﻿Fenner 
  — 
  Stability 
  Relations 
  of 
  Silica 
  Minerals. 
  363 
  

  

  The 
  results 
  first 
  obtained 
  were 
  much 
  higher 
  than 
  those 
  found 
  

   by 
  Mallard 
  and 
  Wright, 
  and 
  I 
  was 
  inclined 
  to 
  attribute 
  the 
  

   differences 
  to 
  foreign 
  material 
  taken 
  up 
  in 
  solid 
  solution, 
  as 
  

   the 
  cristobalite 
  which 
  was 
  being 
  employed 
  had 
  been 
  ob- 
  

   tained 
  from 
  a 
  tungstate 
  melt. 
  It 
  was 
  found, 
  however, 
  that 
  

   the 
  amount 
  of 
  variation 
  was 
  independent 
  of 
  the 
  quantity 
  of 
  

  

  

  

  

  

  

  

  

  

  

  Fig 
  

  

  . 
  6. 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  & 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  »-~?v 
  

  

  

  

  

  

  . 
  _ 
  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  -\ 
  

  

  

  -\ 
  

  

  

  \" 
  

  

  N 
  

  

  \ 
  

  

  

  — 
  < 
  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  

  

  \ 
  

  

  

  \ 
  

  

  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  \ 
  

  

  

  \ 
  

  

  

  

  ) 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  y 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  ^3S-* 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  \ 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  v 
  

  

  I 
  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  

  Fig. 
  6. 
  Heating 
  and 
  cooling 
  curve 
  of 
  the 
  a- 
  (3 
  cristobalite 
  inversion. 
  Tem- 
  

   perature-differences 
  plotted 
  against 
  temperatures. 
  

  

  Horizontal 
  scale 
  ; 
  1 
  square 
  = 
  500 
  microvolts 
  (approximately 
  10°). 
  

   Yertical 
  scale 
  ; 
  1 
  square 
  = 
  20 
  microvolts 
  (approximately 
  0'4°). 
  

  

  impurity 
  shown 
  by 
  analysis, 
  and 
  the 
  possibility 
  was 
  entirely 
  

   eliminated 
  by 
  using 
  cristobalite 
  obtained 
  by 
  the 
  inversion 
  of 
  

   almost 
  chemically 
  pure 
  quartz 
  at 
  a 
  high 
  temperature 
  without 
  

   a 
  flux. 
  The 
  quartz 
  used 
  for 
  this 
  purpose 
  was 
  specially 
  purified 
  

   material 
  supplied 
  by 
  Baker 
  and 
  Adamson, 
  which 
  was 
  shown 
  

  

  Am. 
  Jour. 
  Sci.— 
  Fourth 
  Series, 
  Vol. 
  XXXVI, 
  No. 
  214. 
  -October, 
  1913. 
  

   24 
  

  

  