﻿398 
  C. 
  Bar 
  us 
  — 
  Diffusion 
  of 
  Air 
  through 
  Water. 
  

  

  water, 
  equal 
  to 
  the 
  height 
  of 
  A. 
  There 
  is 
  thus 
  a 
  continual 
  

   tendency 
  to 
  squeeze 
  out 
  the 
  air 
  from 
  v 
  through 
  the 
  pores 
  of 
  

   the 
  water 
  into 
  the 
  atmosphere, 
  which 
  in 
  the 
  lapse 
  of 
  years 
  

   may 
  very 
  well 
  produce 
  a 
  definite 
  effect. 
  

  

  Obviously, 
  merely 
  the 
  head 
  of 
  water 
  stated 
  induces 
  diffu- 
  

   sion 
  ; 
  for 
  change 
  of 
  atmospheric 
  pressure 
  is 
  without 
  differen- 
  

   tial 
  effect. 
  Great 
  pressure-excess 
  at 
  v 
  may 
  be 
  secured 
  by 
  

   this 
  head, 
  but 
  the 
  distance 
  through 
  which 
  the 
  

   imprisoned 
  air 
  will 
  have 
  to 
  diffuse 
  

   to 
  escape 
  into 
  the 
  air 
  is 
  correspond- 
  

   ingly 
  lengthened. 
  Pressure-excess 
  

   at 
  v 
  for 
  a 
  given 
  head 
  may 
  be 
  incre- 
  

   mented 
  by 
  virtually 
  making 
  the 
  

   water 
  heavier 
  ; 
  and 
  this 
  may 
  be 
  

   done 
  by 
  putting 
  the 
  jar 
  A 
  into 
  a 
  

   centrifugal 
  machine, 
  in 
  which 
  case 
  

   the 
  transpiration 
  would 
  still 
  take 
  

   place 
  under 
  known 
  conditions. 
  

  

  2. 
  A 
  method 
  of 
  accurately 
  meas- 
  

   uring 
  the 
  gas 
  in 
  v 
  is 
  next 
  to 
  be 
  

   devised. 
  This 
  is 
  accomplished 
  if 
  a 
  

   sensitive 
  thermometer, 
  T 
  7 
  , 
  and 
  a 
  

   reservoir, 
  R, 
  joined 
  by 
  rubber 
  tub- 
  

   ing 
  to 
  A 
  for 
  producing 
  a 
  variable 
  

   atmosphere, 
  is 
  added 
  as 
  suggested 
  

   in 
  the 
  figure. 
  For 
  reasons 
  stated 
  

   C 
  ! 
  f\W\^r 
  i 
  (diver 
  sunk), 
  the 
  vessel, 
  R, 
  in 
  the 
  

  

  following 
  experiments 
  is 
  depressed. 
  

   This 
  is 
  objectionable 
  since 
  there 
  is 
  

   some 
  danger 
  of 
  parting 
  the 
  column 
  

   of 
  water 
  in 
  the 
  rubber 
  tube 
  with 
  

   influx 
  of 
  air 
  if 
  the 
  thermometer, 
  

   T, 
  is 
  to 
  be 
  moved 
  up 
  and 
  down. 
  

   A 
  floating 
  diver 
  and 
  raised 
  vessel, 
  

   constitute 
  a 
  better 
  disposition. 
  Instead 
  of 
  the 
  divided 
  ves- 
  

   , 
  A, 
  R, 
  a 
  single 
  tube 
  about 
  6 
  feet 
  long 
  and 
  a 
  diver 
  usually 
  

   floating 
  at 
  the 
  middle 
  would 
  answer 
  all 
  requirements; 
  but 
  the 
  

   temperature-distribution 
  in 
  so 
  long 
  a 
  tube 
  would 
  be 
  liable 
  to 
  

   introduce 
  serious 
  errors. 
  Finally 
  R 
  is 
  provided 
  with 
  a 
  special 
  

   piece 
  of 
  rubber 
  tubing, 
  6 
  Y 
  , 
  through 
  which 
  pressure 
  increments 
  

   or 
  decrements 
  may 
  be 
  applied 
  temporarily 
  by 
  the 
  mouth. 
  

  

  Before 
  beginning 
  a 
  series 
  of 
  measurements, 
  suction 
  and 
  

   pressure 
  are 
  applied 
  alternately 
  at 
  C, 
  causing 
  the 
  diver 
  to 
  move 
  

   up 
  and 
  down 
  throughout 
  A 
  (braised), 
  until 
  the 
  temperature, 
  

   nearly 
  uniform 
  before, 
  has 
  become 
  quite 
  so. 
  Care 
  must 
  be 
  

   taken 
  to 
  avoid 
  violent 
  motion. 
  Thus 
  with 
  a 
  differential 
  suc- 
  

   tion 
  of 
  half 
  an 
  atmosphere, 
  the 
  adiabatic 
  change 
  of 
  tempera- 
  

  

  sel 
  

  

  