﻿the 
  Conductivity 
  of 
  Liquids 
  in 
  Thin 
  Layers. 
  257 
  

  

  round 
  the 
  glass 
  rod 
  g±. 
  The 
  lead 
  wire, 
  b, 
  from 
  the 
  lower 
  

   plate 
  is 
  brought 
  up 
  through 
  a 
  thin 
  glass 
  tube. 
  

  

  The 
  micrometer-screw 
  has 
  20 
  threads 
  to 
  the 
  inch, 
  and 
  

   passes 
  through 
  a 
  very 
  long 
  nut 
  ; 
  and 
  as 
  it 
  carries 
  a 
  consider- 
  

   able 
  weight 
  there 
  is 
  no 
  backlash. 
  The 
  micrometer 
  head 
  is 
  

   divided 
  into 
  100 
  parts, 
  so 
  that 
  one 
  division 
  represents 
  20*00 
  

   inch. 
  The 
  screw 
  was 
  examined 
  by 
  observing 
  the 
  edge 
  of 
  

   the 
  upper 
  plate 
  A 
  with 
  a 
  microscope, 
  and 
  the 
  error 
  was 
  found 
  

   to 
  be 
  so 
  small 
  as 
  to 
  be 
  negligible. 
  

  

  Adjustment 
  of 
  the 
  Plates. 
  — 
  (1) 
  As 
  the 
  upper 
  plate 
  A 
  

   rotates 
  with 
  the 
  screw 
  it 
  must 
  be 
  made 
  to 
  move 
  parallel 
  to 
  

   itself. 
  To 
  do 
  this 
  the 
  lower 
  edge 
  of 
  A 
  was 
  observed 
  with 
  a 
  

   horizontal 
  microscope 
  containing 
  a 
  micrometer-scale 
  ; 
  each 
  of 
  

   the 
  three 
  screws 
  s^s^i 
  was 
  brought 
  opposite 
  the 
  microscope 
  

   in 
  turn 
  (the 
  exact 
  position 
  being 
  read 
  on 
  the 
  divided 
  head) 
  

   and 
  adjusted 
  until 
  the 
  downward 
  or 
  upward 
  motion 
  during 
  

   each 
  third 
  of 
  a 
  revolution 
  was 
  the 
  same. 
  

  

  (2) 
  The 
  lower 
  plate 
  B 
  must 
  then 
  be 
  made 
  parallel 
  to 
  the 
  

   upper 
  one. 
  To 
  do 
  this 
  a 
  strong 
  light 
  was 
  placed 
  behind 
  the 
  

   plates 
  and 
  looked 
  at 
  through 
  the 
  space 
  between 
  the 
  plates. 
  

   The 
  screws 
  s 
  2 
  s 
  2 
  s 
  2 
  were 
  adjusted 
  until, 
  on 
  screwing 
  down 
  the 
  

   upper 
  plate, 
  the 
  light 
  was 
  extinguished 
  right 
  across 
  at 
  the 
  

   same 
  instant. 
  The 
  light 
  was 
  moved 
  round 
  the 
  plate, 
  and 
  

   Avhen 
  it 
  disappeared 
  at 
  the 
  same 
  moment, 
  whatever 
  its 
  position, 
  

   the 
  adjustment 
  was 
  assumed 
  to 
  be 
  correct, 
  and 
  the 
  reading 
  of 
  

   the 
  divided 
  head 
  at 
  that 
  instant 
  was 
  taken 
  as 
  the 
  zero. 
  This 
  

   method 
  was 
  tested 
  by 
  making 
  the 
  adjustments 
  with 
  the 
  glass 
  

   vessel 
  removed, 
  and 
  then 
  measuring 
  the 
  thickness 
  of 
  a 
  piece 
  

   of 
  thin 
  glass 
  at 
  several 
  points 
  round 
  the 
  edges 
  of 
  the 
  plates. 
  

   The 
  measurements 
  were 
  found 
  to 
  agree 
  very 
  closely 
  amongst 
  

   themselves 
  and 
  with 
  the 
  thickness 
  of 
  the 
  glass, 
  as 
  measured 
  

   by 
  a 
  spherometer, 
  to 
  *01 
  mm. 
  The 
  adjustment 
  was 
  probably 
  

   not 
  always 
  as 
  good 
  as 
  this 
  ; 
  but 
  it 
  could 
  not 
  be 
  tested 
  when 
  

   the 
  glass 
  vessel 
  was 
  in 
  position. 
  

  

  The 
  Plates. 
  — 
  The 
  plates 
  first 
  used 
  were 
  of 
  brass, 
  10 
  cm. 
  

   in 
  diameter 
  and 
  3 
  mm. 
  thick. 
  They 
  were 
  ground 
  flat 
  on 
  an 
  

   emery-wheel, 
  and 
  then 
  on 
  plate-glass 
  with 
  fine 
  emery-powder. 
  

   They 
  were 
  then 
  carefully 
  amalgamated. 
  It 
  was 
  found 
  im- 
  

   possible 
  to 
  get 
  these 
  plates 
  perfectly 
  flat 
  on 
  account 
  of 
  their 
  

   springiness; 
  both 
  were 
  found 
  to 
  be 
  slightly 
  concave. 
  The 
  

   correction 
  for 
  this 
  was 
  estimated 
  to 
  be 
  0*013 
  turn 
  of 
  the 
  screw 
  

   = 
  0*0165 
  mm., 
  to 
  be 
  added 
  to 
  the 
  distance 
  between 
  the 
  plates 
  

   given 
  by 
  the 
  reading 
  of 
  the 
  divided 
  head. 
  

  

  As 
  these 
  plates 
  were 
  not 
  considered 
  satisfactory 
  two 
  others 
  

   were 
  made. 
  These 
  were 
  5*64 
  cm. 
  in 
  diameter 
  and 
  about 
  

   1 
  cm. 
  thick, 
  so 
  as 
  to 
  avoid 
  springiness. 
  The 
  metal 
  was 
  zinc. 
  

  

  