484  Guar  ana.    Methods  Jor  its  Assay.  {Am'o™!iS!arm 
wards  passing  the  liquid  through  an  ordinary  filter  and  using  the 
wash  water  of  the  first  filter  as  wash  water  for  the  second  filter.  The 
washing  should  be  continued  until  they  cease  to  precipitate  with  phos- 
phomolybdic  acid.  The  operation  of  filtering  and  washing  does  not 
take  much  time  and  if  the  cotton  filter  cloggs,  it  is  easy  to  stir  up 
the  insoluble  fatty  or  waxy  matter,  so  as  to  permit  the  liquid  to  pass 
through.  The  aqueous  liquid,  which  will  usually  measure  40  cc.  to 
50  cc.  is  next  placed  in  a  separator  and  washed  with  successive  por- 
tions of  chloroform,  or  until  the  aqueous  liquid  ceases  to  precipitate 
with  phosphomolybdic  acid.  Two  washings  are  generally  sufficient, 
though  the  writer  always  applies  the  test  before  discontinuing  the 
washing,  and  in  fact  it  is  always  better  for  the  operator,  wherever 
possible,  to  assure  himself  by  some  such  a  test,  that  the  alkaloid  is  all 
contained  in  the  liquid  supposed  to  contain  it.  The  chloroform  so- 
lution is  now  evaporated  in  a  weighed  beaker,  and  the  residue  dried 
at  100°  C.  for  half  an  hour,  cooled  and  again  weighed.  The  weight 
in  grammes  multiplied  by  25  gives  the  percentage  of  alkaloid.  This 
process  gives  somewhat  higher  results  than  that  of  Greene,  which  in  a 
sample  of  guarana  showed  4.14  per  cent.,  duplicate  4.28  per  cent. 
The  same  drug  assayed  by  the  same  process,  showed  by  four  different 
assays,  4.64  per  cent.  4.66  per  cent.  4.84  per  cent,  and  4.89  per  cent, 
of  caffeine.  The  writer  also  tried  Prollius'  fluid  for  the  assay  of  cin- 
chona barks,  but  without  success.  The  same  drug,  as  above,  showing 
only  3.97  per  cent,  of  alkaloid.  In  general,  ether  is  not  so  well 
suited  for  the  extraction  of  caffeine,  as  either  benzol  or  chloroform, 
owing  to  the  comparative  insolubility  of  the  alkaloid  in  that  men- 
struum. Ammoniated  ether  and  ammoniated  chloroform  were  both  tried 
for  the  extraction  of  caffeine  from  colanuts  yielding  in  the  one  case  1.34 
per  cent,  and  in  the  other  1.45  per  cent  of  alkaloid,  which  did  not, 
however,  seem  in  such  a  pure  condition  as  in  the  case  of  guarana. 
In  "The  New  Idea"  the  writer  gave  the  results  of  a  number  of  assays 
of  fluid  extract  of  guarana,  which  will  probably  be  of  interest  here. 
The  process  employed  was  practically  as  follows:  Evaporate  5  cc. 
of  the  fluid  extract  to  dryness,  with  about  one  ounce  of  litharge  pow- 
der, introduce  into  a  flask  of  about  120  cc.  to  150  cc.  capacity,  and 
pour  in  100  cc.  of  chloroform,  and  shake  repeatedly  during  three  or 
four  hours,  and  finally  filter  through  a  dry  filter,  containing  some 
magnesia,  which  removes  suspended  particles  of  insoluble  matter,, 
which  would  be  difficult  to  remove  without  some  such  device  as  the 
