ADo'ctober,^9oaLm'}    American  Pharmaceutical  Association.  493 
two  c.c.  concentrated  sulphuric  acid,  previously  mixed  with  a  few 
c.c.  alcohol,  is  added,  the  mixture  diluted  with  ether  to  about  double 
its  volume  and  the  assay  finished  exactly  as  just  described.  In  the 
case  of  hydrastis,  the  rest  of  the  alcoholic  filtrate  can  be  used  for 
the  estimation  of  hydrastine.  For  this  purpose  the  alcohol  from  25 
c.c.  of  the  alcoholic  extract  is  distilled  off  till  only  a  few  c.c.  are  left, 
the  residue  diluted  with  water  containing  about  one  per  cent,  acetic 
acid  and  a  few  per  cent,  potassium  iodide  to  25  c.c.  The  liquid  is 
then  filtered  and  12.5  c.c.  of  the  filtrate  are  treated  as  described  in 
a  previous  paper.  If  the  liquid  in  which  the  berberine  is  to  be 
estimated  is  a  strong  alcoholic  extract  like  normal  tincture  of  hy- 
drastis, 20  c.c.  are  diluted  with  four  times  its  amount  of  alcohol  to 
100  c.c,  filtered  if  necessary,  and  25  c.c.  of  the  filtrate  are  treated 
exactly  as  above  described. 
This  method  is  not  well  adapted  to  the  assay  of  liquids  contain- 
ing a  considerable  amount  of  water  or  containing  no  alcohol  at  all, 
like  fluid  extract  of  hydrastis  without  alcohol.  From  such  liquids, 
even  after  dilution  with  alcohol  and  filtration,  sulphuric  acid  and 
ether  precipitates  much  coloring  matter  besides  the  berberine  salts, 
so  that  after  the  addition  of  potassium  iodide  and  filtration  as  above 
described,  the  filtrate  is  sufficiently  colored  through  the  presence  of 
the  coloring  matter  to  make  the  final  reaction  lack  in  sharpness. 
Owing  to  the  quickness  and  simplicity  of  this  assay  method,  it 
might  be  adopted  by  many  even  in  those  cases  where  the  final  reac- 
tion is  not  very  sharp,  i.e.,  for  solution  of  berberine  salts  containing 
much  water. 
But  a  much  more  exact  assay  method  which  can  be  used  in  all 
cases  is  as  follows  : 
(2)  Another  method  of  estimating  berberine  in  liquids  containing, 
much  other  matter  is  to  separate  the  berberine  by  precipitating  it 
as  an  insoluble  hydroiodide,  washing  thoroughly  with  water  con- 
taining a  little  potassium  iodide  and  converting  the  moist  hydro- 
iodide  into  the  very  insoluble  and  beautifully  crystalline  berberine 
acetone.  The  latter  can  then  be  thoroughly  washed  with  water  and 
after  drying  at  105 0  C.  to  constant  weight,  weighed.  One  gm.  ber- 
berine acetone  is  equivalent  to  0.8524  gm.  berberine.  In  order  to 
obtain  the  acetone  compound  in  a  crystalline  form  suitable  for  wash- 
ing it  is  necessary  that  the  liquid  should  be  warm  and  should  contain 
about  33  per  cent,  acetone. 
