426  Stability  of  Digitalis  Leaf  Extracts.  {^ffi/™^ 
The  chloroform  was  recovered  and  the  residue  digested  with  ether  and 
petroleum  ether.  The  insoluble  crude  digitoxin  is  recrystallized  from  80  per 
cent,  alcohol,  from  which  it  separated  in  fine  needles  or  leaflets  as  a  crystal 
powder,  melting  at  about  240°.  0=63.6;  H  =  8.5.  It  is  insoluble  in  water, 
benzin,  ether,  cold  alcohol,  but  more  readily  in  hot  alcohol,  and  very  soluble 
in  chloroform. 
Kiliani2  took  up  the  work  twenty-one  years  later.  He  followed 
Schmiedeberg  closely  as  to  extraction.  He  extracted  first  with  water  to 
which  5  per  cent,  alcohol  had  been  added.  The  water  extract  was  shaken 
with  ether  and  the  deep  green  extracts  were  shaken  with  dilute  soda  solu- 
tion. As  he  now  concentrated  the  ether  solution,  crusts  separated,  which 
were  filtered  off  and  washed  with  ether.  The  yield  was  0.015  per  cent.  It 
was  considered  digitoxin  but  as  it  did  not  agree  entirely  with  Schmiedeberg's 
it  is  designated  as  P  digitoxin. 
The  same  substance  was  isolated  in  larger  quantity  (0.085  per  cent.)  by 
extracting  with  50  per  cent,  alcohol  after  first  removing  the  water  extract. 
The  50  per  cent,  alcoholic  extract  was  precipitated  with  lead  subacetate  and 
filtered.  Alcohol  was  recovered  from  the  filtrate  which  was  then  exhausted 
with  ether.  The  ether  extract  was  washed  with  water.  When  concentrated, 
more  "  P  digitoxin  "  separated  in  crusts.  This  material  was  crystallized  from 
5-10  parts  of  boiling  85  per  cent,  alcohol,  whereupon  wartlike  aggregates  of 
leaflike  crystals  separated,  melting  at  145-1500  and  containing  water  of 
crystallization.  When  dissolved  in  methyl  alcohol  -\-  chloroform  (1:1)  and 
crystallized  by  the  addition  of  ether  it  separates  in  small  prisms,  water-free 
and  decomposing  above  2400. 
Concerning  the  identity  of  P  digitoxin  with  Schmiedeberg's  preparation 
as  well  as  with  the  product  of  Merck,  Kiliani  finally  concluded3  that  these 
products  were  identical,  and  possessed  the  formula  CwH^Ou4  0  =  63.9; 
H  =  8.5.    The  experimental  work  recorded  is  not  conclusive  in  our  opinion. 
Kiliani  in  this  work  first  hydrolyzed  his  P  digitoxin  with  dilute  hydro- 
chloric acid  and  separated  digitoxigenin.  This  body  crystallizes  from  boiling 
95  per  cent,  alcohol  upon  the  addition  of  water  in  large  prisms  and  does  not 
melt  at  2200.  (Sinters  at  225-2300.)  This  substance  to  which  the  formula 
CsiHs^O*  was  first  assigned  is  now  corrected  to  C22H32O4.  The  sugar  split 
from  digitoxin  is  digitoxose,  C6H1204,  m.p.  1010,  (ad)=  +  46°.  It  is  an 
interesting  methyltetrose  sugar.  The  presence  of  this  sugar  group  causes 
the  characteristic  color  test  for  digitoxin,  i.e.,  with  H2SO4  containing  a  trace 
of  iron  sulphate,  a  wine  red  coloration  results.  In  hydrochloric  acid  a  yellow 
color  changing  to  green  is  obtained.  Dissolved  in  acetic  acid  containing 
iron  sulphate  plus  sulphuric  acid  containing  iron,  an  indigo  blue  is  obtained. 
Although  there  was  thus  some  agreement  on  this  substance  as  an  ether 
or  chloroform  extractable  product,  Kraft5  has  rather  failed  to  confirm  these 
views.    Kraft  extracted  with  cold  water  clarified  with  lead  acetate  and 
2  Arch.  d.  Pharm.,  233,  311  (1895). 
3  Arch.  d.  Pharm.,  234,  273  (1896). 
*B.  31,  2454  (1898). 
5  Arch.  Pharm.,  250,  118  (1912). 
