﻿Fenner 
  — 
  Stability 
  Relations 
  of 
  Silica 
  Minerals. 
  359 
  

  

  be 
  about 
  825°. 
  Here 
  the 
  conversion 
  of 
  the 
  whole 
  charge 
  takes 
  

   about 
  three 
  days. 
  A 
  lower 
  working 
  limit 
  to 
  the 
  use 
  of 
  sodic 
  

   tungstate 
  is 
  imposed 
  by 
  its 
  solidification 
  at 
  698°. 
  Although 
  

   quartz 
  is 
  the 
  stable 
  form 
  below 
  870°, 
  either 
  amorphous 
  silica 
  or 
  

   cristobalite 
  first 
  yields 
  tridymite 
  and 
  only 
  after 
  much 
  longer 
  

   heating 
  does 
  quartz 
  appear. 
  

  

  Between 
  870° 
  and 
  1470° 
  tridymite 
  is 
  always 
  formed 
  in 
  a, 
  

   tungstate 
  melt. 
  From 
  1300° 
  upward 
  the 
  reaction 
  is 
  fairly 
  

   rapid. 
  At 
  high 
  temperatures 
  (1400° 
  and 
  upward) 
  an 
  alkaline 
  

   silicate 
  glass 
  may 
  be 
  used 
  as 
  a 
  flux. 
  The 
  best 
  crystals 
  of 
  tridy- 
  

   mite 
  have 
  been 
  obtained 
  in 
  a 
  tungstate 
  melt 
  at 
  1300° 
  or 
  there- 
  

   abouts. 
  Within 
  the 
  tridymite 
  range 
  amorphous 
  silica 
  in 
  fused 
  

   sodic 
  tungstate 
  or 
  alkaline 
  silicate 
  yields 
  at 
  first 
  a 
  mixture 
  of 
  

   cristobalite 
  and 
  tridymite, 
  which 
  later 
  becomes 
  entirely 
  tridy- 
  

   mite. 
  

  

  From 
  1470° 
  upward 
  to 
  the 
  melting-point 
  any 
  form 
  of 
  silica 
  

   heated 
  in 
  a 
  tungstate 
  melt 
  is 
  changed 
  into 
  cristobalite. 
  At 
  

   1500° 
  and 
  upward 
  the 
  reaction 
  is 
  fairly 
  rapid. 
  

  

  At 
  high 
  temperatures 
  quartz, 
  heated 
  without 
  a 
  flux, 
  changes 
  

   to 
  cristobalite 
  even 
  below 
  the 
  tridymite-cristobalite 
  inversion- 
  

   point. 
  The 
  upper 
  limit 
  of 
  this 
  reaction 
  is 
  set 
  by 
  the 
  melting- 
  

   point 
  of 
  cristobalite. 
  The 
  lower 
  limit 
  is 
  uncertain. 
  A 
  prac- 
  

   tical 
  experimental 
  limit 
  is 
  set 
  by 
  the 
  length 
  of 
  time 
  required. 
  

   Fine 
  grinding 
  much 
  increases 
  the 
  rate 
  of 
  reaction, 
  but 
  finely 
  

   ground 
  quartz 
  heated 
  108 
  hours 
  at 
  1250° 
  ± 
  showed 
  only 
  a 
  

   small 
  percentage 
  of 
  inversion. 
  After 
  90 
  hours 
  at 
  1360° 
  db 
  the 
  

   product 
  consisted 
  of 
  about 
  f 
  cristobalite 
  and 
  ^ 
  unchanged 
  

   quartz. 
  At 
  1570° 
  the 
  reaction 
  is 
  nearly 
  complete 
  in 
  an 
  hour. 
  

  

  At 
  1570° 
  tridymite 
  heated 
  without 
  a 
  flux 
  is 
  converted 
  to 
  

   cristobalite. 
  

  

  At 
  any 
  temperature 
  below 
  the 
  melting-point 
  of 
  cristobalite 
  

   down 
  to 
  a 
  limit 
  only 
  conditioned 
  by 
  the 
  length 
  of 
  time 
  

   required, 
  amorphous 
  silica 
  (either 
  glass 
  or 
  precipitated 
  silica), 
  

   heated 
  without 
  a 
  flux, 
  changes 
  to 
  cristobalite. 
  At 
  1030° 
  ± 
  

   precipitated 
  silica 
  appeared 
  completely 
  changed 
  after 
  69 
  hours' 
  

   heating. 
  (The 
  change 
  may 
  have 
  been 
  complete 
  in 
  rather 
  less 
  

   time.) 
  

  

  Low 
  Temperature 
  Inversions. 
  

  

  The 
  inversions 
  so 
  far 
  discussed 
  have 
  been 
  characterized 
  by 
  a 
  

   complete 
  change 
  of 
  crystal 
  form. 
  Under 
  the 
  most 
  favorable 
  

   conditions 
  they 
  take 
  place 
  slowly 
  and 
  with 
  difficulty, 
  and 
  any 
  

   of 
  the 
  species 
  can 
  be 
  exposed 
  to 
  temperatures 
  far 
  exceeding 
  

   the 
  limits 
  of 
  stability 
  without 
  any 
  abrupt 
  change 
  occurring. 
  

   There 
  is 
  another 
  class 
  of 
  inversions 
  however, 
  whose 
  character 
  

   is 
  markedly 
  different. 
  There 
  is 
  no 
  noticeable 
  change 
  in 
  the 
  

   outer 
  form 
  of 
  the 
  mineral 
  in 
  question, 
  but 
  some 
  small 
  rearrange- 
  

  

  