

   the 
  freezing- 
  and 
  boiling-points 
  with 
  the 
  thermometer 
  vertical 
  

   and 
  subdividing 
  the 
  fundamental 
  interval 
  in 
  the 
  usual 
  way. 
  

  

  This 
  fact 
  has 
  been 
  pointed 
  out 
  by 
  Prof. 
  Schuster 
  and 
  

   Mr. 
  Gannon 
  * 
  ; 
  it 
  was 
  arrived 
  at 
  independently 
  by 
  myself 
  

   and 
  embodied 
  in 
  a 
  report 
  on 
  thermometry 
  made 
  some 
  years 
  

   ago 
  to 
  the 
  Kew 
  Observatory 
  Committee. 
  

  

  When 
  an 
  ordinary 
  Kew 
  thermometer 
  is 
  used 
  in 
  the 
  

   horizontal 
  position 
  a 
  subtractive 
  correction 
  must 
  be 
  applied 
  

   to 
  obtain 
  the 
  true 
  temperature. 
  Exceptions 
  to 
  this 
  rule 
  exist 
  

   in 
  the 
  shape 
  of 
  meteorological 
  maximum 
  and 
  minimum 
  

   thermometers, 
  which 
  it 
  is 
  the 
  custom 
  to 
  employ 
  in 
  a 
  hori- 
  

   zontal 
  position. 
  To 
  prevent 
  possible 
  misconception, 
  the 
  

   certificates 
  issued 
  state 
  explicitly 
  to 
  which 
  of 
  the 
  two 
  positions 
  

   the 
  table 
  of 
  corrections 
  applies. 
  

  

  * 
  Phil. 
  Trans. 
  A., 
  1895, 
  p. 
  434. 
  

  

  