﻿186 
  PROCEEDINGS 
  OF 
  THE 
  NATIONAL 
  MUSEUM 
  vol.96 
  

  

  material 
  from 
  the 
  known 
  American 
  range 
  of 
  both 
  species 
  : 
  ■mufasciatus, 
  

   from 
  Cape 
  Cod 
  to 
  Uruguay 
  in 
  the 
  Atlantic 
  and 
  from 
  San 
  Diego 
  - 
  

   to 
  Peru 
  in 
  the 
  Pacific 
  ; 
  and 
  hrasiliensis, 
  from 
  New 
  York 
  to 
  Brazil. 
  

  

  Gill 
  (1859, 
  p. 
  131) 
  based 
  Hyporhamphus 
  principally 
  on 
  the 
  tricuspid 
  

   teeth 
  (whence 
  the 
  name 
  of 
  the 
  type 
  species, 
  H. 
  trieuspidatus, 
  a 
  syno- 
  

   nym 
  of 
  unifasciatus) 
  , 
  but 
  he 
  later 
  (1863) 
  found, 
  and 
  Poey 
  (1860, 
  

   p. 
  298) 
  previously 
  had 
  noted, 
  that 
  Eemiramphus 
  likewise 
  has 
  tricus- 
  

   pid 
  teeth. 
  Poey's 
  and 
  Gill's 
  observations 
  on 
  the 
  nature 
  of 
  the 
  teeth 
  

   were 
  correct, 
  and 
  hence 
  I 
  do 
  not 
  agree 
  with 
  Weed 
  (1933, 
  pp. 
  47, 
  57) 
  

   and 
  others 
  who 
  stated 
  that 
  the 
  teeth 
  are 
  simple 
  in 
  Hemiramphus. 
  As 
  

   Smith 
  (1933) 
  has 
  shown, 
  and 
  as 
  I 
  have 
  also 
  observed, 
  the 
  form 
  and 
  ar- 
  

   rangement 
  of 
  the 
  teeth 
  vary 
  with 
  age 
  and 
  with 
  different 
  species. 
  The 
  

   jaws 
  of 
  a 
  single 
  individual 
  may 
  have 
  unicuspid, 
  bicuspid, 
  and 
  tri- 
  

   cuspid 
  teeth, 
  and, 
  in 
  at 
  least 
  one 
  American 
  species, 
  Hyporhamphus 
  

   rosae 
  (Jordan 
  and 
  Gilbert), 
  only 
  the 
  largest 
  individuals 
  appear 
  to 
  

   have 
  tricuspid 
  teeth 
  — 
  hence 
  the 
  frequent 
  statement 
  that 
  H. 
  rosae 
  has 
  

   only 
  unicuspid 
  teeth. 
  

  

  The 
  fundamental 
  characters 
  distinguishing 
  the 
  American 
  species 
  of 
  

   Heiniramphus 
  and 
  Hyporhamphus^ 
  such 
  as 
  the 
  presence 
  or 
  absence 
  

   of 
  scales 
  on 
  the 
  upper 
  jaw, 
  the 
  presence 
  or 
  absence 
  of 
  a 
  bony 
  rim 
  along 
  

   the 
  side 
  of 
  the 
  nasal 
  fossa, 
  and 
  the 
  arrangement 
  of 
  the 
  sensory 
  canal 
  

   and 
  pores 
  on 
  the 
  preorbital 
  (fig. 
  9), 
  may 
  be 
  features 
  that 
  will 
  separate 
  

   world 
  half 
  beaks 
  of 
  this 
  type. 
  This 
  is 
  suggested 
  to 
  me 
  by 
  Smith 
  (1933, 
  

   pp. 
  130-131), 
  who 
  made 
  a 
  primary 
  division 
  in 
  his 
  key 
  on 
  the 
  basis 
  of 
  

   a 
  naked 
  versus 
  a 
  scaled 
  upper 
  jaw, 
  and 
  b}^ 
  the 
  very 
  few 
  Old 
  World 
  

   half 
  beaks 
  I 
  have 
  examined. 
  In 
  Euleptorhamphus 
  Gill, 
  however, 
  the 
  

   upper 
  jaw 
  is 
  scaled 
  as 
  in 
  Hyporha7nj)hus, 
  whereas 
  the 
  rim 
  and 
  the 
  

   form 
  of 
  the 
  nasal 
  fossa 
  and 
  the 
  sensory 
  canal 
  of 
  the 
  preorbital 
  are 
  

   essentially 
  as 
  in 
  Hemiramphus. 
  

  

  The 
  pattern 
  of 
  the 
  scales 
  on 
  the 
  uj^per 
  jaw, 
  the 
  shape 
  of 
  the 
  pre- 
  

   orbital, 
  and 
  the 
  arrangement 
  of 
  the 
  teeth 
  maj^ 
  be 
  found 
  to 
  have 
  generic 
  

   or 
  only 
  specific 
  value. 
  The 
  solution 
  of 
  these 
  problems 
  will 
  necessitate 
  

   a 
  comprehensive 
  review 
  of 
  the 
  halfbeaks 
  of 
  the 
  world. 
  

  

  The 
  form 
  of 
  the 
  sensory 
  canal 
  and 
  the 
  pore 
  on 
  the 
  preorbital 
  are 
  

   usually 
  visible 
  in 
  Hyporhamphus^ 
  but 
  the 
  overlying 
  scales 
  and 
  skin 
  

   must 
  be 
  dissected 
  from 
  this 
  bone 
  in 
  Heniiramphus 
  before 
  the 
  canal 
  

   and 
  pores 
  can 
  be 
  clearly 
  seen. 
  The 
  two 
  pores 
  shown 
  near 
  the 
  upper 
  

   end 
  of 
  the 
  posterior 
  margin 
  of 
  the 
  preopercle 
  in 
  Hyporhamphus 
  (fig. 
  

   9, 
  A) 
  are 
  apparently 
  absent 
  in 
  Hemiramphus^ 
  but 
  this 
  character 
  was 
  

   checked 
  only 
  on 
  a 
  comparatively 
  few 
  individuals 
  of 
  each 
  genus. 
  

  

  In 
  table 
  1 
  I 
  have 
  abandoned 
  the 
  "key" 
  characters 
  — 
  air 
  bladder 
  cel- 
  

   lular 
  or 
  simple, 
  sides 
  of 
  body 
  vertical 
  or 
  convex, 
  position 
  and 
  shape 
  of 
  

  

  2 
  In 
  material 
  from 
  San 
  Diego, 
  Calif. 
  (Stanford 
  Nat. 
  Hist. 
  Mua. 
  No. 
  9912) 
  I 
  found 
  one 
  

   specimen 
  of 
  this 
  species, 
  Which, 
  to 
  my 
  linowledge, 
  represents 
  a 
  northward 
  extension 
  of 
  

   known 
  range 
  on 
  the 
  Paciflc 
  coast. 
  

  

  