﻿Sigh 
  Temperatures, 
  

  

  185 
  

  

  After 
  filling. 
  

  

  Ho 
  

  

  a 
  

  

  t 
  

  

  Ci 
  < 
  

  

  Obs.-Calc 
  

  

  8th 
  day 
  

  

  126-66 
  

  

  0-003678 
  

  

  

  

  

  9th 
  " 
  

  

  

  

  918-3 
  

  

  8435 
  

  

  -4-8 
  

  

  

  

  

  923-9 
  

  

  8502 
  

  

  —5-3 
  

  

  

  

  

  943-6 
  

  

  8704 
  

  

  —3-7 
  

  

  

  

  

  1113-6 
  

  

  10630 
  

  

  -3-5 
  

  

  

  

  

  1117-5 
  

  

  10664 
  

  

  -2-5 
  

  

  

  

  

  Cont. 
  to 
  1215°. 
  

  

  

  

  

  

  

  1066-4 
  

  

  10008 
  

  

  + 
  3-7 
  

  

  

  

  

  1066-8 
  

  

  10018 
  

  

  + 
  3-2 
  

  

  

  

  

  940-9 
  

  

  8622 
  

  

  + 
  1-0 
  

  

  10th 
  " 
  

  

  127-10 
  

  

  0-0036795 
  

  

  

  

  

  Bulb 
  freshly 
  1 
  

  

  ailed. 
  

  

  

  

  

  

  3d 
  day 
  

  

  148-81 
  

  

  

  

  

  

  oth 
  " 
  

  

  148-85 
  

  

  0-003667 
  

  

  

  

  

  8th 
  " 
  

  

  148-88 
  

  

  0-003668 
  

  

  

  

  

  In 
  Table 
  YII 
  are 
  given 
  the 
  results 
  obtained 
  with 
  two 
  of 
  

   these 
  bulbs. 
  One 
  (numbered 
  lY 
  in 
  the 
  table) 
  served 
  for 
  meas- 
  

   urements 
  up 
  to 
  1000° 
  and 
  was 
  intended 
  to 
  furnish 
  a 
  control 
  

   upon 
  the 
  results 
  obtained 
  previously 
  from 
  bulbs 
  glazed 
  on 
  

   both 
  sides. 
  

  

  The 
  values 
  given 
  by 
  the 
  gas 
  thermometer 
  with 
  this 
  bulb 
  are 
  

   all 
  too 
  small 
  when 
  compared 
  with 
  the 
  element 
  T^. 
  It 
  is 
  never- 
  

   theless 
  a 
  matter 
  of 
  doubt 
  whether 
  this 
  difference 
  arises 
  entirely 
  

   from 
  the 
  nature 
  of 
  the 
  bulb 
  ; 
  it 
  may 
  be 
  partly 
  due 
  to 
  the 
  change 
  

   in 
  the 
  volume 
  of 
  the 
  unheated 
  space. 
  

  

  In 
  regard 
  to 
  the 
  changes 
  in 
  the 
  zero 
  point 
  here 
  noticed, 
  it 
  

   should 
  be 
  borne 
  in 
  mind 
  that 
  as 
  the 
  quantity 
  of 
  gas 
  is 
  

   here 
  smaller, 
  a 
  small 
  increase 
  would 
  produce 
  a 
  comparatively 
  

   greater 
  effect 
  upon 
  the 
  pressure 
  at 
  0°. 
  

  

  The 
  other 
  unglazed 
  bulb 
  (numbered 
  Y) 
  was 
  several 
  times 
  

   heated 
  above 
  1200° 
  and 
  observed 
  by 
  rising 
  as 
  well 
  as 
  by 
  falling 
  

   temperature, 
  as 
  a 
  result 
  of 
  which 
  the 
  differences 
  between 
  the 
  

   values 
  obtained 
  and 
  the 
  normal 
  curve 
  are 
  found 
  to 
  change 
  sign 
  

   during 
  the 
  heating, 
  in 
  fact 
  to 
  form 
  a 
  loop 
  about 
  the 
  curve. 
  

   The 
  cause 
  of 
  this 
  phenomenon 
  we 
  were 
  not 
  able 
  to 
  discover, 
  

   for 
  it 
  was 
  not 
  possible, 
  on 
  account 
  of 
  the 
  thin 
  wall 
  of 
  the 
  bulb, 
  

   to 
  heat 
  it 
  above 
  800° 
  under 
  vacuum, 
  and 
  further 
  experiments 
  

   with 
  thick- 
  walled 
  bulbs 
  could 
  not 
  conveniently 
  be 
  undertaken 
  

   at 
  that 
  time. 
  

  

  A 
  series 
  of 
  observations 
  of 
  the 
  ice 
  point 
  and 
  expansion 
  

   coefficient 
  was 
  here 
  also 
  undertaken 
  with 
  a 
  fresh 
  gas 
  filling 
  at 
  

   the 
  close 
  of 
  Table 
  YII, 
  to 
  see 
  that 
  the 
  connections 
  had 
  remained 
  

   tight. 
  

  

  In 
  regard 
  to 
  the 
  cause 
  of 
  the 
  more 
  or 
  less 
  systematic 
  ice 
  

   point 
  variations 
  throughout 
  the 
  observations, 
  it 
  is 
  emphatically 
  

   our 
  opinion 
  that 
  it 
  lies 
  in 
  the 
  porcelain 
  wall 
  itself. 
  The 
  varia- 
  

   tions 
  are 
  always 
  greatest 
  after 
  the 
  first 
  heating, 
  so 
  that 
  quite 
  

  

  