﻿1864.] 
  DUNCAN 
  — 
  WEST 
  INDIAN 
  COEALS. 
  365 
  

  

  Antigua, 
  and 
  is 
  a 
  form 
  well 
  known 
  in 
  European 
  collections. 
  The 
  

   replacement 
  of 
  the 
  original 
  sclerenchyma 
  by 
  silica, 
  sometimes 
  

   transparent, 
  but 
  usually 
  of 
  the 
  exact 
  tint 
  of 
  the 
  original 
  texture, 
  is 
  

   evidently 
  of 
  the 
  same 
  date 
  as 
  the 
  filling 
  up 
  of 
  the 
  interspaces 
  with 
  

   homogeneous 
  and 
  differently 
  tinted 
  silica 
  ; 
  the 
  probable 
  filling 
  up 
  of 
  

   the 
  interseptal 
  loculi 
  and 
  exothecal 
  cells, 
  in 
  the 
  first 
  instance, 
  with 
  

   carbonate 
  of 
  lime, 
  and 
  its 
  gradual 
  silicification 
  by 
  transposition, 
  

   will 
  be 
  noticed 
  in 
  considering 
  the 
  calcareo-siliceous 
  fossils 
  of 
  the 
  

   latest 
  Antiguan 
  formation. 
  

  

  It 
  is 
  very 
  common 
  to 
  find 
  the 
  granules 
  on 
  the 
  septa, 
  their 
  spines, 
  

   and 
  the 
  exothecal 
  and 
  endothecal 
  dissepiments 
  deficient, 
  in 
  some 
  of 
  

   the 
  most 
  perfect 
  specimens, 
  and 
  a 
  general 
  wasting 
  of 
  the 
  walls 
  also. 
  

   The 
  space 
  occupied 
  formerly 
  by 
  these 
  details 
  is 
  filled 
  up 
  with 
  the 
  

   variously 
  coloured 
  silica 
  of 
  the 
  interspaces, 
  so 
  that 
  an 
  originally 
  

   stout 
  Astraean 
  may 
  be 
  presented 
  as 
  a 
  very 
  delicate 
  form*. 
  It 
  is 
  

   impossible 
  to 
  determine 
  how 
  much 
  of 
  this 
  loss 
  has 
  been 
  of 
  prefossil 
  

   date, 
  but 
  some 
  of 
  it 
  occurs 
  during 
  the 
  replacement 
  of 
  the 
  calcareous 
  

   by 
  the 
  siliceous 
  particles. 
  This 
  variety 
  of 
  fossilization 
  affects 
  the 
  

   fossil 
  woods 
  of 
  many 
  "West 
  Indian 
  islands, 
  preserving 
  their 
  details 
  

   with 
  considerable 
  accuracy 
  in 
  some 
  instances, 
  but 
  producing 
  much 
  

   destruction 
  of 
  them 
  in 
  others. 
  

  

  The 
  third 
  variety 
  particularly 
  affects 
  two 
  genera 
  f 
  of 
  the 
  Anti- 
  

   guan 
  corals, 
  and 
  renders 
  their 
  diagnosis 
  very 
  difficult 
  ; 
  in 
  fact, 
  lon- 
  

   gitudinal 
  sections 
  of 
  masses 
  thus 
  sihcified 
  have 
  been 
  mistaken 
  for 
  

   the 
  cut 
  surfaces 
  of 
  Bryozoa. 
  The 
  colouring 
  of 
  the 
  silica 
  which 
  fills 
  

   up 
  the 
  interspaces 
  is 
  very 
  bright 
  and 
  various 
  in 
  some 
  specimens, 
  

   and 
  in 
  others 
  is 
  complicated 
  by 
  the 
  presence 
  of 
  much 
  opaque-white 
  

   chert, 
  as 
  well 
  as 
  of 
  minute 
  collections 
  of 
  crystalline 
  quartz 
  in 
  the 
  

   coloured 
  silica. 
  The 
  silica 
  is 
  rarely 
  transparent 
  when 
  brightly 
  

   coloured, 
  but 
  is 
  jasperoid 
  and 
  granular, 
  or 
  opalescent 
  ; 
  when 
  lighter 
  

   in 
  tint, 
  it 
  is 
  more 
  or 
  less 
  homogeneous 
  and 
  transparent 
  ; 
  and 
  it 
  is 
  

   by 
  comparing 
  the 
  transparent 
  with 
  the 
  opaque 
  varieties 
  of 
  this 
  form 
  

   of 
  silicification, 
  and 
  by 
  mapping 
  out 
  the 
  sclerenchyma, 
  which 
  is 
  

   usually 
  transparent, 
  pale 
  grey, 
  or 
  slightly 
  opaque, 
  that 
  the 
  coral 
  can 
  

   be 
  determined 
  to 
  be 
  a 
  species 
  of 
  a 
  genus 
  well 
  known 
  under 
  a 
  very 
  

   different 
  form 
  of 
  fossilization 
  in 
  European 
  Mid-tertiary 
  collections. 
  

   The 
  transverse 
  endothecal 
  dissepiments 
  which 
  divide 
  the 
  coloured 
  

   rectangular 
  spaces 
  in 
  longitudinal 
  sections, 
  like 
  the 
  steps 
  of 
  a 
  ladder, 
  

   give 
  the 
  coral 
  the 
  appearance 
  of 
  a 
  tabulate 
  form 
  ; 
  but 
  it 
  will 
  be 
  ob- 
  

   served 
  that 
  the 
  dissepiments 
  of 
  one 
  interseptal 
  space 
  are 
  not 
  on 
  the 
  

   same 
  plane 
  as 
  those 
  of 
  the 
  others, 
  and 
  that 
  the 
  supposed 
  tabulae 
  are 
  

   not 
  continued 
  throughout 
  the 
  whole 
  diameter 
  of 
  the 
  corallite. 
  The 
  

   apparent 
  quaternary 
  arrangement 
  of 
  the 
  principal 
  septa 
  is, 
  moreover, 
  

   very 
  curious 
  and 
  deceptive 
  J. 
  There 
  are 
  many 
  impressions 
  of 
  the 
  

   calices 
  of 
  a 
  species 
  of 
  this 
  genus, 
  and 
  some 
  nearly 
  perfect 
  calices 
  also, 
  

   in 
  the 
  collections 
  of 
  the 
  Geological 
  Society 
  and 
  British 
  Museum. 
  

  

  * 
  As 
  in 
  Astrcea 
  endothecata, 
  nobis, 
  vars. 
  1 
  and 
  2, 
  Antigua, 
  

   t 
  Astrocoenia 
  and 
  Stylocoenia. 
  

  

  \ 
  Stylocoenia 
  lobato-rotimdata, 
  Edwards 
  and 
  Haime. 
  

   VOL. 
  XX. 
  PART 
  I. 
  2D 
  

  

  