﻿10 
  W. 
  P. 
  White 
  — 
  Silicate 
  Specific 
  Heats. 
  

  

  heat 
  measured 
  was 
  over 
  8 
  times 
  as 
  great 
  with 
  the 
  load. 
  

   The 
  visible 
  errors 
  of 
  0-5 
  per 
  mille 
  or 
  more, 
  therefore, 
  are 
  

   completely 
  accounted 
  for 
  by 
  causes 
  which 
  give 
  only 
  acci- 
  

   dental 
  errors, 
  and 
  hence 
  do 
  not 
  seem 
  to 
  be 
  associated 
  with 
  

   any 
  concealed 
  or 
  constant 
  error. 
  

  

  On 
  the 
  whole, 
  then, 
  it 
  appears 
  that 
  errors, 
  from 
  all 
  

   causes 
  whatever, 
  greater 
  than 
  3 
  per 
  mille 
  are 
  probably 
  

   exceptional, 
  and 
  this 
  indication 
  comes 
  from 
  an 
  unusually 
  

   extensive 
  combination 
  of 
  checks 
  and 
  tests. 
  

  

  Materials. 
  

  

  Of 
  the 
  materials 
  used, 
  three 
  substances, 
  quartz, 
  micro- 
  

   cline, 
  and 
  albite, 
  were 
  natural, 
  the 
  silica 
  glass 
  was 
  made 
  11 
  

   from 
  quartz 
  of 
  practically 
  the 
  same 
  composition, 
  the 
  

   cristobalite 
  was 
  made 
  by 
  crystallizing 
  the 
  silica 
  glass, 
  the 
  

   rest 
  were 
  synthetic. 
  The 
  making 
  of 
  these 
  synthetic 
  

   forms 
  required 
  in 
  some 
  cases 
  considerable 
  skill 
  and 
  inti- 
  

   mate 
  knowledge 
  of 
  the 
  peculiarities 
  of 
  the 
  substances. 
  

   I 
  am 
  very 
  grateful 
  to 
  several 
  colleagues, 
  namely 
  E. 
  T. 
  

   Allen, 
  N. 
  L. 
  Bowen, 
  J. 
  L. 
  Crenshaw, 
  John 
  Johnston, 
  and 
  

   E. 
  Gr. 
  Zies, 
  for 
  the 
  help 
  they 
  gave 
  in 
  making 
  these 
  prep- 
  

   arations. 
  The 
  solid 
  impurities 
  of 
  the 
  synthetic 
  materials 
  

   are 
  undoubtedly 
  negligible 
  in 
  their 
  effect 
  on 
  the 
  specific 
  

   heat. 
  Probably 
  this 
  is 
  also 
  true 
  of 
  the 
  gaseous 
  impur- 
  

   ities, 
  and 
  since 
  satisfactory 
  facilities 
  for 
  determining 
  

   these 
  are 
  not 
  now 
  at 
  hand, 
  it 
  has 
  seemed 
  best 
  to 
  leave 
  the 
  

   question 
  of 
  such 
  impurity 
  open 
  for 
  the 
  present, 
  meantime 
  

   preserving 
  samples 
  for 
  future 
  testing, 
  if 
  this 
  should 
  ever 
  

   become 
  desirable. 
  The 
  gaseous 
  impurities 
  originally 
  in 
  

   the 
  natural 
  minerals 
  were 
  presumably 
  greater 
  than 
  in 
  the 
  

   synthetic. 
  

  

  Reduction 
  of 
  the 
  Results. 
  

  

  The 
  first 
  results 
  of 
  the 
  present 
  determinations 
  are 
  

   interval 
  12 
  specific 
  heats, 
  i. 
  e., 
  specific 
  heats 
  over 
  considera- 
  

   ble 
  temperature 
  intervals, 
  and 
  as 
  such 
  are 
  similar 
  to 
  most 
  

   of 
  the 
  specific 
  heat 
  determinations 
  made 
  by 
  others. 
  Such 
  

   values 
  are 
  unsatisfactory 
  in 
  two 
  respects: 
  (1) 
  The 
  

  

  "Arthur 
  L. 
  Day 
  and 
  E. 
  S. 
  Shepherd, 
  Quartz 
  Glass, 
  Science, 
  N. 
  S., 
  23, 
  

   670, 
  1906. 
  

  

  "Since 
  it 
  will 
  be 
  necessary 
  in 
  this 
  paper 
  to 
  deal 
  with 
  several 
  kinds 
  of 
  

   mean, 
  it 
  seems 
  desirable 
  to 
  distinguish* 
  by 
  using 
  the 
  unambiguous 
  term 
  

   "interval 
  heat" 
  in 
  this 
  case. 
  

  

  