158  Hydrangea  ArhoresGens.  {^"^A^rii' fssi''^' 
the  best  solvents  for  its  important  constituents,  a  decided  indication  of 
an  alkaloid  was  obtained  by  the  application  of  the  various  alkaloidal 
tests ;  the  following  experiments  were  made  in  order  to  isolate  the 
alkaloid,  if  possible. 
One  thousand  grams  of  the  drug  were  exhausted  with  alcohol  of 
"828  sp.  gr.  by  percolation,  yielded  38*209  grams  of  extract,  which  was 
redissolved  in  a  small  quantity  of  warm  alcohol,  and  poured  into 
one  quart  of  water  previously  acidulated  with  acetic  acid ;  the  resin 
separated  weighed  6'43  grams. 
The  filtrate  (a)  was  set  aside  for  further  investigation.  The  resin 
obtained  had  a  sweet,  mild,  but  afterwards  lastingly  acrid  taste.  It  was 
insoluble  in  amnionic  hydrate,  almost  entirely  soluble  in  potassic 
hydrate,  which  solution  was  precipitated  by  hydric  chloride.  Approxi- 
mately one-half  the  resin  was  soluble  ii)  ether.  The  insoluble  por- 
tion was  almost  insoluble  in  amylic  alcohol.  The  alcoholic  solution  of 
the  resin  was  precipitated  by  solutions  of  neutral  and  basic  plumbic 
acetate. 
The  filtrate  (a),  ol^tained  on  separating  the  resin,  was  evaporated  to 
an  extract  which  had  a  saccharine  odor  and  a  bitterish  taste,  and 
weighed  31*566  grams.  This  extract  was  dissolved  in  a  small  quantity 
of  warm  water  and  filtered ;  the  next  morning  the  solution  contained 
a  few  crystals  which  it  was  found  impossible  to  separate.  Solutions  of 
neutral  and  basic  plumbic  acetate  were  added  to  thoroughly  precipitate 
tannin,  yielding  a  bluish-black  precipitate  with  iron  salts ;  the  excess  of 
lead  was  removed  by  sulphuretted  hydrogen,  the  filtrate  evaporated  to 
small  bulk,  acidulated  with  hydric  chloride  and  shaken  with  ether; 
tlie  ethereal  layer  evaporated  spontaneousl}'  left  but  a  slight  residue,  in 
which  no  alkaloid  was  present. 
The  aqueous  portion  was  then  shaken  with  pure  amylic  alcohol,  the 
latter  separated  and  evaporated  to  dryness,  the  slight  residue  remaining 
gave  in  solution  the  following  reactions :  with  Mayer's  solution,  a  light, 
whitish,  fiocculent  precipitate  ;  with  iodine  in  potassium  iodide,  a  brown 
precipitate.  Phospho-molybdic  acid  and  tannin  both  gave  precipitates. 
Platinic  chloride  gave  a  yellowish  precipitate  having  a  crystalline 
appearance  on  standing  a  short  time.    Picric  acid  gave  no  precipitate. 
One  hundred  and  twenty-five  grams  of  the  drug  were  exhausted  by 
maceration  and  percolation  with  pure  coal  tar  benzol.  The  greater 
portion  of  the  benzol  was  recovered  and  an  extract  weighing  1*43  gram 
was  obtained.    Warm  acidulated  water  (with  HCl)  was  shaken  with 
