﻿4 
  W. 
  P. 
  White 
  — 
  Silicate 
  Specific 
  Heats. 
  

  

  tested 
  by 
  raising 
  or 
  lowering 
  them 
  in 
  the 
  fnrnace. 
  A 
  

   special 
  fnrnace 
  of 
  small 
  diameter 
  was 
  constrncted 
  to 
  

   assist 
  in 
  this 
  part 
  of 
  the 
  work. 
  The 
  gradient 
  was 
  kept 
  

   as 
  constant 
  as 
  possible 
  in 
  the 
  platinnm-wonnd 
  fnrnace 
  

   by 
  keeping 
  the 
  depth 
  of 
  insertion 
  of 
  the 
  thermoelement 
  

   constant 
  to 
  2 
  mm. 
  or 
  so; 
  in 
  the 
  lead-bath 
  fnrnace 
  the 
  

   thermoelement 
  was 
  not 
  moved 
  at 
  all 
  from 
  day 
  to 
  day. 
  

  

  The 
  standards 
  were 
  nsed 
  only 
  occasionally 
  in 
  the 
  plati- 
  

   nnm-wonnd 
  fnrnace. 
  The 
  working 
  element 
  in 
  that 
  fnr- 
  

   nace 
  was 
  compared 
  with 
  the 
  standards 
  15 
  times 
  dnring 
  

   the 
  ten 
  months 
  in 
  which 
  that 
  fnrnace 
  was 
  in 
  nse. 
  It 
  

   changed 
  less 
  than 
  0-3° 
  at 
  800° 
  in 
  all 
  that 
  time. 
  Two 
  

   working 
  thermoelements, 
  in 
  different 
  parts 
  of 
  the 
  lead- 
  

   bath 
  fnrnace, 
  were 
  safe 
  from 
  change 
  at 
  500° 
  and 
  under, 
  

   and 
  served 
  to 
  check 
  each 
  other 
  and 
  to 
  test 
  the 
  nnif 
  ormity 
  

   of 
  the 
  fnrnace 
  temperature. 
  It 
  seems 
  fairly 
  certain, 
  

   therefore, 
  that 
  in 
  applying 
  the 
  temperatnre 
  scale 
  to 
  the 
  

   present 
  work 
  there 
  was 
  no 
  error 
  as 
  great 
  as 
  0-3° 
  at 
  

   1000°, 
  or 
  0-3 
  per 
  mille. 
  This, 
  of 
  course, 
  is 
  less 
  than 
  the 
  

   other 
  errors 
  in 
  the 
  work 
  and 
  possibly 
  in 
  the 
  high 
  tempera- 
  

   ture 
  scale 
  itself. 
  

  

  The 
  High 
  Temperature 
  Scale. 
  

  

  The 
  Day 
  and 
  Sosman 
  high 
  temperature 
  scale 
  was 
  

   applied 
  in 
  the 
  form 
  presented 
  by 
  Adams' 
  thermoelement 
  

   curve. 
  4 
  

  

  The 
  100-degree 
  Reading. 
  

  

  The 
  temperature, 
  always 
  close 
  to 
  100°, 
  in 
  the 
  steam 
  

   heater, 
  was 
  read 
  by 
  a 
  thermoelement 
  which 
  was 
  calibrated 
  

   in 
  the 
  heater 
  by 
  several 
  times 
  determining 
  with 
  a 
  barom- 
  

   eter 
  the 
  pressure 
  of 
  the 
  steam. 
  Comparison 
  was 
  also 
  

   made 
  with 
  a 
  regular 
  hypsometer, 
  and 
  3 
  thermoelements 
  

   were 
  used. 
  In 
  6 
  standardizations 
  of 
  the 
  best 
  one 
  of 
  these 
  

   the 
  maximum 
  discrepancy 
  was 
  only 
  0-008°. 
  

  

  The 
  Calorimeter 
  Temperature 
  Scale. 
  

  

  The 
  calorimeter 
  temperature 
  reading 
  was 
  based 
  on 
  the 
  

   thermoelement 
  calibration 
  made 
  in 
  1907, 
  and 
  described 
  

   in 
  1910, 
  5 
  but 
  this 
  use 
  of 
  that 
  calibration 
  was 
  merely 
  a 
  

   convenience, 
  for 
  the 
  real 
  standardization 
  was 
  by 
  an 
  elec- 
  

  

  *L. 
  H. 
  Adams, 
  Calibration 
  Tables 
  for 
  Copper-Constantan 
  and 
  Platinum- 
  

   Platinrhodium 
  Thermoelements, 
  J. 
  Am. 
  Chem. 
  Soc., 
  36, 
  72, 
  1914. 
  

  

  B 
  W. 
  P. 
  White, 
  H. 
  C. 
  Dickinson 
  and 
  E. 
  F. 
  Mueller. 
  The 
  Calibration 
  of 
  

   Copper-Constantan 
  Thermoelements, 
  Phys. 
  Eev., 
  31, 
  159, 
  1910. 
  

  

  