X/ember'^Y92^:}         The  Theory  of  Percolatton.  775 
THK  PKRCOIvATOR. 
The  design  of  the  apparatus  in  which  percolation  is  to  be  con- 
ducted has  received  a  great  deal  of  intelligent  attention  and  in  this 
narrow  field,  our  knowledge  may  be  regarded  as  very  nearly  com- 
plete. Nearly  every  thoughtful  pharmacist  has  studied  the  subject 
and  the  result  is  that  scores  of  new  percolators  have  been  suggested 
during  the  past  eighty  years.  Many  of  these  have  involved  in- 
novations or  radical  changes  in  the  process  of  extraction  and  are  thus 
more  related  to  some  specific  method  than  to  general  percolation. 
For  this  reason  those  percolators  which  substitute  apparatus  with- 
out afi^ecting  the  process  seriously  will  alone  be  considered  in  this  sec- 
tion; the  others  will  be  analyzed  when  we  discuss  the  unusual  and 
special  methods  of  percolation. 
The  early  apparatus  of  the  Boullays  was  essentially  the  cylin- 
drical percolator  sometimes  seen  .to-day.  Emil  Mouchon  of  Lyons^ 
substituted  a  funnel  and  the  concial  shape  is  first  recorded  in  Gil- 
bertson's  apparatus. ^  The  very  tall,  nearly  cylindrical  percolator 
was  advocated  by  Oscar  Oldberg  in  1884.^  The  funnel  shaped,  the 
conical,  and  Oldberg' s  percolators  are  in  common  use  to-day,  the 
conical  being  the  form  most  widely  used.  The  essential  differ- 
ence between  the  three  forms  is,  of  course,  the  taper  upon  which 
depends  the  height  of  the  column  of  drug  and  the  number  of  contacts 
between  each  particle  of  drug  and  each  aggregate  of  menstruum  as 
well  as  the  relative  number  of  such  contacts  for  the  drug  in  the  widest 
part  and  that  in  the  narrowest  part  of  the  percolator. 
Lloyd  has  shown ^  that  the  longer  the  column  of  drug  is,  the  more 
concentrated  the  first  portions  of  percolate  are  and  he  states  two 
rules:  first,  that  the  height  of  both  liquid  and  powder  increase  in- 
versely as  the  square  of  the  diameter  of  the  percolator,  and  second, 
the  contact  between  the  liquid  and  powder  increases  inversely  as  the 
fourth  power  of  the  diameter  of  the  percolator.  The  efficiency  of  the 
Oldberg  design  has  also  been  confirmed  by  E.  Moor,  Jr.^ 
The  importance  of  securing  as  much  concentration  in  the  early 
portions  of  the  percolate  as  possible,  other  things  being  equal,  is 
obvious.    Nevertheless,  for  operations  involving  large  amounts  of 
^  Squibb,  Proc.  A.  Ph.  A.  1867,  391. 
^  Pharm.  Jour,  i,  591,  (1842). 
3  Proc.  A.  Ph.  A.  1884,  388. 
4  Proc.  A.  Ph.  A.  1879,  679. 
^  This  Journal,  Vol.  62,  333,  (1890). 
