﻿370 
  Mr. 
  N. 
  E. 
  Gilbert 
  : 
  Experiments 
  upon 
  the 
  

  

  Each 
  reading 
  consists 
  of 
  a 
  zero, 
  a 
  deflexion 
  given 
  by 
  throwing 
  

   the 
  reversing 
  switch, 
  and 
  another 
  zero 
  given 
  by 
  throwing 
  the 
  

   reversing 
  switch 
  back. 
  In 
  the 
  first 
  row, 
  11 
  is 
  the 
  sum 
  of 
  40 
  

   readings 
  to 
  the 
  west 
  on 
  the 
  galvanometer-scale, 
  and 
  55 
  is 
  the 
  

   sum 
  of 
  the 
  corresponding 
  readings 
  to 
  the 
  east 
  ; 
  the 
  sixth 
  

   column 
  gives 
  the 
  sums 
  of 
  differences 
  in 
  the 
  various 
  sets 
  ; 
  the 
  

   last 
  tw 
  T 
  o 
  columns 
  give 
  the 
  connexions 
  of 
  the 
  wires. 
  The 
  

   galvanometer 
  circuit 
  was 
  kept 
  closed, 
  thus 
  eliminating 
  thermal 
  

   currents 
  which 
  may 
  have 
  appeared 
  in 
  the 
  earlier 
  readings. 
  The 
  

   galvanometer 
  terminals 
  were 
  reversed 
  by 
  a 
  single 
  mercury 
  

   switch. 
  Care 
  was 
  taken 
  to 
  avoid 
  touching 
  any 
  part 
  of 
  the 
  

   circuit 
  during 
  a 
  series 
  of 
  readings. 
  The 
  sensibility 
  of 
  the 
  

   galvanometer 
  was 
  10 
  — 
  9 
  , 
  The 
  signs 
  have 
  been 
  so 
  adjusted 
  in 
  

   the 
  columns 
  of 
  differences, 
  that 
  positive 
  differences 
  always 
  

   mean 
  deflexions 
  in 
  the 
  direction 
  looked 
  for. 
  The 
  wheel 
  

   during 
  this 
  series 
  was 
  wound 
  with 
  8000 
  cms. 
  of 
  No. 
  36 
  

   copper 
  wire, 
  giving 
  a 
  resistance 
  of 
  180 
  ohms 
  in 
  the 
  circuit. 
  

   The 
  speed 
  was 
  6000 
  cms. 
  per 
  second. 
  The 
  average 
  deflexion 
  

   of 
  '01 
  mm. 
  then 
  corresponds 
  to 
  K 
  = 
  15xl0 
  — 
  17 
  volt 
  or 
  to 
  

   p 
  = 
  *012 
  ohm, 
  a 
  resistance 
  only 
  130 
  times 
  as 
  great 
  as 
  that 
  of 
  

   mercury. 
  In 
  the 
  series 
  recorded 
  in 
  Table 
  II., 
  taken 
  eight 
  

   months 
  later 
  with 
  a 
  galvanometer 
  ten 
  times 
  more 
  sensitive, 
  a 
  

   deflexion 
  of 
  '175 
  mm. 
  in 
  the 
  opposite 
  direction 
  was 
  obtained, 
  

   and 
  this 
  is 
  about 
  sufficient 
  to 
  balance 
  the 
  last. 
  It 
  seems 
  

   certain, 
  then, 
  that 
  a 
  resistance 
  as 
  small 
  as 
  130 
  times 
  the 
  re- 
  

   sistance 
  of 
  mercury 
  would 
  be 
  required 
  for 
  the 
  interior 
  of 
  the 
  

   earth 
  if 
  terrestrial 
  magnetism 
  were 
  due 
  to 
  this 
  cause. 
  Schuster* 
  

   has 
  shown 
  that, 
  to 
  account 
  for 
  the 
  earth-currents 
  induced 
  by 
  

   the 
  currents 
  in 
  the 
  upper 
  atmosphere, 
  which 
  are 
  undoubtedly 
  

   the 
  primary 
  cause 
  of 
  the 
  short-period 
  variations, 
  the 
  average 
  

   specific 
  resistance 
  of 
  the 
  earth 
  must 
  be 
  of 
  the 
  order 
  of 
  mag- 
  

   nitude 
  p 
  = 
  l-'23 
  x 
  10 
  13 
  c.G.s. 
  units, 
  a 
  value 
  1,000,000 
  times 
  as 
  

   great 
  as 
  that 
  derived 
  above. 
  

  

  It 
  is 
  evident 
  from 
  this 
  comparison 
  that 
  the 
  largest 
  value 
  

   obtained 
  for 
  the 
  hypothetical 
  electromotive 
  force 
  is 
  far 
  too 
  

   small 
  to 
  account 
  for 
  the 
  magnetism 
  of 
  the 
  earth. 
  The 
  second 
  

   series 
  of 
  readings 
  gives 
  an 
  average 
  which 
  just 
  balances 
  the 
  

   first, 
  reducing 
  the 
  final 
  average 
  to 
  zero. 
  The 
  result 
  of 
  the 
  

   entire 
  research 
  is 
  then 
  purely 
  negative. 
  If 
  any 
  such 
  cause 
  

   for 
  terrestrial 
  magnetism 
  exists, 
  it 
  has 
  not 
  been 
  detected 
  by 
  

   this 
  experiment. 
  

  

  As 
  has 
  been 
  mentioned, 
  an 
  attempt 
  was 
  made 
  in 
  some 
  of 
  

   the 
  experiments 
  to 
  reduce 
  possible 
  convection-currents 
  in 
  the 
  

   sether 
  by 
  placing 
  a 
  heavy 
  copper 
  shield 
  around 
  the 
  wheel. 
  

  

  * 
  Phil. 
  Trans, 
  of 
  Eoyal 
  Soc. 
  1889, 
  p. 
  496. 
  

  

  