﻿360 
  Fenner 
  — 
  Stability 
  Relations 
  of 
  Silica 
  Minerals. 
  

  

  ment 
  of 
  the 
  internal 
  structure 
  befalls, 
  attended 
  by 
  slight 
  

   changes 
  in 
  the 
  optical 
  properties, 
  and 
  the 
  transformation 
  occurs 
  

   promptly 
  and 
  at 
  a 
  definite 
  temperature 
  when 
  the 
  mineral 
  is 
  

   heated. 
  On 
  cooling, 
  reversion 
  follows 
  with 
  similar 
  promptness, 
  

   usually 
  not 
  at 
  exactly 
  the 
  same 
  temperature, 
  but 
  at 
  only 
  a 
  few 
  

   degrees 
  lower. 
  The 
  fact 
  that 
  each 
  of 
  the 
  minerals 
  in 
  question 
  

   exhibited 
  a 
  phenomenon 
  of 
  this 
  kind 
  has 
  been 
  known 
  for 
  some 
  

   time, 
  but 
  further 
  investigation 
  has 
  brought 
  out 
  some 
  new 
  facts, 
  

   which 
  suggest 
  an 
  explanation 
  of 
  the 
  radical 
  differences 
  between 
  

   the 
  two 
  types 
  of 
  phenomena 
  and 
  have 
  a 
  bearing 
  upon 
  theories 
  

   of 
  the 
  internal 
  structure 
  of 
  the 
  crystals. 
  

  

  The 
  new 
  phenomena 
  are 
  exhibited 
  most 
  prominently 
  by 
  

   cristobalite 
  ; 
  hence 
  the 
  investigation 
  of 
  this 
  mineral 
  will 
  first 
  

   be 
  taken 
  up, 
  and 
  later 
  the 
  same 
  sort 
  of 
  inquiry 
  extended 
  to 
  

   quartz 
  and 
  tridymite. 
  

  

  The 
  a-(S 
  Cristobalite 
  Inversions. 
  — 
  Mallard* 
  determined 
  that 
  

   upon 
  heating 
  cristobalite 
  crystals 
  the 
  birefringence 
  disappeared 
  

   abruptly 
  at 
  a 
  certain 
  temperature, 
  to 
  reappear 
  upon 
  cooling. 
  

   Above 
  the 
  transition 
  point 
  the 
  crystals 
  possessed 
  the 
  isotropic 
  

   character 
  consistent 
  with 
  their 
  crystaliographic 
  form. 
  The 
  

   temperature 
  was 
  placed 
  by 
  him 
  at 
  175°. 
  Later, 
  F. 
  E. 
  Wrightf 
  

   made 
  a 
  redetermination 
  by 
  heating 
  in 
  the 
  thermal 
  microscope 
  

   plates 
  of 
  cristobalite 
  cut 
  from 
  spherulites 
  crystallizing 
  in 
  silica 
  

   glass, 
  and 
  found 
  that 
  Mallard's 
  determination 
  was 
  too 
  low. 
  He 
  

   placed 
  the 
  temperature 
  at 
  approximately 
  225°. 
  Recently, 
  

   Endell 
  and 
  Eieke^: 
  have 
  found 
  by 
  dilatometric 
  methods 
  a 
  

   noticeable 
  volume-change 
  taking 
  place 
  at 
  about 
  230°. 
  

  

  In 
  my 
  own 
  work 
  the 
  problem 
  appeared 
  at 
  first 
  to 
  be 
  merely 
  

   the 
  determination 
  of 
  the 
  inversion-point 
  by 
  methods 
  as 
  exact 
  

   as 
  possible. 
  By 
  the 
  application 
  of 
  a 
  very 
  simple 
  device 
  (to 
  be 
  

   described 
  immediately) 
  it 
  was 
  found 
  possible 
  to 
  determine 
  the 
  

   inversion-point 
  in 
  a 
  manner 
  which 
  left 
  nothing 
  to 
  be 
  desired, 
  

   but 
  it 
  was 
  soon 
  observed 
  that 
  the 
  results 
  obtained 
  differed 
  

   from 
  each 
  other 
  most 
  remarkably. 
  Whereas 
  the 
  inversion- 
  

   point 
  of 
  any 
  given 
  preparation 
  determined 
  under 
  varying 
  con- 
  

   ditions 
  gave 
  results 
  which 
  agreed 
  most 
  satisfactorily, 
  another 
  

   preparation 
  might 
  differ 
  by 
  twenty 
  or 
  thirty 
  degrees 
  from 
  the 
  

   first. 
  There 
  was 
  thus 
  revealed 
  a 
  line 
  of 
  investigation 
  which 
  

   demanded 
  attention, 
  namely, 
  the 
  determination 
  of 
  the 
  limits 
  

   within 
  which 
  the 
  temperature 
  of 
  inversion 
  could 
  be 
  made 
  to 
  

   vary, 
  and. 
  the 
  factors 
  influencing 
  the 
  variation; 
  finally, 
  the 
  

   development 
  of 
  an 
  explanation 
  which 
  would 
  be 
  in 
  harmony 
  

   with 
  the 
  observed 
  facts, 
  and 
  the 
  elimination 
  of 
  all 
  other 
  pos- 
  

   sibilities 
  which 
  could 
  be 
  conceived 
  by 
  showing 
  that 
  they 
  were 
  

   inconsistent 
  in 
  one 
  way 
  or 
  another 
  when 
  put 
  to 
  the 
  test. 
  

  

  *E. 
  Mallard, 
  Bull. 
  Soc. 
  Min., 
  xiii, 
  176, 
  1890. 
  

  

  f 
  F. 
  E. 
  Wright, 
  J. 
  Ind. 
  Eng. 
  Chem., 
  iii, 
  4, 
  223, 
  1911. 
  

  

  JK. 
  Endell 
  and 
  E. 
  Rieke, 
  Zs. 
  anorg. 
  Chem., 
  lxxix, 
  239, 
  1912. 
  

  

  