ON  THE  BYTTERIA  FEBMFUGA. 
335 
M.  Gerardias  gives  the  following  description  of  the  bytteria, 
from  the  samples  which  he  procured  in  the  neighborhood  of  Fort 
de  France. 
The  bytteria  febrifuga  is  a  tree  from  15  to  20  metres  in 
height,  whose  diameter  is  about  60  centimetres  ;  the  bark  is  a 
brownish  grey,  slightly  furrowed ;  the  wood  is  light,  white  with 
yellow  veins ;  the  leaves  are  alternate,  they  are  formed  of  from 
four  to  eight  pairs  of  large  leaflets,  rather  unequilateral,  almost 
sessile,  lanceolated,  and  of  one  terminal  petiolated  leaflet. 
The  flowers  form  a  compound  corymb,  they  are  of  a  greenish 
white,  hermaphrodite  and  complete,  the  calix  has  usually  four  free 
sepals,  the  corolla  has  four  open,  free  petals,  four  stamens  are 
inserted  in  a  hypogenous  disc.  In  a  very  few  flowers  we  find 
five  sepals,  petals  and  stamens.  The  carpels  are  always  two  in 
those  flowers  having  four  petals,  and  in  most  of  them  with  five, 
but  in  some  few  of  the  latter  we  find  three.  In  all  cases  the  car- 
pels are  attached  to  a  very  developed  disc,  the  ovaries  are  per- 
fectly free,  the  styles  only  are  united  from  their  origin,  and  are 
only  divided  towards  their  extremities,  to  form  as  many  stigmas 
as  there  are  ovaries.  One  carpel  in  general  arrives  at  maturity, 
sometimes  two  may  be  found,  but  never  three.  The  fruits  are 
globular  and  dry.    The  seeds  have  no  endosperm. 
According  to  these  characters,  M.  Gerardias  thinks  the  bytteria 
should  be  reckoned  as  approaching  the  tribe  of  simarubas  of  the 
family  of  rutacese  ;  that  if,  notwithstanding  the  characters  which 
distinguish  it  from  them,  we  do  not  form  a  new  group  for  this 
plant,  and  still  unite  it  to  the  simarubas,  it  should  at  least  form 
a  distinct  genus. 
M.  Gerardias  says,  that  the  bytteria  cannot  be  the  same  tree 
as  the  quassia  of  Jamaica ;  this  latter  indeed  (the  simaruba  ex- 
celsa  of  Candolle,  pierasma  of  Lindley,  and  Endlicher,  Jamaica 
quassia  of  M.  Guibourt,)  is  always  described  as  a  plant  with 
polygamous  flowers  and  bivalve  fruits,  whereas  in  the  bytteria 
the  fruits  are  always  hermaphrodite  and  the  fruit  indehiscent. 
The  fruits  show  no  trace  of  separation,  and  if  we  press  them  hard 
when  ripe,  they  break  irregularly,  which  never  occurs  with  other 
simarubas. 
Finding  that  the  bytteria  so  nearly  approached  the  simarubas, 
M.  Gerardias  wished  to  ascertain  whether  bytterine  were  not 
