I 
338  ON  FRACTIONAL  PERCOLATION. 
they  have  proved  reliable,  as  far  as  we  are  enabled  to  judge 
without  actual  analysis  of  the  resulting  preparations.  Yet 
several  objections  to  the  process  exist ;  the  first  being  the  one 
mentioned  above  with  regard  to  the  concentration  of  hydro-alco- 
holic percolates  ;  the  next,  and  most  important  one  is,  that,  if  a 
fluid  extract  is  at  all  liable  to  be  injured  by  heat,  a  portion  at 
least  is  sacrificed,  and  we  cannot  obtain  an  absolute  representa- 
tive of  a  normal  quantity  of  the  drug.  Dr.  Squibb,  by  his  ex- 
periments upon  the  fluid  extracts  of  colchicum  seed  and  of 
buchu,  is  satisfied^that,  by  even  ordinary  care  in  percolation,  the 
first  twelve  fluidounces  of  percolate  from  sixteen  troyounces  of 
the  drug,  will  contain  two-thirds  of  the  soluble  matter  contained 
therein,  and  that  the  remaining  one-third  of  soluble  matter  is 
weight  for  weight  less  active  than  the  ^v^i  two-thirds.  Granting 
this  to  be  true  of  drugs  in  general,  and  that  the  deficiency  of 
activity  in  the  last  one-third  will  bring  its  actual  value  to  one- 
fourth,  there  still  occurs  a  great  loss,  if  the  drug  is  readily  altered 
by  heating.  This  must,  for  instance,  be  true  of  fluid  extract 
of  senna,  in  the  formula  for  which  all  due  regard  has  been  paid 
to  the  changeable  nature  of  the  active  principle ;  but  as  the  first 
pint  of  percolate  does  not  contain  all  the  active  matter  of  the 
senna,  it  follows  that,  in  evaporating  the  subsequent  percolate, 
a  portion  of  activity  is  sacrificed,  and  that  therefore  a  pint  of 
fluid  extract  can  not  absolutely  represent  sixteen  troy-ounces  of 
senna.  Similar  objections  pertain  to  the  process  for  fluid  extract 
of  rhubarb.  It  was  with  a  view  to  improving  the  process  for 
these  preparations,  that  I  availed  myself  of  the  valuable  hints 
contained  in  the  papers  of  Dr.  Squibb,  modifying  his  sugges- 
tions as  appeared  to  me  desirable.  In  explanation  of  the 
heading  to  this  paper  I  will  say  that  I  regarded  the  term  frac' 
tional  percolation  more  to  the  point  than  repereolation,  in  desig- 
nating a  process  depending  upon  the  percolation  of  fractional 
portions  of  a  drug  by  one  and  the  same  menstruum. 
Experiment  1. — Twenty-four  troyounces  of  Alexandria  senna  in  mod- 
erately fine  powder  was  divided  into  portions  of  ten,  eight  and  six  ounces. 
The  first  portion  (of  ten  ounces)  was  moistened  with  five  fluidounces  of 
diluted  alcohol,  and,  after  standing  a  short  time,  packed  appropriately  in 
a  glass  funnel,  the  surface  covered  with  a  disk  of  filtering  paper,  and  suf- 
ficient diluted  alcohol  poured  on,  as  required.  At  the  expiration  of  about 
