﻿High 
  Temjperatures. 
  181 
  

  

  the 
  room 
  to 
  a 
  minimuin, 
  the 
  coil 
  was 
  first 
  surrounded 
  with 
  

   one, 
  later 
  with 
  two 
  concentric 
  tubes 
  of 
  claj, 
  separated 
  from 
  

   each 
  other 
  as 
  well 
  as 
  from 
  the 
  coil 
  by 
  a 
  thin 
  air 
  jacket 
  (1^°^). 
  

   These 
  spaces 
  were 
  closed 
  at 
  the 
  ends 
  and 
  the 
  tubes 
  held 
  in 
  

   place 
  by 
  a 
  tight 
  packing 
  of 
  asbestos. 
  Each 
  tube 
  was 
  provided 
  

   at 
  the 
  ends 
  with 
  a 
  clay 
  cover^ 
  fitting 
  tightly, 
  and 
  between 
  the 
  

   covers 
  was 
  also 
  left 
  an 
  air 
  space 
  of 
  about 
  a 
  centimeter. 
  Open- 
  

   ings 
  were 
  filed 
  in 
  the 
  covers 
  to 
  admit 
  the 
  stem 
  of 
  the 
  ther- 
  

   mometer 
  bulb 
  and 
  the 
  insulating 
  tubes 
  carrying 
  the 
  thermo- 
  

   elements. 
  This 
  oven, 
  while 
  maintaining 
  a 
  temperature 
  of 
  

   1000°, 
  consumed 
  about 
  1400 
  watts. 
  

  

  The 
  thermometer 
  bulb 
  lay 
  symmetrically 
  in 
  the 
  oven 
  tube 
  

   resting 
  upon 
  three 
  very 
  small 
  pieces 
  of 
  fire-clay, 
  especial 
  care 
  

   being 
  taken 
  that 
  it 
  nowhere 
  touched 
  the 
  oven. 
  For 
  each 
  set 
  

   of 
  measurements 
  two 
  or 
  three 
  thermoelements 
  were 
  used, 
  

   stretched 
  through 
  the 
  oven 
  from 
  end 
  to 
  end 
  so 
  that 
  the 
  junc- 
  

   tion 
  could 
  be 
  brought 
  to 
  any 
  point 
  along 
  the 
  bulb 
  whose 
  

   temperature 
  was 
  required. 
  The 
  fall 
  of 
  temperature 
  along 
  the 
  

   bulb 
  amounted 
  to 
  not 
  more 
  than 
  6°, 
  in 
  the 
  mean 
  perhaps 
  to 
  

   half 
  that. 
  The 
  results 
  obtained 
  from 
  the 
  several 
  elements 
  

   were 
  all 
  referred 
  to 
  the 
  element 
  T,, 
  a 
  careful 
  comparison 
  with 
  

   Tj 
  over 
  the 
  entire 
  range 
  of 
  temperature 
  having 
  been 
  made 
  

   previously 
  ; 
  the 
  mean 
  of 
  these 
  is 
  the 
  value 
  given 
  in 
  the 
  table. 
  

   Inasmuch 
  as 
  this 
  mean 
  gave 
  an 
  excellent 
  agreement 
  with 
  the 
  

   gas 
  thermometer 
  in 
  two 
  so 
  different 
  sets 
  of 
  conditions 
  (nitre 
  

   hath 
  and 
  electric 
  oven), 
  and 
  the 
  temperature 
  conditions 
  within 
  

   the 
  oven 
  can 
  only 
  improve 
  as 
  the 
  temperature 
  rises 
  (compare 
  § 
  

   6), 
  no 
  more 
  accurate 
  determination 
  of 
  the 
  error 
  due 
  to 
  varia- 
  

   tions 
  of 
  temperature 
  along 
  the 
  bulb 
  was 
  made. 
  

  

  Table 
  YI 
  contains 
  two 
  sets 
  of 
  results 
  obtained 
  in 
  this 
  oven. 
  

   In 
  one 
  respect 
  the 
  above 
  described 
  arrangement 
  made 
  a 
  distinct 
  

   gain 
  over 
  the 
  previous 
  one 
  ; 
  with 
  the 
  bath 
  the 
  bulb 
  had 
  to 
  be 
  

   provided 
  with 
  a 
  long 
  bent 
  stem 
  for 
  dipping 
  into 
  the 
  kettle 
  

   without 
  unduly 
  heating 
  the 
  sealed 
  joint 
  where 
  the 
  manometer 
  

   capillary 
  was 
  attached 
  ; 
  here 
  a 
  short 
  straight 
  one 
  sufficed. 
  The 
  

   " 
  unheated 
  space 
  " 
  was 
  thus 
  materially 
  diminished, 
  and 
  the 
  

   error 
  from 
  this 
  cause 
  also. 
  The 
  effect 
  of 
  the 
  unheated 
  space 
  

   upon 
  t 
  was 
  here 
  9*7 
  per 
  cent 
  at 
  800° 
  and 
  12-3 
  per 
  cent 
  at 
  1100°. 
  

  

  The 
  temperature 
  was 
  not 
  often 
  carried 
  over 
  1100°, 
  only 
  once 
  

   a 
  temperature 
  of 
  1161*5° 
  was 
  reached, 
  but 
  directly 
  following 
  a 
  

   large 
  error 
  in 
  the 
  ice 
  point 
  (1-30"'"') 
  was 
  recorded, 
  which 
  shows 
  

   1100° 
  to 
  be 
  about 
  the 
  limit 
  of 
  the 
  practical 
  application 
  of 
  

   these 
  bulbs. 
  To 
  be 
  sure, 
  differences 
  in 
  the 
  zero 
  point 
  appear 
  

   below 
  this 
  temperature; 
  these 
  in 
  the 
  calculation 
  are 
  distributed 
  

  

  * 
  These 
  clay 
  covers 
  (made 
  by 
  ITaldenwauger, 
  Charlottenburg), 
  though 
  perfectly 
  

   able 
  to 
  withstand 
  the 
  high 
  temperatures, 
  were 
  and 
  remained 
  of 
  almost 
  the 
  con- 
  

   sistency 
  of 
  chalk 
  and 
  could 
  be 
  filed 
  or 
  sawed 
  at 
  will. 
  

  

  Am. 
  Jour. 
  Sci.— 
  Fourth 
  Series, 
  Vol. 
  Till, 
  No. 
  45.— 
  September, 
  1899. 
  

   13 
  

  

  