﻿Dr. 
  C. 
  Chree's 
  Notes 
  on 
  Thermometry, 
  313 
  

  

  the 
  ordinary 
  way 
  from 
  calibration 
  — 
  by 
  a 
  constant 
  amount 
  

   6t", 
  where 
  & 
  has 
  its 
  previous 
  meaning, 
  while 
  i" 
  is 
  the 
  number 
  

   of 
  degree-lengths 
  by 
  which 
  the 
  stem 
  is 
  shortened 
  through 
  

   the 
  existence 
  of 
  the 
  chamber. 
  

  

  In 
  all 
  possible 
  combinations 
  of 
  auxiliary 
  chambers 
  the 
  

   degree-spaces 
  must 
  be 
  equal 
  in 
  volume 
  throughout 
  the 
  entire 
  

   stem, 
  if 
  the 
  scale 
  is 
  to 
  be 
  that 
  natural 
  to 
  the 
  glass. 
  

  

  § 
  30. 
  In 
  all 
  ordinary 
  mercury-thermometers 
  some 
  trace 
  of 
  

   air 
  * 
  exists, 
  and 
  if 
  this 
  collects 
  in 
  the 
  stem 
  when 
  the 
  thermo- 
  

   meter 
  is 
  exposed 
  to 
  a 
  high 
  temperature 
  the 
  mercury-column 
  

   is 
  apt 
  to 
  break. 
  Partly 
  to 
  prevent 
  this, 
  and 
  partly 
  to 
  increase 
  

   the 
  range, 
  limited 
  by 
  the 
  boiling-point 
  of 
  mercury, 
  it 
  is 
  

   becoming 
  common 
  to 
  introduce 
  nitrogen 
  or 
  carbonic-acid 
  gas 
  

   above 
  the 
  mercury, 
  sometimes 
  under 
  very 
  considerable 
  

   pressure. 
  In 
  such 
  thermometers 
  the 
  internal 
  pressure 
  from 
  

   the 
  gas 
  may 
  much 
  exceed 
  that 
  arising 
  from 
  the 
  weight 
  of 
  the 
  

   mercury-column. 
  If 
  there 
  were 
  a 
  chamber 
  at 
  the 
  top 
  of 
  

   the 
  tube 
  similar 
  in 
  size 
  to 
  the 
  thermometer's 
  bulb, 
  and 
  if 
  this 
  

   chamber 
  and 
  its 
  gaseous 
  contents 
  remained 
  at 
  a 
  constant 
  

   temperature, 
  the 
  internal 
  pressure 
  contributed 
  by 
  the 
  gas 
  

   would 
  increase 
  very 
  approximately 
  directly 
  as 
  the 
  length 
  of 
  

   the 
  mercury-column. 
  Thus, 
  supposing 
  the 
  tube 
  calibrated 
  

   before 
  the 
  introduction 
  of 
  the 
  gas, 
  the 
  experimental 
  deter- 
  

   mination 
  of 
  two 
  fixed 
  temperatures 
  and 
  the 
  subdivision 
  of 
  the 
  

   tube 
  into 
  equal 
  volumes 
  would 
  supply, 
  at 
  least 
  theoretically, 
  

   an 
  independent 
  standard 
  thermometer. 
  In 
  practice, 
  however, 
  

   the 
  top 
  chamber 
  is 
  usually 
  of 
  restricted 
  size, 
  and 
  its 
  tempera- 
  

   ture 
  tends 
  to 
  rise 
  with 
  that 
  of 
  the 
  thermometer. 
  For 
  both 
  

   reasons 
  the 
  gaseous 
  pressure 
  may 
  become 
  dangerously 
  large 
  

   when 
  the 
  thermometer 
  is 
  recording 
  near 
  the 
  top 
  of 
  its 
  scale. 
  

  

  In 
  practice 
  it 
  is 
  advisable 
  to 
  regard 
  a 
  gas-filled 
  thermo- 
  

   meter 
  as 
  deriving 
  its 
  authority 
  from 
  a 
  direct 
  comparison 
  with 
  

   an 
  ordinary 
  standard. 
  And 
  the 
  conditions 
  of 
  this 
  comparison 
  

   should 
  approximate 
  as 
  closely 
  as 
  possible 
  to 
  the 
  circumstances 
  

   under 
  which 
  the 
  thermometer 
  is 
  to 
  be 
  used. 
  

  

  Under 
  ordinary 
  conditions 
  the 
  external 
  pressure- 
  correction 
  

   is 
  of 
  trifling 
  importance 
  relative 
  to 
  the 
  internal 
  pressure- 
  

   correction. 
  Yariations 
  of 
  barometric 
  pressure 
  at 
  sea-level 
  

   may 
  give 
  rise 
  to 
  corrections 
  of 
  a 
  few 
  thousandths 
  of 
  a 
  degree, 
  

   and 
  similar 
  consequences 
  may 
  follow 
  from 
  immersion 
  in 
  a 
  

   water-bath. 
  In 
  a 
  heavy 
  liquid, 
  however, 
  such 
  as 
  mercury, 
  

   a 
  comparatively 
  shallow 
  immersion 
  produces 
  effects 
  which 
  

   cannot 
  be 
  disregarded 
  even 
  for 
  accuracy 
  to 
  o, 
  01 
  C 
  

  

  * 
  In 
  the 
  Reich 
  sanstalt's 
  Wiss. 
  Abhandl. 
  vol. 
  i. 
  p. 
  71, 
  there 
  is 
  a 
  

   reference 
  to 
  a 
  supposed 
  influence 
  of 
  the 
  small 
  trace 
  of 
  air 
  or 
  vapour 
  in 
  

   ordinary 
  thermometers 
  on 
  the 
  internal 
  pressure. 
  

  

  