318  ON  PERCOLATION  OR  DISPLACEMENT. 
Unfortunately,  many  persons  who  employ  percolation  habitu- 
ally, do  not  understand  its  principles  and  details  sufficiently 
clear  to  gain  all  the  advantages  it  affords,  or  even  to  get  pro- 
ducts, in  many  cases,  equal  to  those  by  the  old  methods.  My 
attention  has  been  reattracted  to  this  subject,  partly  by  the 
near  approach  of  anew  edition  of  our  Pharmacopoeia,  and  partly 
by  the  excellent  paper  of  Prof.  Grahame,  read  at  the  last  meet- 
ing of  the  Association,  at  Washington,  which  has  been  reprinted 
on  page  354  of  the  present  number  from  the  published  Proceed- 
ings ;  and  I  propose  first  to  consider  what  the  displacement  pro- 
cess was  intended  to  accomplish,  and  the  conditions  to  be  at- 
tended to,  and  afterwards  remark  on  the  results  of  Prof. 
Grahame. 
The  object  to  be  obtained  in  practice  by  Boullay's  theory  is 
this  :  a  solvent,  poured  on  the  top  of  a  powder  consisting  par- 
tially of  soluble  matter  contained  in  a  cylindrical  vessel  and  sup- 
ported on  a  porous  diaphragm  descends  from  layer  to  layer  by 
capillary  attraction,  and  its  own  gravity,  exerting  its  solvent 
power  on  each  successive  layer  until  its  power  of  solution  is  ex- 
hausted, after  which  it  continues  to  descend  by  the  pressure  of 
the  superincumbent  fluid,  until  forced  out  through  the  diaphragm 
into  the  vessel  below,  a  saturated  solution  ;  this  process  continu- 
ing until  the  soluble  matter  is  so  far  removed  from  the  powder 
that  the  liquid  by  the  contact  becomes  less  and  less  charged 
with  the  soluble  matter  until  exhausted.  But  to  gain  this  result 
it  is  absolutely  necessary  that  the  substance  treated  shall  be  in 
a  uniform  powder,  and  that  the  capillarity  or  porousness  of  the 
mass  of  powder  be  not  destroyed  by  any  cause  whatever,  for,  on 
the  fact  of  the  slow,  regular,  and  even  descent  of  the  solvent,  from 
one  horizontal  layer  to  the  next  without  side  channels  or  circuits, 
caused  by  irregular  powdering  or  imperfect  packing,  depends 
the  success  of  the  process.  Now  what  are  the  causes  which 
interfere  with  this  important  condition  ?  They  are  several : — 1. 
The  imperfect  preparation  of  the  powder  owing  to  the  difference 
of  opinion  as  to  the  degree  of  fineness  it  should  be  made  to 
assume,  and  to  the  careless  dislike  of  the  trouble  it  causes.  To 
say  that  all  substances  should  be  equally  fine  would  be  incorrect, 
as  in  substances  containing  a  large  proportion  of  matter  very 
soluble  in  the  liquid  used,  the  powder  should  not  be  quite  so  fine 
