MR.  R.  OWEN  ON  THE  ANATOMY  OF  THE  BRACHIOPODA. 
149 
Ter.  dorsata,  Ter.  dentata,  Lam.,  and  Ter.  Soiverbii,  the  reflected  portion  is  consider- 
able. The  loop,  besides  being  fixed  by  its  origins  or  crura,  is  commonly  attached  to 
two  processes  going  off  at  right  angles  from  the  sides,  or  formed  by  a  bifurcation  of  the 
extremity,  of  a  central  process,  which  is  continued  forwards  to  a  greater  or  less  extent 
from  the  hinge ;  but  it  is  sometimes  entirely  free,  except  at  its  origins,  as,  e.  in 
Ter.  vitrea.  This  reflected  loop,  forming  two  arches  on  either  side  the  mesial  plane, 
towards  which  their  concavities  are  directed,  I  have  figured  as  it  exists  in  Ter.  Chi- 
lensis^  and  Ter.  Sowerhii^.  It  is  represented  of  a  similarly  perfect  form  in  Ter.  dentata, 
by  M.  de  Blainville  in  his  '  Malacologie'-^:  and  the  same  apparatus  in  Ter.  dorsata  is 
very  well  figured  by  Chemnitz* ;  by  Sowerby^ ;  and  more  recently  by  G.  Fischer  de 
Waldheim^.   A  similar  form  is  also  figured  in  another  species  of  Terebratula  by  Poli^. 
The  arches  of  the  loop  are  so  slender,  that,  notwithstanding  their  calcareous  nature, 
they  possess  a  slight  degree  of  elasticity,  and  yield  a  little  to  pressure ;  but,  for  the 
same  reason,  they  readily  break  if  the  experiment  be  not  made  with  due  caution.  The 
interspace  between  the  two  folds  of  the  calcareous  loop  is  filled  up  by  a  strong  but 
extensile  membrane,  w^hich  binds  them  together,  and  forms  a  protecting  wall  to  the 
viscera :  the  space  between  the  bifurcated  process  in  Ter.  Chilensis  is  also  similarly  oc- 
cupied by  a  strong  aponeurosis.  In  this  species  the  muscular  stejn  of  each  arm  is 
attached  to  the  outer  sides  of  the  loop  and  the  intervening  membrane.  They  com- 
mence at  the  pointed  processes  at  the  origins  of  the  loop,  advance  along  the  lower 
portion,  turn  round  upon  the  upper  one,  and  are  continued  along  it  till  they  reach  the 
transverse  connecting  bar,  where  they  advance  again  forwards,  and  terminate  by  making 
a  half  spiral  twist  in  front  of  the  mouth.  It  is  these  free  extremities  which  form  the 
third  arm  mentioned  by  Cuvier^.  These  arms  are  ciliate  on  their  outer  side  for  their 
entire  length  ;  but  the  cilia  are  longer  and  much  finer  than  the  brachial  fringes  of  Lin- 
gula;  and  except  at  the  extreme  ends,  which  have  a  slight  incurvation,  they  are  uni-^ 
formly  straight.  There  is  thus  an  important  difference  between  Lingula  and  those 
species  of  Terebratula  which  resemble  Ter.  Chilensis  in  the  powers  of  motion  with  which 
the  arms  are  endowed ;  since  from  their  attachment  to  the  calcareous  loop  they  are 
fixed,  and  cannot  be  unfolded  outwards  as  in  Lingula.  Owing  to  this  mode  of  con- 
nexion, and  their  ciliated  structure,  their  true  nature  was  much  more  liable  to  be  mis- 
taken by  the  early  observers,  though  it  appears  not  to  have  escaped  the  discrimination  of 
Linnaeus,  who,  as  Cuvier  has  observed,  founded  his  character^  of  the  animal  of  Anomia 
on  the  organization  of  one  of  the  Terebratula,  which  he  included  in  that  genus. 
'  Fig.  4,  2  Fig.  16.  3  PI.  li.  1  fl.  Conchyl.  Cabinet,  band  viii.  tab.  Ixxviii.  fig.  711. 
5  Genera  of  Shells.  Notice  sur  le  Charpente  Osseux  des  Terebratules,  fig.  3. 
^  Testacea  utriusque  Sicilise,  vol.  ii.  pi.  16.  R^gne  Anim.,  (nouv.  ed.)  torn.  iii.  p.  171. 
^  Anomia,  Animal  Corpus  Ligula  emarginata  cUiata,  ciliis  valvular  sujyeriori  affixis.  Brachiis  2,  linearihus,  cor- 
pore  longioribus,  conniventibus,  porrectis,  valvul<e  alternis,  utrinque  ciliatis,  ciliis  affixis  valvulis  utrisque. — Syst. 
Nat.,  (ed,  xii.)  vol.  i.  pars  2.  p.  1150, 
X  2 
