﻿220 
  Dr. 
  C. 
  Chree's 
  Notes 
  on 
  Thermometry. 
  

  

  to 
  100° 
  C. 
  the 
  difference 
  8 
  between 
  the 
  natural 
  scale 
  of 
  the 
  

   air-thermometer 
  and 
  Jena 
  glass 
  59 
  in 
  is 
  represented 
  by 
  

  

  8 
  = 
  *(100-0 
  {•00000487--0000002638(100-0}*. 
  

   This 
  appears 
  to 
  be 
  the 
  algebraic 
  excess 
  of 
  the 
  reading 
  on 
  the 
  

   air-thermometer, 
  though 
  I 
  do 
  not 
  see 
  an 
  explicit 
  statement 
  to 
  

   that 
  effect. 
  At 
  least 
  it 
  makes 
  

  

  S=-0°-021 
  at 
  50° 
  0., 
  and 
  +0°'064 
  at 
  -20° 
  C. 
  

  

  High- 
  Temperature 
  Results. 
  

  

  § 
  14. 
  At 
  temperatures 
  above 
  100° 
  C. 
  the 
  most 
  exact 
  method 
  of 
  

   comparing 
  thermometers 
  is 
  probably 
  to 
  use 
  the 
  boiling-points 
  

   of 
  a 
  series 
  of 
  liquids. 
  By 
  modifying 
  the 
  pressure 
  under 
  which 
  

   the 
  liquid 
  boils, 
  a 
  single 
  liquid 
  may 
  suffice 
  for 
  comparisons 
  

   over 
  a 
  very 
  considerable 
  range. 
  Thus 
  Crafts 
  f, 
  in 
  1882, 
  

   covered 
  a 
  range 
  of 
  70° 
  C. 
  with 
  water- 
  vapour 
  alone. 
  According 
  

   to 
  Crafts, 
  the 
  method 
  was 
  suggested 
  primarily 
  by 
  Regnault. 
  

   A 
  liquid 
  bath 
  has 
  certain 
  advantages 
  from 
  its 
  simplicity 
  ; 
  but 
  

   the 
  stirring 
  must 
  be 
  very 
  good, 
  and 
  the 
  whole 
  of 
  the 
  mercury 
  

   in 
  the 
  thermometer-stem 
  should, 
  if 
  possible, 
  be 
  immersed. 
  

  

  In 
  addition 
  to 
  instrumental 
  difficulties, 
  troubles 
  are 
  apt 
  to 
  

   arise 
  from 
  the 
  behaviour 
  of 
  the 
  thermometer-glass. 
  A 
  ther- 
  

   mometer 
  exposed 
  for 
  some 
  hours, 
  or 
  even 
  minutes, 
  to 
  a 
  high 
  

   temperature, 
  in 
  excess 
  of 
  what 
  it 
  has 
  been 
  exposed 
  to 
  since 
  its 
  

   construction, 
  usually 
  experiences 
  a 
  permanent 
  rise 
  in 
  zero. 
  

   This 
  is 
  more 
  especially 
  true 
  of 
  temperatures 
  of 
  300° 
  C. 
  and 
  

   upwards. 
  In 
  some 
  experiments 
  by 
  Guillaume*, 
  exposure 
  for 
  

   ten 
  hours 
  to 
  a 
  temperature 
  of 
  between 
  360° 
  and 
  380° 
  C. 
  

   raised 
  the 
  zeros 
  of 
  three 
  thermometers 
  of 
  verre 
  dur, 
  Jena 
  

   glass, 
  and 
  French 
  crystal 
  by 
  1°*8, 
  3°0, 
  and 
  7 
  o, 
  C. 
  respec- 
  

   tively. 
  Further 
  heating 
  for 
  periods 
  of 
  twelve 
  and 
  ten 
  hours 
  

   at 
  the 
  same 
  temperature 
  produced 
  a 
  considerable 
  further 
  rise. 
  

   Probably 
  the 
  most 
  complete 
  experiments 
  on 
  this 
  subject 
  have 
  

   been 
  those 
  by 
  Crafts 
  nearly 
  twenty 
  years 
  ago. 
  In 
  1879-80 
  

   he 
  exposed 
  eight 
  thermometers, 
  four 
  of 
  French 
  crystal 
  and 
  

   four 
  of 
  a 
  German 
  soda-glass, 
  to 
  a 
  variety 
  of 
  temperatures, 
  

   the 
  highest 
  being 
  355° 
  C.§ 
  The 
  exposures 
  were 
  of 
  various 
  

   durations, 
  from 
  20 
  minutes 
  to 
  several 
  months. 
  The 
  rise 
  in 
  

   the 
  zero 
  varied 
  from 
  10° 
  to 
  18° 
  — 
  or, 
  including 
  rise 
  under 
  

   preliminary 
  treatment, 
  even 
  26° 
  C. 
  The 
  conclusions 
  drawn 
  

   by 
  Crafts 
  were 
  that 
  the 
  rise 
  is 
  much 
  most 
  rapid 
  at 
  the 
  begin- 
  

  

  * 
  " 
  Thatigkeit 
  der 
  Phys. 
  Tech. 
  Keichs." 
  (Zeits. 
  f. 
  Instr. 
  1894, 
  p. 
  305). 
  

  

  t 
  Comptes 
  Eendus 
  of 
  the 
  French 
  Academy, 
  vol. 
  xcv. 
  1882, 
  p. 
  838. 
  

   (Cf. 
  also 
  Kamsay 
  and 
  Young, 
  Phil. 
  Trans, 
  for 
  1884, 
  p. 
  37, 
  and 
  W. 
  

   Watson, 
  Phil. 
  Mag. 
  July 
  1897, 
  p. 
  116). 
  

  

  \ 
  Thermometries 
  p. 
  147. 
  

  

  § 
  Comptes 
  Rendus 
  of 
  the 
  French 
  Academy, 
  vol. 
  xci. 
  1880, 
  p. 
  291. 
  

  

  