Pressures  of  Alcoholic  Solutions. 
597 
quantity  of  the  liquid  absorbed  per  unit  area  of  the  mem- 
brane. In  general,  work  with  non-aqueous  solutions  seems 
so  far  to  have  been  confined  to  electrical  conductivity  and 
the  depression  of  the  freezing-point. 
I  turned  to  pure  indiarubber  as  a  possible  membrane  when 
it  was  clear  that  the  ferrocyanide  cells  were  unserviceable. 
The  pure  substance  I  was  unable  to  obtain,  so  I  made  use  of 
guttapercha  tissue.  Separating  alcohol  and  water,  this  tissue 
gave  a  pressure  on  the  side  of  the  water.  I  shall  say  here 
that  1  am  not  aware  that  guttapercha  has  been  used  for  the 
direct  measurement  of  osmotic  pressures,  though  I  think  it 
has  been  used  to  show  that  with  it,  the  osmotic  current  is 
towards  the  water.  I  may  also  say  in  passing,  that  thin  and 
slightly  vulcanized  indiarubber  tissue  has  been  tried  with 
these  solutions,  but  no  pressures  were  obtained  with  the 
samples  used. 
The  experiments  that  remain  to  be  recorded  are  those  in 
which  guttapercha  tissue  was  the  membrane.  The  first  series 
of  readings  with  solutions  of  lithium  chloride  was  obtained 
with  cell  B  I.  Only  three  pressures  were  obtained  in  three 
weeks  ;  that  is,  including  the  cleaning  of  the  cell,  each  reading 
below  took  more  than  a  week  to  obtain.  The  maximum 
pressure  did  not  require  all  that  time  to  be  set  up,  but  the  cell 
was  left  untouched  for  some  time  to  make  certain  that  a 
steady  value  had  been  reached. 
Guttapercha 
membrane. 
Theoretical 
Pressure. 
Observed 
Pressure. 
Temperature. 
By  Cell  B  i 
11         ii     ■ 
ii         ii     
„       Bn.     ... 
312  mm. 
937 
1562 
1875 
4*4  ram. 
12 
37-6 
32-6 
11°-3C. 
ll°-6 
115-6 
ll°-6 
No  LiCl  could  be  detected  on  the  outside  after  the  experiments. 
Plotting  the  observed  pressures  as  ordinates  and  the  theo- 
retical ones  (which  represent  the  concentrations)  as  abscissae, 
we  get  the  following  curve  (p.  598). 
The  falling  away  indicated  by  the  pressure  given  by  B  II. 
is  unexpected  ;  it  will  be  referred  to  later.  Here  we  need 
only  consider  the  curve  as  given  by  the  values  for  cell  B  I. 
The  curve  produced  to  the  left  cuts  the  y-axis  at  a  point 
