﻿High 
  Temperatures. 
  175 
  

  

  Table 
  I 
  gives 
  the 
  results 
  obtained 
  from 
  two 
  separate 
  

   hydrogen 
  fillings 
  between 
  200° 
  and 
  500°. 
  It 
  contains 
  the 
  

   ice 
  point 
  (H^), 
  coejfficient 
  of 
  expansion 
  of 
  the 
  gas 
  between 
  0° 
  

   and 
  100° 
  (a), 
  the 
  temperature 
  as 
  observed 
  with 
  the 
  gas 
  ther- 
  

   mometer 
  and 
  calculated 
  according 
  to 
  above 
  formula 
  (i)^ 
  with 
  

   the 
  corresponding 
  E. 
  M. 
  F. 
  of 
  the 
  therrao-element 
  T^ 
  in 
  micro- 
  

   volts 
  {e^. 
  

  

  For 
  this 
  element 
  a 
  curve 
  has 
  been 
  drawn 
  based 
  upon 
  the 
  

   mean 
  of 
  all 
  the 
  observations 
  made 
  with 
  it 
  between 
  200° 
  and 
  

   1150°. 
  Its 
  form 
  is 
  given 
  by 
  the 
  following 
  table. 
  

  

  

  

  Iable 
  

  

  11. 
  

  

  

  

  

  Thermo-element 
  Ti. 
  

  

  

  

  Microvolts. 
  

  

  

  

  

  Microvolts 
  

  

  200° 
  

  

  1334 
  

  

  

  

  700° 
  

  

  6044 
  

  

  250 
  . 
  

  

  1752 
  

  

  

  

  750 
  

  

  6566 
  

  

  300 
  

  

  2186 
  

  

  

  

  800 
  

  

  7096 
  

  

  350 
  

  

  2635 
  

  

  

  

  850 
  

  

  7634 
  

  

  400 
  

  

  3096 
  

  

  

  

  900 
  

  

  8180 
  

  

  450 
  

  

  3566 
  

  

  

  

  95 
  C 
  

  

  8734 
  

  

  500 
  

  

  4044 
  

  

  

  

  1000 
  

  

  9296 
  

  

  550 
  

  

  4531 
  

  

  

  

  1050 
  

  

  9863 
  

  

  600 
  

  

  5027 
  

  

  

  

  1100 
  

  

  10434 
  

  

  650 
  

  

  5531 
  

  

  

  

  1150 
  

  

  11009 
  

  

  The 
  last 
  column 
  of 
  Table 
  I 
  (" 
  Obs.-CalcuL") 
  contains 
  the 
  

   difference 
  between 
  any 
  particular 
  observed 
  temperature 
  and 
  

   the 
  corresponding 
  point 
  of 
  this 
  curve 
  in 
  degrees. 
  The 
  mean 
  

   value 
  of 
  this 
  error 
  for 
  this 
  set 
  of 
  observations 
  amounts 
  to 
  

   db 
  0'64:° 
  and 
  is 
  to 
  be 
  charged 
  rather 
  to 
  the 
  tliermo-element 
  

   than 
  to 
  the 
  gas 
  thermometer. 
  

  

  The 
  differences 
  between 
  the 
  separate 
  values 
  of 
  the 
  zero 
  point 
  

   (HQ)and 
  of 
  the 
  coefficient 
  of 
  expansion 
  (a) 
  are 
  not 
  greater 
  with 
  

   the 
  first 
  filling 
  than 
  the 
  maximum 
  error 
  of 
  observation, 
  which 
  in 
  

   reading 
  barometer 
  and 
  manometer 
  might 
  in 
  unfavorable 
  cases 
  

   combine 
  to 
  amount 
  to 
  0-05"'°^. 
  The 
  ice 
  point 
  in 
  the 
  second 
  

   series 
  of 
  observations 
  showed 
  a 
  tendency 
  to 
  fall, 
  the 
  amount 
  of 
  

   which, 
  if 
  we 
  may 
  except 
  the 
  last 
  heatiug 
  when 
  the 
  temperature 
  

   was 
  forced 
  up 
  to 
  550°, 
  was 
  0-14™'° 
  (maximum). 
  

  

  Jf.. 
  JEh^periments 
  loith 
  porcelain 
  bulbs 
  glazed 
  inside 
  and 
  out. 
  

  

  Bulbs. 
  — 
  Temperatures 
  above 
  500° 
  were 
  next 
  measured 
  with 
  

   a 
  bulb 
  of 
  this 
  type. 
  These 
  bulbs, 
  in 
  a 
  form 
  adapted 
  for 
  

   this 
  purpose, 
  cannot, 
  on 
  account 
  of 
  the 
  fine 
  capillary 
  tube, 
  be 
  

   glazed 
  by 
  simply 
  dipping 
  a 
  finished 
  bulb 
  in 
  the 
  glazing 
  solution 
  

   and 
  burning. 
  Deville 
  and 
  Troost 
  succeeded 
  in 
  glazing 
  their 
  

   bulbs 
  on 
  the 
  inside, 
  only 
  by 
  preparing 
  bulb 
  and 
  capillary 
  in 
  

   separate 
  pieces 
  and 
  melting 
  them 
  together 
  in 
  the 
  oxy-hydro- 
  

   gen 
  flame. 
  

  

  