﻿374 
  Mr. 
  N. 
  E. 
  Gilbert 
  : 
  Experiments 
  upon 
  the 
  

  

  positive 
  direction. 
  To 
  determine 
  in 
  which 
  direction 
  this 
  

   would 
  deflect 
  the 
  galvanometer, 
  a 
  small 
  test 
  battery 
  was 
  used. 
  

   This 
  consisted 
  of 
  two 
  copper 
  wires 
  to 
  one 
  end 
  of 
  each 
  of 
  which 
  

   brass 
  plates 
  were 
  soldered 
  ; 
  to 
  the 
  other 
  end 
  of 
  one 
  wire 
  a 
  

   zinc 
  plate 
  was 
  soldered 
  ; 
  the 
  brass 
  plates 
  were 
  separated 
  by 
  

   a 
  piece 
  of 
  paper 
  and 
  inserted 
  between 
  the 
  silver 
  spring 
  and 
  

   the 
  tip 
  in 
  the 
  axle 
  ; 
  on 
  the 
  zinc 
  plate 
  was 
  placed 
  a 
  moistened 
  

   piece 
  of 
  paper, 
  and 
  this 
  was 
  touched 
  by 
  the 
  copper 
  wire 
  

   which 
  was 
  connected 
  to 
  the 
  brass 
  plate 
  nearer 
  the 
  galvano- 
  

   meter. 
  Thus 
  the 
  current 
  from 
  this 
  battery 
  passed 
  through 
  

   the 
  galvanometer 
  in 
  the 
  same 
  direction 
  as 
  a 
  positive 
  current 
  

   from 
  the 
  wheel. 
  This 
  direction 
  was 
  such, 
  during 
  the 
  first 
  

   readings, 
  as 
  to 
  give 
  a 
  "negative^' 
  deflexion, 
  and 
  these 
  

   readings, 
  while 
  very 
  variable, 
  were 
  uniformly 
  in 
  the 
  direction 
  

   looked 
  for. 
  The 
  sensibility 
  of 
  the 
  galvanometer 
  was 
  3 
  x 
  10~ 
  9 
  , 
  

   length 
  of 
  wire 
  8000 
  cms., 
  velocity 
  3900 
  cms. 
  per 
  second, 
  and 
  

   resistance 
  of 
  circuit 
  100 
  ohms. 
  The 
  mean 
  deflexion 
  was 
  

   *335 
  mm. 
  in 
  each 
  direction, 
  which 
  would 
  give 
  K 
  = 
  3x 
  10 
  -15 
  

   volts. 
  Substituting 
  this 
  in 
  equation 
  (3) 
  we 
  find 
  /j 
  = 
  *34 
  ohm, 
  

   a 
  resistance 
  about 
  3000 
  times 
  that 
  of 
  mercury. 
  Such 
  a 
  con- 
  

   ductivity 
  is 
  easily 
  possible 
  considering 
  the 
  high 
  temperature 
  

   existing 
  in 
  the 
  interior 
  of 
  the 
  earth 
  ; 
  and 
  at 
  this 
  stage 
  the 
  

   experiment 
  appeared 
  most 
  promising. 
  

  

  The 
  experiment 
  was 
  next 
  varied 
  by 
  crossing 
  the 
  lead 
  wires 
  

   on 
  the 
  wheel 
  so 
  as 
  to 
  bring 
  them 
  out 
  at 
  opposite 
  ends 
  of 
  the 
  

   shaft, 
  with 
  the 
  result 
  that 
  the 
  readings 
  failed 
  to 
  reverse 
  

   properly. 
  The 
  galvanometer 
  terminals 
  were 
  also 
  reversed 
  

   occasionally, 
  the 
  readings 
  being 
  consistent 
  in 
  every 
  case. 
  

   The 
  wires 
  were 
  reversed 
  on 
  the 
  wheel 
  several 
  times, 
  and 
  the 
  

   readings 
  were 
  so 
  inconsistent 
  that 
  a 
  reversing 
  switch 
  was 
  

   devised 
  which 
  was 
  placed 
  directly 
  on 
  the 
  shaft 
  and 
  could 
  be 
  

   reversed 
  while 
  the 
  wheel 
  was 
  in 
  motion. 
  This 
  device 
  removed 
  

   the 
  large 
  variations 
  in 
  thermal 
  currents 
  due 
  to 
  stopping 
  the 
  

   wheel, 
  and 
  eliminated 
  the 
  effects 
  of 
  all 
  electromotive 
  forces 
  in 
  

   the 
  circuit 
  except 
  those 
  on 
  the 
  wheel 
  itself, 
  only 
  these 
  being 
  

   commuted. 
  The 
  switch 
  w 
  T 
  ith 
  its 
  connexions 
  is 
  shown 
  in 
  fig. 
  3. 
  

   The 
  wires 
  from 
  the 
  wheel 
  and 
  from 
  the 
  silver 
  tips 
  in 
  the 
  

   shaft 
  end 
  in 
  copper 
  springs 
  a 
  a. 
  Contact 
  is 
  made 
  between 
  

   these 
  springs 
  by 
  means 
  of 
  two 
  copper 
  plates 
  b 
  b, 
  mounted 
  on 
  

   a 
  fibre 
  collar 
  which 
  can 
  be 
  moved 
  along 
  the 
  shaft 
  by 
  a 
  small 
  

   rod 
  placed 
  in 
  the 
  groove 
  in 
  the 
  movable 
  collar 
  c. 
  The 
  copper 
  

   springs 
  were 
  properly 
  bent 
  and 
  were 
  controlled 
  by 
  adjusting- 
  

   screws, 
  so 
  that 
  good 
  contact 
  was 
  assured. 
  The 
  complete 
  

   switch 
  shown 
  plane 
  in 
  the 
  figure 
  was 
  cylindrical 
  and 
  was 
  only 
  

   3 
  cms. 
  in 
  diameter, 
  so 
  that 
  the 
  various 
  parts 
  were 
  kept 
  as 
  

  

  