﻿Siphonophora 
  of 
  the 
  Ganary 
  Islands. 
  2 
  41 
  

  

  the 
  internodes, 
  i. 
  e. 
  in 
  the 
  interval 
  between 
  two 
  polypites, 
  

   in 
  all 
  stages 
  of 
  development, 
  which 
  originate 
  irregularly 
  

   dispersed 
  upon 
  the 
  stem. 
  Hackel 
  himself 
  says 
  : 
  — 
  " 
  Finally 
  

   all 
  order 
  ceases 
  and 
  the 
  whole 
  stem 
  appears 
  beset 
  with 
  

   hundreds 
  or 
  thousands 
  of 
  different 
  appendages 
  (siphons, 
  

   palpones, 
  gonophores, 
  bracts, 
  &c.) 
  irregularly 
  grouped 
  ; 
  so 
  

   that 
  it 
  is 
  impossible 
  to 
  discriminate 
  the 
  different 
  correlated 
  

   constituents 
  of 
  the 
  dissolved 
  cormidia." 
  

  

  I 
  hope, 
  however, 
  to 
  be 
  able 
  to 
  demonstrate 
  that 
  in 
  Hali- 
  

   stemma 
  (probably 
  also 
  in 
  the 
  other 
  Agalmidse 
  and 
  the 
  

   Forskalia?) 
  a 
  strict 
  rule 
  prevails 
  in 
  the 
  production 
  of 
  the 
  

   group-appendages. 
  In 
  Ilalistemma, 
  as 
  is 
  well 
  known, 
  the 
  

   internodial 
  appendages 
  of 
  the 
  stem 
  (i. 
  e. 
  those 
  situated 
  between 
  

   two 
  polypites) 
  are 
  grouped 
  in 
  such 
  a 
  manner 
  that 
  a 
  

   palp 
  with 
  its 
  tactile 
  filaments, 
  a 
  bract, 
  and 
  a 
  male 
  and 
  

   a 
  female 
  gonophore 
  concur 
  to 
  form 
  a 
  group. 
  As 
  I 
  may 
  

   mention 
  in 
  passing, 
  the 
  female 
  gonophores 
  developing 
  near 
  

   the 
  palp 
  constantly 
  occupy 
  a 
  proximal, 
  the 
  male 
  ones 
  

   on 
  the 
  contrary 
  a 
  distal 
  position 
  on 
  the 
  stem. 
  In 
  the 
  individual 
  

   internodes 
  the 
  number 
  of 
  the 
  groups 
  constantly 
  increases 
  to- 
  

   wards 
  the 
  end 
  of 
  the 
  stem 
  ; 
  thus, 
  while 
  in 
  the 
  proximal 
  inter- 
  

   nodes 
  we 
  find 
  only 
  two 
  or 
  three 
  groups, 
  there 
  are 
  12-15 
  in 
  

   the 
  distal 
  internodes. 
  The 
  stem 
  therefore 
  grows 
  internodially, 
  

   and 
  indeed 
  in 
  such 
  a 
  way 
  that 
  in 
  each 
  individual 
  internode 
  the 
  

   groups 
  continually 
  decrease 
  in 
  size 
  in 
  a 
  proximal 
  direction. 
  

   Hence, 
  if 
  we 
  trace 
  the 
  groups 
  of 
  the 
  stem 
  in 
  a 
  distal 
  direction, 
  

   we 
  find 
  behind 
  the 
  polypite 
  with 
  its 
  tentacles 
  the 
  buds 
  of 
  

   the 
  youngest 
  group, 
  and 
  then 
  the 
  other 
  groups 
  gradually 
  

   increasing 
  in 
  size. 
  The 
  youngest 
  group 
  consists 
  of 
  a 
  bud 
  for 
  

   the 
  poiypite 
  and 
  tentacles, 
  and 
  two 
  buds 
  situated 
  close 
  to 
  it, 
  

   one 
  of 
  which, 
  as 
  the 
  female 
  primordial 
  bud, 
  originates 
  the 
  

   female 
  gonophores, 
  and 
  the 
  other, 
  as 
  the 
  male 
  primordial 
  bud, 
  

   the 
  male 
  gonophores. 
  Somewhat 
  more 
  distally 
  comes 
  the 
  bud 
  

   for 
  the 
  bract. 
  

  

  To 
  show 
  clearly 
  the 
  further 
  conditions 
  of 
  growth 
  in 
  the 
  

   internode, 
  I 
  indicate 
  by 
  A, 
  B, 
  C 
  . 
  . 
  . 
  the 
  polypites 
  with 
  the 
  

   tentacles 
  belonging 
  to 
  them, 
  and 
  by 
  a^b,c 
  . 
  . 
  . 
  the 
  bud-groups 
  

   situated 
  in 
  the 
  internode 
  between 
  two 
  polypites. 
  If, 
  then, 
  

   A 
  and 
  a 
  signify 
  the 
  oldest 
  groups, 
  and 
  13 
  and 
  b 
  &c. 
  the 
  suc- 
  

   cessive 
  younger 
  ones, 
  we 
  should 
  obtain 
  for 
  the 
  last 
  internode 
  

   of 
  the 
  stem 
  the 
  following 
  formula 
  : 
  — 
  

  

  B 
  . 
  h 
  . 
  g 
  ./. 
  e.d.c.b.aA. 
  

  

  At 
  this 
  stage, 
  however, 
  only 
  the 
  younger 
  specimens 
  of 
  

   Halistemma 
  remain. 
  The 
  number 
  of 
  groups 
  in 
  the 
  internode, 
  

   which 
  in 
  our 
  special 
  case 
  we 
  have 
  assumed 
  to 
  be 
  8 
  {a-h) 
  } 
  

  

  