596  Mr.  P.  S.  Barlow  on  the  Osmotic 
colloid  that  would  modify  its  behaviour  towards  alcohol  (as 
observed  by  Tammann).  A  considerable  time  was  devoted  to 
this  work,  a  full  account  of  which  would  be  needlessly  long  ; 
it  need  only  be  said  here  that  the  results  give  fresh  negative 
evidence  in  support  of  the  theory  that  for  the  osmotic  pressure 
to  show  itself,  the  membrane  must  be  able  to  dissolve  the 
solvent  and  have  a  distinct  "  attraction  of  solution  "  for  it. 
Every  opportunity  was  allowed  as  regards  time,  some  of  the 
experiments  continuing  for  several  months,  and  all  for  more 
than  seven  days.  The  solutions  used  were  of  lasvulose, 
lithium  chloride,  lithium  sulphate  (anhyd),  camphor,  shellac, 
methyl  oxalate,  ferric  chloride  (anhyd),  glycerine.  In 
addition  to  the  cylindrical  cell  of  the  usual  type,  cells  were 
also  made  of  thin  porous  plates  in  which  the  membrane  was 
formed.     All  the  cells  worked  normally  with  water. 
The  copper-ferrocyanide  membranes  being  unserviceable, 
other  membranes  were  tried.  Of  the  vegetable  and  animal 
parchments,  the  bladder  membrane  was  the  only  one  to  show 
any  pressure.  The  cell  was  made  by  stretching  the  prepared 
bladder  across  a  brass  cylinder.  The  cylinder  consisted  of 
two  parts  which  could  be  joined  by  means  of  flanges  and 
screws.  The  bladder  was  held  between  the  flanges,  thus 
dividing  the  interior  into  two  compartments.  The  side  con- 
nected to  the  gauge  was  filled  with  the  solution  ;  the  other 
side  was  connected  to  a  small  glass  reservoir  and  was  filled 
with  the  solvent.  The  bladder  was  backed  on  the  side  of  the 
solvent  (the  outside)  by  perforated  zinc. 
The  bladder  cell  (B  n.)  was  used  with  solutions  of  lithium 
chloride,  methyl  oxalate,  and  camphor  in  absolute  alcohol. 
Of  these,  the  lithium  chloride  and  camphor  solutions  gave 
osmotic  pressures,  The  methyl  oxalate  showed  no  sign  of 
pressure.  No  quantitative  values  were  expected  from  the 
experiments  with  bladder,  it  being  a  very  imperfect  semi- 
permeable membrane.  The  result,  however,  was  satisfactory 
after  the  experiments  with  the  ferrocyanide  membrane  as 
showing  that,  under  conditions  depending  on  the  membrane, 
these  solutions  would  give  pressures  with  the  solvent. 
I  am  not  aware  that  any  quantitative  results  have  been 
published  of  experiments  made  with  non-aqueous  solutions 
to  measure  directly  the  osmotic  pressure.  Flusin"*  has  made 
quantitative  measurements  of  the  rate  of  inflow,  using  vul- 
canized indiarubber  and  bladder.  He  found  that  the  amount 
of   liquid   entering  in  a  given  time  is   proportional  to  the 
*   Comptes  Bendus,  cxxxi.  p.  1308,  Dec.  31,  1900. 
