﻿1-iO 
  C. 
  Barus 
  — 
  Motion 
  of 
  a 
  Submerged 
  Index 
  Thread 
  

  

  and 
  of 
  the 
  normal 
  viscosity 
  of 
  water, 
  its 
  thickness 
  would 
  have 
  

   to 
  be 
  much 
  below 
  even 
  the 
  diameter 
  of 
  the 
  molecule 
  of 
  water. 
  

  

  This 
  surprising 
  result 
  shows 
  that 
  the 
  conditions 
  assumed 
  are 
  

   not 
  given. 
  If 
  water 
  contributes 
  to 
  the 
  motion, 
  it 
  must 
  do 
  so 
  

   with 
  so 
  thin 
  a 
  film 
  as 
  to 
  possess 
  a 
  viscosity 
  enormously 
  greater 
  

   than 
  the 
  normal 
  rj 
  assumed, 
  or 
  in 
  other 
  words, 
  the 
  difference 
  

   between 
  superficial 
  and 
  internal 
  viscosity 
  is 
  here 
  paramount. 
  

   Again 
  the 
  result 
  is 
  a 
  mixed 
  phenomenon 
  and 
  it 
  must 
  be 
  the 
  

   object 
  of 
  experiment 
  to 
  disentangle 
  the 
  part 
  played 
  by 
  capil- 
  

   larity 
  from 
  the 
  part 
  due 
  to 
  viscosity. 
  

  

  2. 
  When 
  in 
  a 
  room 
  of 
  variable 
  temperature 
  the 
  thread 
  of 
  

   mercury 
  is 
  moved 
  up 
  and 
  down 
  in 
  consequence 
  of 
  the 
  thermal 
  

   expansion 
  of 
  the 
  water 
  in 
  the 
  bulb, 
  the 
  effect 
  is 
  a 
  tendency 
  to 
  

   moisten 
  the 
  tube 
  or 
  keep 
  up 
  a 
  supply 
  of 
  water 
  between 
  mer- 
  

   cury 
  and 
  glass. 
  In 
  the 
  absence 
  of 
  such 
  motion 
  the 
  film 
  will 
  

   gradually 
  be 
  squeezed 
  out 
  between 
  mercury 
  and 
  glass 
  where- 
  

   upon 
  the 
  thread 
  of 
  mercury 
  will 
  hermetically 
  seal 
  the 
  contents 
  

   of 
  the 
  bulb. 
  One 
  would 
  expect, 
  therefore, 
  that 
  if 
  the 
  liquid 
  

   is 
  kept 
  at 
  constant 
  or 
  very 
  slowly 
  varying 
  temperature, 
  the 
  

   mercury 
  thread 
  will 
  be 
  stationary. 
  To 
  test 
  this 
  I 
  placed 
  the 
  

   above 
  piezometer 
  in 
  a 
  space 
  of 
  approximately 
  constant 
  tem- 
  

   perature 
  (i. 
  e. 
  varying 
  very 
  gradually 
  from 
  season 
  to 
  season). 
  

   The 
  results 
  obtained 
  were 
  

  

  1896, 
  Feb. 
  

  

  14 
  

  

  Temp 
  

  

  . 
  145° 
  C. 
  

  

  Top 
  4-4 
  cm 
  

  

  Bottom 
  7 
  cm 
  Eeduced 
  mean 
  ) 
  5-9 
  

   7*1 
  positionat 
  14° 
  C. 
  ) 
  58 
  

  

  

  17 
  

  

  

  14-2 
  

  

  4-4 
  

  

  

  23 
  

  

  

  130 
  

  

  4-9 
  

  

  7-6 
  5-8 
  

  

  Mar. 
  

  

  15 
  

  

  

  14-5 
  

  

  5-3 
  

  

  7 
  9 
  6-8 
  

  

  Apr. 
  

  

  19 
  

  

  

  160 
  

  

  4-9 
  

  

  7-5 
  7-0 
  

  

  Aug. 
  

  

  5 
  

  

  

  19-9 
  

  

  5-2 
  

  

  7-8 
  8-9 
  

  

  1897, 
  Feb, 
  

  

  6 
  

  

  

  14-0 
  

  

  13-0 
  

  

  Lost 
  14-3 
  

  

  Comparing 
  this 
  table 
  with 
  the 
  last, 
  it 
  is 
  seen 
  that 
  between 
  

   Feb. 
  8, 
  1896, 
  and 
  Feb. 
  6, 
  1897, 
  both 
  at 
  14° 
  C, 
  the 
  descent 
  is 
  

   about 
  12 
  cra 
  for 
  the 
  year. 
  The 
  mean 
  rate 
  of 
  descent 
  for 
  the 
  

   year, 
  however, 
  is 
  but 
  *0014 
  cm 
  per 
  hour 
  as 
  compared 
  with 
  -OIT 
  0111 
  

   per 
  hour 
  at 
  the 
  outset. 
  Indeed, 
  if 
  the 
  mean 
  positions 
  of 
  the 
  

   thread 
  be 
  reduced 
  to 
  a 
  common 
  temperature 
  (see 
  table), 
  the 
  

   curve 
  obtained 
  indicates 
  an 
  initial 
  retardation 
  of 
  motion 
  with 
  

   time, 
  succeeded 
  by 
  a 
  marked 
  acceleration 
  of 
  the 
  motion 
  on 
  

   passing 
  through 
  the 
  region 
  of 
  high 
  summer 
  temperatures; 
  

   although 
  the 
  thread 
  is 
  but 
  2'6 
  cm 
  long 
  and 
  the 
  bulb 
  quite 
  sub- 
  

   merged, 
  some 
  reason 
  already 
  exists 
  to 
  lead 
  one 
  to 
  suspect 
  vis- 
  

   cous 
  dilatations 
  of 
  the 
  bulb. 
  

  

  3. 
  At 
  the 
  circular 
  edge 
  of 
  the 
  meniscus 
  three 
  surfaces 
  meet, 
  

   and 
  the 
  difference 
  of 
  surface 
  tensions 
  of 
  mercury-glass 
  and 
  

   water-glass 
  must 
  be 
  equal 
  to 
  the 
  resolved 
  part 
  of 
  the 
  surface 
  

   tension 
  of 
  mercury-water. 
  In 
  other 
  words, 
  the 
  resultant 
  of 
  

   surface 
  tensions 
  is 
  normal 
  to 
  the 
  glass 
  surface 
  at 
  the 
  edge 
  in 
  

  

  