598  Mr.  P.  S.  Barlow  on  the  Osmotic 
above  the  origin  ;  that  is,  the  pure  solvent  gives  an  osmotic 
pressure — which  is  impossible.  According  to  the  curve,  the 
osmotic  pressure  in  the  early  stages  does  not  increase  as 
rapidly  as  the  concentration  ;  while  later,  it  increases  more 
rapidly.  Hence  the  curve  is  below  the  theoretical  one  for  a 
water  solution. 
Fig.  1 . — Curve  I. 
/ 
V 
/ 
\ 
j 
/ 
Obsd 
Pft. 
jf     ~ 
/v 
Thi 
r^/?.z 
P#. 
The  curve  can  only  be  taken  as  a  preliminary  one.  But 
one  thing  seems  evident :  that  there  is  some  substance  present 
which  produces  an  osmotic  pressure  in  addition  to  that  due 
to  the  lithium  chloride.  The  suggestion  that  naturally  occurs 
is  that  the  extremely  hygroscopic  lithium  chloride  introduced 
a  little  water  into  the  solution  ;  and  previous  experiment 
has  shown  that,  with  the  guttapercha  membrane,  water 
behaves  as  the  solute.  I£  this  explanation  is  legitimate,  the 
zero  of  the  curve  for  the  lithium  chloride  will  be  raised  and 
the  curve  should  touch  or  cut  the  ^-axis  on  the  right  of  the 
origin. 
In  a  paper  on  experiments  with  alcohols  and  water  (part 
of  which  appeared  in  the  Phil.  Mag.  July  1905)  it  was 
suggested  that,  il  the  solution  of  the  solvent  by  the  membrane 
was  an  essential  condition  of  osmotic  phenomena,  then  the 
tendency  of  the  membrane  to  retain  the  solvent  would  oppose 
the  dilution  of  the  solution  and  so  cut  down  the  osmotic 
pressure.      The    effect    of   such    depression    would   be    most 
