﻿the 
  Thermoeleetromotive 
  Force 
  of 
  Nickel 
  and 
  Iron. 
  179 
  

  

  thermo-couple 
  was 
  balanced 
  against 
  that 
  due 
  to 
  two 
  accumu- 
  

   lators 
  M. 
  Before 
  each 
  reading 
  of 
  the 
  E.M.F. 
  of 
  a 
  thermo- 
  

   couple 
  the 
  standard 
  cadmium 
  cell 
  C* 
  was 
  balanced 
  on 
  a 
  

   definite 
  resistance 
  AB 
  o£ 
  the 
  box 
  by 
  adjusting 
  r, 
  fine 
  ad- 
  

   justment 
  being 
  made 
  by 
  the 
  rheostat. 
  Alterations 
  in 
  the 
  

   E.M.F. 
  of 
  the 
  cadmium 
  cell 
  with 
  temperature 
  were 
  too 
  small 
  

   to 
  affect 
  the 
  observations 
  to 
  1 
  part 
  in 
  1000. 
  Since 
  the 
  iron- 
  

   copper 
  couple 
  gives 
  an 
  E.M.F. 
  about 
  10 
  times 
  smaller 
  in 
  

   places 
  than 
  the 
  nickel-copper, 
  a 
  shunt 
  equivalent 
  to 
  1/9 
  

   of 
  the 
  bridge-wire 
  resistance 
  was 
  constructed, 
  and 
  arranged 
  

   so 
  that 
  it 
  could 
  be 
  inserted 
  at 
  will 
  between 
  the 
  ends 
  of 
  the 
  

   bridge-wire. 
  

  

  Fig. 
  1. 
  

  

  Cd. 
  Cell 
  

  

  

  

  Cold 
  Junction 
  

  

  Reader 
  

  

  /OOO 
  ohms 
  (™>minal)Q 
  

  

  Pe 
  ten 
  tiom 
  eter 
  

  

  Rheostat 
  

  

  Zero 
  P 
  

  

  M 
  

  

  The 
  bridge-wire 
  was 
  calibrated 
  by 
  the 
  Carey 
  Foster 
  method. 
  

   A 
  galvanometer 
  of 
  the 
  Wiedemann 
  pattern 
  was 
  used 
  with 
  

   a 
  pair 
  of 
  1 
  ohm 
  coils, 
  and 
  with 
  the 
  usual 
  scale 
  and 
  telescope 
  

   arrangement 
  for 
  reading 
  deflexions. 
  Thermoelectric 
  effects 
  

   at 
  the 
  slide-wire 
  contact 
  were 
  avoided 
  by 
  always 
  depressing 
  

   the 
  key 
  with 
  a 
  glass 
  cap. 
  With 
  the 
  above 
  arrangement 
  

   of 
  apparatus, 
  the 
  position 
  of 
  the 
  key 
  on 
  the 
  potentiometer- 
  

   scale 
  could 
  be 
  adjusted 
  in 
  the 
  case 
  of 
  copper-nickel 
  junction 
  

   to 
  the 
  fifth 
  of 
  a 
  millimetre, 
  in 
  the 
  case 
  of 
  copper-iron 
  junction 
  

   to 
  half 
  a 
  millimetre. 
  Readings 
  of 
  E.M.F. 
  of 
  copper-nickel 
  

   couples 
  were 
  accurate 
  to 
  1*8 
  microvolts, 
  while 
  those 
  of 
  copper- 
  

   iron 
  couples 
  were 
  accurate 
  to 
  less 
  than 
  1 
  microvolt 
  at 
  

   moderate 
  temperatures. 
  

  

  * 
  E.M.F. 
  10J9 
  millivolts. 
  

  

  N2 
  

  

  