Am'jui7ri9Pi4arm' }  Estimation  of  Morphine  in  Pills,  Etc.  309 
Two  further  assays  of  the  same  solution  were  made,  but  after 
adding  the  phenyl-ethyl  alcohol  and  benzene  mixture  and  shaking, 
the  separators  were  allowed  to  stand  over  night. 
The  morphine  was  precipitated  through  the  liquid  and  on  the 
sides  of  the  separators.  The  crystals  of  morphine  were  washed  of! 
with  alcohol  and  the  assay  completed.  A,  using  cochineal  as  in- 
dicator, gave  0.1872  Gm.  morphine  sulphate  =  93.6  per  cent.  B, 
with  hematoxylin  as  indicator,  gave  0.1805  Gm.  morphine  sul- 
phate =  90.25  per  cent.  2  c.c.  n/io  acid  and  4  or  5  drops  of  cochi- 
neal were  added  to  B  and  again  titrated,  giving  0.1841  Gm.  morphine 
sulphate  =  92.05  per  cent. 
It  was  thought  that  perhaps  by  replacing  the  benzene  in.  the 
phenyl-ethyl  alcohol  and  benzene  mixture  with  benzene  or  petro- 
leum ether  a  sharper  separation  would  be  obtained.  This  was  tried 
on  a  solution  of  1  Gm.  of  morphine  alkaloid  and  50  c.c.  n/io  acid 
in  100  c.c.  Ten  cubic  centimetres  of  this  solution  titrated  direct 
showed  the  presence  of  0.0981  Gm.  morphine  alkaloid.  Ten  cubic 
centimetres  extracted  by  the  above  modified  method,  replacing  the 
benzene  with  petroleum  ether,  gave  a  sharp  separation  but  low  re- 
sults, 0.08729  Gm.  =  87.29  per  cent,  being  obtained. 
Judging  from  the  limited  number  of  assays  made,  the  method 
is  unsatisfactory  both  as  to  accuracy,  time  required,  and  cost  and 
availability  of  material. 
As  above  stated,  the  morphine  does  not  seem  to  be  dissolved  in 
the  phenyl-ethyl  alcohol  and  benzene  mixture — at  least  not  in  the 
quantity  of  solvent  specified. 
The  U.  S.  P.  gives  the  solubility  of  morphine  in  alcohol  as  1-168 
and  in  chloroform  as  1-1800,  but  in  a  mixture  of  these  two  solvents 
it  will  dissolve  far  more  freely  than  in  either  of  them  separately. 
In  order  to  ascertain  the  relative  solubility  of  the  freshly-pre- 
cipitated alkaloid  in  these  solvents,  morphine  sulphate  was  added 
in  excess  to  alcohol,  chloroform,  and  a  mixture  of  alcohol  1  part 
and  chloroform  2  parts,  respectively.  Sufficient  ammonia  was  added 
to  liberate  the  alkaloid,  and  the  flasks  shaken  for  two  or  three 
hours.  Ten  cubic  centimetres  of  the  filtered  liquid  were  then  evap- 
orated to  dryness  and  the  alkaloid  estimated  volumetrically  as  crys- 
talline morphine.  The  averages  of  several  estimations,  weight  to 
volume,  were  as  follows : 
