Amyu°y^i9^.rm"  )  Stability  of  Digitalis  Leaf  Extracts.  427 
sodium  phosphate,  and  precipitated  all  the  products  in  the  water  extracts 
with  tannin.  From  this  he  liberated  the  glucosides  in  the  usual  manner  with 
ZnO.  These  were  taken  up  as  far  as  possible  in  water  and  the  water  solu- 
tion was  extracted  with  chloroform.  The  activity  was  in  the  chloroform 
extract  while  the  water  soluble  substances  which  greatly  predominated  (5  per 
cent.)  are  saponins,  which  upon  hydrolysis  yield  glucose  and  a  pentose. 
Interest  therefore  centers  in  the  chloroform  extract,  but  at  this  point  Kraft 
dropped  the  work  and  instead  shook  out  fresh  aqueous  extracts  with  chloro- 
form omitting  the  tannin  procedure.  From  this  chloroform  extract  he  pre- 
cipitated -with  ligroin  a  substance  termed  by  him  "  gitalin  "  but  which  appears 
to  be  a  mixture.6    The  yield  is  0.7  per  cent. 
Now  Kiliani7  had  stated  that  the  digitalis  leaf  extracts  which  had  been 
ether-extracted  would  yield  when  subsequently  chloroform-extracted  a  second 
active  substance,  not  digitoxin  but  "  digitophyllin."  Later8  in  reviewing 
digitalis  chemistry  he  omits  this  substance  from  consideration. 
Kraft  continued  his  work  by  extracting  the  water-extracted  leaves  with 
50  per  cent,  alcohol.  This  extract  was  clarified  with  lead  acetate  filtered  and 
concentrated  to  a  small  volume  in  the  presence  of  calcium  carbonate.  A 
heavy  precipitate  separated  as  with  Schmiedeberg.  The  precipitate  is  washed 
with  2  per  cent,  soda  solution  and  dried,  and  the  dried  precipitate  is  ex- 
hausted with  chloroform,  which  dissolved  a  glucoside  agreeing,  but  not  con- 
clusively, with  digitoxin. 
After  exhausting  the  dry  precipitate  with  chloroform  Kraft  boiled  the 
residue  with  alcohol,  and  a  glucoside,  "  gitin  "  'yielding  galactose  on  hydroly- 
sis was  isolated  C  =  52.2;  H  =  8j. 
It  therefore  appears  that  the  chloroform  soluble  glucoside  pres- 
ent in  water  solutions  and  in  50  per  cent,  alcoholic  extracts  is  not 
definitely  known,  but  there  is  a  certain  quite  reasonable  hypothesis 
that  there  is  a  chloroform  soluble  glucoside,  yielding  digitoxose  on 
hydrolysis,  and  which  is  insoluble  in  water  when  once  it  has  been 
separated  to  a  certain  degree  of  purity.  To  this  substance  the  term 
digitoxin  applies.  What  the  composition  of  "gitalin"  is  has  not 
been  studied  further  except  that  when  one  purifies  it  further  a  part 
becomes  insoluble  in  chloroform,  ether  and  water.  Merck9  has 
found  in  producing  digitoxin  that  a  part  of  the  glucoside  originally 
taken  out  by  means  of  chloroform  becomes  insoluble.  Kraft  desig- 
nates his  insoluble  product  "  anhydrogitalin,"  a  digitoxose  glucoside 
and  it  appears  to  be  a  pure  product. 
On  the  whole  the  chemistry  of  these  substances  is  in  an  unsatis- 
factory condition.    When  the  leaves  have  been  exhausted  with 
6  Kiliani,  Arch.  Pharm.,  252,  13  (1914). 
7  Arch  d.  Pharm.,  235,  425  (1897). 
8  Am.  J.  Pharm.,  85,  223  (1913). 
»B.  48,  334  (1915). 
