﻿222 
  Dr. 
  C. 
  Chree's 
  Notes 
  on 
  Thermometry. 
  

  

  As 
  the 
  " 
  Proces-Verbaux 
  " 
  remarks, 
  " 
  cette 
  fonction 
  ne 
  se 
  

   raccorde 
  pas 
  tres 
  bien 
  avec 
  celle 
  qui 
  exprime 
  la 
  difference 
  de 
  

   marche 
  entre 
  0° 
  et 
  100°" 
  (cf. 
  § 
  11). 
  

  

  The 
  Bureau 
  do 
  not 
  regard 
  the 
  results 
  as 
  very 
  satisfactory, 
  

   owing 
  to 
  the 
  fact 
  that 
  the 
  differences 
  between 
  different 
  verve 
  

   duv 
  thermometers 
  " 
  peuvent 
  a 
  200° 
  atteindre 
  5 
  a 
  6 
  centiemes 
  

   de 
  degre, 
  quantite 
  du 
  meme 
  ordre 
  de 
  grandeur 
  que 
  la 
  differ- 
  

   ence 
  de 
  marche 
  par 
  rapport 
  au 
  thermometre 
  a 
  hyclrogene 
  " 
  

   (I. 
  c. 
  p. 
  45) 
  . 
  For 
  practical 
  purposes, 
  however, 
  it 
  is 
  interesting 
  

   to 
  know 
  that 
  even 
  up 
  to 
  200° 
  C. 
  the 
  (movable 
  zero) 
  tempera- 
  

   ture-reading 
  from 
  a 
  verve 
  dur 
  thermometer 
  is 
  likely 
  to 
  differ 
  

   from 
  the 
  hydrogen 
  scale 
  by 
  less 
  than 
  o, 
  2 
  C. 
  

  

  At 
  the 
  Reichsanstalt 
  - 
  * 
  comparisons 
  of 
  Jena-glass 
  thermo- 
  

   meters 
  with 
  the 
  air-thermometer 
  in 
  an 
  oil-bath 
  over 
  the 
  range 
  

   100° 
  to 
  300° 
  C. 
  are 
  assigned 
  an 
  accuracy 
  of 
  0°-02 
  C. 
  Even 
  at 
  

   500° 
  C. 
  the 
  comparison 
  of 
  thermometers 
  of 
  Jena 
  glass 
  59 
  IH 
  — 
  

   with 
  the 
  mercury 
  prevented 
  from 
  boiling 
  by 
  gaseous 
  pres- 
  

   sure 
  — 
  is 
  apparently 
  assigned 
  an 
  accuracy 
  of 
  the 
  order 
  o, 
  l 
  C, 
  

   when 
  auxiliary 
  thermometers 
  (Faden-thermometer) 
  of 
  a 
  

   special 
  pattern 
  are 
  used 
  to 
  supply 
  a 
  correction 
  for 
  the 
  emer- 
  

   gent 
  column 
  (see 
  § 
  38). 
  

  

  For 
  the 
  difference 
  8 
  between 
  the 
  natural 
  scale 
  of 
  59 
  ni 
  and 
  

   that 
  of 
  the 
  air-thermometer 
  over 
  the 
  range 
  100° 
  to 
  300° 
  C. 
  the 
  

   formula 
  proposed 
  is 
  

  

  B=t[l00-t)\ 
  -•000007233--0000004259(100-0 
  \, 
  

  

  This 
  differs 
  somewhat 
  conspicuously 
  from 
  the 
  corresponding 
  

   formula 
  for 
  the 
  range 
  —20° 
  to 
  100° 
  C. 
  quoted 
  in 
  § 
  13. 
  

  

  According 
  to 
  the 
  formula, 
  the 
  two 
  scales 
  differ 
  bv 
  0*7° 
  C. 
  at 
  

   200° 
  C, 
  and 
  by 
  4°'7 
  C. 
  at 
  300° 
  C. 
  

  

  § 
  16. 
  Of 
  earlier 
  comparisons 
  the 
  best 
  known 
  are 
  those 
  of 
  

   Regnault, 
  whose 
  results 
  are 
  quoted 
  in 
  several 
  English 
  works 
  

   (e. 
  g. 
  Balfour 
  Stewart's 
  ' 
  Heat,' 
  Lord 
  Kelvin's 
  ' 
  Mathematical 
  

   and 
  Physical 
  Papers,' 
  vol. 
  iii., 
  Smithsonian 
  ' 
  Physical 
  Tables/ 
  

   1896). 
  Two 
  useful 
  warnings 
  with 
  respect 
  to 
  these 
  data 
  are 
  

   given 
  by 
  Crafts 
  f. 
  Firstly, 
  Regnault 
  used 
  a 
  fixed 
  zero 
  

   method 
  ; 
  secondly, 
  French 
  crystal 
  glass 
  for 
  thermometers 
  has 
  

   been 
  considerably 
  altered 
  since 
  his 
  time. 
  The 
  Choisy-le-Roi 
  

   crystal 
  glass 
  used 
  by 
  Regnault 
  in 
  particular 
  contained 
  nearly 
  

   twice 
  as 
  much 
  lead 
  as 
  modern 
  French 
  crystal. 
  Crafts 
  himself, 
  

   in 
  1882, 
  compared 
  a 
  number 
  of 
  thermometers 
  of 
  modern 
  

   French 
  crystal 
  glass, 
  from 
  two 
  different 
  makers, 
  and 
  likewise 
  

  

  * 
  "Die 
  Thatigkeit 
  derPhys. 
  tech. 
  Beichsanstalt" 
  (Zeits.f. 
  Instr. 
  Sept, 
  

   1894, 
  p. 
  303). 
  

  

  t 
  Comptes 
  JRendus, 
  vol. 
  xcv. 
  1882, 
  p. 
  886. 
  

  

  