﻿Resistance 
  of 
  Metals 
  when 
  placed 
  in 
  a 
  Magnetic 
  Field. 
  645 
  

  

  All 
  the 
  junctions 
  (a, 
  fig. 
  2) 
  were 
  soldered 
  and 
  mounted 
  on 
  

   ebonite. 
  For 
  convenience, 
  these 
  junctions 
  were 
  connected 
  

   to 
  mercury 
  cups 
  {c, 
  fig. 
  2) 
  with 
  copper 
  wire. 
  The 
  bridge 
  

   was 
  balanced 
  by 
  placing 
  one 
  of 
  the 
  arms, 
  either 
  B 
  or 
  C 
  as 
  the 
  

   experiment 
  required, 
  in 
  multiple 
  arc 
  with 
  a 
  large 
  variable 
  

   resistance 
  W. 
  

  

  The 
  current 
  was 
  furnished 
  by 
  a 
  single 
  Leclanche 
  cell 
  with 
  

   a 
  variable 
  resistance 
  R 
  in 
  its 
  circuit. 
  This 
  allowed 
  the 
  

   current 
  in 
  the 
  bridge 
  to 
  be 
  adjusted 
  so 
  as 
  not 
  to 
  produce 
  

   appreciable 
  heating. 
  

  

  A 
  D'Arsonval 
  galvanometer, 
  G, 
  was 
  used 
  as 
  it 
  was 
  

   not 
  affected 
  by 
  a 
  magnet. 
  It 
  had 
  a 
  resistance 
  of 
  95 
  ohms 
  

   and, 
  by 
  using 
  a 
  telescope 
  and 
  scale, 
  it 
  gave 
  a 
  deflexion 
  of 
  

   1 
  mm. 
  for 
  2*4 
  x 
  10~ 
  8 
  ampere, 
  or 
  a 
  deflexion 
  of 
  1 
  mm. 
  on 
  the 
  

   scale 
  would 
  correspond 
  to 
  a 
  change 
  of 
  resistance 
  in 
  one 
  arm 
  

   of 
  the 
  bridge 
  to, 
  at 
  most, 
  one 
  part 
  in 
  50,000. 
  

  

  There 
  was 
  an 
  extra 
  wire 
  from 
  r' 
  to 
  the 
  variable 
  resistance 
  

   W, 
  and 
  also 
  a 
  mercury 
  key 
  in 
  the 
  battery-circuit 
  at 
  A. 
  

   This 
  arrangement 
  permitted 
  the 
  resistance 
  W 
  to 
  be 
  used 
  as 
  an 
  

   ordinary 
  Wheatstone 
  bridge 
  by 
  simply 
  changing 
  the 
  mercury 
  

   keys 
  to 
  the 
  dotted 
  position 
  as 
  shown 
  in 
  figure. 
  

  

  All 
  the 
  circuits 
  in 
  the 
  Wheatstone 
  bridge 
  were 
  carefully 
  

   insulated 
  and 
  supported 
  on 
  paraffin 
  blocks, 
  so 
  that 
  there 
  could 
  

   be 
  no 
  leaking 
  to 
  the 
  galvanometer 
  from 
  extraneous 
  sources. 
  

  

  The 
  spiral 
  and 
  the 
  other 
  resistances 
  forming 
  the 
  bridge 
  and 
  

   the 
  connecting 
  wires 
  to 
  the 
  mercury 
  cups 
  were 
  well 
  covered 
  

   with 
  paraffin-wax, 
  so 
  that 
  they 
  could 
  be 
  immersed 
  in 
  a 
  water- 
  

   bath. 
  

  

  To 
  produce 
  the 
  magnetic 
  field, 
  the 
  large 
  ring-magnet 
  of 
  the 
  

   Du 
  Bois 
  type 
  in 
  the 
  Cavendish 
  Laboratory 
  was 
  used. 
  The 
  

   poles, 
  P 
  (fig. 
  2), 
  were 
  about 
  2 
  mm. 
  apart. 
  It 
  gave 
  a 
  very 
  

   concentrated 
  field, 
  so 
  that 
  the 
  other 
  arms 
  of 
  the 
  bridge, 
  which 
  

   were 
  about 
  20 
  cm. 
  away, 
  were 
  not 
  affected 
  by 
  the 
  magnet. 
  

   The 
  fields 
  were 
  measured 
  with 
  a 
  carefully 
  calibrated 
  bismuth 
  

   spiral 
  for 
  known 
  currents 
  through 
  the 
  magnet 
  : 
  these 
  fields 
  

   were 
  taken 
  as 
  constant 
  throughout 
  the 
  experiment 
  for 
  that 
  

   position 
  of 
  the 
  poles. 
  

  

  The 
  poles 
  of 
  the 
  magnet 
  were 
  surrounded 
  by 
  a 
  water-bath, 
  

   which 
  was 
  kept 
  at 
  the 
  temperature 
  of 
  the 
  room. 
  The 
  spiral 
  

   and 
  the 
  resistances 
  were 
  all 
  immersed 
  in 
  this 
  bath. 
  

  

  The 
  first 
  experiments 
  were 
  made 
  with 
  german-silver 
  wire 
  

   0*004 
  inch 
  diameter. 
  The 
  resistances 
  of 
  the 
  two 
  spirals 
  were 
  

   44*035 
  and 
  44*06 
  ohms 
  respectively. 
  The 
  other 
  two, 
  which 
  

   were 
  also 
  of 
  german-silver, 
  had 
  42*41 
  and 
  43*58 
  ohms 
  respect- 
  

   ively. 
  With 
  this 
  arrangement, 
  no 
  change 
  could 
  be 
  detected 
  

   with 
  a 
  field 
  of 
  26,600 
  lines 
  per 
  sq. 
  cm. 
  

  

  