﻿W. 
  P. 
  White 
  — 
  Silicate 
  Specific 
  Heats. 
  21 
  

  

  lower 
  on 
  quartz.) 
  Impurity 
  is, 
  of 
  course, 
  an 
  evident 
  

   possible 
  cause 
  of 
  discrepancy, 
  but 
  it 
  does 
  not 
  seem 
  a 
  very 
  

   probable 
  general 
  cause 
  in 
  quartz, 
  and 
  still 
  less 
  so 
  in 
  silica 
  

   glass. 
  Apparently 
  considerably 
  more 
  accurate 
  work 
  can 
  

   be 
  done, 
  but 
  this 
  is 
  perhaps 
  not 
  worth 
  while 
  for 
  the 
  

   majority 
  of 
  rather 
  impure 
  materials. 
  A. 
  Magnus 
  21 
  has 
  

   determined 
  quartz 
  and 
  silica 
  glass 
  to 
  550°. 
  His 
  results 
  

   to 
  100° 
  are 
  for 
  quartz 
  about 
  6 
  per 
  mille 
  lower 
  than 
  ours, 
  

   for 
  the 
  glass 
  about 
  the 
  same. 
  The 
  difference 
  is 
  not 
  mate- 
  

   rial 
  at 
  present. 
  For 
  the 
  interval 
  17°-550°, 
  however, 
  he 
  

   is 
  12 
  per 
  mille 
  lower 
  for 
  quartz, 
  and 
  22 
  for 
  the 
  glass. 
  

   In 
  reducing 
  his 
  results 
  he 
  takes 
  the 
  interval 
  heat 
  as 
  equal 
  

   to 
  the 
  true 
  heat 
  at 
  the 
  middle 
  temperature 
  of 
  the 
  interval, 
  

   remarking 
  that 
  this 
  causes 
  only 
  a 
  very 
  small 
  error, 
  since 
  

   the 
  course 
  of 
  the 
  specific 
  heat 
  is 
  nearly 
  linear 
  at 
  high 
  

   temperatures. 
  In 
  fact, 
  however, 
  it 
  follows 
  from 
  equa- 
  

   tions 
  4 
  and 
  5 
  that 
  by 
  doing 
  this 
  for 
  the 
  500-degree 
  inter- 
  

   val 
  he 
  introduces 
  gratuitous 
  errors 
  of 
  3 
  per 
  cent 
  or 
  more 
  

   for 
  quartz 
  and 
  the 
  glass, 
  making 
  his 
  results 
  lower. 
  

  

  Conclusions. 
  

  

  1. 
  In 
  General. 
  Considered 
  as 
  constants 
  of 
  nature, 
  

   or 
  as 
  data 
  for 
  geological 
  calculations, 
  the 
  present 
  results 
  

   seem 
  to 
  need 
  no 
  further 
  discussion 
  here. 
  They 
  have, 
  

   however, 
  a 
  bearing 
  on 
  various 
  theoretical 
  questions 
  

   which 
  deserves 
  comment. 
  

  

  The 
  least 
  variable 
  magnitude 
  in 
  connection 
  with 
  

   specific 
  heat 
  is 
  the 
  atomic 
  heat, 
  which 
  is 
  a 
  universal 
  con- 
  

   stant 
  according 
  to 
  the 
  classical 
  form 
  of 
  the 
  kinetic 
  theory. 
  

   The 
  simplest 
  way, 
  apparently, 
  to 
  treat 
  the 
  subject 
  is 
  to 
  

   consider 
  the 
  variations 
  from 
  this 
  constant. 
  The 
  varia- 
  

   tions 
  may 
  be 
  classified 
  as 
  due, 
  in 
  part, 
  to 
  secondary 
  ther- 
  

   mal 
  effects 
  connected 
  with 
  expansion, 
  change 
  of 
  state, 
  or 
  

   chemical 
  action, 
  and 
  in 
  part, 
  to 
  the 
  universal 
  tendency 
  

   toward 
  a 
  gradual 
  development 
  (that 
  is, 
  an 
  increase 
  

   toward 
  a 
  maximum 
  value) 
  of 
  the 
  atomic 
  heat 
  with 
  rising 
  

   temperature, 
  a 
  phenomenon 
  which 
  is 
  at 
  present 
  most 
  fre- 
  

   quently 
  formulated 
  in 
  terms 
  of 
  the 
  quantum 
  hypothesis. 
  

  

  21 
  Messungen 
  spezifische 
  Warmen 
  fester 
  Korper 
  bei 
  hohen 
  Temperaturen, 
  

   Phys. 
  Zs., 
  14, 
  55, 
  1913. 
  

  

  