﻿the 
  Iher 
  mo 
  electromotive 
  Force 
  of 
  Nickel 
  and 
  Iron. 
  189 
  

  

  and 
  in 
  order 
  to 
  investigate 
  these 
  changes 
  observations 
  were 
  

   made 
  with 
  the 
  Fe-Cu 
  couple 
  under 
  the 
  following 
  four 
  con- 
  

   ditions 
  : 
  — 
  

  

  (i.) 
  New 
  couple 
  in 
  air-vacuum. 
  

  

  (ii.) 
  Same 
  couple 
  after 
  continued 
  heating 
  in 
  air-vacuum, 
  

   (iii.) 
  New 
  couple 
  in 
  hydrogen 
  at 
  about 
  860 
  mm. 
  pressure, 
  

   (iv.) 
  Same 
  couple 
  after 
  continued 
  heating 
  in 
  hydrogen. 
  

  

  (i.) 
  On 
  heating 
  a 
  new 
  couple 
  rapidly 
  to 
  the 
  highest 
  tem- 
  

   perature 
  required, 
  and 
  then 
  taking 
  observations 
  throughout 
  

   the 
  range 
  as 
  quickly 
  as 
  is 
  consistent 
  with 
  accuracy, 
  Curve 
  V 
  1. 
  

   was 
  obtained 
  (fig. 
  5). 
  

  

  The 
  E.M.F 
  increases 
  almost 
  as 
  a 
  linear 
  function 
  of 
  t 
  up 
  to 
  

   about 
  895° 
  C, 
  when 
  an 
  abrupt 
  change 
  of 
  slope 
  occurs 
  and 
  the 
  

   E.M.F. 
  increases 
  more 
  rapidly 
  up 
  to 
  the 
  limit 
  of 
  the 
  

   experiments. 
  

  

  (ii.) 
  Using 
  the 
  same 
  couple 
  after 
  10 
  or 
  15 
  hours' 
  heating, 
  

   the 
  abrupt 
  change 
  of 
  slope 
  previously 
  noticed 
  no 
  longer 
  

   occurs, 
  but 
  as 
  is 
  seen 
  by 
  V 
  II., 
  the 
  curve 
  is 
  continuous 
  over 
  

   the 
  whole 
  range 
  so 
  that 
  above 
  895° 
  the 
  curve 
  consists 
  of 
  two 
  

   branches, 
  the 
  lower 
  one 
  obtained 
  with 
  a 
  once-heated 
  couple 
  

   and 
  the 
  upper 
  one 
  with 
  a 
  couple 
  that 
  had 
  undergone 
  more 
  

   prolonged 
  heating, 
  while 
  below 
  895° 
  the 
  curve 
  is 
  common 
  to 
  

   the 
  two 
  branches, 
  and 
  appears 
  to 
  be 
  independent 
  of 
  the 
  

   length 
  of 
  time 
  the 
  couple 
  has 
  been 
  heated. 
  Measurements 
  

   were 
  then 
  made 
  with 
  continuously 
  rising 
  and 
  continuously 
  

   falling 
  temperatures 
  to 
  find 
  if 
  hysteresis 
  occurred 
  above 
  890°. 
  

   Evidence 
  was 
  obtained 
  of 
  small 
  effects 
  of 
  this 
  nature, 
  slightly 
  

   greater 
  values 
  of 
  E.M.F. 
  being 
  obtained 
  with 
  falling 
  tem- 
  

   perature, 
  though 
  any 
  exact 
  estimation 
  of 
  the 
  phenomenon 
  was 
  

   made 
  somewhat 
  difficult 
  by 
  the 
  change 
  which 
  was 
  going 
  on 
  

   in 
  the 
  value 
  of 
  the 
  E.M.F. 
  owing 
  to 
  continued 
  heating. 
  

  

  " 
  Lag 
  " 
  errors 
  due 
  to 
  a 
  temperature 
  which 
  was 
  changing 
  

   too 
  rapidly 
  would 
  not 
  account 
  for 
  these 
  effects, 
  for 
  such 
  

   errors 
  would 
  affect 
  the 
  result 
  in 
  the 
  opposite 
  direction 
  — 
  for 
  

   instance, 
  with 
  a 
  too 
  rapidly 
  falling 
  temperature 
  the 
  curve 
  

   would 
  be 
  shifted 
  upwards 
  owing 
  to 
  " 
  lag 
  ? 
  ' 
  in 
  the 
  recording 
  

   of 
  the 
  temperature. 
  

  

  The 
  variation 
  of 
  E.M.F. 
  above 
  900° 
  is 
  about 
  20 
  microvolts 
  

   per 
  degree 
  for 
  the 
  freshly 
  heated 
  couple. 
  Probably 
  the 
  

   effects 
  of 
  continuous 
  heating 
  are 
  going 
  on 
  the 
  whole 
  time 
  in 
  

   Curve 
  VI.: 
  hence 
  the 
  rapid, 
  though 
  continuous 
  change 
  of 
  

   slope. 
  At 
  895° 
  a 
  change, 
  most 
  likely 
  chemical, 
  occurs 
  in 
  the 
  

   Fe, 
  and 
  below 
  this 
  temperature 
  the 
  couple 
  is 
  in 
  a 
  new 
  state, 
  

   and 
  secular 
  changes 
  go 
  on 
  more 
  slowly, 
  if 
  at 
  all. 
  

  

  