﻿258 
  

  

  THE 
  EXTINCT 
  VERTEBRATE 
  ANIMALS 
  

  

  Smith 
  Woodward 
  (47, 
  p. 
  363). 
  But, 
  as 
  we 
  shall 
  see 
  further 
  on 
  

   (p. 
  270), 
  I 
  have 
  shown 
  that 
  the 
  tooth 
  which 
  Agassiz 
  adopted 
  as 
  

   the 
  type 
  of 
  Cricodus 
  is 
  not 
  a 
  Rhizodont 
  tooth, 
  but 
  belongs, 
  where 
  

   Duff 
  originally 
  referred 
  it, 
  to 
  the 
  genus 
  Dendrodus 
  of 
  Owen, 
  that 
  

   is, 
  Holoptychins 
  of 
  Agassiz. 
  The 
  genus 
  Cricodus 
  therefore 
  falls, 
  and 
  

   Pander's 
  name 
  Polyplocodus 
  must 
  come 
  into 
  use 
  for 
  the 
  Eussian 
  

   remains 
  in 
  question, 
  with 
  which 
  the 
  Nairn 
  specimens 
  are 
  evidently 
  

   closely 
  allied, 
  and 
  for 
  which 
  I 
  now 
  propose 
  the 
  name 
  of 
  Poly- 
  

   plocodus 
  leptognathus. 
  

  

  Family 
  COCCOSTEIDjE. 
  

  

  Genus 
  COCCOSTJSUS, 
  Agassiz. 
  

   Coccosteus 
  magnus, 
  sp. 
  nov. 
  

  

  (Plate 
  vi. 
  fig. 
  10.) 
  

  

  Among 
  the 
  fish-remains 
  from 
  Kingsteps 
  in 
  my 
  collection 
  is 
  a 
  unique 
  

   plate 
  represented 
  in 
  Plate 
  vi. 
  fig. 
  10. 
  It 
  is 
  bilaterally 
  symmetri- 
  

   cal, 
  and 
  measures 
  2J 
  inches 
  in 
  length 
  by 
  the 
  same 
  in 
  greatest 
  

   breadth. 
  Its 
  broad 
  anterior 
  margin 
  is 
  gently 
  rounded, 
  and 
  passes 
  

   out 
  on 
  each 
  side 
  into 
  a 
  sharp 
  angle, 
  which 
  immediately 
  falls 
  away 
  

   behind 
  into 
  the 
  lateral 
  margin 
  ; 
  the 
  two 
  lateral 
  margins 
  then 
  

   gently 
  converge 
  to 
  the 
  very 
  short 
  posterior 
  one, 
  which 
  truncates 
  

   the 
  bone 
  behind. 
  The 
  surface 
  exhibited, 
  obviously 
  eroded, 
  is 
  

   slightly 
  concave 
  in 
  front 
  and 
  flat 
  behind, 
  and 
  exhibits 
  in 
  its 
  middle 
  

   third 
  a 
  slightly 
  elevated 
  central 
  longitudinal 
  ridge. 
  

  

  That 
  this 
  is 
  the 
  anterior 
  median 
  ventral 
  plate 
  of 
  a 
  large 
  Coccos- 
  

   tean, 
  seen 
  from 
  the 
  inner 
  side, 
  is 
  beyond 
  all 
  doubt. 
  It 
  differs, 
  how- 
  

   ever, 
  not 
  only 
  in 
  its 
  great 
  size 
  from 
  that 
  of 
  C. 
  decipiens, 
  but 
  also 
  in 
  its 
  

   greater 
  proportional 
  breadth, 
  and 
  in 
  the 
  fact 
  that 
  along 
  its 
  anterior 
  

   aspect 
  no 
  facet 
  is 
  seen 
  for 
  taking 
  part 
  in 
  the 
  articulation 
  with 
  the 
  

   interlateral 
  plate. 
  But 
  the 
  plate 
  is 
  evidently 
  considerably 
  eroded, 
  

   besides 
  which, 
  in 
  this 
  species, 
  the 
  interlateral 
  articulation, 
  which 
  

   apparently 
  takes 
  place 
  on 
  both 
  sides 
  of 
  the 
  anterior 
  margin 
  of 
  the 
  

   corresponding 
  plate 
  in 
  C. 
  decipiens 
  and 
  C. 
  minor, 
  probably 
  here 
  

   were 
  limited 
  to 
  the 
  outer 
  surface. 
  It 
  may 
  be 
  noted 
  that 
  the 
  

   contour 
  of 
  the 
  plate 
  is 
  quite 
  similar 
  to 
  that 
  of 
  the 
  one 
  figured 
  by 
  

   Pander 
  from 
  the 
  Aa 
  in 
  Livonia 
  (33, 
  PI. 
  B., 
  fig. 
  1). 
  

  

  Possibly 
  the 
  access 
  of 
  additional 
  material 
  may 
  show 
  that 
  these 
  

   Scottish 
  and 
  Russian 
  Upper 
  Devonian 
  Coccostean 
  remains 
  are 
  generi- 
  

  

  