﻿372 
  Mr. 
  N. 
  E. 
  Gilbert 
  : 
  Experiments 
  upon 
  the 
  

  

  of 
  wire 
  moving 
  with 
  unit 
  velocity) 
  would 
  probably 
  have 
  been 
  

   detected, 
  though 
  this 
  is 
  by 
  no 
  means 
  certain. 
  

  

  The 
  form 
  of 
  the 
  above 
  coil 
  was 
  not 
  satisfactory. 
  It 
  was 
  

   inclosed 
  on 
  three 
  sides 
  by 
  brass, 
  and 
  the 
  turns 
  on 
  the 
  interior 
  

   were 
  so 
  shielded 
  by 
  those 
  on 
  the 
  exterior 
  that 
  comparatively 
  

   few 
  would 
  be 
  exposed 
  directly 
  to 
  the 
  action 
  of 
  the 
  aether 
  if 
  

   the 
  aether 
  were 
  dragged 
  along 
  as 
  a 
  viscous 
  fluid 
  would 
  be. 
  A 
  

   new 
  wheel 
  was 
  therefore 
  built 
  whose 
  periphery 
  was 
  a 
  cylinder 
  

   4*2 
  cms. 
  broad 
  and 
  7*3 
  cms. 
  radius. 
  On 
  this 
  was 
  wound 
  in 
  

   a 
  single 
  layer 
  175 
  turns 
  of 
  No. 
  33 
  copper 
  wire 
  giving 
  a 
  

   resistance 
  of 
  57 
  ohms. 
  A 
  test 
  similar 
  to 
  the 
  above 
  was 
  made 
  

   with 
  this 
  wheel 
  with 
  the 
  result 
  that 
  no 
  reversible 
  E.M.F. 
  as 
  

   large 
  as 
  K=l-2xl0~ 
  14 
  volts 
  was 
  detected. 
  The 
  practical 
  

   result 
  of 
  these 
  two 
  tests 
  was 
  to 
  show 
  that 
  any 
  E.M.F. 
  which 
  

   might 
  exist 
  could 
  be 
  detected 
  only 
  by 
  averaging 
  a 
  long 
  series 
  

   of 
  readings, 
  and 
  this 
  was 
  next 
  undertaken. 
  

  

  The 
  sources 
  of 
  difficulty 
  in 
  making 
  these 
  measurements 
  

   were 
  numerous. 
  In 
  order 
  to 
  insure 
  smooth 
  running 
  at 
  the 
  

   high 
  speeds 
  attained, 
  the 
  most 
  delicate 
  balancing 
  of 
  the 
  wheel 
  

   was 
  necessary. 
  The 
  wheel 
  was 
  mounted 
  on 
  a 
  long 
  slender 
  

   steel 
  shaft 
  which 
  was 
  sufficiently 
  flexible 
  to 
  allow 
  rotation 
  

   about 
  a 
  true 
  principal 
  axis 
  when 
  a 
  high 
  speed 
  was 
  once 
  

   reached 
  ; 
  but 
  trouble 
  was 
  experienced 
  in 
  passing 
  the 
  point 
  at 
  

   which 
  the 
  speed 
  of 
  the 
  wheel 
  equalled 
  the 
  period 
  of 
  vibration 
  

   of 
  the 
  wheel 
  on 
  the 
  shaft. 
  Here 
  vibration 
  became 
  so 
  excessive 
  

   that 
  the 
  motor 
  was 
  sometimes 
  unable 
  to 
  increase 
  the 
  speed 
  

   beyond 
  that 
  point. 
  The 
  difficulty 
  was 
  overcome 
  by 
  allowing 
  

   the 
  cast 
  iron 
  base 
  of 
  the 
  machine 
  to 
  stand 
  nnclampecl 
  on 
  

   cotton 
  or 
  on 
  several 
  thicknesses 
  of 
  cardboard, 
  and 
  by 
  most 
  

   careful 
  balancing. 
  Another 
  source 
  of 
  trouble 
  lay 
  in 
  the 
  

   thermal 
  currents 
  in 
  the 
  circuit 
  caused 
  by 
  the 
  heating 
  of 
  

   the 
  copper-silver 
  junctions 
  by 
  the 
  heat 
  flowing 
  away 
  from 
  the 
  

   bearings 
  when 
  the 
  wheel 
  was 
  running. 
  The 
  terminals 
  of 
  

   the 
  coil 
  were 
  led 
  out 
  through 
  the 
  end 
  of 
  the 
  shaft 
  in 
  order 
  to 
  

   avoid 
  as 
  far 
  as 
  possible 
  any 
  friction 
  and 
  consequent 
  heating 
  

   at 
  the 
  brushes, 
  and 
  to 
  reduce 
  to 
  a 
  minimum 
  any 
  alternating 
  

   currents 
  due 
  to 
  motion 
  of 
  conductors 
  in 
  the 
  earth's 
  field. 
  

   After 
  running 
  for 
  some 
  minutes 
  the 
  temperature 
  would 
  

   become 
  approximately 
  steady 
  ; 
  but 
  at 
  best 
  the 
  variations 
  in 
  

   the 
  thermal 
  E.M.F. 
  at 
  the 
  junctions 
  of 
  the 
  wire 
  were 
  much 
  

   greater 
  than 
  the 
  E.M.F. 
  for 
  which 
  we 
  were 
  looking. 
  These 
  

   thermal 
  currents 
  were 
  reduced 
  to 
  a 
  minimum 
  by 
  using 
  the 
  

   same 
  wire, 
  so 
  far 
  as 
  possible, 
  for 
  all 
  connexions 
  ; 
  but 
  this 
  could 
  

   not 
  be 
  done 
  where 
  the 
  circuit 
  passed 
  from 
  the 
  tip 
  of 
  the 
  axis 
  

   to 
  the 
  brush, 
  for 
  it 
  was 
  necessary 
  to 
  use 
  silver 
  to 
  insure 
  good 
  

   contact. 
  Silver 
  wires 
  were 
  used 
  for 
  several 
  inches 
  in 
  both 
  

  

  