﻿116 
  PROCEEDINGS 
  OF 
  THE 
  NATIONAL 
  MUSEUM 
  vol.98 
  

  

  tongue; 
  cheek 
  and 
  operculum 
  scaled; 
  posterior 
  and 
  lower 
  margins 
  

   of 
  orbit 
  denticulate 
  and 
  two 
  denticulate 
  ridges 
  on 
  preorbital 
  bone 
  

   above 
  maxillary, 
  the 
  upper 
  one 
  forming 
  part 
  of 
  the 
  serrae 
  around 
  

   orbit 
  ; 
  preoperculum 
  with 
  double 
  edge, 
  the 
  outer 
  one 
  serrated 
  on 
  both 
  

   posterior 
  and 
  ventral 
  borders, 
  the 
  inner 
  one 
  serrated 
  at 
  lower 
  angle 
  

   and 
  a 
  little 
  on 
  each 
  side 
  of 
  the 
  angle 
  above 
  and 
  forward; 
  inter- 
  

   operculum 
  denticulate 
  ; 
  operculum 
  not 
  spinate 
  ; 
  gill 
  membranes 
  united 
  

   far 
  forward, 
  free 
  from 
  isthmus; 
  branchiostegals 
  6; 
  pseudobranchiae 
  

   present 
  ; 
  gill 
  rakers 
  of 
  moderate 
  length, 
  of 
  rather 
  heavy 
  build 
  ; 
  a 
  single 
  

   dorsal 
  fin 
  with 
  the 
  spiny 
  portion 
  scarcely 
  longer 
  than 
  soft 
  portion 
  of 
  

   VIIT 
  or 
  IX 
  spines, 
  the 
  second 
  or 
  third 
  longest, 
  the 
  next 
  to 
  the 
  last 
  

   shortest; 
  spinous 
  dorsal 
  fin 
  preceded 
  by 
  a 
  short 
  recumbent 
  spine, 
  

   directed 
  forward, 
  hidden 
  in 
  the 
  skin 
  beneath 
  the 
  scales 
  ; 
  anal 
  with 
  3 
  

   spines 
  and 
  usually 
  10 
  soft 
  rays 
  ; 
  both 
  fins 
  with 
  a 
  sheath 
  of 
  scales 
  along 
  

   their 
  bases, 
  one 
  scale 
  in 
  width 
  along 
  dorsal 
  and 
  one 
  or 
  two 
  along 
  anal, 
  

   these 
  fins 
  partly 
  depressible 
  between 
  the 
  sheath 
  ; 
  pectorals 
  ii, 
  13, 
  asym- 
  

   metrical, 
  upper 
  2 
  or 
  3 
  branched 
  rays 
  longest 
  ; 
  pelvic 
  fins, 
  I, 
  5, 
  inserted 
  

   under 
  base 
  of 
  pectorals 
  and 
  a 
  little 
  in 
  front 
  of 
  dorsal 
  origin; 
  pelvic 
  

   spine 
  slender, 
  not 
  reaching 
  to 
  anus 
  and 
  about 
  two-thirds 
  length 
  of 
  

   first 
  branched 
  pelvic 
  ray, 
  which 
  ends 
  in 
  a 
  short 
  filament; 
  axillary 
  

   scale 
  of 
  pelvic 
  small, 
  about 
  length 
  of 
  diameter 
  of 
  pupil; 
  caudal 
  fin 
  

   forked, 
  lobes 
  more 
  or 
  less 
  pointed. 
  

  

  Other 
  characters 
  are 
  those 
  of 
  the 
  genotype, 
  X. 
  honessi, 
  described 
  

   below. 
  

  

  This 
  new 
  genus 
  is 
  related 
  to 
  the 
  percoid 
  fishes 
  usually 
  referred 
  to 
  

   the 
  family 
  Centropomidae 
  but 
  sometimes 
  separated 
  from 
  them 
  and 
  

   grouped 
  in 
  the 
  family 
  Ambassidae 
  or 
  Chandidae. 
  Xenainbassis 
  is 
  

   especially 
  close 
  to 
  Tetra^entrmn 
  Macleay, 
  Synecliopterus 
  Norman, 
  

   and 
  Amhassis 
  Cuvier 
  and 
  Valenciennes, 
  differing 
  from 
  them 
  as 
  in- 
  

   dicated 
  in 
  the 
  accompanying 
  key. 
  Ambassis 
  differs 
  from 
  the 
  new 
  

   genus 
  and 
  from 
  Tetracentrum 
  and 
  Synechopterus 
  by 
  having 
  the 
  first 
  

   dorsal 
  fin 
  made 
  up 
  of 
  VII 
  spines 
  and 
  joined 
  to 
  the 
  base 
  of 
  the 
  first 
  

   spine 
  of 
  the 
  second 
  dorsal, 
  whereas 
  in 
  the 
  other 
  three 
  genera 
  the 
  first 
  

   dorsal 
  is 
  joined 
  at 
  least 
  halfway 
  out 
  the 
  last 
  spine 
  so 
  that 
  the 
  two 
  

   fins 
  are 
  continuous. 
  Chanda 
  Buchanan-Hamilton, 
  1822, 
  with 
  G. 
  lala 
  

   as 
  the 
  type 
  as 
  restricted 
  by 
  Fowler 
  (Proc. 
  Acad. 
  Nat. 
  Sci. 
  Philadel- 
  

   phia, 
  1905, 
  p. 
  500) 
  differs 
  from 
  the 
  above-mentioned 
  genera 
  chiefly 
  in 
  

   having 
  14 
  to 
  17 
  soft 
  rays 
  in 
  the 
  anal 
  fin. 
  Pseudoarribassis 
  Castelnau, 
  

   1878 
  {— 
  Austrochanda 
  Whitley, 
  1935), 
  a 
  substitute 
  name 
  supposed 
  by 
  

   Whitley 
  to 
  be 
  preoccupied 
  by 
  Pseudarribassis 
  Bleeker, 
  1876; 
  and 
  Vel- 
  

   anihassis, 
  proj)osed 
  by 
  Whitley 
  (Rec. 
  South 
  Australian 
  Mus., 
  vol. 
  5, 
  

   No. 
  3, 
  pp. 
  356-365, 
  1935) 
  , 
  are 
  here 
  considered 
  as 
  subgenera 
  of 
  Amhassis 
  

   Cuvier 
  and 
  Valenciennes. 
  Indeed, 
  they 
  may 
  even 
  be 
  synonyms 
  of 
  

   Ambassis, 
  since 
  the 
  generic 
  differences 
  appear 
  so 
  slight. 
  I 
  also 
  refer 
  

  

  