Am'Apinr'i9ih3arm' }      Constituents  of  Taraxacum  Root.  167 
The  question  regarding  the  occurrence  of  mannitol  in  taraxacum 
root  was  investigated  many  years  ago  by  T.  and  H.  Smith  (Pharm. 
1 1849,  8,  480),  who  conclusively  proved  that  this  compound  does 
not  pre-exist  therein,  but  that  it  is  formed  when  an  extract  of  the 
root  undergoes  the  so-called  mucous  or  viscous  fermentation.  Its 
formation  under  these  conditions  would  appear  to  permit  of 
explanation  by  the  fact  that  taraxacum  root  contains  an  abund- 
ance of  inulin,  which,  on  hydrolysis,  is  converted  into  lsevulose, 
and  the  latter,  by  the  special  fermentative  process  referred  to, 
becomes  reduced  to  mannitol.  The  above  observation  has  now  also 
been  confirmed  by  the  present  authors,  inasmuch  as  no  trace  of 
mannitol  could  be  isolated  directly  from  the  root  employed  for  this 
research. 
While  the  present  investigation  was  in  progress  it  has  been 
recorded  (Brit.  Med.  J.,  May  25th,  1912,  p.  1181)  that  the  use  of 
taraxacum  in  cases  of  cancer  has  been  attended  with  remarkably 
beneficial  results,  and  shortly  afterwards  (ibid.,  July  13th,  1912;. 
p.  97)  attention  was  directed  to  the  use  of  choline  in  the  treatment 
of  this  disease.  Additional  interest  is  imparted  to  these  two  quite 
independent  observations,  especially  when  considered  conjointly, 
by  the  fact  that  taraxacum  root  has  now  been  found  to  contain 
choline.  The  various  other  well-defined  compounds  which  have 
been  isolated  are  summarized  at  the  end  of  this  paper. 
Experimental. 
The  material  employed  for  this  investigation  consisted  of  the 
best  quality  of  English  taraxacum  root,  which  was  collected  in 
the  autumn  of  191 1,  and  kindly  supplied  to  us  by  Messrs.  W. 
Ransom  and  Son,  of  Hitchin. 
A  small  portion  (25  grams)  of  the  ground  root  was  treated  with 
Prollius'  fluid,  and  the  resulting  product  tested  for  an  alkaloid 
with  the  usual  reagents.  The  reactions  obtained  were  so  slight 
as  to  indicate  the  presence  of  not  more  than  traces  of  an  alkaloidal 
substance. 
Another  portion  (20  grams)  of  the  ground  material  was  succes- 
sively extracted  in  a  Soxhlet  apparatus  with  various  solvents,  and 
the  resulting  extracts  dried  in  a  water-oven  until  of  constant 
weight : 
