246  Fluid  Extracts  by  Reper eolation.  {Am£y%8^ 
no  stopcock  could  be  found  which  would  continue  to  run  at  so  slow  a 
rate  of  dropping  with  any  degree  of  uniformity  through  so  long  a  time. 
The  smallest  particle  of  solid  matter  would  diminish  the  rate,  or  stop 
it  altogether,  so  that  it  required  to  be  continually  watched  or  readjusted,, 
and  every  change  that  was  made  disturbed  the  whole  of  the  delicate 
balances  and  motions  of  the  process.  Again  all  the  metals  of  which 
ordinary  stopcocks  are  made  are  attacked  by  the  liquids  in  so  long  a 
process,  and  have  the  same  objection  that  metallic  percolators  have. 
Glass  stopcocks  were  tried,  but  they  were  found  equally  liable  to  all 
the  objections  except  the  last.  This  experience  led  the  writer  directly 
to  the  well-tube,  and  to  the  use  of  a  syphon,  f\  as  best  fulfilling  all  the 
conditions  required,  since  it  can  easily  be  set  lower  or  higher,  to  adjust 
the  levels  to  the  desired  rate  of  motion,  and  having  a  free  flow  it  works 
automatically  and  with  certainty.  This  syphon,  f,  is  made  of  glass 
tubing  of  about  '125  inch  =  3  millimetres  bore,  bent  twice  at  right 
angles,  the  two  legs  being  about  12*5  inches  =  31  centimetres  long. 
The  outer  leg  is  a  little  longer  than  the  inner  one,  and  turned  up  upon 
itself  for  about  '750  inch  =  2  centimetres,  as  shown  in  the  cuts.  The 
legs  should  have  only  such  a  difference  in  length  that  the  inner  one 
should  reach  the  bottom  of  the  well-tube  when  required,  and  when 
measured  upon  the  outer  one,  should  reach  to  about  midway  of  the 
turned  up  end  of  the  outer  leg.  This  construction  prevents  the  syphon 
from  emptying  itself  at  any  time,  for,  when  the  liquid  is  drawn  over 
by  the  syphon  until  the  surface  of  liquid  in  the  well-tube  falls  to  a  level* 
with  the  end  of  the  turned  up  portion,  as  shown  by  the  lines  in  the  cut 
of  the  larger  percolator,  the  columns  of  liquid  in  the  syphon  will  be  of 
equal  length,  and  will  counterbalance  each  other,  and  therefore  the 
flow  will  cease  without  emptying  the  syphon.  But  as  soon  as  the 
level  of  liquid  in  the  well  is  raised  by  fresh  additions  of  menstruum 
on  to  the  substance,  the  flow  will  recommence  at  a  rate  proportionate 
to  the  difference  of  levels,  and  may  be  readjusted  to  the  required  rate 
by  slipping  it  up  or  down  in  the  cork,  i,  in  the  upper  end  of  the  well- 
tube.  This  cork,  ^,  should  be  bored  to  fit  the  syphon  so  tightly  as  to 
hold  it  in  any  position  and  should  have  a  groove  filed  longitudinally  on 
its  outer  side  so  as  to  allow  free  entrance  and  exit  of  air  to  the  well- 
tube.  A  receiving  bottle,  /,  upon  which  a  strip  of  paper  is  pasted  to 
receive  any  graduation  marks  that  may  be  desired,  completes  the  appa- 
ratus.   The  strip  of  paper  on  the  bottle  should  have  the  tare  of  the 
