AmMay,r'i8>7h8arm"}      Fluid  Extracts  by  Reper eolation.  245 
past.  And  it  can  only  be  effectually  done  by  means  of  maintaining  a 
difference  of  levels  in  the  liquid  inside  and  outside  of  the  well-tube. 
The  moment  a  drop  of  liquid  is  taken  from  the  well-tube,  the  liquid 
outside  tends  to  supply  its  place,  and  will  supply  it  in  a  certain  definite 
time  ;  and  if  the  entire  mass  of  liquid  be  under  the  same  tension,  and 
equally  free  to  move  through  the  short  distance  required  at  the  slow 
rate  required,  then  as  liquids  are  practically  inelastic,  every  drop 
throughout  the  whole  mass,  whether  in  the  interstices  of  the  solid  par- 
ticles or  between  those  particles,  will  have  moved  downward  through 
the  distance  and  at  the  rate  required  to  supply  its  share  of  the  drop 
required  to  replace  the  one  taken  from  the  well.  And  if  the  volume 
of  the  drop  be  compared  with  the  total  volume  of  liquid  set  in  motion, 
and  the  rate  of  its  movement  as  taken  out  be  multiplied  into  the  mean 
horizontal  area  of  the  percolator,  a  good  indication  is  obtained  of  the 
almost  infinite  slowness  with  which  the  vertical  columns  of  molecules 
of  the  liquid  descend  through  the  mass,  and  the  physical  laws  which 
govern  percolation  may  be  better  understood. 
The  proposition  then  is  to  keep  the  total  mass  under  a  uniform  ten- 
sion throughout,  and  to  do  this  by  maintaining  a  difference  in  levels 
between  the  liquid  inside  and  outside  the  well ;  and  this  difference  of 
levels,  which  varies  with  each  substance,  and  with  every  stage  of  the 
percolation  of  the  same  substance,  is  to  be  maintained  by  drawing 
liquid  from  the  well  at  so  slow  a  rate  as  to  disturb  the  uniformity  of 
tension  throughout  the  whole  mass  as  little  as  possible,  so  as  to  have 
the  whole  liquid  in  motion  at  a  uniform  rate  like  a  slowly  descending 
piston.  Of  course  the  simplest  way  of  taking  the  liquid  from  the  bot- 
tom of  the  percolator  is  the  old  often-used  stop-cock  in  an  opening  in 
the  bottom  ;  but  this  proves  objectionable  for  several  reasons.  First, 
it  is  difficult  though  not  impossible,  by  means  of  a  stop-cock  at  the 
bottom,  to  keep  the  whole  mass  of  matter  at  a  uniform  tension,  or  in 
uniform  maceration.  Channels  of  liquid  in  more  rapid  motion  are 
more  apt  to  form,  and  the  packing  has  to  be  much  more  carefully  done 
in  order  to  prevent  this  tendency  to  currents  in  the  mass.  Next,  in 
actual  practice  it  was  found  that  no  stopcock  could  be  arranged  to  do 
its  work  automatically  according  to  the  natural  requirements  and  vary- 
ing conditions  of  each  case.  It  must  be  arbitrarily  set  to  run  at  same 
rate  of  dropping  that  would  be  decided  by  the  judgment  of  the  operator 
rather  than  by  the  natural  conditions  and  laws  of  the  process.  Then 
