Am.  jour.^Pharm. }       Determination  of  Morphine. 
691 
way.  Experiment  3  shows  that  the  reaction  is  strictly  quantitative, 
and  that  the  iodine  solution  need  not  be  standardized  directly  on 
morphine. 
1  odo-Acidimetrlc  Titration  of  Morphine. — This  method,  pro- 
posed by  Gordin  (Pharm.  Archives,  1899,  2,  313-318)  appears  so 
plausible  that  it  finds  a  place  in  works  on  alkaloidal  assays,  despite 
Kippenberger's  destructive  criticism  (Zeit.  Anal  Chem.  42,  101- 
108).  Experiment  4  leads  to  the  belief  that  the  sodium  tetra- 
thionate  formed  in  the  course  of  the  reaction  is  partly  responsible 
for  its  entire  unreliability. 
Extraction  of  Morphine  by  Solvents  Immiscible  with  Water- 
Chloroform- Alcohol  Mixtures. — Many  solvents  have  been  proposed 
for  the  extraction  of  morphine.  Most  offer  great  practical  disad- 
vantages. Some  are  extremely  apt  to  form  emulsions  with  alkaline 
liquids,  others  are  specifically  lighter  than  water,  so  that  their  use 
entails  greater  manipulative  losses.  Those  of  high  boiling-point  are 
very  objectionable,  as  the  morphine  can  only  be  quantitatively  sepa- 
rated from  them  by  re-solution  in  acid.  None  of  these  disadvantages 
adhere  to  chloroform-alcoholic  mixtures.  The  writer  prefers  to 
use  2  volumes  of  chloroform  to  1  volume  of  alcohol,  as  advocated 
by  J.  B.  Williams  (Amer.  Jour.  Pharm.,  1914,  86,  308-312).  That 
four  extractions,  made  under  the  conditions  there  defined,  will  com- 
pletely separate  morphine,  is  shown  by  Experiments  5  and  6.  The 
very  slight  deficiency  of  morphine  at  the  end  of  these  experiments 
is  within  the  limit  of  error  due  to  manipulation,  etc.,  and  direct  tests 
establish  that  no  alkaloid  was  left  in  the  aqueous  solution.  W.  A. 
Puckner  (Jour.  Amer.  Chem.  Soc,  1901,  XXIII,  470-473)  depre- 
cates the  use  of  ammonium  hydroxide  for  the  liberation  of  mor- 
phine, but  it  should  be  remarked  that  his  gravimetric  determinations 
were  admittedly  of  no  value,  while  the  errors  attributed  to  the  use 
of  ammonium  hydroxide  rather  than  sodium  bicarbonate  are  less 
than  the  smallest  error  that  his  volumetric  method  could  detect.  He 
worked  with  chloroform-alcohol  mixtures  in  which  the  proportion 
was  4:1.  Kippenberger  has  published  some  observations  on  this 
subject  (Zeit.  Anal  Chem.,  XXXIX,  290),  but  his  paper  is  not  ac- 
cessible to  the  present  writer.  Chloroform-alcohol  mixtures  extract 
small  quantities  of  morphine  from  solution  even  in  the  presence  of 
alkaline  hydroxides  or  free  mineral  acid.  In  the  latter  case  the 
morphine  is  extracted  in  the  form  of  a  salt,  as  proved  by  Exper- 
iment 7. 
