﻿48 
  W. 
  P. 
  White 
  — 
  Specific 
  Heat 
  Determination. 
  

  

  that 
  actual 
  experiment 
  indicated 
  no 
  appreciable 
  irregu- 
  

   larity 
  within 
  the 
  used 
  portions 
  of 
  the 
  chamber. 
  Since 
  

   he 
  treated 
  as 
  negligible 
  discrepancies 
  of 
  1 
  per 
  cent, 
  7 
  his 
  

   argument, 
  while 
  a 
  sufficient 
  justification 
  of 
  his 
  practice, 
  is 
  

   not 
  inconsistent 
  with 
  my 
  own 
  observation 
  8 
  that 
  in 
  work- 
  

   ing 
  to 
  1 
  per 
  thousand 
  appreciable 
  error 
  may 
  result 
  if 
  

   the 
  ends 
  of 
  the 
  chamber 
  are 
  not 
  kept 
  near 
  100°. 
  In 
  the 
  

   1910 
  heater 
  steam 
  actually 
  flowed 
  over 
  the 
  top 
  of 
  the 
  

   chamber, 
  and 
  the 
  bottom 
  had 
  several 
  cork 
  partitions. 
  

   Cup-shaped 
  copper 
  partitions 
  are 
  much 
  better 
  than 
  cork, 
  

   and 
  have 
  proved 
  perfectly 
  satisfactory 
  in 
  the 
  top 
  of 
  the 
  

   modified 
  Eegnault 
  heater 
  also. 
  A 
  pair 
  was 
  far 
  more 
  

   than 
  sufficient, 
  since 
  even 
  the 
  outer 
  one 
  was 
  only 
  0-05° 
  

   from 
  the 
  steam 
  temperature. 
  The 
  1910 
  steam 
  heater 
  

   also 
  was 
  improved 
  by 
  attaching 
  a 
  water 
  condenser, 
  and 
  

   substituting 
  a 
  specially 
  made 
  electric 
  heating 
  plate 
  for 
  

   the 
  flame 
  first 
  used. 
  These 
  changes, 
  besides 
  saving 
  

   trouble, 
  promoted 
  speed; 
  the 
  electric 
  heating 
  could 
  

   safely 
  be 
  set 
  so 
  as 
  to 
  keep 
  the 
  water 
  just 
  below 
  boiling 
  

   all 
  night 
  (without 
  any 
  need 
  of 
  running 
  the 
  condenser) 
  

   ready 
  to 
  start 
  operations 
  almost 
  instantly 
  in 
  the 
  

   morning. 
  

  

  4. 
  Time 
  of 
  Heating 
  Charges. 
  — 
  The 
  time 
  required 
  to 
  

   bring 
  any 
  charge 
  to 
  uniform 
  temperature 
  of 
  course 
  

   varies 
  according 
  to 
  fineness 
  of 
  grain, 
  heat 
  conductivity, 
  

   volume 
  specific 
  heat, 
  and 
  especially 
  the 
  space 
  between 
  

   the 
  container 
  and 
  the 
  wall 
  of 
  the 
  heating 
  chamber. 
  It 
  is 
  

   not 
  altogether 
  easy 
  to 
  determine, 
  in 
  advance 
  of 
  experi- 
  

   ment, 
  the 
  resultant 
  of 
  all 
  these 
  factors; 
  so 
  it 
  seems 
  

   desirable 
  to 
  call 
  attention 
  to 
  the 
  probable 
  necessity, 
  in 
  

   many 
  cases, 
  at 
  least, 
  of 
  rather 
  numerous 
  determinations 
  

   of 
  this 
  important 
  quantity. 
  Fortunately 
  the 
  determina- 
  

   tion 
  is 
  very 
  easy 
  to 
  make, 
  by 
  thermoelectric 
  means, 
  as 
  

   described 
  in 
  the 
  previous 
  paper. 
  

  

  II. 
  Transferring 
  the 
  Charge 
  to 
  the 
  Calorimeter. 
  

  

  1. 
  In 
  the 
  platinum-wound 
  furnace. 
  — 
  At 
  first 
  Harker's 
  

   method 
  of 
  dropping 
  the 
  charge 
  was 
  used, 
  in 
  which 
  a 
  

   platinum 
  suspending 
  wire 
  was 
  fused 
  by 
  a 
  strong 
  electric 
  

   current. 
  Above 
  1000° 
  great 
  trouble 
  arose 
  from 
  arcing 
  

  

  7 
  Loc. 
  cit., 
  p. 
  265. 
  

  

  "Obtained 
  by 
  direct 
  thermoelectric 
  measurement 
  of 
  the 
  temperature 
  

   differences. 
  

  

  