﻿On 
  Temperature 
  Compensation 
  of 
  Standard 
  Cells. 
  275 
  

  

  a— 
  48*46 
  ohms. 
  

   na 
  = 
  AS2'7 
  „ 
  

  

  b 
  = 
  63-63 
  „ 
  

  

  ^6 
  = 
  495-0 
  „ 
  

  

  As 
  the 
  absolute 
  values 
  of 
  a 
  and 
  b 
  are 
  of 
  no 
  importance 
  so 
  

   long 
  as 
  they 
  are 
  high 
  enough 
  to 
  keep 
  the 
  auxiliary 
  Leclanche 
  

   cell 
  from 
  running 
  down, 
  the 
  copper 
  coils 
  a 
  and 
  pb 
  did 
  not 
  

   require 
  to 
  be 
  adjusted 
  but 
  were 
  merely 
  measured 
  ; 
  their 
  

   resistances 
  were 
  reduced 
  to 
  15° 
  C. 
  and 
  the 
  manganin 
  coils 
  na 
  

   and 
  b 
  were 
  then 
  adjusted 
  to 
  give 
  n 
  = 
  9*96, 
  p 
  = 
  7'18*. 
  It 
  is 
  

   clearly 
  a 
  great 
  saving 
  of 
  trouble 
  not 
  to 
  have 
  to 
  adjust 
  the 
  coils 
  

   which 
  have 
  a 
  high 
  temperature-coefficient. 
  The 
  four 
  coils 
  were 
  

   wound 
  on 
  a 
  brass 
  cylinder 
  7 
  or 
  8 
  centim. 
  in 
  diameter, 
  and 
  then 
  

   well 
  protected 
  with 
  silk 
  ribbon, 
  four 
  leads 
  being 
  brought 
  out. 
  

   To 
  test 
  the 
  whole 
  directly 
  the 
  cylinder 
  well 
  wrapped 
  in 
  

   wadding 
  was 
  placed 
  in 
  a 
  tin 
  canister, 
  the 
  lid 
  of 
  which 
  was 
  

   then 
  soldered 
  on 
  so 
  as 
  to 
  be 
  watertight. 
  The 
  leads 
  were 
  

   brought 
  out 
  through 
  a 
  long 
  tube 
  soldered 
  to 
  the 
  lid. 
  A 
  ther- 
  

   mometer 
  also 
  passed 
  down 
  this 
  tube 
  into 
  the 
  canister. 
  A 
  

   Leclanche 
  cell 
  was 
  connected 
  as 
  in 
  fig. 
  1 
  . 
  The 
  tin 
  was 
  then 
  

   well 
  immersed 
  in 
  water 
  at 
  various 
  temperatures, 
  and 
  steady 
  

   temperatures 
  inside 
  the 
  tin 
  were 
  got 
  by 
  stirring 
  the 
  water 
  

   until 
  a 
  thermometer 
  in 
  it 
  showed 
  the 
  same 
  temperature 
  as 
  the 
  

   one 
  inside 
  the 
  tin. 
  The 
  potential-difference 
  from 
  P 
  to 
  Q 
  was 
  

   measured 
  by 
  a 
  potentiometer 
  method 
  at 
  various 
  temperatures 
  

   with 
  the 
  following 
  results 
  : 
  — 
  

  

  Temperature. 
  

  

  P.-D., 
  volt. 
  

  

  155 
  C. 
  

   34-2 
  C. 
  

  

  48 
  "5 
  C. 
  

  

  0-0338 
  

   001472 
  

  

  o-ocoo 
  

  

  It 
  will 
  be 
  seen 
  that 
  the 
  actual 
  observed 
  values 
  of 
  the 
  auxiliary 
  

   voltage 
  are 
  just 
  what 
  would 
  be 
  required 
  along 
  with 
  the 
  

   Clark 
  cell 
  to 
  give 
  1*400 
  volts 
  at 
  all 
  ordinary 
  temperatures. 
  

  

  The 
  arrangement 
  in 
  fig. 
  1 
  was 
  chosen 
  so 
  as 
  to 
  give 
  the 
  

   maximum 
  possibility 
  of 
  correction. 
  Mr. 
  C. 
  W. 
  S. 
  Crawley 
  

   has 
  pointed 
  out 
  to 
  me 
  that 
  some 
  trouble 
  is 
  saved 
  in 
  the 
  con- 
  

   struction 
  of 
  the 
  coils 
  by 
  making 
  only 
  one 
  of 
  the 
  four 
  of 
  

  

  * 
  The 
  numerical 
  values 
  of 
  p 
  and 
  n 
  given 
  in 
  my 
  paper 
  mentioned 
  

   above 
  are 
  not 
  quite 
  right 
  owing 
  to 
  a 
  slip 
  in 
  signs 
  in 
  the 
  equation 
  there 
  

   given. 
  

  

  