﻿W. 
  P. 
  White 
  — 
  Specific 
  Heat 
  Determination. 
  57 
  

  

  container, 
  and 
  since 
  they 
  do 
  not 
  vary 
  with 
  anything 
  else, 
  

   and 
  are 
  too 
  large 
  to 
  be 
  the 
  effect 
  of 
  external 
  conditions, 
  

   the 
  loss 
  from 
  the 
  container 
  in 
  dropping 
  seems 
  almost 
  

   the 
  only 
  possible 
  cause 
  of 
  most 
  of 
  the 
  accidental 
  errors. 
  

  

  The 
  simple 
  dropping 
  method 
  chosen 
  thus 
  appears 
  to 
  

   have 
  been 
  relatively 
  unsatisfactory, 
  though 
  the 
  experi- 
  

   ence 
  of 
  the 
  present 
  work 
  seems 
  to 
  have 
  been 
  necessary 
  

   to 
  prove 
  this, 
  since 
  the 
  preliminary 
  data 
  gave 
  a 
  different 
  

   indication. 
  The 
  heat 
  given 
  out 
  by 
  the 
  empty 
  container 
  

   was 
  so 
  near 
  that 
  of 
  platinum 
  itself 
  as 
  to 
  indicate 
  that 
  the 
  

   whole 
  drop 
  loss 
  was 
  under 
  3 
  per 
  mille 
  18 
  of 
  the 
  total 
  heat 
  

   of 
  an 
  average 
  charge. 
  Moreover, 
  the 
  steam 
  produced 
  in 
  

   a 
  drop 
  was 
  usually 
  almost 
  imperceptible. 
  It 
  was 
  there- 
  

   fore 
  reasonable 
  to 
  expect 
  that 
  the 
  irregularity, 
  which 
  

   with 
  the 
  method 
  used 
  is 
  evidently 
  the 
  only 
  drop 
  error, 
  

   would 
  always 
  be 
  well 
  under 
  1 
  per 
  mille. 
  

  

  But 
  though 
  the 
  drop 
  errors 
  were 
  unexpected 
  and 
  may 
  

   be 
  larger 
  than 
  some 
  other 
  methods 
  might 
  give, 
  they 
  are 
  

   really 
  small 
  in 
  consideration 
  of 
  the 
  precision 
  of 
  most 
  

   work 
  in 
  this 
  field. 
  They 
  are 
  large 
  only 
  in 
  comparison 
  

   with 
  the 
  performance 
  secured 
  in 
  the 
  furnaces 
  and 
  cal- 
  

   orimeter. 
  

  

  Variation 
  in 
  the 
  heat 
  radiated 
  to 
  the 
  calorimeter 
  from 
  

   the 
  furnace 
  evidently 
  may 
  contribute 
  to 
  the 
  variation 
  in 
  

   the 
  observed 
  drop 
  losses. 
  The 
  opened 
  furnace 
  was 
  con- 
  

   stant 
  enough 
  for 
  this 
  radiated 
  heat 
  to 
  be 
  easily 
  meas- 
  

   ured. 
  19 
  It 
  was 
  about 
  1 
  calory 
  per 
  second 
  at 
  1100° 
  ; 
  0-3 
  

   calory 
  at 
  700°. 
  The 
  total 
  time 
  of 
  exposure 
  was 
  probably 
  

   never 
  over 
  a 
  second, 
  but 
  a 
  variation 
  of 
  all 
  that 
  would 
  have 
  

   produced 
  scarcely 
  1/6 
  of 
  the 
  average 
  heat 
  variation 
  

   observed. 
  Hence 
  the 
  error 
  was 
  due 
  to 
  the 
  losses, 
  and 
  

   thus 
  almost 
  certainly 
  to 
  irregularity 
  in 
  the 
  splash 
  — 
  most 
  

   of 
  it, 
  probably, 
  through 
  the 
  resulting 
  effect 
  on 
  the 
  pro- 
  

   duction 
  .of 
  steam. 
  The 
  radiation 
  time 
  error 
  is 
  clearly 
  

   quite 
  accidental. 
  It 
  can 
  be 
  eliminated 
  by 
  an 
  automatic 
  

   shutter, 
  and 
  this 
  might 
  possibly 
  be 
  desirable 
  in 
  more 
  

   accurate 
  work. 
  

  

  Plato 
  has 
  shown 
  how 
  to 
  prevent 
  the 
  escape 
  of 
  steam 
  20 
  

   by 
  an 
  arrangement 
  subsequently 
  described, 
  without 
  

  

  18 
  Put 
  at 
  6 
  per 
  mille 
  in 
  1909 
  for 
  a 
  very 
  wide 
  container, 
  giving 
  a 
  large 
  

   splash. 
  "Specific 
  Heat 
  of 
  Silicates, 
  etc.," 
  op. 
  cit., 
  p. 
  336. 
  

  

  19 
  The 
  change 
  of 
  a 
  thermoelement 
  in 
  an 
  empty 
  platinum 
  crucible 
  was 
  only 
  

   20° 
  in 
  the 
  first 
  2 
  minutes 
  after 
  opening 
  at 
  1100°. 
  

  

  20 
  W. 
  Plato, 
  Erstarrungserscheinungen 
  an 
  Anorganischen 
  Salzen 
  und 
  

   Salzgemischen, 
  I, 
  Zs. 
  Phys. 
  Chem., 
  55, 
  735, 
  1906. 
  

  

  