﻿382 
  Fenner 
  — 
  Stability 
  Relations 
  of 
  Silica 
  Minerals. 
  

  

  were 
  placed 
  was 
  filled 
  with 
  an 
  oxidizing 
  atmosphere. 
  Thus 
  

   the 
  contaminating 
  effect 
  of 
  a 
  reducing 
  atmosphere 
  upon 
  the 
  

   thermoelement 
  was 
  obviated. 
  The 
  charge 
  consisted 
  of 
  a 
  small 
  

   pinch 
  of 
  cristobalite 
  powder 
  (made 
  from 
  specially 
  purified 
  

   quartz) 
  wrapped 
  in 
  platinum 
  foil 
  and 
  attached 
  by 
  wires 
  to 
  

   a 
  standard 
  Pt-Pt 
  Rh 
  thermoelement 
  in 
  direct 
  contact 
  with 
  the 
  

   junction. 
  Any 
  desired 
  temperature 
  could 
  be 
  attained 
  in 
  a 
  few 
  

   minutes 
  and 
  kept 
  nearly 
  constant 
  for 
  as 
  long 
  a 
  time 
  as 
  wished. 
  

   After 
  exposure 
  to 
  the 
  desired 
  temperature 
  the 
  thermoelement 
  

   and 
  attached 
  charge 
  were 
  withdrawn 
  and 
  the 
  powder 
  was 
  

   examined 
  under 
  the 
  microscope 
  for 
  indications 
  of 
  the 
  forma- 
  

   tion 
  of 
  glass. 
  

  

  The 
  reluctance 
  in 
  fusion 
  was 
  so 
  great 
  that 
  length 
  of 
  heating 
  

   was 
  found 
  to 
  be 
  an 
  important 
  factor. 
  At 
  first 
  a 
  period 
  of 
  

   fifteen 
  minutes 
  was 
  supposed 
  to 
  be 
  ample 
  and 
  a 
  temperature 
  

   between 
  1680° 
  and 
  1690° 
  was 
  determined 
  upon 
  as 
  that 
  of 
  

   fusion. 
  Then 
  the 
  effect 
  of 
  longer 
  heating 
  was 
  tried 
  and 
  the 
  

   previous 
  determination 
  was 
  shown 
  to 
  be 
  too 
  high. 
  In 
  all 
  the 
  

   later 
  experiments 
  heating 
  was 
  prolonged 
  for 
  half 
  an 
  hour 
  and 
  

   the 
  lowest 
  temperature 
  at 
  which 
  a 
  minute 
  quantity 
  of 
  glass 
  

   could 
  be 
  detected 
  was 
  considered 
  to 
  be 
  the 
  temperature 
  of 
  

   fusion. 
  At 
  a 
  temperature 
  of 
  1640° 
  (uncorrected) 
  the 
  amount 
  

   of 
  glass 
  was 
  small, 
  but 
  could 
  be 
  plainly 
  seen 
  ; 
  at 
  1630° 
  most 
  of 
  

   the 
  grains 
  were 
  unaltered, 
  but 
  a 
  few 
  showed 
  a 
  little 
  glass 
  in 
  

   minute 
  ridges 
  ; 
  at 
  1620° 
  glass 
  was 
  believed 
  to 
  be 
  visible, 
  but 
  

   there 
  was 
  a 
  little 
  uncertainty. 
  

  

  A 
  check 
  on 
  the 
  accuracy 
  of 
  the 
  thermoelement 
  was 
  main- 
  

   tained 
  by 
  two 
  calibrations, 
  one 
  made 
  during 
  the 
  course 
  of 
  the 
  

   experiments 
  and 
  one 
  immediately 
  after 
  their' 
  close. 
  The 
  

   temperature 
  of 
  the 
  furnace 
  was 
  run 
  up 
  to 
  the 
  point 
  at 
  which 
  

   the 
  platinum 
  wire 
  of 
  the 
  element 
  fused, 
  and 
  the 
  observed 
  read- 
  

   ing 
  compared 
  with 
  the 
  correct 
  reading 
  for 
  the 
  melting 
  point 
  

   of 
  platinum. 
  In 
  one 
  case 
  the 
  reading 
  was 
  10° 
  low 
  and 
  in 
  the 
  

   other 
  13° 
  low. 
  It 
  is 
  evident, 
  however, 
  that 
  the 
  correction 
  to 
  

   be 
  applied 
  is 
  not 
  of 
  this 
  magnitude, 
  for 
  the 
  same 
  contamina- 
  

   tion 
  which 
  caused 
  the 
  reading 
  to 
  be 
  too 
  low 
  caused 
  the 
  wire 
  to 
  

   melt 
  below 
  the 
  true 
  melting 
  temperature 
  of 
  platinum. 
  It 
  

   seems 
  probable, 
  on 
  the 
  whole, 
  that 
  the 
  readings 
  were 
  not 
  more 
  

   than 
  5°-6° 
  low, 
  and 
  I 
  have, 
  therefore, 
  placed 
  the 
  fusing 
  point 
  

   of 
  cristobalite 
  at 
  1625° 
  as 
  derived 
  from 
  these 
  experiments. 
  

  

  From 
  theoretical 
  considerations 
  the 
  temperature 
  of 
  fusion 
  of 
  

   quartz 
  should 
  be 
  less 
  than 
  that 
  of 
  cristobalite. 
  There 
  was 
  a 
  

   question, 
  however, 
  whether 
  the 
  velocity 
  of 
  transformation 
  to 
  

   cristobalite 
  might 
  not 
  equal 
  that 
  of 
  fusion 
  and 
  prevent 
  the 
  

   experimental 
  realization 
  of 
  the 
  melting-point 
  of 
  quartz. 
  To 
  try 
  

   this 
  a 
  quantity 
  of 
  ground 
  quartz 
  (Laboratory 
  stock 
  — 
  0*04-0 
  # 
  07 
  

   per 
  cent 
  impurity) 
  was 
  placed 
  in 
  a 
  platinum 
  bucket 
  suspended 
  

   from 
  a 
  Marquardt 
  tube. 
  A 
  thermoelement 
  was 
  imbedded 
  in 
  

  

  