﻿t 
  274 
  ] 
  

  

  XXX. 
  On 
  Apparatus 
  for 
  Self-acting 
  Temperature 
  Compensa- 
  

   tion 
  of 
  Standard 
  Cells. 
  By 
  Albert 
  Campbell, 
  B.A* 
  

  

  IN 
  ordinary 
  testing 
  work 
  when 
  using 
  a 
  Clark 
  cell 
  as 
  the 
  

   working 
  standard 
  of 
  difference 
  of 
  potential, 
  it 
  is 
  desirable 
  

   to 
  get 
  rid 
  of 
  the 
  trouble 
  of 
  continually 
  having 
  to 
  make 
  the 
  

   temperature 
  correction. 
  Some 
  time 
  ago 
  (' 
  The 
  Electrician/ 
  

   pp. 
  601-603, 
  September 
  6, 
  1895) 
  I 
  proposed 
  and 
  investigated 
  

   three 
  methods 
  by 
  which 
  a 
  self-acting 
  correction 
  could 
  be 
  

   attained. 
  Quite 
  recently 
  I 
  have 
  constructed 
  the 
  apparatus 
  

   for 
  two 
  of 
  the 
  methods, 
  and 
  I 
  have 
  found 
  it 
  an 
  easy 
  matter 
  

   to 
  reach 
  the 
  accuracy 
  aimed 
  at 
  (viz. 
  maximum 
  error 
  1 
  in 
  

   2000). 
  

  

  The 
  arrangement 
  of 
  the 
  first 
  piece 
  of 
  apparatus 
  is 
  shown 
  

   in 
  fig, 
  1. 
  

  

  Fig. 
  1. 
  

  

  /WW 
  COPPER 
  

  

  MANCANIN 
  

  

  S.C. 
  is 
  the 
  standard 
  cell. 
  Four 
  resistances 
  of 
  values 
  a, 
  Z>, 
  

   na, 
  and 
  pb 
  are 
  connected 
  as 
  in 
  a 
  Wheatstone's 
  bridge 
  with 
  an 
  

   auxiliary 
  cell 
  X, 
  whose 
  voltage 
  must 
  be 
  known 
  within 
  5 
  or 
  6 
  

   per 
  cent. 
  A 
  Leclanche 
  cell 
  is 
  found 
  suitable 
  for 
  this. 
  The 
  

   resistances 
  a 
  and 
  pb 
  are 
  of 
  a 
  metal 
  such 
  as 
  copper 
  or 
  iron 
  

   with 
  a 
  large 
  temperature-coefficient, 
  while 
  b 
  and 
  na 
  are 
  of 
  

   manganin 
  or 
  a 
  similar 
  alloy. 
  As 
  shown 
  in 
  the 
  paper 
  cited 
  

   above, 
  we 
  can 
  choose 
  n 
  and^ 
  so 
  as 
  to 
  have 
  a 
  constant 
  voltage 
  of, 
  

   say, 
  1*400 
  volts 
  between 
  X 
  and 
  Q 
  at 
  all 
  temperatures 
  from 
  

   0° 
  to 
  20° 
  C. 
  when 
  S. 
  0. 
  is 
  a 
  Clark 
  cell. 
  The 
  arrangement 
  

   thus 
  not 
  only 
  corrects 
  for 
  temperature 
  but 
  can 
  be 
  made 
  to 
  

   give 
  a 
  more 
  even 
  figure 
  by 
  cutting 
  off 
  the 
  0'034 
  volts 
  at 
  the 
  

   same 
  time. 
  In 
  one 
  of 
  the 
  coils 
  exhibited 
  the 
  resistances 
  

   have 
  the 
  following 
  values 
  at 
  15° 
  C 
  : 
  — 
  

  

  * 
  Communicated 
  by 
  the 
  Physical 
  Society 
  : 
  read 
  December 
  10, 
  1897. 
  

  

  