﻿Holborn 
  da 
  Day 
  — 
  TJiermoelectricity 
  in 
  Certain 
  Metals. 
  305 
  

  

  junction 
  and 
  the 
  electromotive 
  force 
  of 
  any 
  particular 
  metal 
  

   then 
  being 
  measured 
  with 
  respect 
  to 
  the 
  pure 
  platinum 
  

   which 
  with 
  the 
  platin-rhodium 
  formed 
  the 
  standard 
  element 
  

   and 
  indicated 
  the 
  temperature 
  of 
  the 
  hot 
  junction. 
  

  

  The 
  temperature 
  was 
  held 
  as 
  nearly 
  constant 
  as 
  possible 
  at 
  

   intervals 
  of 
  about 
  50° 
  and 
  the 
  electromotive 
  force 
  of 
  the 
  metals 
  

   in 
  various 
  combinations 
  measured 
  ; 
  the 
  values 
  for 
  the 
  exact 
  

   intervals 
  contained 
  in 
  the 
  table 
  being 
  obtained 
  from 
  these 
  by 
  

   graphical 
  interpolation. 
  

  

  The 
  table 
  contains 
  first 
  the 
  coefficients 
  <2^, 
  100 
  h^ 
  and 
  10000 
  c 
  

   of 
  the 
  formula 
  

  

  e 
  =r 
  — 
  a-\-ht 
  + 
  cf 
  

  

  the 
  thermoelectric 
  potential 
  difference 
  {e) 
  being 
  expressed 
  in 
  

   microvolts 
  ; 
  second, 
  the 
  temperature 
  limits 
  t^ 
  and 
  t^, 
  within 
  

   which 
  the 
  formula 
  has 
  been 
  found 
  to 
  apply 
  ; 
  third, 
  the 
  mean 
  

   temperature 
  coefficient 
  a 
  of 
  the 
  electrical 
  resistance 
  between 
  

   0° 
  and 
  100° 
  ; 
  and 
  finally, 
  the 
  differences 
  (calculated 
  — 
  ob- 
  

   served) 
  expressed 
  in 
  degrees 
  between 
  the 
  values 
  obtained 
  from 
  

   the 
  formula 
  and 
  the 
  corresponding 
  observed 
  values 
  of 
  the 
  

   electromotive 
  force. 
  Where 
  the 
  formula 
  no 
  longer 
  applies 
  the 
  

   differences 
  are 
  enclosed 
  in 
  parentheses. 
  

  

  It 
  will 
  be 
  seen 
  that 
  with 
  the 
  exception 
  of 
  palladium, 
  the 
  

   Avenarius 
  formula 
  applies 
  between 
  a 
  definite 
  comparatively 
  

   low 
  temperature 
  and 
  the 
  highest 
  temperature 
  observed, 
  the 
  dif- 
  

   ference 
  between 
  observed 
  and 
  calculated 
  values 
  being 
  in 
  the 
  

   mean 
  less 
  than 
  1°, 
  smaller, 
  in 
  fact, 
  than 
  the 
  differences 
  observ- 
  

   able 
  in 
  the 
  same 
  element 
  under 
  different 
  experimental 
  con- 
  

   ditions. 
  In 
  the 
  case 
  of 
  palladium, 
  however, 
  two 
  distinct 
  

   equations 
  are 
  distinguishable, 
  the 
  one 
  for 
  the 
  lower 
  tempera- 
  

   tures 
  (below 
  400°) 
  and 
  the 
  other 
  for 
  the 
  higher 
  (above 
  600°), 
  

   and 
  between 
  the 
  two 
  lies 
  a 
  domain 
  of 
  200° 
  where 
  neither 
  

   equation 
  applies. 
  Otherwise 
  the 
  agreement 
  between 
  observed 
  

   and 
  calculated 
  values 
  is 
  nearly 
  as 
  good 
  as 
  with 
  the 
  other 
  ele- 
  

   ments 
  except 
  in 
  the 
  case 
  of 
  the 
  alloy 
  90 
  Pd, 
  10 
  Pt, 
  where 
  at 
  

   the 
  lower 
  temperatures 
  the 
  sensitiveness 
  is 
  very 
  small. 
  This 
  

   element 
  in 
  particular 
  is 
  also 
  the 
  only 
  one 
  having 
  a 
  minimum 
  

   within 
  the 
  temperature 
  domain 
  of 
  these 
  observations, 
  the 
  change 
  

   in 
  direction 
  falling 
  within 
  the 
  limits 
  where 
  neither 
  of 
  the 
  equa- 
  

   tions 
  applies. 
  

  

  Regarding 
  the 
  difference 
  between 
  the 
  value 
  of 
  the 
  constants 
  

   a, 
  l 
  and 
  c 
  for 
  various 
  specimens 
  of 
  the 
  same 
  material, 
  no 
  very 
  

   useful 
  conclusions 
  can 
  be 
  drawn 
  from 
  the 
  material 
  at 
  hand. 
  

   In 
  the 
  case 
  of 
  gold 
  the 
  value 
  of 
  a 
  seems 
  to 
  decrease 
  with 
  

  

  * 
  It 
  is 
  evident 
  that 
  when 
  the 
  formula 
  contains 
  a 
  constant 
  term, 
  it 
  cannot 
  apply 
  

   downwards 
  as 
  far 
  as 
  0°. 
  If, 
  as 
  is 
  nearly 
  always 
  the 
  case, 
  this 
  constant 
  is 
  negative 
  

   (for 
  which 
  reason 
  it 
  appears 
  with 
  the 
  negative 
  sign 
  in 
  the 
  formula), 
  it 
  indicates 
  

   that 
  the 
  curve 
  is 
  too 
  steep 
  to 
  pass 
  through 
  0°. 
  

  

  