﻿W. 
  P. 
  White 
  — 
  Silicate 
  Specific 
  Heats. 
  39 
  

  

  non-equilibrium 
  ratio 
  of 
  A 
  and 
  B 
  in 
  the 
  uninverted 
  

   substance, 
  and 
  (4) 
  occurs 
  because 
  this 
  causes 
  a 
  solu- 
  

   bility-equilibrium 
  line 
  NP 
  to 
  be 
  cut 
  in 
  an 
  unusual 
  place. 
  

   At 
  any 
  rate 
  something 
  of 
  this 
  sort 
  must 
  be 
  the 
  case 
  if 
  

   cristobalite 
  is 
  to 
  bring 
  to 
  the 
  theory 
  anything 
  beyond 
  a 
  

   trivial 
  assistance. 
  But 
  the 
  minute 
  we 
  examine 
  the 
  con- 
  

   sequences 
  of 
  thus 
  fitting 
  Smits 
  ' 
  theory 
  to 
  cristobalite, 
  we 
  

   find 
  them 
  to 
  be 
  really 
  extraordinary. 
  For 
  we 
  must 
  then 
  

   suppose: 
  (1) 
  That 
  the 
  change 
  from 
  A 
  to 
  B, 
  whose 
  

   progress 
  is 
  measured 
  only 
  in 
  hours 
  at 
  temperatures 
  

   nearly 
  high 
  enough 
  to 
  melt 
  the 
  substance, 
  goes 
  with 
  

   almost 
  instantaneous 
  rapidity 
  at 
  a 
  low 
  temperature 
  when 
  

   combined 
  with 
  the 
  production 
  of 
  a 
  new 
  crystalline 
  form 
  ; 
  

   and 
  not 
  only 
  this, 
  but 
  (2) 
  that 
  the 
  ratio 
  of 
  the 
  A 
  and 
  B 
  

   components 
  before 
  inversion, 
  which 
  is 
  far 
  from 
  an 
  equi- 
  

   librium 
  ratio, 
  is 
  nevertheless 
  restored 
  after 
  the 
  complete 
  

   shaking 
  up 
  of 
  the 
  A-B 
  composition 
  involved 
  in 
  the 
  inver- 
  

   sion. 
  Or, 
  to 
  express 
  the 
  matter 
  in 
  terms 
  of 
  Smits' 
  

   theory, 
  the 
  composition, 
  for 
  a 
  very 
  sluggish 
  change 
  of 
  

   chemical 
  equilibrium, 
  runs 
  sharply 
  along 
  a 
  non-equilib- 
  

   rium 
  line, 
  as 
  W 
  ? 
  X 
  2 
  (fig. 
  4), 
  both 
  going 
  and 
  coming, 
  quite 
  

   contrary 
  to 
  the 
  behavior 
  of 
  the 
  analogous 
  liquids 
  in 
  such 
  

   a 
  case. 
  This 
  occurrence 
  is 
  not 
  mathematically 
  impos- 
  

   sible, 
  but 
  the 
  necessity 
  of 
  assuming 
  it 
  is 
  anything 
  but 
  a 
  

   point 
  in 
  favor 
  of 
  an 
  unproved 
  theory. 
  

  

  The 
  fact 
  that 
  the 
  expansion 
  and 
  specific 
  heats 
  of 
  solids 
  

   are 
  very 
  much 
  less 
  anomalous 
  and 
  irregular 
  than 
  those 
  

   of 
  liquids 
  seems 
  to 
  show 
  that 
  molecular 
  changes 
  are 
  at 
  

   least 
  of 
  a 
  different 
  order 
  of 
  magnitude 
  in 
  solids. 
  To 
  the 
  

   same 
  effect 
  is 
  Bridgman's 
  demonstration 
  50 
  that 
  nuclei 
  

   will 
  not 
  form 
  and 
  the 
  reaction 
  from 
  existing 
  nuclei 
  will 
  

   not 
  even 
  run 
  appreciably 
  until 
  the 
  difference 
  in 
  the 
  free 
  

   energy 
  of 
  the 
  two 
  forms 
  exceeds 
  a 
  certain 
  threshold 
  

   value. 
  

  

  Bridsman, 
  51 
  who 
  has 
  investigated 
  numerous 
  inversions 
  

   over 
  wide 
  pressure 
  intervals, 
  finds 
  the 
  greatest 
  diversity 
  

   among 
  them. 
  The 
  hi^rh 
  temperature 
  form 
  has 
  in 
  some 
  

   cases 
  a 
  greater, 
  in 
  others 
  a 
  less 
  volume 
  ; 
  in 
  some 
  cases 
  the 
  

   higher, 
  in 
  others 
  the 
  lower 
  specific 
  heat 
  ; 
  the 
  denser 
  form 
  

   often 
  has 
  a 
  smaller 
  cohesion, 
  etc. 
  There 
  is 
  therefore 
  no 
  

   special 
  significance 
  in 
  the 
  fact 
  that 
  the 
  specific 
  heat 
  of 
  

  

  50 
  P. 
  W. 
  Bridgman, 
  The 
  Velocity 
  of 
  Polymorphic 
  Changes 
  between 
  Solids, 
  

   Proc. 
  Am. 
  Acad., 
  52, 
  p. 
  86, 
  1916. 
  

   61 
  Ibid., 
  p. 
  172. 
  

  

  