﻿2<58 
  Mr. 
  Gr. 
  B. 
  Bryan 
  on 
  the 
  Determination 
  of 
  

  

  three 
  ebonite 
  pings 
  e 
  2 
  , 
  e^ 
  e 
  2 
  . 
  After 
  adjusting 
  the 
  lower 
  

   plate 
  B 
  parallel 
  to 
  the 
  upper 
  one 
  the 
  levelling-screws 
  are 
  

   fixed 
  by 
  lock-nuts, 
  and 
  the 
  plate 
  can 
  then 
  be 
  removed 
  and 
  

   replaced 
  without 
  impairing 
  the 
  adjustment. 
  The 
  glass 
  

   vessel 
  V 
  1 
  consists 
  of 
  a 
  cylinder 
  ground 
  to 
  fit 
  the 
  shoulder 
  on 
  

   the 
  lower 
  plate, 
  and 
  of 
  such 
  internal 
  diameter 
  that 
  the 
  upper 
  

   plate 
  just 
  slides 
  easily 
  inside 
  it. 
  Just 
  before 
  pouring 
  in 
  the 
  

   liquid 
  to 
  be 
  tested 
  this 
  cylinder 
  was 
  fastened 
  down 
  with 
  

   clean 
  paraffin 
  wax. 
  

  

  The 
  Plates. 
  — 
  These 
  were 
  of 
  cast 
  brass, 
  turned 
  to 
  the 
  shape 
  

   shown 
  in 
  the 
  figure. 
  The 
  surfaces 
  were 
  very 
  carefully 
  

   scraped 
  up 
  to 
  a 
  good 
  surface-plate, 
  until, 
  after 
  cleaning, 
  one 
  

   would 
  hold 
  the 
  other 
  up 
  (weight 
  of 
  each 
  plate 
  about 
  1 
  lb.). 
  

   They 
  were 
  then 
  gilded, 
  great 
  care 
  being 
  taken 
  not 
  to 
  damage 
  

   the 
  surfaces 
  in 
  any 
  way, 
  and 
  only 
  a 
  very 
  thin 
  coating 
  of 
  gold 
  

   was 
  put 
  on, 
  just 
  enough 
  to 
  prevent 
  oxidation. 
  These 
  plates 
  

   were 
  very 
  satisfactory, 
  and 
  the 
  error 
  due 
  to 
  irregularities 
  of 
  

   the 
  surfaces 
  negligibly 
  small. 
  Area 
  of 
  plates 
  = 
  20'82 
  sq. 
  cm. 
  

  

  To 
  realize 
  condition 
  (4) 
  a 
  new 
  commut°tor 
  was 
  made. 
  It 
  

   is 
  similar 
  to 
  the 
  one 
  shown 
  in 
  fig. 
  3, 
  but 
  each 
  drum 
  has 
  eight 
  

   segments 
  instead 
  of 
  two, 
  thus 
  increasing 
  the 
  number 
  of 
  

   reversals 
  per 
  revolution 
  four 
  times. 
  If 
  the 
  drum 
  were 
  rolled 
  

   out 
  it 
  would 
  appear 
  as 
  in 
  fig. 
  4. 
  Two 
  brushes 
  press 
  on 
  the 
  

   outer 
  rims, 
  and 
  two 
  others 
  on 
  two 
  consecutive 
  teeth. 
  

  

  This 
  commutator 
  was 
  driven 
  by 
  an 
  electromotor, 
  and 
  the 
  

   speed 
  could 
  be 
  varied 
  so 
  that 
  from 
  100 
  to 
  250^ 
  per 
  second 
  

   of 
  the 
  current 
  could 
  be 
  obtained. 
  The 
  speed 
  used 
  gave 
  about 
  

   200 
  ~ 
  per 
  second. 
  

  

  The 
  galvanometer 
  used 
  was 
  a 
  dead-beat 
  one 
  of 
  1430 
  ohms 
  

   resistance, 
  and 
  figure 
  of 
  merit, 
  with 
  the 
  interrupted 
  current, 
  

   of 
  10 
  -7 
  ampere. 
  

  

  The 
  connexions 
  were 
  the 
  same 
  as 
  in 
  fig. 
  3. 
  The 
  adjustment 
  

   of 
  the 
  lower 
  plate 
  parallel 
  to 
  the 
  upper 
  one 
  was 
  made 
  in 
  the 
  

   same 
  w 
  T 
  ay 
  as 
  before 
  : 
  but 
  could 
  now 
  be 
  easily 
  tested 
  by 
  

   placing 
  a 
  piece 
  of 
  glass 
  of 
  known 
  thickness 
  between 
  the 
  

   plates. 
  The 
  error 
  was 
  found 
  to 
  be 
  about 
  *01 
  mm., 
  but 
  was 
  

   probably 
  sometimes 
  less 
  than 
  this. 
  

  

  The 
  micrometer-screw 
  s 
  has 
  30 
  threads 
  per 
  inch, 
  and 
  the 
  

   head 
  H 
  is 
  divided 
  into 
  100 
  parts. 
  The 
  screw 
  was 
  compared 
  

   with 
  a 
  standard. 
  

  

  