146  On  the  Fruit  of  Magnolia  Tripetala.  Si^ 
is  cone-shaped,  from  three  to  five  inches  in  length,  and  one  and  a 
half  to  two  inches  in  diameter  ;  its  green  color  changes  to  a  beautiful 
rose  color  in  the  autumn. 
This  fruit  is  a  compound  succulent  capsule,  consisting  of  numerous 
dehiscent  carpels,  arranged  in  a  sort  of  imbricated  spike,  in  each  of 
which  is  a  rod  seed  attached  to  the  carpel  by  a  silky  thread ;  these, 
when  the  carpels  dehisce,  fall  out  and  remain  suspended.  The  seeds 
are  red,  irregularly  shaped,  somewhat  flattened  and  angular,  and 
quite  bitter  and  acrid  when  chewed,  and  consist,  when  fresh,  of  a 
soft  outer  portion,  with  smooth  epidermis  and  a  soft  oily  kernel  en- 
closed in  a  hard  shell. 
The  fruit  med  in  the  first  experiments  was  gathered  September  7th, 
1871,  from  a  tree  about  twenty  feet  high,  growing  in  the  vicinity  of 
Mt.  Holly,  N.  J.,  while  yet  unripe;  it  was  purplish  colored  and  fleshy. 
On  grating,  the  firm  juicy  exterior  was  removed,  exposing  the  seeds 
and  cellular  structure  of  the  capsule  beneath.  xVbout  three  weeks 
later  the  remainder  of  the  ripe  fruit  was  collected,  when  most  of  the 
carpels  were  opened  with  the  seeds  yet  attached. 
Experiment  1. — A  quantity  of  the  fresh  capsules,  deprived  of  the 
seed,  were  sliced  and  macerated  in  alcohol  95  per  cent,  for  five  days, 
the  tincture  filtered,  and  the  alcohol  recovered  by  distillation  with  a 
water-bath  still,  leaving  a  reddish-brown,  semifluid,  resinous  extract 
with  a  separate  watery  portion.  This  extract  was  set  aside  for  several 
days,  until  a  deposition  of  crystals  occurred,  when  the  dark  mother 
liquid  was  drained  off  and  set  aside ;  it  solidified  after  a  time  by  the 
formation  of  other  crystals  of  the  same  kind. 
This  crystalline  matter  was  found  to  be  readily  soluble  in  alcohol, 
ether  and  chloroform,  but  owing  to  the  fact  that  the  resin  was  also 
taken  up,  it  was  difficult  to  purify.  Seeking  to  overcome  this  diffi- 
culty, I  tried  purified  commercial  benzine,  and  was  gratified  to  find 
that  it  would  dissolve  the  crystals  without  acting  much  on  the  resin- 
ous and  coloring  matters  of  the  fruit  and  seeds. 
The  black  crystalline  mass  was  then  digested  in  boiling  benzine, 
and,  on  cooling,  the  filtered  solution  deposited  the  crystals  almost 
pure,  and  quite  free  from  coloring  matter.  A  quantity  of  the  dried 
capsules,  deprived  of  the  seeds,  was  powdered,  exhausted  with  alcohol 
of  95  per  cent,  by  percolation,  and  the  tincture  allowed  to  evaporate 
spontaneously.  The  dark  brown  soft  extract  resulting  was  well  washed 
with  hot  water,  dried  and  treated  with  boiling  benzine  repeatedly 
