AaAigustPi898"m'}     Valuation  of  Crude  Carbolic  Acid.  373 
diminished  in  volume  slightly,  a  small  amount  remaining  perma- 
nently in  suspension  in  the  aqueous  liquid,  with  a  portion  of  the 
precipitated  impurities  from  the  acid. 
Various  changes  in  the  details  of  both  procedures  were  tried  with- 
out success.    It  seems  likely  that  at  the  time  they  were  devised 
crude  carbolic  acid  was  quite  different  in  composition  from  that  now 
met  with.    The  writer  gives  preference  to  the.following  expedient, 
but  realizes  that  it  is  far  from  accurate,  the  results  probably  being 
always  too  low  :  mix  10  grammes  (or  10  c.c.)  of  the  sample  with  50 
c.c.  of  5  per  cent,  caustic  soda  solution  in  a  separator.    Add  to  the 
mixture,  after  cooling  to  room  temperature,  1 5  c.c.  of  ether,  and  shake 
actively  for  a  few  minutes,  then  put  it  aside  to  separate.    Draw  off 
the  aqueous  liquid  and  transfer  the  ethereal  solution,  to  a  weighed 
porcelain  dish.    Extract  the  alkaline  solution  with  two  additional 
portions  of  10  c.c.  each,  of  ether,  and  add  these  to  the  contents  of 
the  dish.    Allow  the  ether  to  evaporate  spontaneously  and  weigh 
the  residue,  which  represents  the  tar-oils  in  the  sample.    As  nearly 
all  of  these  are  more  or  less  volatile  at  ordinary  temperatures,  a 
considerable  proportion  of  them  is  probably  lost  before  all  ether  is 
vaporized,  and  since  the  amounts  of  the  more  volatile  constituents 
vary,  it  is  difficult  to  estimate  this  loss.    In  a  warm  and  humid 
atmosphere,  water  condenses  in  the  dish  during  evaporation  of  the 
ether,  and  additional  time  must  be  allowed  in  such  cases  for  the 
evaporation  of  this  water  before  weighing,  which  entails  an  addi- 
tional loss  of  the  oils.    Under  any  circumstances  the  percentage  of 
tar-oils  obtained  by  this  or  other  available  means  must  be  accepted 
with  considerable  reservation. 
The  uncertainty  in  the  estimation  of  those  constituents  of  crude 
carbolic  acid  not  phenols,  would  render  Koppeschaar's  Bromine 
Method  inapplicable  to  the  determination  of  the  ratio  of  phenol  and 
cresols,  even  if  there  were  no  other  objections,  but  there  are  two 
other  reasons,  which  alone  would  be  sufficient  to  reject  it.  One  is 
the  fact  that  crude  carbolic  acid  nearly  always  contains  still  higher 
homologues  of  the  phenol  series  than  the  cresols,  each  group  of 
isomers  combining  with  bromine  in  a  different  ratio.  The  other 
reason  is  that  some  of  the  impurities  in  the  acid  also  form  substitu- 
tion compounds  of  bromine  under  the  conditions  of  the  assay,  as  the 
writer  determined  by  experiment.  The  tar-oils  obtained  from  two 
samples  by  extraction  with  ether  were  found  in  one  case  to  require 
