126  DE.  T.  WILLIAMS’S  EESEAECHES  ON  THE  STEHCTHRE  AXD  HOMOLOGY 
to  the  ciliated  tube  as  that  which  has  already  been  explained  in  the  Sabellidae  and 
Nereis. 
These  Annelids  present  a favourable  opportunity  for  determining  the  question  whether 
the  large  masses  (fig.  17  a c and  a c)  of  ova  which  in  the  female  apparently  occupy  the 
perigastric  chamber,  are  really  ix  that  cavity,  or  are  contained  in  the  areolee  of  a dehcate 
stromatous  tissue  as  already  described  in  Nereis,  8abella  and  Serpula.  'Well-defined  and 
repeatedly  conducted  observations  have  convinced  the  author  that  the  latter,  and  not  the 
former,  is  the  case.  In  the  Ariciadae  especially,  it  may  be  observed  that  the  peculiarly 
areolated  tissue  in  which  the  ova  are  contained  is  distinguishable  into  two  parts,  one  of 
which  is  densely  and  intimately  connected  with  the  blood-vessels  (see  fig.  16  ; fig.  IT  cli), 
the  other  forming  a mere  utricular  receptacle  to  the  generative  elements  (fig.  17  a,a\ 
fig.  15  e).  There  is  no  doubt  that  that  portion  (fig.  17  c h;  fig.  16  db)  which  is  ulti- 
mately bound  up  with  the  vascular  system,  is  designed  to  supply  a fiuid  secretion,  from 
the  blood-proper,  which  is  necessary  to  the  further  growth  of  the  generative  products. 
This  illustration  confirms  what  has  been  constantly  observed*  dui’ing  the  foregoing 
descriptions,  that  the  blood-proper  system  in  the  Annelids  is  much  more  intimately 
connected  with  the  process  and  the  organs  of  reproduction  than  the  chylaqueous  fluid. 
^This  important  fact  implies  evidently  a difference  of  vital  and  chemical  composition  in 
these  two  fluids.  The  higher  fluid  is  used  for  a higher  purpose,  the  lower  for  a lower. 
The  segmental  organs  in  the  Ariciadm  occupy  the  floor  of  each  annular  compartment 
(as  shown  in  fig.  17  ij  and  ij). 
The  genus  Nerine,  first  defined  by  Dr.  Johnson,  affords  by  far  the  most  favourable 
conditions  for  the  study  of  the  segmental  and  reproductive  systems  of  any  of  the  Xereid 
families.  The  coast  near  Swansea  is  inhabited  by  three  species  of  this  genus  (^iz.  N.  vul- 
garis, N comocephala,  N.  beata  (mihi).  The  last,  and  unfortunately  the  rarest,  is  beyond 
comparison  the  best  adapted  for  the  purposes  of  the  present  investigations.  In  this  ele- 
gant worm  the  entire  apparatus  (as  represented  in  fig.  Y^ab  c and  abc)  may  be  readily  and 
clearly  defined  by  the  eye.  It  is  here  quite  easy  to  demonstrate  the  connexion  between 
the  ciliated  tube  (fig.  a b)  and  the  appended  mass  [c  c c)  of  the  germinal  products. 
This  mass  fills  completely  the  hollow  bases  of  the  foot  and  the  entire  chamber  of  its  own 
segment.  It  embraces  the  ciliated  tube  {a  b).  It  is  evidently  a part  of  the  latter ; and  yet 
in  the  most  transparent  specimen  [Nerine  beata,  mihi)  it  is  utterly  impossible  to  define 
how  the  germ-products  escape  from  the  tube  [a  b)  into  the  mass  (c  c c).  But  the  author 
is  most  anxious  to  draw  especial  attention  to  the  fact  of  the  presence  of  this  o-^  arian  or 
sperm-mass  (according  to  the  sex)  in  constant  and  visible  association  until  the  ciliated 
looped  tube.  It  serves  to  reflect  light  upon  the  association  of  the  same  parts  in  speci- 
mens in  which  the  demonstration  is  necessarily  obscure.  It  illustrates  the  principle 
contended  for  in  this  paper,  that  the  “ segmental  organ  ” is  the  true  and  real  source  of 
the  generative  products ; in  other  words,  that  the  latter  are  the  organs  of  reproduction ; 
the  masses  appended  to  the  latter,  hitherto  mistaken  by  De  Quateepages  for  oiuria  and 
testes,  themselves  being  only  the  utricular  or  spermatic  receptacles  of  their  products. 
