﻿326 
  Posnjak 
  and 
  Merwin 
  — 
  Hydrated 
  Ferric 
  Oxides. 
  

  

  According 
  to 
  Le 
  Chatelier 
  34 
  the 
  rate 
  of 
  heating 
  influ- 
  

   ences 
  the 
  decomposition 
  temperature 
  but 
  little. 
  This, 
  

   however, 
  is 
  true 
  only 
  for 
  the 
  comparatively 
  rapid 
  heating 
  

   such 
  as 
  is 
  used 
  in 
  making 
  temperature-time 
  curves. 
  Only 
  

   under 
  such 
  conditions 
  are 
  reproducible 
  results 
  obtained. 
  

   These, 
  however, 
  give 
  much 
  higher 
  temperatures 
  for 
  the 
  

   decomposition 
  than 
  those 
  obtained 
  by 
  heating 
  the 
  sub- 
  

   stance 
  to 
  constant 
  weight 
  at 
  definite 
  temperatures, 
  as 
  will 
  

   be 
  stated 
  later. 
  Such 
  higher 
  decomposition 
  temperatures 
  

   are 
  probably 
  caused 
  by 
  such 
  factors 
  as 
  rate 
  of 
  reaction 
  

   and 
  physical 
  character 
  of 
  substance, 
  including 
  the 
  size 
  of 
  

   grains. 
  

  

  While, 
  for 
  that 
  reason, 
  the 
  decomposition 
  temperature 
  

   found 
  by 
  this 
  method 
  could 
  not 
  be 
  relied 
  upon 
  for 
  the 
  defi- 
  

   nite 
  identification 
  of 
  hydrates, 
  the 
  curves 
  obtained 
  are 
  

   quite 
  valuable 
  for 
  a 
  first 
  orientation 
  and 
  quick 
  informa- 
  

   tion 
  as 
  to 
  the 
  way 
  water 
  is 
  held. 
  Water 
  which 
  enters 
  

   into 
  composition 
  of 
  such 
  substances 
  in 
  a 
  definite 
  ratio, 
  is 
  

   readily 
  distinguished 
  by 
  these 
  curves 
  from 
  dissolved 
  or 
  

   adsorbed 
  water. 
  The 
  former 
  gives 
  a 
  curve 
  with 
  sharp 
  

   heat 
  effects 
  due 
  to 
  rapid 
  decomposition, 
  while 
  the 
  latter, 
  

   passing 
  off 
  gradually, 
  produces 
  fairly 
  smooth 
  curves. 
  

  

  The 
  results 
  for 
  a 
  number 
  of 
  decomposition 
  curves 
  are 
  

   given 
  in 
  fig. 
  2. 
  The 
  substance 
  was 
  heated 
  by 
  means 
  of 
  an 
  

   electric 
  resistance 
  furnace. 
  The 
  curves 
  were 
  taken 
  by 
  

   Eoberts- 
  Austen's 
  differential 
  device 
  35 
  against 
  copper 
  

   oxide 
  as 
  a 
  neutral 
  body 
  and 
  the 
  temperature 
  was 
  meas- 
  

   ured 
  with 
  a 
  platinum-rhodium 
  thermoelement. 
  The 
  num- 
  

   bers 
  of 
  the 
  curves 
  refer 
  to 
  analyses 
  in 
  Tables 
  I 
  and 
  III. 
  

  

  Except 
  turgite, 
  all 
  the 
  specimens 
  of 
  hydrated 
  ferric 
  

   oxide 
  which 
  have 
  been 
  investigated 
  gave 
  a 
  strong 
  heat 
  

   effect. 
  This 
  was 
  due 
  to 
  the 
  decomposition 
  of 
  the 
  sub- 
  

   stance, 
  and 
  no 
  heat 
  effect 
  was 
  produced 
  on 
  reheating. 
  

   Some 
  of 
  these 
  substances 
  also 
  show 
  a 
  slight 
  heat 
  effect 
  

   just 
  above 
  100°. 
  This 
  effect 
  gradually 
  increases 
  with 
  the 
  

   total 
  amount 
  of 
  water 
  held 
  by 
  the 
  hydrated 
  ferric 
  oxides 
  

   and 
  is 
  undoubtedly 
  due 
  to 
  adsorbed 
  water, 
  the 
  greater 
  

   part 
  of 
  which 
  would 
  come 
  off 
  at 
  about 
  this 
  temperature. 
  

   The 
  temperature 
  at 
  which 
  decomposition 
  takes 
  place 
  can 
  

   be 
  reproduced 
  for 
  each 
  sample 
  very 
  closely. 
  However, 
  

   as 
  seen 
  from 
  the 
  curves 
  in 
  fig. 
  2, 
  there 
  is 
  a 
  continuous 
  

   gradation 
  in 
  the 
  decomposition 
  temperatures 
  of 
  the 
  

  

  84 
  Loc. 
  cit. 
  

  

  85 
  See 
  G. 
  K. 
  Burgess, 
  Bull. 
  Bureau 
  of 
  Standards, 
  5, 
  p. 
  210, 
  1908; 
  W. 
  P. 
  

   White, 
  this 
  Jour. 
  18, 
  453, 
  1909. 
  

  

  