THE  AMERICAN 
JOURNAL  OF  PHARMACY. 
JULY,  1877. 
EXTRACTION  OF  CAFFEINA  FROM  GUARANA. 
By  Francis  V.  Greene,  M.D.,  U.S.N. 
In  determining  the  percentage  of  caffeina  in  guarana,  Stenhouse 
("Pharm.  Jour.,"  1st  ser.,  vol.  xvi,  p.  212,  "Amer.  Jour.  Phar.,"  1857, 
p.  68)  employed  the  following  process  :  The  finely  powdered  substance 
was  boiled  for  some  time  with  distilled  water,  the  insoluble  portion  re- 
moved by  filtration,  and  a  slight  excess  of  basic  acetate  of  lead  added  to 
the  filtrate.  The  resulting  brownish  red  precipitate  was  thoroughly 
washed  with  boiling  water,  and  sulphuretted  hydrogen  gas  passed  through 
the  filtered  liquid.  The  sulphide  of  lead  was  filtered  off",  the  filtrate 
evaporated  to  drynesss  on  a  water  bath,  and  the  residue  dissolved  in  a 
small  quantity  of  boiling  alcohol  and  filtered.  On  evaporating  this 
liquid  nearly  to  dryness,  yellowish  crystals  were  obtained,  which,  after 
being  pressed  between  folds  of  bibulous  paper,  were  dissolved  in  diluted 
alcohol.  By  the  evaporation  of  this  menstruum,  the  crystals  of  cafFeina 
were  obtained  perfectly  free  from  color. 
In  repeating  this  process  for  the  purpose  of  estimating  the  quantity 
of  cafFeina  contained  in  a  specimen  of  guarana  procured  from  the  Bra- 
zilian collection  at  the  Centennial  Exhibition,  the  difficulty  of  separa- 
ting the  solution  from  the  portion  insoluble  in  boiling  water,  and  the 
tediousness  of  the  process  of  washing  the  mass  precipitated  by  the 
acetate  of  lead  solution  led  me  to  seek  some  other  method  by  which 
these  impediments  might  be  avoided,  and  I  therefore  determined  to 
attempt  the  separation  by  means  of  litharge,  which  substance  has  been 
highly  recommended  ("  Amer.  Jour.  Pharm.,"  1875,  p.  135)  by  Prof- 
E.  S.  Wayne,  for  the  extraction  of  cafFeina  from  tea  and  coffee.  The 
result,  confirmed  by  several  trials  of  the  process,  proves  that  this  com- 
pound of  lead  answers  equally  well  for  guarana,  and  that  by  its  employ- 
22 
