174 
Constituents  of  Taraxacum  Root.      /Am  jour  Pharm. 
'  l       April,  1913. 
filtered.  The  filtered  liquid  was  then  concentrated  under 
diminished  pressure  to  the  consistency  of  a  thin  syrup.  It  gave  a 
dark  brown  coloration  with  ferric  chloride,  but  no  precipitate  with 
gelatin,  thus  indicating  the  absence  of  tannin,  and  it  also  gave  no 
reaction  with  potassium-mercuric  iodide.  As  nothing  separated 
from  the  liquid  on  keeping,  it  was  heated  to  boiling,  neutralised 
with  potassium  hydroxide,  and  sufficient  of  a  concentrated  solution 
of  the  latter  added  to  represent  about  10  per  cent,  of  the  mixture, 
after  which  it  was  boiled  for  about  five  minutes.  The  liquid  was 
then  poured  into  dilute  sulphuric  acid,  and,  after  cooling,  the 
mixture  was  extracted  many  times  with  ether,  the  combined  ethereal 
liquids  being  subsequently  extracted  with  aqueous  ammonium 
carbonate  and  sodium  hydroxide.  Nothing  of  interest  was  removed 
by  the ,  last-mentioned  alkali,  and  on  finally  evaporating  the  ether 
only  a  trace  of  amorphous  material  remained.  The  ammonium 
carbonate  extract,  however,  after  acidification  and  extraction  with 
ether,  yielded  a  product  which  gave  with  ferric  chloride  a  dark 
green  color.  From  this  product,  after  several  crystallisations  from 
hot  water  containing  a  little  alcohol,  a  small  amount  (about  0.2 
gram)  of  an  acid  was  obtained,  which  melted  and  decomposed 
at  214 — 2 1 6°  with  evolution  of  gas.  It  was  dried  at  no0,  and 
analysed : 
0.0643  gave  0.1405  C02  and  0.0265  H20.    C  =59.6;  H  =  4.6. 
C9H804  requires  C  ==  60.0 ;  H  ==  4.4  per  cent. 
The  above-described  substance  was  thus  definitely  identified  as 
3  :4-dihydroxycinnamic  acid,  a  smaller  amount  of  which  had  previ- 
ously been  obtained  from  the  amyl-alcoholic  extract  of  the  original 
aqueous  liquid. 
The  filtrate  from  the  precipitate  produced  by  basic  lead  acetate 
was  treated  with  hydrogen  sulphide  for  the  removal  of  the  excess 
of  lead,  again  filtered,  and  concentrated  under  diminished  pressure 
to  the  consistency  of  a  syrup.  It  evidently  contained  an  abundance 
of  sugar,  since  it  readily  yielded  d-phenylglucosazone,  melting  at 
204 — 2060.  A  portion  of  the  syrup  was  acetylated,  but  as  nothing 
crystalline  separated  frorm  the  product,  even  after  long  keeping,  it 
was  finally  hydrolysed.  The  regenerated  sugar  was  then  found  to 
be  strongly  lsevorotatory,  thus  indicating  that  it  must  have  con- 
sisted, to  a  large  extent  at  least,  of  lsevulose.  Another  portion  of 
the  syrup  was  heated  for  about  two  hours  with  5  per  cent,  sulphuric 
