Am\prnN9?5arm' }  Separation  of  0  xy  methyl  anthraquinones*  147 
To  obtain  the  mixed  color  principles,  hot  extraction  with  benzol 
was  employed,  as  recommended  by  De  la  Rue  and  Muller  8  in  their 
work  on  rhubarb.  Twenty-five  Cc.  of  fluidextract  of  the  various 
drugs  examined  were  evaporated  to  remove  alcohol,  diluted  with 
25  Cc.  of  water,  treated  with  an  excess  of  normal  lead  acetate,  and 
filtered  through  paper  pulp  on  a  Buchner  funnel  by  the  aid  of 
suction.  The  lead  precipitate,  together  with  the  paper  pulp,  was 
transferred  to  a  beaker  and  digested  for  one  hour  with  10  per  cent, 
sulphuric  acid  in  a  boiling  water  bath.  The  solution  was  filtered 
and  the  filtrate  extracted,  while  still  hot,  with  hot  benzol.  In  the 
case  of  powdered  drugs,  three  grammes  of  material  were  boiled  with 
alcoholic  potassium  hydroxide  under  a  reflux  condenser  for  one 
hour.  The  solution  was  then  evaporated  to  remove  alcohol,  diluted 
with  50  Cc.  of  water,  acidified  with  dilute  sulphuric  acid,  and 
extracted  directly  with  hot  benzol. 
By  washing  the  benzol  solution  first  with  5  per  cent,  sodium 
carbonate  solution  and  then  with  5  per  cent,  sodium  hydroxide, 
emodin  and  chrysophanic  acid  can  be  separated.  The  separation  can 
be  made  quite  sharp  if  one  bears  in  mind  that  emodin  is  very  readily 
soluble  in  sodium  carbonate,  and  that  chrysophanic  acid  is  slightly 
soluble  also.  Two  or  three  washings  of  the  benzol  solution  with 
sodium  carbonate  are  usually  sufficient  to  remove  emodin,  and  the 
aqueous  solution  is  of  deep  red  color.  As  the  benzol  is  further 
washed  with  this  reagent,  the  washings  become  pink-colored,  due  to 
the  slight  solubility  of  chrysophanic  acid.  Treatment  with  sodium 
carbonate,  then,  should  be  discontinued  when  the  washings  become 
pink.  The  chrysophanic  acid  can  be  readily  removed  by  one  or  two 
washings  with  sodium  hydroxide. 
In  a  few  preliminary  experiments  it  was  found  that  an  initial 
treatment  of  the  benzol  solution  with  5  per  cent,  ammonium  car- 
bonate removed  a  considerable  amount  of  coloring  matter  which  did 
not  subsequently  behave  like  either  emodin  or  chrysophanic  acid,  but 
did,  in  most  cases,  give  the  test  for  oxymethylanthraquinones.  On 
account  of  this  fact,  and  since  chrysophanic  acid  is  practically  in- 
soluble in  dilute  ammonium  carbonate,  and  emodin  but  slightly 
soluble,  the  addition  of  ammonium  carbonate  to  the  series  of  reagents 
for  the  fractional  washing  of  the  benzol  extract  suggested  itself. 
The  procedure  adopted,  then,  was  to  wash  the  benzol  solution 
8  Jour.  Chem.  Soc.  10,  298. 
