﻿A 
  MONOGRAPH 
  OF 
  THE 
  VICTORIAN 
  SPONGES. 
  11 
  

  

  In 
  many 
  cases 
  the 
  collared 
  cells 
  appear 
  to 
  be 
  but 
  very 
  feebly 
  attached 
  to 
  the 
  

   underlying 
  layer 
  of 
  mesoderm, 
  and 
  frequently 
  they 
  come 
  away 
  over 
  large 
  areas 
  

   in 
  the 
  course 
  of 
  preparation. 
  Hence, 
  sometimes, 
  they 
  cannot 
  be 
  found 
  at 
  all. 
  

  

  Occasionally, 
  however, 
  the 
  collared 
  cells 
  are 
  exceptionally 
  well 
  preserved 
  and 
  

   only 
  partially 
  retracted, 
  even 
  in 
  ordinary 
  sections 
  prepared 
  by 
  the 
  paraffin 
  method. 
  

   This 
  is 
  the 
  case 
  in 
  one 
  of 
  my 
  preparations 
  of 
  Leucosolenia 
  prox'una. 
  Figure 
  3, 
  PI. 
  VIII. 
  

   shows 
  three 
  collared 
  cells 
  placed 
  side 
  by 
  side 
  in 
  a 
  row, 
  and 
  each 
  one 
  exhibits 
  the 
  

   characteristic 
  collar 
  still 
  expanded. 
  The 
  collars 
  are 
  funnel-shaped 
  or 
  tubular, 
  and 
  

   come 
  into 
  contact 
  at 
  their 
  margins, 
  without, 
  however, 
  forming 
  a 
  distinct 
  Sollas's 
  

   membrane. 
  Of 
  course 
  in 
  the 
  case 
  of 
  collars 
  which 
  thus 
  come 
  into 
  contact 
  at 
  their 
  

   margins, 
  a 
  temporary 
  fusion 
  may 
  take 
  place 
  as 
  between 
  any 
  two 
  naked 
  masses 
  of 
  

   protoplasm, 
  and 
  in 
  this 
  manner 
  Sollas's 
  membrane 
  in 
  all 
  probability 
  first 
  arose. 
  

   Between 
  the 
  bodies 
  of 
  the 
  cells, 
  filling 
  up 
  the 
  interstices, 
  a 
  transparent 
  substance 
  is 
  

   seen, 
  which 
  is 
  probably 
  part 
  of 
  the 
  gelatinous 
  ground 
  substance 
  of 
  the 
  mesoderm 
  in 
  

   which 
  the 
  bodies 
  of 
  the 
  collared 
  cells 
  are 
  partially 
  embedded. 
  Figure 
  4, 
  Pl.YIII. 
  shows 
  

   another 
  collared 
  cell 
  from 
  the 
  same 
  preparation 
  which 
  has 
  become 
  entirely 
  separated 
  

   from 
  the 
  rest 
  of 
  the 
  sponge 
  and 
  yet 
  distinctly 
  exhibits 
  its 
  tubular 
  collar. 
  None 
  of 
  

   these 
  cells 
  show 
  the 
  flagellum, 
  which 
  always 
  seems 
  to 
  be 
  retracted 
  before 
  the 
  collar. 
  

   The 
  transverse 
  diameter 
  of 
  the 
  body 
  of 
  the 
  collared 
  cell 
  is, 
  in 
  this 
  species, 
  scarcely 
  

   0'005mm., 
  while 
  the 
  length 
  of 
  the 
  body 
  and 
  collar 
  taken 
  together 
  is 
  0-013mni. 
  

  

  Even 
  in 
  the 
  small 
  grouj) 
  of 
  the 
  Calcarea 
  Homoccela 
  at 
  least 
  one 
  interesting 
  

   deviation 
  from 
  what 
  maybe 
  regarded 
  as 
  the 
  tyjncal 
  condition 
  of 
  the 
  collared 
  cells 
  is 
  

   found, 
  and 
  probably 
  more 
  will 
  subsequently 
  be 
  shown 
  to 
  occur 
  when 
  the 
  subject 
  

   receives 
  the 
  attention 
  which 
  it 
  deserves. 
  The 
  i)eculiarity 
  to 
  which 
  I 
  now 
  refer 
  occurs 
  

   in 
  the 
  very 
  remarkable 
  S})ecies 
  Leucosolenia 
  tripodifera, 
  Carter 
  sp., 
  and 
  is 
  illustrated 
  by 
  

   Figures 
  5 
  and 
  6, 
  PL 
  VIII. 
  In 
  the 
  first 
  place 
  there 
  is 
  a 
  well-developed 
  Sollas's 
  membrane 
  

   in 
  the 
  ordinary 
  position. 
  The 
  liagella 
  of 
  the 
  cells 
  are 
  retracted, 
  but 
  the 
  collars 
  are 
  in 
  

   some 
  cases 
  well 
  preserved 
  and 
  funnel-shaped. 
  Their 
  margins 
  do 
  not 
  come 
  into 
  direct 
  

   contact, 
  but 
  are 
  united 
  by 
  the 
  membrane, 
  which 
  runs 
  from 
  one 
  to 
  another 
  at 
  right 
  angles 
  

   to 
  the 
  long 
  axes 
  of 
  the 
  collars. 
  In 
  vertical 
  sections 
  this 
  membrane 
  appears 
  as 
  a 
  thin, 
  

   sharp 
  line 
  running 
  parallel 
  to 
  the 
  row 
  of 
  collared 
  cells, 
  and 
  at 
  a 
  little 
  distance 
  (about 
  

   the 
  length 
  of 
  a 
  collared 
  cell 
  body) 
  from 
  it 
  (Fig. 
  (')). 
  Not 
  only, 
  however, 
  is 
  Sollas's 
  

   membrane 
  present, 
  but 
  it 
  presents 
  in 
  itself 
  a 
  very 
  peculiar 
  nioditication. 
  Instead 
  of 
  

   being 
  i)erfectly 
  smooth 
  the 
  outer 
  surface 
  is 
  thickly 
  studded 
  with 
  delicate, 
  rod-like 
  

   processes 
  of 
  uniform 
  length 
  and 
  i)rojectiug 
  at 
  right 
  angles 
  from 
  it 
  into 
  the 
  gastral 
  

   cavity. 
  These 
  processes 
  are 
  about 
  0*007 
  mm. 
  in 
  length 
  and 
  O'OOlli 
  mm. 
  in 
  (lianu>ter 
  ; 
  

   they 
  stain 
  lightly 
  and 
  have 
  a 
  distinctly 
  granular 
  ai)pearance 
  ; 
  they 
  are 
  arranged 
  with 
  

   great 
  regularity 
  and 
  occur 
  in 
  specimens 
  from 
  dillercnt 
  localities, 
  so 
  that 
  tlu\v 
  would 
  

   appear 
  to 
  be 
  of 
  constant 
  occurrence 
  in 
  the 
  si)ccieH. 
  From 
  their 
  general 
  ai)pearauce 
  one 
  

  

  