﻿38 
  W. 
  P. 
  White 
  — 
  Silicate 
  Specific 
  Heats. 
  

  

  described 
  by 
  Smits; 
  our 
  results 
  also 
  indicated 
  that 
  if 
  

   this 
  had 
  taken 
  place 
  the 
  re-inversion 
  Smits 
  mentions 
  

   would 
  not 
  have, 
  so 
  there 
  seems 
  little 
  likelihood 
  that 
  we 
  

   overlooked 
  such 
  a 
  thing. 
  Tammann's 
  suggestion 
  as 
  to 
  

   an 
  error 
  in 
  Smits 
  ' 
  work 
  from 
  a 
  momentary 
  exposure 
  to 
  

   the 
  air 
  also 
  appears 
  to 
  be 
  quite 
  off 
  the 
  target 
  ; 
  we 
  found 
  

   the 
  material 
  too 
  sluggish 
  to 
  be 
  affected 
  in 
  this 
  way. 
  We 
  

   also 
  note 
  that 
  Smits' 
  1916 
  diagram 
  48 
  for 
  this 
  case 
  is 
  

   inconsistent 
  with 
  the 
  result 
  from 
  225° 
  as 
  given 
  in 
  both 
  

   his 
  record 
  and 
  his 
  diagram 
  49 
  of 
  1910. 
  

  

  (3c) 
  Smits 
  claims 
  that 
  the 
  remarkable 
  behavior 
  of 
  

   cristobalite, 
  discovered 
  by 
  Fenner, 
  is 
  a 
  confirmation 
  of 
  

   the 
  theory 
  we 
  are 
  discussing. 
  Here 
  there 
  is 
  little 
  ques- 
  

   tion 
  as 
  to 
  the 
  facts: 
  cristobalite 
  formed 
  at 
  higher 
  tem- 
  

   peratures, 
  say 
  1500°, 
  inverts 
  from 
  p 
  to 
  a 
  at 
  a 
  temperature 
  

   70° 
  or 
  so 
  above 
  the 
  inversion 
  of 
  cristobalite 
  formed 
  near 
  

   1200°, 
  and 
  this 
  a-p 
  inversion 
  temperature 
  is 
  permanent 
  

   as 
  long 
  as 
  no 
  heating 
  again 
  to 
  high 
  temperature 
  occurs. 
  

   The 
  question 
  is 
  as 
  to 
  the 
  interpretation. 
  It 
  seems 
  clear, 
  

   in 
  the 
  first 
  place, 
  that 
  the 
  mere 
  existence 
  of 
  different 
  

   inversion 
  temperatures 
  in 
  material 
  of 
  the 
  same 
  ultimate 
  

   composition 
  and 
  crystalline 
  form 
  is 
  evidence 
  of 
  a 
  change 
  

   with 
  temperature 
  of 
  internal 
  equilibrium 
  in 
  the 
  solid 
  

   which 
  may 
  well 
  be 
  of 
  the 
  sort 
  for 
  which 
  Smits 
  contends 
  ; 
  

   and 
  this 
  tends 
  to 
  confirm 
  and 
  be 
  confirmed 
  by 
  the 
  evi- 
  

   dence 
  of 
  sulphur 
  and 
  other 
  substances. 
  But 
  proof 
  of 
  

   such 
  changes, 
  though 
  necessary 
  to 
  establish 
  Smits 
  f 
  

   theory, 
  is 
  not 
  sufficient 
  to 
  do 
  so, 
  and 
  Smits 
  almost 
  cer- 
  

   tainly 
  sees 
  in 
  the 
  behavior 
  of 
  cristobalite 
  a 
  more 
  import- 
  

   ant 
  argument 
  for 
  his 
  theory 
  than 
  would 
  come 
  from 
  

   merely 
  demonstrating 
  a 
  change 
  of 
  internal 
  equilibrium. 
  

   He 
  does 
  not 
  trouble 
  himself 
  to 
  state 
  this 
  argument, 
  

   apparently 
  thinking 
  it 
  would 
  be 
  more 
  obvious 
  to 
  his 
  

   readers 
  than 
  one 
  reader 
  has 
  found 
  it 
  to 
  be, 
  but 
  he 
  treats 
  

   cristobalite 
  in 
  connection 
  with 
  mercury 
  iodide, 
  from 
  

   which 
  it 
  seems 
  fair 
  to 
  suppose 
  that 
  he 
  considers 
  cristo- 
  

   balite 
  to 
  present 
  the 
  same 
  hypothetical 
  internal 
  changes 
  

   as 
  he 
  has 
  carefully 
  described 
  for 
  mercury 
  iodide 
  ; 
  that 
  is, 
  

   that 
  it 
  shows 
  an 
  inversion, 
  the 
  p-a 
  inversion 
  which 
  (1) 
  is 
  

   dependent 
  on 
  a 
  considerable 
  change 
  in 
  the 
  ratio 
  of 
  two 
  

   components 
  A 
  and 
  B, 
  which 
  (2) 
  occurs 
  prematurely 
  

   on 
  cooling, 
  which 
  (3) 
  so 
  occurs 
  because 
  there 
  is 
  a 
  

  

  48 
  K. 
  Akart. 
  Wetenschappen, 
  articles 
  cited. 
  

   40 
  K. 
  Akart. 
  Wetenschappen, 
  articles 
  cited. 
  

  

  