Yol.  49.] 
OP  THE  GrEXTTS  ASTROCCEXIA. 
5  73 
the  central  corallites,  they  may  probably,  as  in  the  foregoing  species, 
be  free  from  stereoplasm. 
It  is  obvious  that  all  the  definitions  of  the  genus  AstvococniG,  are 
defective,  because  the  internal  structure  has  not  been  studied  by 
means  of  sections,  and  that  they  are  superseded  by  the  results  of 
the  foregoing  investigations.  Under  those  circumstances  a  new 
definition  of  the  genus  becomes  imperative.  I  define  it  as  follows, 
but  x  may  observe  that  I  do  not  at  present  include  in  it  any  species 
of  earlier  date  than  the  Cretaceous  period,  having  concluded  that  all 
the  so-called  J urassic  Astrocoenice  are  referable  to  other  and  quite 
distinct  genera. 
O 
Definition  of  tlie  Genus  Astroccexia. 
Corallum  compact,  spreading  or  dendroid,  and  composed  of  pris¬ 
matic  corallites  intimately  united  by  their  walls,  which  are  thin 
and  sometimes  rudimentary.  Corallites  greatly  lessened,  and  ren- 
deied  moieoi  less  cylindrical  internally  by  a  considerable  deposit  of 
stereoplasm.  Septa  straight,  alternately  long  and  short,  denticulated, 
the  longer  ones  blending  with  the  columella.  The  sides  of  the  septa 
are  without  growth  of  any  kind.  There  are  no  dissepiments,  but 
the  loculi  are  probably  filled  up  interiorly  by  stereoplasm.  Columella 
variable  in  size,  but  not  prominent. 
Of  the  mode  of  increase  I  am  unable  to  speak  definitely,  but  should 
suppose  that  gemmation  must  take  place  on  the  top  0 f  the  °reatlv 
thickened  wall.  0  > 
At  present  I  refrain  from  a  precise  opinion  on  the  classificatory 
position  of  this  very  peculiar  genus,  but  that  it  will  be  removed  from 
the  place  it  has  hitherto  held  there  can  be  no  doubt.  Styloccenia , 
with  which  it  was  associated  by  At.  Alilne-Edwards,  'has  few 
characters  m  common  "with  it,  and  neither  have  any  of  the  genera 
with  which  it  has  been  associated  by  Dr.  de  Uromentel.  The 
late  Prof.  Duncan  made  of  it  an  4  alliance  ’  in  which  he  also 
included  Cycithoccenia,  Stephanocoenia ,  Narcissastrceci ,  Hcddonia,  and 
Bcithyccenia ,  to  none  of  which  is  it  at  all  nearly  related. 
In  conclusion,  I  may  offer  a  few  remarks  on  the  apparent  discre¬ 
pancies  relative  to  the  thickness  of  the  walls  enclosing  the  corallites. 
As  I  have  shown,  both  At.  Alilne-Edwards  and  Prof.  Duncan 
have  made  what  appear  to  be  most  contradictory  statements  on 
the  subject.  The  explanation  of  their  apparently  (but  not  really) 
inconsistent  assertions  is,  however,  perfectly  simple  and  easv.  In 
many  calices  the  attenuated  wall  has  become  rudimentary,*  while 
in  some  it  has  wholly  disappeared.  "When  this  is  the  case,  the 
stereoplasm  takes  its  place,  though  always,  so  far  as  I  have  been  able 
to  ascertain,  with  some  remaining  indication  of  its  proper  position 
and  the  contracted  and  circular  calices,  being  defined  by  it,  appear 
to  possess  what  ADI.  Alilne-Edwards  and  Haime,  and  the  late  Prof. 
Duncan,  respectively  designated  ‘  pseudo-ccenenchyma  ’  and  ‘  dense 
ccenenchyma.  The  stereoplasm  in  Astrocoenice  is  not  unlike 
dermic  ccenenchyma  in  appearance,  from  which,  however,  it  differs 
radically,  being  wholly  within  the  walls  of  the  corallites. 
