ECONOMY  IN  THE  USE  OF  ALCOHOL  IN  PERCOLATION. 
9 
The  process  of  repercolation  is  very  well  adapted  to  this 
powder,  and,  when  carefully  and  skillfully  performed,  saves  at 
least  one-third  of  the  alcohol  otherwise  required.  The  details  of 
this  process  given  here  are  such  as  especially  apply  to  this 
powder,  and  here  again  the  slowness  of  the  process  is  indispensa- 
ble to  success.  Indeed,  the  farther  the  writer's  experience 
extends  upon  this  subject  of  percolation  and  repercolation,  the 
more  evidence  is  obtained  as  to  the  primary  importance  of  time 
in  effecting  the  exhaustion  of  the  powders.  Nine  or  ten  days 
is  required  for  the  proper  performance  of  this  repercolation, 
and  the  difficulty  of  avoiding  loss  of  alcohol  by  evaporation 
during  this  prolonged  period  is  a  serious  objection  to  the  process. 
When  the  alcohol  has  nearly  ceased  to  come  over,  the  ex- 
tract in  the  still  measures  five  fluidounces,  the  water  added 
brings  this  measure  up  to  about  eight  fluidounces,  and  the  addi- 
tional alcohol  brings  this  up  to  twelve  fluidounces.  There  is 
therefore  about  six  fluidounces  of  alcohol  in  the  extract  when 
poured  in  for  precipitation,  and  this  added  to  the  two  and  a  half 
pints  received,  leaves  the  loss  by  evaporation  to  be  estimated  at 
two  pints  or  more  of  the  five.  By  special  apparatus  and  particu- 
lar care  this  can  be  much  reduced,  but  the  aim  here  is  to  pre- 
sent the  process  as  it  would  be  conducted  in  common  practice. 
The  addition  of  water  to  the  concentrated  extract  in  the  still 
is  useful,  first,  by  rendering  the  extract  more  liquid,  and  further, 
by  dissolving  the  large  portion  which  is  soluble  in  water  to  render 
it  more  easily  separable  from  the  resin  in  the  precipitation.  The 
addition  of  this  water  precipitates  much  of  the  resin  and  renders 
the  extract  muddy,  even  when  heated,  but  the  after  addition  of 
the  alcohol  makes  the  whole  a  perfect  and  clear  solution  as  it 
should  be  before  being  poured  into  the  acidulated  water.  This 
pouring  in  should  be  slow,  and  the  stirring  active,  to  avoid 
clotting  of  the  precipitate.  With  the  proportion  of  acid  indi- 
cated, the  precipitate  settles  out  promptly,  so  that  in  two  hours 
more  than  one  half  of  the  mother  water  may  be  poured  off  and 
thrown  away.  Another  water  put  on  and  decanted  in  the  same 
way,  and  then  a  displacement  washing  in  the  filter  are  considered 
quite  sufficient.  The  resin  is  rather  nicer  and  lighter  the  slower 
it  is  dried,  and  when  it  does  not  mould  is  perhaps  best  dried 
