﻿Osmosis 
  and 
  Osmotic 
  Pressure. 
  253 
  

  

  unit 
  will 
  be 
  altered 
  on 
  account 
  of 
  the 
  dilatation 
  o£ 
  the 
  

   glass. 
  For 
  the 
  temperatures 
  actually 
  used 
  this 
  change 
  need 
  

   not 
  be 
  considered. 
  

  

  The 
  whole 
  apparatus 
  as 
  it 
  was 
  put 
  up 
  at 
  the 
  Physical 
  

   Institute 
  at 
  Christiania 
  is 
  shown 
  in 
  the 
  photograph 
  fig. 
  4 
  

   (PI. 
  VIII.). 
  

  

  §2. 
  

   The 
  Semipermeable 
  Membrane, 
  

  

  5. 
  The 
  membrane 
  used 
  in 
  these 
  measurements 
  is 
  the 
  usual 
  

   one 
  consisting 
  of 
  copper-ferrocyanide. 
  In 
  preparing 
  the 
  

   colls 
  I 
  used 
  electrolysis, 
  practically 
  the 
  same 
  method 
  as 
  used 
  

   by 
  Morse 
  and 
  Frazer, 
  and 
  later 
  by 
  Berkeley 
  and 
  Hartley. 
  

  

  It 
  proved, 
  however, 
  to 
  be 
  very 
  difficult 
  to 
  get 
  a 
  cell 
  which 
  

   gave 
  a 
  good 
  membrane, 
  and 
  I 
  tried 
  several 
  sorts, 
  till 
  I 
  found 
  

   one 
  which 
  gave 
  satisfactory 
  results. 
  

  

  At 
  first 
  I 
  tried 
  cells 
  made 
  of 
  porcelain, 
  but 
  these 
  proved 
  

   either 
  too 
  weak 
  or 
  else 
  so 
  hard 
  as 
  not 
  to 
  permit 
  of 
  the 
  fluid 
  

   passing 
  through. 
  

  

  Some 
  cells 
  of 
  fayence 
  made 
  at 
  the 
  Fayence 
  Factory, 
  

   Egersund, 
  showed 
  better 
  qualities, 
  but 
  were 
  not 
  fully 
  

   satisfactory. 
  

  

  At 
  last 
  I 
  got 
  cells, 
  made 
  of 
  " 
  Pukalische 
  Masse," 
  from 
  

   C. 
  Desaga, 
  Heidelberg, 
  that 
  proved 
  very 
  good. 
  

  

  For 
  the 
  electrolysis 
  a 
  somewhat 
  variable 
  potential 
  was 
  

   used, 
  arranging 
  matters 
  so 
  that 
  the 
  voltage 
  increased 
  with 
  

   increasing 
  resistance 
  in 
  the 
  cell 
  until 
  for 
  resistances 
  of 
  about 
  

   50,000 
  ohms 
  it 
  kept 
  a 
  fairly 
  constant 
  value 
  of 
  115 
  volts. 
  This 
  

   method 
  was 
  originally 
  chosen, 
  as 
  I 
  had 
  to 
  use 
  currents 
  of 
  230 
  

   volts 
  ; 
  but 
  I 
  also 
  consider 
  it 
  good 
  for 
  the 
  forming 
  of 
  the 
  mem- 
  

   brane, 
  as 
  the 
  voltage 
  is 
  small 
  as 
  long 
  as 
  the 
  membrane 
  is 
  weak. 
  

  

  The 
  arrangement 
  is 
  shown 
  in 
  fig. 
  5. 
  R 
  t 
  and 
  R 
  2 
  are 
  two 
  

   equal 
  resistances, 
  consisting 
  of 
  two 
  16 
  candle-power 
  incan- 
  

   descent 
  lamps. 
  By 
  means 
  of 
  two 
  commutators, 
  K 
  x 
  and 
  K 
  2 
  , 
  

   the 
  cell 
  C 
  can 
  be 
  put 
  as 
  a 
  shunt 
  to 
  one 
  of 
  the 
  lamps. 
  The 
  

   resistance 
  of 
  the 
  membrane 
  as 
  well 
  as 
  the 
  voltage 
  at 
  a 
  certain 
  

   moment 
  could 
  be 
  measured 
  by 
  means 
  of 
  a 
  torsion 
  galva- 
  

   nometer 
  G, 
  furnished 
  with 
  a 
  resistance-box 
  with 
  resistances 
  

   of 
  900, 
  9900, 
  and 
  99,900 
  ohms, 
  the 
  resistance 
  of 
  the 
  galva- 
  

   nometer 
  being 
  100 
  ohms. 
  

  

  I 
  have 
  not 
  treated 
  more 
  than 
  four 
  cells 
  of 
  this 
  type, 
  and 
  

   all 
  of 
  them 
  have 
  attained 
  an 
  electric 
  resistance 
  of 
  more 
  than 
  

   125,000 
  ohms; 
  one 
  of 
  them 
  especially, 
  which 
  I 
  shall 
  call 
  S, 
  

   has 
  shown 
  very 
  good 
  qualities 
  as 
  the 
  resistance 
  has 
  gone 
  up 
  

   to 
  400,000 
  ohms. 
  This 
  resistance, 
  however, 
  was 
  first 
  obtained 
  

   after 
  it 
  had 
  been 
  long 
  in 
  use, 
  and 
  after 
  it 
  had 
  several 
  times 
  

   been 
  tested 
  for 
  osmosis. 
  This, 
  my 
  first 
  cell, 
  has 
  been 
  used 
  

   in 
  all 
  my 
  measurements. 
  

  

  