5i8  Assay  Processes  of  the  U.  S.  P.  {^S^ov^XrfJJiT' 
struum  than  by  percolating.  A  percolator  is  perhaps  our  most 
unscientific  piece  of  apparatus.  A  channel  might  be  formed  in  the 
packed  drug,  the  parts  adjoining  this  channel  may  be  exhausted, 
while  other  parts  of  the  drug  come  in  contact  with  the  menstruum 
only  superficially.  The  method,  besides  that,  is  very  tedious,  espe- 
cially when  such  a  fine  powder  (No.  60)  as  prescribed  by  the  U.S. P. 
is  employed,  and  also  requires  a  larger  amount  of  menstruum  for  the 
exhaustion. 
We,  therefore,  strongly  recomniend  the  adoption  of  the  aliquot 
part  method,  having  proven  by  numerous  experiments  that  the 
results  by  this  method  compare  favorably  with  those  obtained  by 
exhausting  the  drug  completely  by  percolation. 
For  the  final  shaking  out  of  the  alkaloids,  we  recommend  to  use, 
whenever  possible,  simple  menstrua,  viz.,  ether  or  chloroform  and 
not  mixtures  of  both  in  various  proportions.  As  a  rule,  simple 
menstrua  are  less  liable  to  produce  emulsions  than  mixtures,  and  the 
menstrua  are  more  easily  recovered  for  future  use  than  mixtures, 
which  always  require  tiresome  adjusting. 
For  the  extraction  of  the  drugs,  however,  a  mixture  of  ether- 
chloroform  is  to  be  preferred.  Such  a  mixture  seems  tO'  penetrate 
the  cell-walls  better  than  a  simple  menstruum,  and  consequently 
to  extract  the  alkaloids  more  thoroughly.  It  is  to  be  recommended 
to  allow  the  drug  to  stand  with  the  menstruum  for  at  least  one- 
quarter  hour  before  adding  the  ammonia,  as  the  results  obtained  by 
doing  so  are  somewhat  higher,  in  our  opinion,  than  those  obtained 
by  adding  ether-chloroform  and  ammonia  to  the  drug  together  at 
once. 
Whenever  possible  the  alkaloids  should  be  estimated  by  titra- 
tion ;  only,  in  some  cases  when  hydrolysis  is  liable  to  take  place,  as 
in  aconite,  coca  leaves,  etc.,  a  check  by  gravimetric  estimation  rnight 
be  of  advantage. 
Of  all  the  indicators  for  alkaloids,  we  have  found  cochineal  to 
be  the  best,  since  only  in  titrating  the  alkaloids  of  ipecac  is  any 
difficulty  experienced  with  this  indicator.  lodeosin,  at  present  used 
in  the  U.S. P.,  is  rather  unreliable  since  the  aqueous  liquid  is  not 
always  colored  red  when  the  end  point  is  reached,  but  at  times  a 
red  scum  is  formed  at  the  contact  of  the  two  layers,  the  color  of  this 
scum  increasing  in  intensity  with  the  addition  of  the  alkali.  It  is 
difficult  to  judge,  in  case  this  happens,  when  the  end  point  is 
reached. 
