﻿24 
  PROCEEDINGS 
  OF 
  THE 
  NATION 
  AL 
  MUSEUM. 
  vol.44. 
  

  

  precise 
  and 
  less 
  crowded, 
  and 
  they 
  are 
  supplemented 
  by 
  a 
  number 
  of 
  

   intervening 
  folds 
  at 
  least 
  over 
  part 
  of 
  the 
  manubrium 
  (pi. 
  2, 
  fig. 
  8), 
  

   though 
  the 
  gonads 
  as 
  a 
  whole 
  show 
  considerable 
  variation 
  in 
  different 
  

   specimens, 
  and 
  even 
  in 
  different 
  interradii 
  of 
  a 
  given 
  specimen. 
  The 
  

   (Ufference 
  between 
  this 
  gonad-type 
  and 
  that 
  of 
  the 
  typa-simulans 
  

   group 
  is 
  so 
  striking 
  that 
  this 
  character 
  alone 
  is 
  sufficient 
  to 
  separate 
  

   them. 
  But 
  the 
  lip 
  affords 
  an 
  even 
  more 
  diagnostic 
  feature. 
  In 
  all 
  the 
  

   other 
  members 
  of 
  the 
  genus 
  this 
  structure 
  is 
  simple, 
  quadrate, 
  but 
  

   slightly 
  folded, 
  if 
  at 
  all, 
  with 
  a 
  smooth 
  margin 
  (Bigelow, 
  19096). 
  

   In 
  netnatopTiora 
  it 
  is 
  extremely 
  extensible; 
  capable 
  of 
  being 
  thrown 
  

   into 
  complex 
  folds, 
  and, 
  most 
  important, 
  its 
  margin 
  is 
  thickly 
  set 
  

   with 
  a 
  very 
  large 
  number 
  of 
  stalked 
  nematocyst 
  knobs 
  (pi. 
  3, 
  fig. 
  3). 
  

  

  Canal 
  system. 
  — 
  The 
  specimens 
  are 
  all 
  far 
  advanced 
  in 
  development, 
  

   and 
  m 
  most 
  of 
  them 
  all 
  the 
  canals 
  are 
  united 
  with 
  the 
  cruciform 
  base 
  

   of 
  the 
  manubrium. 
  But 
  in 
  one, 
  from 
  station 
  4766, 
  21 
  mm. 
  high, 
  

   one 
  of 
  the 
  18 
  canals 
  is 
  still 
  blind, 
  and 
  reaches 
  only 
  to 
  about 
  one-third 
  

   the 
  height 
  of 
  the 
  bell 
  cavity, 
  and 
  in 
  another 
  there 
  are 
  two 
  very 
  short 
  

   blind 
  canals 
  close 
  together 
  (pi. 
  3, 
  fig. 
  1). 
  In 
  several 
  instances 
  two 
  

   canals 
  unite, 
  the 
  union 
  being 
  either 
  close 
  to 
  the 
  margin 
  or 
  higher, 
  

   and 
  in 
  one 
  instance 
  a 
  canal 
  bifurcates 
  close 
  to 
  the 
  margin, 
  the 
  two 
  

   resultant 
  trunks 
  reuniting 
  just 
  below 
  thp 
  base 
  of 
  the 
  manubrium; 
  

   furthermore, 
  canals 
  may 
  be 
  connected 
  by 
  a 
  transverse 
  bridge, 
  as 
  also 
  

   happens 
  in 
  C. 
  simulans 
  (Bigelow, 
  1909&). 
  Evidence 
  of 
  the 
  readiness 
  

   with 
  which 
  canals 
  may 
  send 
  out 
  such 
  branches 
  is 
  afforded 
  by 
  the 
  fact 
  

   that 
  their 
  margins, 
  particularly 
  near 
  the 
  circular 
  canal, 
  are 
  sometimes 
  

   jagged 
  (pi. 
  3, 
  fig. 
  1), 
  sometimes 
  smooth. 
  The 
  position 
  of 
  each 
  canal 
  

   corresponds 
  to 
  the 
  center 
  of 
  a 
  longitudinal 
  band 
  of 
  subumbral 
  muscle 
  

   fibers 
  (pi. 
  2, 
  fig. 
  8). 
  No 
  such 
  condition 
  has 
  ever 
  been 
  observed 
  in 
  

   any 
  of 
  the 
  specimens 
  of 
  Calycopsis 
  previously 
  studied. 
  

  

  Tentacles. 
  — 
  As 
  a 
  rule 
  every 
  canal 
  which 
  reaches 
  the 
  margin 
  of 
  the 
  

   bell 
  is 
  associated 
  with 
  a 
  tentacle; 
  but 
  in 
  the 
  specimen 
  noted 
  above 
  

   the 
  two 
  short 
  centripetal 
  canals 
  bear 
  no 
  definite 
  relation 
  to 
  the 
  neigh- 
  

   bormg 
  tentacles. 
  This 
  fact 
  suggests 
  that 
  these 
  canals 
  are 
  merely 
  

   sporadic 
  outgrowtlis 
  from 
  the 
  circular 
  canal, 
  and 
  that 
  they 
  would 
  

   never 
  have 
  attained 
  much 
  greater 
  length. 
  Judging 
  from 
  this, 
  we 
  must 
  

   assume 
  either 
  that 
  tentacles 
  precede 
  canals 
  m 
  development 
  or 
  that 
  

   the 
  number 
  of 
  canals 
  present, 
  16-18, 
  is 
  about 
  the 
  final 
  one, 
  and 
  that 
  

   the 
  intermediate 
  tentacles 
  would 
  never 
  be 
  associated 
  with 
  canals. 
  

  

  The 
  different 
  sizes 
  of 
  the 
  canalar 
  tentacles 
  gives 
  us 
  an 
  idea 
  of 
  the 
  

   relative 
  ages 
  of 
  the 
  canals 
  with 
  which 
  they 
  are 
  associated. 
  Thus 
  in 
  

   one 
  example 
  only 
  4 
  of 
  the 
  canals 
  have 
  large 
  tentacles, 
  12 
  have 
  

   small 
  ones, 
  and 
  if 
  we 
  follow 
  the 
  series 
  through, 
  from 
  the 
  data 
  given 
  

   above, 
  we 
  find 
  a 
  specimen 
  with 
  6 
  large 
  and 
  1 
  1 
  small 
  ; 
  2 
  with 
  8 
  large, 
  

   8 
  and 
  10 
  small 
  — 
  that 
  is, 
  large 
  and 
  small 
  roughly 
  alternating 
  (pi. 
  3, 
  

  

  