﻿Coefficient 
  of 
  Diffusion. 
  

  

  867 
  

  

  3. 
  Apparatus. 
  

   A 
  and 
  B 
  are 
  two 
  hollow 
  spherical 
  glass 
  bulbs 
  respectively 
  

   5 
  cms. 
  and 
  9 
  cms. 
  in 
  diameter, 
  connected 
  together 
  as 
  shown 
  

   by 
  two 
  parallel 
  glass 
  quill-tubes 
  t, 
  t 
  (fig. 
  2). 
  

  

  On 
  the 
  upper 
  side 
  of 
  A 
  the 
  bulb 
  

   is 
  pierced 
  by 
  an 
  aperture 
  into 
  which 
  

   the 
  diffusion-tube 
  D 
  is 
  hermetically 
  

   sealed. 
  

  

  The 
  ends 
  o£ 
  this 
  tube 
  were 
  ground 
  

   off 
  perfectly 
  plane 
  and 
  perpendicular 
  

   to 
  its 
  axis, 
  and 
  its 
  dimensions 
  were, 
  

   in 
  most 
  of 
  the 
  experiments, 
  4 
  cms. 
  

   long, 
  and 
  1*694 
  sq. 
  cms. 
  in 
  cross- 
  

   section. 
  

  

  The 
  complete 
  apparatus 
  AB 
  was 
  

   filled 
  with 
  the 
  solution 
  under 
  inves- 
  

   tigation, 
  and 
  suspended 
  in 
  a 
  cylin- 
  

   drical 
  copper 
  vessel 
  C 
  of 
  about 
  7 
  litres 
  

   capacity, 
  containing 
  distilled 
  water, 
  

   the 
  temperature 
  of 
  which 
  may 
  be 
  

   kept 
  constant 
  by 
  means 
  of 
  a 
  sur- 
  

   rounding 
  ice-jacket 
  I, 
  protected 
  by 
  

   a 
  layer 
  of 
  cork-dust 
  E, 
  several 
  inches 
  

   thick. 
  The 
  mercury 
  thermometer 
  T 
  

   had 
  a 
  small 
  range 
  on 
  each 
  side 
  of 
  

   0° 
  C. 
  and 
  was 
  divided 
  into 
  r 
  ^Q° 
  C. 
  

  

  The 
  Sartorius 
  balance 
  used 
  was 
  

   mounted 
  upon 
  a 
  slate 
  shelf 
  firmly 
  

   cemented 
  into 
  the 
  stone 
  wall 
  of 
  the 
  

   laboratory. 
  

   Two 
  holes 
  were 
  sand-blasted 
  through 
  the 
  base 
  of 
  the 
  

   balance-case, 
  and 
  a 
  light 
  frame, 
  //, 
  passed 
  through 
  them 
  as 
  

   shown, 
  and 
  to 
  this 
  was 
  attached 
  a 
  fine 
  wire 
  w 
  which, 
  passing- 
  

   through 
  a 
  third 
  hole, 
  in 
  the 
  slate 
  shelf, 
  sustained 
  the 
  appa- 
  

   ratus 
  already 
  described. 
  The 
  movement 
  of 
  the 
  balance 
  beam 
  

   was 
  limited 
  by 
  means 
  of 
  stops 
  shown 
  at 
  s, 
  so 
  that 
  the 
  pointer 
  

   could 
  only 
  move 
  over 
  1 
  scale-division, 
  and 
  the 
  vertical 
  

   this 
  permitted 
  to 
  the 
  suspended 
  system 
  

   so 
  that 
  any 
  disturbance 
  which 
  might 
  

  

  movement 
  

   was 
  only 
  

  

  move 
  

   which 
  

   12 
  

  

  mm., 
  

  

  arise 
  from 
  this 
  small 
  and 
  very 
  slow 
  movement 
  will 
  be 
  quite 
  

   negligible. 
  

  

  4. 
  Method 
  of 
  Experiment. 
  

  

  As 
  the 
  salt 
  diffuses 
  up 
  the 
  tube, 
  the 
  stronger 
  solution 
  in 
  

   the 
  upper 
  larger 
  bulb, 
  B, 
  sinks 
  by 
  gravity, 
  by 
  means 
  of 
  the 
  

   side 
  tubes 
  1 
  1, 
  and 
  maintains 
  the 
  initial 
  concentration 
  of 
  the 
  

  

  3L2 
  

  

  