﻿Dr. 
  C. 
  Chree's 
  Notes 
  on 
  Thermometry. 
  209 
  

  

  is 
  conspicuous 
  when 
  we 
  compare 
  ice-readings 
  taken 
  immedi- 
  

   ately 
  before 
  and 
  after 
  exposure 
  to 
  temperatures 
  between 
  

   50° 
  C. 
  and 
  100° 
  C. 
  ; 
  the 
  "depression 
  of 
  zero," 
  as 
  it 
  is 
  called, 
  

   is 
  greater 
  the 
  higher 
  the 
  previous 
  temperature. 
  Exposure 
  

   for 
  only 
  a 
  minute 
  or 
  two 
  to 
  a 
  high 
  temperature 
  is 
  only 
  

   partially 
  effective 
  ; 
  but 
  20 
  or 
  25 
  minutes 
  exposure 
  usually 
  

   produces 
  practically 
  the 
  full 
  effect. 
  On 
  the 
  other 
  hand, 
  the 
  

   depression 
  takes 
  a 
  considerable 
  time 
  to 
  disappear 
  ; 
  it 
  is 
  a 
  

   question 
  of 
  days 
  or 
  even 
  weeks 
  if 
  the 
  high 
  temperature 
  has 
  

   approached 
  100° 
  C. 
  

  

  Hitherto 
  at 
  the 
  Bureau 
  International 
  attention 
  has 
  been 
  

   mainly 
  directed 
  to 
  a 
  standard 
  glass, 
  French 
  verve 
  dur, 
  at 
  

   temperatures 
  between 
  —30° 
  C. 
  and 
  100° 
  C. 
  At 
  the 
  Reichs- 
  

   anstalt 
  certain 
  Jena 
  glasses, 
  notably 
  16 
  m 
  and 
  59 
  m 
  , 
  have 
  

   been 
  examined 
  in 
  similar 
  detail. 
  The 
  conclusion 
  reached 
  for 
  

   these 
  glasses 
  throughout 
  at 
  least 
  this 
  range, 
  is 
  that 
  the 
  

   difficulty 
  can 
  be 
  met 
  by 
  treating 
  a 
  reading 
  z 
  t 
  in 
  ice 
  immedi- 
  

   ately 
  after 
  the 
  measurement 
  of 
  any 
  temperature 
  t 
  as 
  the 
  zero 
  

   of 
  the 
  thermometer 
  for 
  the 
  preceding 
  observation. 
  This 
  is 
  

   equivalent 
  to 
  the 
  conclusion 
  that 
  the 
  residual 
  or 
  after-effect, 
  

   whose 
  existence 
  is 
  made 
  manifest 
  by 
  the 
  depression 
  of 
  zero, 
  

   remains 
  practically 
  unchanged 
  during 
  the 
  brief 
  interval 
  

   required 
  to 
  observe 
  both 
  t 
  and 
  z 
  t 
  . 
  If 
  t 
  has 
  been 
  preceded 
  by 
  

   a 
  higher 
  temperature, 
  the 
  residual 
  effect 
  is 
  in 
  reality 
  larger, 
  

   but 
  it 
  depresses 
  equally 
  the 
  readings 
  t 
  and 
  z 
  t 
  . 
  This 
  method 
  

   of 
  a 
  u 
  movable 
  zero 
  " 
  requires 
  that 
  in 
  determining 
  the 
  

   fundamental 
  interval 
  the 
  boiling-point 
  100° 
  C. 
  be 
  first 
  

   determined, 
  and 
  that 
  the 
  ice-reading 
  e 
  10 
  o 
  be 
  taken 
  immediately 
  

   after. 
  

  

  The 
  ordinary 
  practice 
  in 
  this 
  country 
  proceeds 
  on 
  the 
  

   hypothesis 
  of 
  a 
  " 
  fixed 
  zero/'' 
  In 
  determining 
  the 
  funda- 
  

   mental 
  interval 
  the 
  observation 
  of 
  the 
  ice-point 
  precedes 
  that 
  

   of 
  the 
  boiling-point 
  ; 
  and 
  in 
  ordinary 
  use 
  the 
  observed 
  

   departure 
  from 
  0° 
  in 
  a 
  preliminary 
  observation 
  in 
  ice 
  is 
  

   applied 
  as 
  a 
  constant 
  correction 
  at 
  all 
  points 
  of 
  the 
  scale. 
  

   The 
  relative 
  merics 
  of 
  the 
  movable 
  zero 
  and 
  fixed 
  zero 
  

   methods 
  will 
  be 
  discussed 
  later 
  ; 
  in 
  the 
  mean 
  time 
  let 
  us 
  

   consider 
  what 
  modifications 
  are 
  required 
  in 
  the 
  theory 
  

   developed 
  for 
  the 
  ideal 
  thermometer. 
  

  

  § 
  4. 
  In 
  ordinary 
  investigations 
  of 
  thermal 
  expansion 
  the 
  

   temperature 
  is 
  gradually 
  raised, 
  and 
  the 
  volume 
  compared 
  at 
  

   intervals 
  with 
  that 
  originally 
  possessed 
  by 
  the 
  material 
  at 
  

   0° 
  C. 
  This 
  is 
  a 
  fixed 
  zero 
  method 
  ; 
  thus 
  supposing 
  a 
  lf 
  a 
  2) 
  

   &c. 
  to 
  have 
  the 
  values 
  obtained 
  in 
  ordinary 
  experiments, 
  (6) 
  

   is 
  immediately 
  applicable 
  in 
  ordinary 
  British 
  thermometry, 
  

   provided 
  readings 
  be 
  always 
  taken 
  with 
  the 
  temperature 
  

  

  