46 
Digitalis  Glucosides  and  Allied  Drugs.  {A^2J£ ^fg™- 
purposes.     In  this  case  a  biological  examination  is  absolutely 
necessary. 
Besides  the  qualitative  tests,  the  quantitative  estimation  of  digi- 
talis gluocosides  in  digitalis  leaves  is  of  more  general  interest. 
So  far  the  estimation  of  digitoxin,  as  worked  out  by  Keller,  is 
the  only  one  deserving  of  consideration.  It  can  be  applied  in  a 
slightly  modified  form  in  the  following  way. 
28  grammes  of  air-dried,  powdered  digitalis  leaves  are  placed 
in  a  suitable  glass-stoppered  flask  of  500  c.c.  capacity;  over  these 
are  poured  280  grammes  of  alcohol  60  p.c.  (by  weight)  and  the 
mixture  is  left  to  stand  for  3  to  4  hours,  shaking  it  frequently. 
It  is  then  filtered  and  207  grammes  of  the  filtrate  are  evaporated  to 
about  25  grammes  on  a  water-bath.  Sufficient  water  is  added  to 
the  residue  to  bring  the  total  weight  to  222  grammes,  and  while 
stirring,  25  grammes  of  official  liq.  plumbi  subacetatis  fort,  are 
added.  The  mixture  is  immediately  filtered  and  to  132  grammes 
of  the  filtrate,  in  an  Erlenmeyer  flask,  a  solution  of  5  grammes  of 
sodium  sulphate  in  8  grammes  of  water  is  added.  When  the 
precipitate  has  settled,  130  grammes  of  the  clear  fluid  are 
poured  into  a  separator,  2  grammes  of  solution  of  ammonia  ( 10  p.c. 
NHS)  are  added  and  the  mixture  is  shaken  5  times,  each  time  with 
30  c.c.  of  chloroform.  The  chloroformic  solutions  are  filtered 
and  then  evaporated,  the  dry  residue  is  dissolved  in  3  grammes  of 
chloroform,  and,  in  order  to  precipitate  the  digitoxin,  7  grammes 
of  ether  and  50  grammes  of  petroleum  ether  are  added.  The 
flocculent  digitoxin  which  separates  is  collected  on  a  small  filter 
(5  cm.  diameter)  and  dissolved  on  the  filter  by  the  addition  of  hot 
absolute  alcohol.  The  alcoholic  solution  which  runs  through  is 
collected  in  a  glass  capsule,  evaporated  to  dryness  and  the  residue 
dried  until  the  weight  is  constant.  This  multiplied  by  10  gives  the 
percentage  of  digitoxin  contained  in  the  leaves  analysed.  But, 
according  to  J.  Burmann,29  this  so-called  digitoxin  is  pseudodigi- 
toxin,  for  in  contradistinction  to  true  digitoxin  it  is  amorphous 
and  soluble  in  water  and  ether.  Kraft 30  declares  that  the  product 
obtained  by  Keller's  method  consists  chiefly  of  gitalin  (or  gitalin 
hydrate  and  anhydrogitalin,  in  addition  to  digitoxin.  He  agrees 
with  Burmann  in  that  he  also  considers  Keller's  digitoxin  to  con- 
tain only  a  very  small  amount  of  digitoxin. 
29  Burmann,  Bulletin  de  la  societe  chimique  1910,  p.  973.  —  Chemiker- 
Zeitung  191 1,  Repert,  p.  31. 
30  Kraft,  Schweizer  Wochenschrift  fiir  Chemie  und  Pharmazie  1911,  p.  174. 
