462    Estimation  of  Morphine  and  Lloyd's  Reagent.  {A^tJX.rr"  Kf*' 
saturated  with  potassium  carbonate,  either  ordinary  or  previously 
washed  with  alcohol,  and  then  shaken  with  a  definite  volume  of  equal 
volumes  of  a  mixture  of  alcohol  and  chloroform.  After  complete 
separation  of  the  liquid  into  two  layers,  an  aliquot  portion  of  the 
alcoholic  layer  was  drawn  off  and  evaporated  to  dryness.  The  residue 
was  washed  with  a  saturated  solution  of  morphine  in  water  until  the 
washings  gave  no  test  for  potassium  carbonate  with  phenolphthalein, 
and  the  morphine  determined  alkalimetrically,  using  N/25  H2S04 
and  N/50  KOH.   The  indicator  was  methyl-red. 
The  experiments  showed  that  in  all  cases  the  amount  of  morphine 
found  exceeded  the  amount  originally  taken,  the  variation  being  from 
2  to  15  per  cent.  Hence  the  method,  at  least  in  the  form  here  de- 
scribed, is  not  reliable. 
2.  Extraction  of  alkaloids  by  means  of  Lloyd's  reagent. 
Owing  to  the  facility  and  completeness  of  precipitation  of  al- 
kaloids by  Lloyd's  reagent,  it  was  thought  that  this  reagent  could  be 
advantageously  used  for  the  quantitative  extraction  of  alkaloids  from 
their  original  sources  or,  in  general,  from  the  solution  of  their  salts 
in  water. 
It  is  evident  that  in  order  to  attain  this  aim  it  is  necessary  to 
prove  that  the  alkaloids,  once  precipitated  by  Lloyd's  reagent,  can  be 
readily  and  completely  recovered  from  the  precipitate  containing 
alkaloid  and  reagent.  With  a  view  of  determining  this  point,  the 
following  experiments  were  carried  out : 
A  definite  amount  of  morphine  was  dissolved  in  an  excess  of 
dilute  sulfuric  acid,  the  alkaloid  completely  precipitated  with  an  ex- 
cess of  Lloyd's  reagent,  and  the  precipitate  washed  with  water  till  the 
washings  gave  no  test  for  sulfuric  acid.  The  precipitate  was  dried 
at  6o°  and  then  repeatedly  extracted  with  boiling  methyl  alcohol, 
which  is  a  very  fair  solvent  for  morphine.  The  solvent  was  evapo- 
rated to  dryness,  and  the  residue  weighed.  This  residue  was  free 
of  sulfuric  acid,  showing  it  to  be  probably  free  morphine,  but  its 
amount  was  less  than  4  per  cent,  of  the  morphine  originally  taken. 
The  precipitate  was  then  again  extracted  with  methyl  alcohol  to 
which  a  small  amount  of  ammonia  had  been  added,  and  the  residue 
left  after  again  evaporating  the  solvent  weighed.  The  total  amount 
of  alkaloid  recovered  by  the  two  successive  extractions  was  about  90 
per  cent,  of  the  morphine  taken. 
Another  experiment  was  made  with  strychnine,  using  chloroform, 
which  is  an  exceptionally  good  solvent  for  this  alkaloid.    A  dilute 
