462         ON  A  PROCESS  OF  FRACTIONAL  CONDENSATION. 
It  is  advisable  to  boil  the  retort  as  rapidly  as  possible  without 
choking  the  lower  end  of  the  heated  worm  with  the  returning 
liquid.  As  this  choking  would  give  rise  to  additional  pressure 
in  the  retort,  and  consequently  occasion  abnormal  elevation  of 
the  temperature,  and  possibly  a  rush  of  liquid  into  the  receiver, 
and  thus  introduce  irregularities  in  the  work,  excessive  heat 
under  the  retort  should  be  avoided.  The  first  indication  of 
choking  of  the  worm  is  a  partial  or  entire  stoppage  of  the  stream 
of  liquid  which  normally  flows  steadily  from  the  end  of  the  worm 
into  the  retort.  Any  interruption  or  unsteadiness  of  this  flow 
would  indicate  too  rapid  ebullition. 
As  a  rule,  other  things  being  equal,  the  greater  the  difference 
between  the  temperature  of  the  bath  and  that  of  the  retort,  the 
slower  the  products  will  come  off,  and  the  more  effectual  will  be 
the  separation.  I  think  it  possible,  however,  that  the  earlier 
fractionings  may  be  conducted  so  slowly  that  the  loss  of  time 
would  more  than  counterbalance  what  might  be  gained  by  more 
thorough  separation,  and  that  equally  good  results  may  be  more 
economically  obtained  by  more  frequent  operations,  somewhat 
more  rapidly  conducted. 
A  striking  illustration  of  the  advantage  to  be  gained  by  this 
process  is  presented  by  the  fact  that,  during  the  first  fractioning 
of  a  crude  mixture,  such  as  American  petroleum  or  coal-tar 
naphtha,  for  example,  the  difference  between  the  temperature  of 
the  bath  and  that  of  the  retort  may  sometimes  be  as  much  as  35° 
C,  or  even  more.  While,  as  the  products  become  purer,  this 
difference  between  the  temperatures  of  the  bath  and  retort  pro- 
portionally decreases,  till  finally,  in  operating  on  a  pure  product, 
the  temperature  of  the  bath  must  be  brought  to  within  a  few 
degrees  of  that  of  the  retort,  in  order  to  bring  the  vapors  through. 
But  the  amount  of  this  difference  is  variable  for  different  bodies 
of  equal  purity. 
The  first  fractionings  must  necessarily  be  quite  arbitrary  ;  for, 
as  a  general  rule,  when  operating  on  such  mixtures  as  those  just 
mentioned,  neither  the  thermometer  nor  the  quantities  obtained 
for  any  given  range  of  temperature  will  indicate  any  decided 
preponderance  of  any  one  substance.  On  the  contrary,  the 
temperature  rises  uniformly,  and  about  the  same  quantity  is 
