26o 
Hydrastine  Hexiodide. 
Am.  Jour.  Pharm. 
June.  1899. 
PLAN  OF  THE  ASSAY  OF  HYDRASTIS  CANADENSIS. 
The  estimation  of  hydrastine  and  berberine  in  this  root  is  based 
upon  the  following  principles : 
(1)  Hydrastine  is  quite  soluble  in  absolute  ether,  and  forms,  as 
we  have  just  shown,  a  definite  hexiodide  when  a  weak  solution  of 
any  of  its  salts  is  added  to  a  large  excess  of  iodine  dissolved  in  a 
potassium  iodide  solution. 
(2)  Berberine,  on  the  contrary,  is  completely  insoluble  in  abso- 
lute ether.  This  can  be  shown  by  rubbing  up  some  pure  berberine 
in  a  mortar  with  absolute  ether,  filtering,  evaporating  the  ether  and 
taking  up  with  acidulated  v/ater.  The  liquid  thus  obtained  is  per- 
fectly colorless,  and  no  trace  of  turbidity  is  produced  in  it  by  Mayer's 
reagent,  Wagner's  reagent  or  picric  acid,  nor  does  chlorine  water 
produce  the  characteristic  rose  band. 
(3)  Of  all  the  difficultly  soluble  berberine  salts,  the  hydriodide 
seems  to  be  the  least  soluble,  particularly  so  in  a  very  large  excess 
of  potassium  iodide.  If  to  a  solution  of  berberine  in  water  slightly 
acidulated  with  acetic  or  sulphuric  acid  a  large  excess  of  a  potas- 
sium iodide  solution  be  added,  the  precipitation  is  so  complete  that 
the  filtrate  is  almost  entirely  colorless,  and  no  alkaloid  can  be  de- 
tected in  it  by  any  of  the  above-mentioned  reagents.  Even  chlorine 
water,  reacting  with  a  delicacy  said  to  be  1  in  250,000,  does  not 
give  any  coloration,  after  removal  of  potassium  iodide  from  the  fil- 
trate by  silver  nitrate,  and  of  the  excess  of  the  latter  by  hydrochloric 
acid. 
(4)  If  to  a  very  dilute  solution  of  a  salt  of  pure  berberine  about 
ten  to  fifteen  times  its  amount  of  acetone  be  added  and  the  solution 
then  made  strongly  alkaline  with  solution  of  sodium  hydrate,  ber- 
berine-acetone  is  so  fully  precipitated,  particularly  after  ten  or  fifteen 
minutes'  shaking,  that  the  filtrate  is  almost  completely  colorless,  and 
no  traces  of  alkaloid  can  be  detected  in  it  after  acidulation,  by 
means  of  the  above-mentioned  reagents.1    This  berberine-acetone 
1  Besides  berberine  and  hydrastine,  one  other  alkaloid  has  been  found  in 
hydrastis,  but  in  proportions  too  small  to  have  any  bearing  upon  the  assay 
percentages,  either  of  berberine  or  hydrastine.  That  the  assay  process  here 
proposed  leaves  no  appreciable  quantity  of  any  alkaloid  behind,  we  have  veri- 
fied by  operating  upon  a  considerable  quantity  of  the  drug.  After  removing 
the  hydrastine  with  the  ether,  and  the  berberine  with  potassium  iodide,  there 
was  nothing  left  that  gave  indications  of  alkaloidal  character.  What  has  been 
learned  of  canadine  makes  it  probable  that  it  would  adhere  to  berberine  or  to 
