AFebJr0aar7Pi^ra-}        A  Ba^c  Reaction  of  Aldehydes.  8$ 
are  not  extracted  at  all,  or  else  in  smaller  proportion  by*  the  pro- 
cesses of  expression.  It  is  recorded  as  a  fact  that  the  gravity  of 
machine-expressed  oil  is  greater  than  hand-pressed,  and  these  oils 
are  heavier  than  either. 
With  the  use  of  sodium  bisulphite  for  determining  the  aldehydes 
the  addition  compound  forms  partly  in  the  aqueous  layer  and  partly 
in  the  oily  layer,  and  further  work  is  very  discouraging  when  this 
stage  is  reached.  I  soon  tried  sulphite  of  soda  instead  of  the  bisul- 
phite. On  heating  a  solution  of  this  with  the  oil,  the  dihydrodisul- 
phonate  of  sodium  is^  formed,  which  is  soluble  in  water,  and  two 
molecules  of  sodium  hydrate  are  split  off.  Tiemann,  in  his  work  on 
the  aldehydes  of  lemon-grass  oil,  pointed  out  that  the  reaction  could 
be  indicated  by  phenolphtalein,  and  by  adding  as  olution  of  sulphur, 
ous  acid  from  time  to  time,  the  end  of  the  reaction  could  be  known 
and  the  difference  of  volume  of  the  oil  could  be  told.  This,  however, 
is  not  a  method  that  is  capable  of  the  greatest  accuracy,  and  requires 
a  special  form  of  flask. 
It  occurred  to  me  that  the  alkali  formed  in  this  reaction  might  be 
titrated  for  with  standard  acid,  omitting  the  bicarbonate  of  sodium 
used  by  Tiemann.  I  first  tried  this  with  a  standarized  solution  of 
sulphurous  acid,  but  later  found  N /  2  HC1,  or  normal  HC1,  very 
satisfactory  and  more  convenient.  It  seems  to  require  the  use  of 
more  sulphite  solution,  however,  as  the  sulphurous  acid  reforms 
sulphite  from  the  NaOH  split  off.  I  will  try,  however,  at  a  future 
time,  to  take  up  all  the  details  necessary  in  the  determination  of 
citral  as  a  matter  by  itself,  and  speak  of  the  general  procedure  more 
especially  at  present. 
From  other  tests  on  the  oils  in  question,  I  found  that  they  con- 
tained considerable  quantities  of  acid  resin,  which  would,  unless 
previously  neutralized,  interfere  with  the  results.  In  fact,  I  found  the 
acid  resin  of  the  orange  oil  almost  sufficient  to  neutralize  all  the 
alkali  formed.  This  difficulty  was  overcome  by  neutralizing  these 
acids  first  with  dilute  caustic  potash  solution  (N /2KOH  being  con- 
venient for  the  purpose,  and  if  done  quantitatively,  might  be  interest- 
ing as  part  of  a  more  or  less  complete  examination  of  the  oil)  and 
rosolic  acid  solution  as  indicator,  which  will  serve  as  indicator  in  the 
subsequent  reaction. 
Analyses  were  made  with  oil  of  orange,  but  until  the  mixtures  of 
aldehydes  of  this  oil  are  more  fully  studied,  I  will  not  refer  to  figures 
