^ApnVL9ih3arm' }      Constituents  of  Taraxacum  Root.  171 
shown  that  when  ^-hydroxyphenylethylamine  is  administered  by 
the  mouth  to  an  animal,  it  is  transformed  to'  a  large  extent  into 
/>-hydroxyphenylacetic  acid,  which  may  subsequently  be  extracted 
from  the  urine. 
In  order  completely  to  establish  the  identity  of  the  substance 
above  described  with  /'-hydroxyphenylacetic  acid,,  it  was  deemed 
desirable  to  compare  it  with  the  synthetic  acid,  especially  as  it  had 
been  stated  by  Salkowski  {Ber^  1879.  12,  1438; .  who  first  effected 
its  synthesis.,  that  its  aqueous  solution  gives  with  ferric  chloride 
a  slight  greyish-violet  coloration,  which  immediately  changes  to  a 
dirty  greyish-green.  It  was,  moreover,  thought  possible  that  the 
coloration  given  by  the  synthetic  product  might  be  due  to  a  slight 
contamination  with  the  corresponding  ortho-compound,  which  is 
known  to  produce  a  violet  color  with  ferric  chloride.  Baumann 
(Ber.,  1880,  13,  279),  who  obtained  ^-hydroxyphenylacetic  acid 
from  human  urine,  has,  however,  also  noted  that  its  aqueous  solu- 
tion gives  with  ferric  chloride  a  slight  violet  coloration. 
A  quantity  of  the  synthetic  acid  was  accordingly  prepared  from 
phenylacetic  acid,  the  latter  having  first  been  nitrated  according 
to  the  method  of  Maxwell  (Bcr.,  1879.  12,  1765).  After  very 
prolonged  fractional  crystallisation  from  warm  water,  a  product 
was  obtained  which  melted  at  1520,  the  pure  />-nitrophenylacetic 
acid  having  been  observed  by  Maxwell  (loc.  cit.)  to  melt  at  1500, 
and  by  Bedson  [Trans.,  1880,  37,  91)  at  150 — 151°.  This  nitro-acid 
was  reduced  by  tin  and  hydrochloric  acid,  and  then,  by  means  of 
the  diazo- compound,  converted  into  the  corresponding  hydroxy-acid. 
The  acid  thus  prepared  was  found  to  have  the  same  melting  point 
as  that  obtained  from  taraxacum  root,  and  when  the  two  products 
were  mixed  no  depression  of  the  melting  point  ensued.  The  reaction 
with  Millon's  reagent  was  precisely  the  same  as  that  previously 
mentioned.  On  comparing  the  behaviour  of  the  natural  and  the 
synthetic  acid  towards  ferric  chloride,  it  was  observed  in  both  cases 
that  if  a  fairly  concentrated  solution  of  the  acid  were  employed  a 
faint  and  exceedingly  fugitive  violet  coloration  was  produced, 
rapidly  changing  to  greenish-brown,  thus  confirming  the  observa- 
tions of  Salkowski  and  of  Baumann  (loc.  cit.) 
The  above  results  thus  completely  established  the  identity  of  the 
acid  from  taraxacum  root  with  ^-hydroxyphenylacetic  acid. 
After  the  extraction  of  the  original  aqueous  liquid  with  ether, 
as  above  described,  it  was  shaken  with  eighteen  successive  portions 
