Pressures  of  Alcoliolie  Solutions. 
603 
the  same  tissue,  and  therefore  the  presence  of  lithium  chloride 
outride  is  not  due  to  a  leak  in  the  membrane  caused  by  over- 
strain.    Obviously  above  a  certain  pressure  the  guttapercha 
Kg 
.  5.— C 
urvTe  IV. 
&- 
0036 
X 
?<¥~ 
y 
/ 
C( 
i 
5NDUCT- 
VITY 
/ 
/■■ 
! 
! 
1 
•00)8 
/ 
/ 
. 
- 
/ 
] 
r 
sJrensti- 
\ 
•067Z, 
•1272, 
membrane  is  only  imperfectly  semi-permeable  for  the  case  of 
lithium  chloride.  It  may  be  that  this  membrane  is  permeable 
to  chlorides  as  the  copper  ferroevanide  one  is  ;  and,  like  the 
latter,  would  act  more  perfectly  with  sulphates. 
The  general  character  of  Curves  II.  and  III.  agrees  with 
theory  :  the  observed  pressure  decreases  more  rapidly  than  the 
strength  of  the  solution,  ov\  ing  to  the  counter  attraction  exerted 
by  the  membrane  on  the  solvent.  The  fall  of  the  curves  towards 
the  right  is  unexpected.  (A  similar  fall  was  indicated  by 
Curve  I.)  The  permeability  of  the  membrane  to  lithium 
chloride  will  not  account  for  this  :  at  most  it  could  only  make 
that  portion  of  the  curve  horizontal.  The  only  explanation 
that  seems  reasonable,  supposing  the  upper  part  of  the  curve 
is  reliable,  is  that  beyond  a  certain  strength  of  solution  there 
is  polymerization  of  the  solute  or  a  more  complicated  grouping 
of  the  solnte  and  solvent  molecules.  This  is  supported  by 
the  form  of  Curve  IV. 
Curves  II.  and  III.  cut  the  ;r-axis  at  points  denoting  very 
weak  solutions  ;  and  therefore  indicating  that  the  equivalent 
strength  of  the  membrane,  in   its  attraction  for  the  solvent, 
2  n  •> 
