A^Sr"i9i5rin"}  Separation  of  Oxymethylanthraquinones:  151 
of  soda,  and  the  color  was  not  discharged  by  boiling  with  zinc  dust, 
but  rather  intensified.  The  hot  solution  was  filtered,  and  the  filtrate 
was  of  a  brilliant  deep  red  color.  This  color  faded  slowly,  and  the 
process  was  hastened  by  the  addition  of  H202.  After  some  minutes 
the  solution  became  water-white,  and  a  light-colored  or  white  pre- 
cipitate separated.  At  somewhat  above  200 ~  the  material  sublimed. 
Xo  liquefaction  had  taken  place  at  290°,  but  a  copious  yellow  sub- 
limate had  collected  in  the  upper  part  of  the  capillary  tube.  This 
behavior  is  strikingly  like  that  of  anthraquinone  itself.12 
Sodium  Carbonate  Fraction. — A  considerable  amount  of  color 
was  removed  from  the  benzol  by  this  reagent.  When  passed  into 
ether  and  evaporated,  the  residue  was  orange  red  in  color,  but  not 
crystalline.  The  melting-point  was  not  sharp,  but  liquefaction  was 
distinct  at  about  230 2 .  The  color  test  gave  the  reactions  observed 
for  emodin  as  derived  from  buckthorn  and  rhubarb. 
Sodium  Hydroxide  Fraction. — The  ether  residue  from  this 
fraction  was  golden  yellow  in  color.  It  fused  at  185°  to  190°,  and 
was  distinctly  liquid  at  about  200°.  By  the  color  test  reactions  were 
obtained  corresponding  to  those  of  chrysophanic  acid  as  derived 
from  chrysarobin  and  rhubarb. 
Aloes. 
The  experimental  material  consisted  of  Socotrine  and  Barbadoes 
varieties  in  powder  form  and  a  commercial  fluidextract. 
Ammonium  Carbonate  Fraction. — A  considerable  amount  of 
color  was  removed  from  benzol,  the  washings  being  orange-colored. 
The  ether  residue  was  white  or  pale  yellow,  and  not  crystalline. 
The  general  test  for  oxymethylanthraquinones  was  positive,  but 
faint  and  unsatisfactory,  some  foreign  coloring  matter  evidently  in- 
terfering. It  is  to  this  fraction  that  the  difficulty  noted  by  Warren.13 
on  applying  the  test  to  aloes,  is  due.  By  the  color  test  the  following 
reactions  were  obtained : 
Residue  +  H2S04  =  purple ;  +  HXOs  =  yellow  ;  +  H20  =  yellow 
solution. 
This  behavior  corresponds  to  that  obtained  for  the  unidentified 
oxymethylanthraquinone  derived  in  the  corresponding  fraction  from 
senna. 
12  Mulliken.  vol.  1,  p.  211  ;  also  p.  216.  test  1011. 
13  Am.  Jour.  Pharm.,  86,  10,  444  (1914). 
