﻿368 
  Fenner 
  — 
  Stability 
  Relations 
  of 
  Silica 
  Minerals. 
  

  

  the 
  point 
  K'. 
  For 
  this 
  reason, 
  also, 
  only 
  a 
  small 
  portion 
  of 
  the 
  

   compositions 
  represented 
  by 
  AB 
  can 
  be 
  realized, 
  that 
  is, 
  the 
  

   compositions 
  corresponding 
  to 
  very 
  high 
  temperatures. 
  It 
  has 
  

   been 
  found, 
  however, 
  that 
  if 
  material 
  corresponding 
  in 
  com- 
  

   position 
  to 
  I', 
  whose 
  /3-a 
  inversion 
  temperature 
  is 
  determined, 
  

   is 
  placed 
  again 
  in 
  the 
  furnace 
  at 
  the 
  temperature 
  K/, 
  a 
  well- 
  

   marked 
  tendency 
  to 
  change 
  to 
  the 
  composition 
  K' 
  is 
  shown 
  by 
  

   a 
  shift 
  in 
  the 
  inversion-point 
  toward 
  the 
  right. 
  Likewise 
  

   material 
  of 
  the 
  composition 
  K 
  r 
  can 
  have 
  its 
  composition 
  shifted 
  

   toward 
  the 
  left 
  by 
  a 
  second 
  heating 
  at 
  a 
  lower 
  temperature. 
  

  

  In 
  another 
  respect 
  the 
  course 
  of 
  the 
  $-a 
  inversion 
  differs 
  

   from 
  the 
  theoretical 
  course 
  first 
  outlined. 
  Upon 
  cooling 
  

   material 
  of 
  the 
  composition 
  represented 
  by 
  K 
  the 
  transforma- 
  

   tion 
  from 
  the 
  /3 
  to 
  the 
  a 
  form 
  might 
  be 
  supposed 
  to 
  start 
  at 
  

   W 
  and 
  be 
  spread 
  throughout 
  the 
  range 
  of 
  temperature 
  from 
  

   K" 
  to 
  K 
  //7 
  . 
  There 
  is, 
  however, 
  some 
  undercooling, 
  as 
  is 
  often 
  

   found 
  in 
  transformations 
  in 
  solids, 
  and 
  the 
  temperature 
  K"' 
  or 
  

   possibly 
  even 
  a 
  little 
  lower 
  temperature 
  is 
  reached 
  before 
  

   transformation 
  starts 
  and 
  then 
  the 
  whole 
  transformation 
  goes 
  

   off 
  at 
  once. 
  Upon 
  the 
  heating-curve 
  also 
  there 
  is 
  a 
  similar 
  

   delay 
  and 
  the 
  temperature 
  rises 
  to 
  W 
  or 
  above 
  before 
  trans- 
  

   formation 
  occurs. 
  

  

  These 
  variations 
  from 
  the 
  simple 
  diagram 
  are 
  easily 
  explained 
  

   and 
  do 
  not 
  obscure 
  the 
  broad, 
  general 
  fact 
  of 
  great 
  variations 
  

   in 
  the 
  temperatures 
  of 
  inversion 
  produced 
  by 
  previous 
  heat- 
  

   treatment. 
  

  

  In 
  Table 
  I, 
  I 
  have 
  collected 
  the 
  experimental 
  data 
  which 
  

   have 
  a 
  bearing 
  upon 
  the 
  matter. 
  

  

  These 
  tabulated 
  figures 
  bring 
  out 
  in 
  a 
  striking 
  manner 
  the 
  

   general 
  decrease 
  of 
  the 
  a-/3 
  inversion 
  temperature 
  with 
  de- 
  

   crease 
  of 
  the 
  temperature 
  to 
  which 
  the 
  material 
  has 
  been 
  pre- 
  

   viously 
  exposed. 
  The 
  irregularity 
  introduced 
  by 
  the 
  tendency 
  

   of 
  material 
  made 
  from 
  amorphous 
  silica 
  to 
  give 
  higher 
  figures 
  

   than 
  quartz 
  is 
  also 
  quite 
  evident. 
  Cristobalite 
  formed 
  in 
  a 
  

   tungstate 
  melt, 
  it 
  may 
  be 
  noticed, 
  shows 
  similarity 
  to 
  that 
  from 
  

   amorphous 
  silica 
  rather 
  than 
  to 
  that 
  from 
  quartz. 
  

  

  A 
  certain 
  source 
  of 
  irregularity 
  requires 
  a 
  few 
  words 
  of 
  

   explanation 
  ; 
  that 
  is, 
  that 
  two 
  preparations 
  which 
  give 
  closely 
  

   coincident 
  inversion-points 
  on 
  heating 
  may 
  differ 
  a 
  number 
  of 
  

   degrees 
  in 
  the 
  cooling 
  temperature. 
  I 
  think 
  this 
  may 
  be 
  due, 
  

   in 
  part 
  at 
  least, 
  to 
  some 
  slight 
  variation 
  in 
  composition 
  (propor- 
  

   tion 
  of 
  X 
  and 
  yu, 
  molecules) 
  among 
  different 
  crystals 
  in 
  the 
  same 
  

   preparation. 
  Frequently 
  a 
  sample 
  of 
  cristobalite 
  formed 
  by 
  

   the 
  inversion 
  of 
  quartz 
  shows 
  still 
  an 
  occasional 
  grain 
  of 
  un- 
  

   changed 
  quartz, 
  and 
  it 
  seems 
  likely 
  that 
  there 
  should 
  be 
  various 
  

   degrees 
  of 
  transition 
  between 
  such 
  an 
  unchanged 
  remnant 
  and 
  

   crystals 
  whose 
  composition 
  is 
  far 
  advanced 
  toward 
  the 
  equili- 
  

   brium 
  demanded 
  by 
  the 
  temperature 
  to 
  which 
  they 
  have 
  been 
  

  

  