﻿Theo 
  

  

  y 
  of 
  the 
  Thermoelectric 
  Forces. 
  

  

  

  tr 
  a 
  ■ 
  10 
  C 
  

   observed. 
  

  

  Tj3 
  a 
  . 
  10 
  G 
  . 
  

  

  (iy^).io*. 
  

  

  Bi 
  ord 
  

  

  Sbord 
  

  

  Cdord 
  

  

  Zn 
  

  

  +24-5 
  

   4-21-6 
  

   + 
  10-2 
  

   + 
  3 
  6 
  

   4- 
  P8 
  

   4- 
  0-6 
  

   

  

  - 
  0-04 
  

  

  - 
  69 
  

  

  - 
  7-8 
  

   -10-2 
  

  

  4-28-1 
  

   4-15-68 
  

   4- 
  9-8 
  

   4- 
  6-21 
  

  

  

   4- 
  3-59 
  

  

  -272 
  

  

  -284 
  

   -290 
  

   -294 
  

  

  -296 
  

   -297 
  

  

  As 
  

  

  Ou 
  

  

  Pb 
  .... 
  

  

  Sn 
  

  

  Al 
  

  

  Pt 
  

  

  Hs 
  

  

  Sb 
  pure 
  

  

  Bi 
  pure 
  

  

  781 
  

  

  Only 
  the 
  values 
  of 
  commercial 
  bismuth 
  and 
  of 
  mercury 
  

   correspond 
  to 
  absolute 
  measurements, 
  and 
  for 
  these 
  two 
  

   metals 
  the 
  second 
  coefficient 
  /3a 
  of 
  the 
  thermoelectric 
  force 
  

   is 
  unknown. 
  A 
  larger 
  number 
  of 
  absolute 
  determinations 
  

   of 
  the 
  Thomson 
  effect 
  in 
  the 
  same 
  time 
  as 
  the 
  exact 
  measure- 
  

   ments 
  of 
  fta 
  are 
  very 
  necessary. 
  The 
  value 
  T/3« 
  is 
  not 
  much 
  

   greater 
  than 
  the 
  specific 
  heat 
  of 
  electricity, 
  being 
  at 
  any 
  

   rate 
  of 
  the 
  same 
  order 
  of 
  magnitude 
  as 
  <r 
  a 
  . 
  The 
  theoretical 
  

  

  value 
  T/9 
  a 
  - 
  

  

  —- 
  , 
  ^however, 
  is 
  not 
  even 
  of 
  the 
  same 
  order 
  of 
  

   eA 
  

  

  magnitude 
  as 
  the 
  observed 
  value 
  of 
  the 
  specific 
  heat 
  of 
  

   electricity. 
  There 
  is 
  a 
  serious 
  disagreement 
  between 
  the 
  

   theory 
  and 
  the 
  experimental 
  result. 
  

  

  We 
  assumed 
  in 
  paragraph 
  5 
  that 
  a 
  corpuscle, 
  moving 
  from 
  

   x 
  + 
  dx 
  to 
  x 
  or 
  from 
  T 
  4-dT 
  to 
  T, 
  will 
  lose 
  a 
  part 
  of 
  its 
  kinetic 
  

   energy 
  equal 
  to 
  udT 
  which 
  is 
  communicated 
  to 
  the 
  metal 
  in 
  

   the 
  form 
  of 
  heat. 
  This 
  transference 
  of 
  heat 
  would 
  take 
  place 
  

   if 
  the 
  corpuscles 
  were 
  drifting, 
  but 
  under 
  the 
  action 
  of 
  the 
  

   difference 
  of 
  temperature. 
  The 
  heat 
  motion 
  of 
  the 
  corpuscles 
  

   not 
  beino- 
  much 
  affected 
  by 
  electric 
  forces 
  which 
  we 
  are 
  able 
  

   to 
  apply 
  on 
  conductors, 
  the 
  heat 
  communicated 
  by 
  the 
  cor- 
  

   puscles 
  to 
  the 
  metal 
  in 
  the 
  interval 
  of 
  temperature 
  dT 
  as 
  

   consequence 
  of 
  the 
  electric 
  force, 
  is 
  not 
  adT, 
  but 
  adT 
  . 
  0, 
  

   where 
  C 
  denotes 
  a 
  certain 
  constant, 
  smaller 
  than 
  unity. 
  

   Thus 
  we 
  find 
  a 
  to 
  be 
  equal 
  to 
  

  

  -:M'¥)- 
  

  

  From 
  the 
  numbers 
  of 
  the 
  last 
  table 
  we 
  derived 
  the 
  following 
  

  

  