ON  PERCOLATION  OR  DISPLACEMENT. 
321 
half  its  weight  of  water,  which  does  not  destroy  its  pulveru- 
lent condition,  and  treat  it  in  a  conical  percolator.  Prof. 
Grahame  having  met  with  no  difficulty  on  the  score  of  compac- 
tion by  swelling,  overlooks  this  fruitful  source  of  trouble  in  di- 
rect displacement  with  the  ordinary  cylinder,  owing  to  his  hav- 
ing accidentally  employed  the  funnel  as  a  convenient  percolator, 
and  his  success  with  those  substances  that  ordinarily  require 
previous  maceration  to  swell  them  is  largely  attributable  to  this 
cause.  When  a  damp  vegetable  powder  is  packed  in  the  conical 
cavity  of  a  funnel  and  covered  with  paper  to  avoid  the  derange- 
ment of  its  strata,  and  water  is  poured  carefully  on,  the  liquid 
is  absorbed,  the  insoluble  cellular  structure  of  the  particles  be- 
comes swollen,  and  the  whole  mass,  stratum  after  stratum,  ex- 
panding laterally  and  vertically,  finds  no  difficulty  in  relieving 
itself,  owing  to  the  inclination  of  the  sides  ;  but  in  a  cylinder 
the  lateral  expansion  gaining  no  relief  except  vertically,  the 
particles  are  compacted  together  often  to  such  a  degree  as  to 
prevent  the  percolation  altogether.  I  believe  that,  philosophi- 
cally speaking,  the  cylinder  is  the  proper  shape  of  a  percolator 
in  cases  where  expansion  presents  no  difficulty,  but  where  this 
occurs,  the  use  of  a  cone,  as  employed  by  Prof.  Grahame,  has 
very  important  advantages. 
Whatever  form  of  apparatus  is  used,  or  however  the  packing 
may  be  affected,  the  operator  should  either  by  a  disc  of  paper, 
muslin,  or  lint,  or  a  layer  ofsand,  cover  the  surface  of  the  ingre- 
dients so  that  the  addition  of  fluid  shall  not  disturb  the  stratifi- 
cation of  the  powders.  Boullay  employed  an  upper  metallic 
diaphragm.  If  paper  or  muslin,  etc.,  is  U5ed,  it  should  be 
soaked  in  the  liquid,  and  carefully  pressed  into  its  position  so 
as  to  remove  the  air  beneath  before  adding  the  menstruum,  else 
it  will  be  displaced  by  its  buoyancy. 
But  the  peculiar  merit  of  Prof.  Grahame's  paper  is,  that  it 
insists  on  the  invariable  production  of  a  highly  concentrated 
solution  at  first,  (a  result  always  attained  most  effectually  by 
direct  displacement,)  which  enables  the  operator  to  fraction  his 
product,  to  make  less  menstruum  suffice  for  exhaustion,  and  in 
cases  requiring  evaporation,  shielding  the  principles  from  the  in- 
jury of  prolonged  heating.  He  also  can  avail  himself  of  it  in 
making  fluid  extracts,  by  reserving  the  densest  first  liquid,  eva- 
