*&8Ebg5£}       Active  Principle  of  Digitalis.  451 
ether,  and  50  c.c.  petroleum  spirit  added.  The  digitoxin  then  sep- 
arates in  white  flocks,  and  on  shaking,  the  liquid  becomes  quite 
clear.  For  weighing  the  digitoxin  it  may  be  dissolved  off  the 
filter  with  hot  alcohol  and  the  solution  evaporated  in  a  suitable 
vessel,  or  the  petroleum  spirit  may  be  decanted  off  and  the 
digitoxin  weighed  in  the  state  of  powder. 
Digitoxin  thus  obtained  dissolves  in  strong  hydrochloric  acid 
with  a  yellowish  color,  and  the  solution  when  warmed  becomes 
greenish,  then  greenish-brown  ;  on  adding  water  the  color  becomes 
greenish-yellow,  and  after  some  time  flocks  are  separated. 
A  solution  of  digitoxin.  in  glacial  acetic  acid  containing  ferric 
chloride,  gives  Keller's  reaction  when  floated  on  strong  sulphuric 
acid.  At  the  line  of  contact  a  dark  zone  appears,  and  after  a  few 
minutes  the  acetic  acid  liquor  becomes  dark  blue.  This  reaction 
takes  place  with  one-tenth  of  a  milligramme  of  digitoxin  in  1  c.c. 
of  acetic  acid. 
The  complete  separation  of  digitalin  is  difficult,  as  it  is  sufficiently 
soluble  in  chloroform  for  traces  to  be  taken  up,  and  to  that  circum- 
stance must  be  ascribed  the  red  coloration  of  digitoxin  when  mixed 
with  strong  sulphuric  acid. 
The  watery  liquid  from  which  digitoxin  has  been  extracted  by 
shaking  with  chloroform  has  a  bitter  taste  and  contains  digitonin, 
which  can  be  separated  by  precipitation  with  tannin;  but  as  it  does 
not  possess  the  peculiar  efficacy  of  digitalis,  its  separation  is  of  little 
importance.  Keller  was  unable  to  obtain  any  substance  correspond- 
ing to  the  description  of  digitalein,  and  he  considers  it  is  merely 
digitonin  mixed  with  traces  of  digitoxin  and  digitalin. 
After  separating  digitonin  tannate,  the  filtered  liquid  still  contains 
digitalin. 
The  general  conclusion  arrived  at  by  Keller  is  that  digitalis  leaves 
contain  digitoxin,  digitonin  and  digitalin  identical  with  the  products 
from  digitalis  seeds,  but  in  somewhat  different  proportions,  the 
amount  of  digitoxin  in  the  seeds  being  much  smaller  than  that  in 
good  leaves,  but  it  varies  very  much  in  different  samples  of  leaves,  or 
from  0  26  to  O  62  per  cent.  A  still  greater  variation  was  found  in  the 
pharmaceutical  preparations  of  digitalis,  and  Keller  strongly  recom- 
mends the  adoption  of  means  for  ascertaining  the  medicinal  value 
of  the  drug  on  the  basis  of  the  amount  of  digitoxin  it  contains. 
