﻿Posnjak 
  and 
  Merwin 
  — 
  Hydrated 
  Ferric 
  Oxides. 
  331 
  

  

  The 
  temperatures 
  of 
  the 
  dehydration 
  intervals 
  are 
  con- 
  

   siderably 
  lower 
  than 
  the 
  corresponding 
  decomposition 
  

   temperatures 
  obtained 
  by 
  the 
  rapid 
  heating. 
  This 
  dif- 
  

   ference 
  is 
  probably 
  caused 
  by 
  the 
  slowness 
  of 
  the 
  rate 
  of 
  

   reaction. 
  In 
  a 
  comparative 
  determination 
  of 
  the 
  decom- 
  

   position 
  temperatures 
  of 
  several 
  specimens, 
  other 
  factors, 
  

   such 
  as 
  physical 
  structure, 
  size 
  of 
  grains, 
  etc., 
  will 
  also 
  

   prove 
  of 
  considerable 
  influence, 
  and 
  it 
  seems 
  certain, 
  

   •therefore, 
  that 
  for 
  purposes 
  of 
  identification 
  the 
  determi- 
  

   nation 
  of 
  decomposition 
  temperatures 
  is 
  wholly 
  unreli- 
  

   able. 
  

  

  As 
  is 
  seen 
  from 
  the 
  figures, 
  the 
  dehydration 
  curves 
  

   consist 
  of 
  three 
  distinct 
  parts. 
  The 
  middle 
  portion 
  is 
  

   formed 
  by 
  loss 
  of 
  water 
  due 
  to 
  decomposition 
  of 
  the 
  sub- 
  

   stance. 
  The 
  upper 
  and 
  lower 
  portions 
  of 
  the 
  curves 
  show 
  

   the 
  losses 
  in 
  water 
  not 
  combined 
  in 
  a 
  definite 
  ratio. 
  This 
  

   is 
  in 
  all 
  probability 
  adsorbed 
  and 
  capillary 
  water 
  and 
  the 
  

   amount 
  held 
  will 
  depend 
  largely 
  on 
  the 
  structure 
  of 
  the 
  

   material. 
  Comparison 
  of 
  curves 
  for 
  coarsely 
  crystallized 
  

   material 
  with 
  those 
  for 
  the 
  more 
  fibrous 
  specimens 
  

   strongly 
  supports 
  this 
  view. 
  It 
  is 
  also 
  confirmed 
  by 
  

   rehydration 
  experiments. 
  

  

  The 
  results 
  of 
  rehydration 
  are 
  indicated 
  by 
  thin 
  lines 
  

   in 
  a 
  number 
  of 
  curves 
  given. 
  These 
  experiments 
  were 
  

   carried 
  out 
  as 
  follows 
  : 
  after 
  the 
  first 
  dehydration 
  up 
  to 
  

   the 
  highest 
  temperature 
  the 
  substance 
  was 
  left 
  to 
  come 
  

   to 
  constancy 
  in 
  the 
  room 
  ; 
  38 
  then 
  it 
  was 
  dehydrated 
  in 
  the 
  

   same 
  manner 
  as 
  before. 
  The 
  amount 
  of 
  water 
  taken 
  up 
  

   in 
  the 
  room 
  by 
  the 
  different 
  samples 
  was 
  roughly 
  equal 
  

   to 
  the 
  amount 
  lost 
  in 
  the 
  first 
  portion 
  of 
  the 
  curves. 
  

   Apparently 
  dehydration 
  did 
  not 
  appreciably 
  change 
  their 
  

   original 
  gross 
  structure 
  : 
  thus 
  the 
  formerly 
  crystallized 
  

   goethite 
  took 
  up 
  very 
  little, 
  while 
  the 
  originally 
  fibrous 
  

   and 
  especially 
  the 
  ' 
  ' 
  amorphous 
  ' 
  ' 
  samples 
  took 
  up 
  cor- 
  

   respondingly 
  larger 
  amounts. 
  

  

  The 
  second 
  dehydration 
  gave 
  perfectly 
  smooth 
  curves, 
  

   which 
  in 
  the 
  beginning 
  approximately 
  parallel 
  the 
  first 
  

   portions 
  of 
  the 
  corresponding 
  original 
  dehydration 
  curves 
  

   and 
  later 
  are 
  practically 
  identical 
  with 
  the 
  third 
  portion. 
  

   The 
  slight 
  deviation 
  is 
  probably 
  accounted 
  for 
  by 
  shrink- 
  

   age. 
  39 
  These 
  experiments 
  strongly 
  suggest 
  that 
  a 
  single 
  

   cause 
  must 
  be 
  responsible 
  for 
  the 
  character 
  of 
  the 
  first 
  

  

  38 
  The 
  humidity 
  of 
  the 
  air 
  was 
  taken 
  into 
  account 
  and 
  the 
  data 
  refer 
  to 
  

   a 
  relative 
  humidity 
  of 
  approximately 
  40. 
  

  

  39 
  Silica 
  probably 
  originally 
  holds 
  some 
  of 
  the 
  water, 
  which 
  after 
  heating 
  

   is 
  not 
  so 
  readily 
  taken 
  up 
  again. 
  

  

  