﻿524 
  

  

  PROCEEDINGS 
  OF 
  THE 
  NATIONAL 
  MUSEUM 
  

  

  Samoan 
  Islands 
  should 
  be 
  reexamined. 
  U. 
  tragula 
  has 
  not 
  been 
  

   reported 
  from 
  the 
  Hawaiian 
  faunal 
  area. 
  

  

  Remarks. 
  — 
  The 
  change 
  and 
  variation 
  in 
  length 
  of 
  barbel 
  with 
  in- 
  

   crease 
  in 
  body 
  length 
  was 
  investigated 
  (table 
  12) 
  but 
  no 
  appreciable 
  

   difference 
  was 
  found. 
  

  

  This 
  species 
  appears 
  constant 
  in 
  body 
  proportions, 
  meristic 
  counts, 
  

   and 
  coloration 
  over 
  its 
  range. 
  Counts 
  of 
  the 
  number 
  of 
  gillrakers 
  and 
  

   vertical 
  scale 
  rows 
  were 
  segregated 
  by 
  locality 
  (table 
  12) 
  and 
  these 
  

   data 
  do 
  not 
  even 
  suggest 
  any 
  population 
  divergence. 
  

  

  U. 
  tragula 
  is 
  most 
  closely 
  related 
  to 
  oligospilus 
  and 
  luzonius 
  and 
  

   compared 
  with 
  them 
  in 
  their 
  descriptive 
  accounts. 
  Table 
  3 
  shows 
  the 
  

   relationship 
  of 
  the 
  number 
  of 
  oblique 
  bars 
  on 
  the 
  caudal 
  fin 
  with 
  

   increase 
  in 
  body 
  size 
  in 
  these 
  three 
  species. 
  

  

  Table 
  12. 
  — 
  Number 
  of 
  gillrakers 
  and 
  vertical 
  scale 
  rows, 
  by 
  locality, 
  and 
  length 
  of 
  

   chin 
  barbel 
  in 
  four 
  size-groups 
  in 
  Upeneus 
  tragula 
  

  

  Locality 
  

  

  Number 
  of 
  gillrakers 
  

  

  21 
  22 
  23 
  24 
  25 
  

  

  Number 
  of 
  vertical 
  scale 
  rows 
  

  

  28 
  29 
  30 
  31 
  32 
  

  

  Zanzibar 
  

  

  East 
  Indies. 
  

   Philippines. 
  

  

  China 
  

  

  Japan 
  

  

  Palau 
  

  

  A 
  specimen 
  of 
  U. 
  suhvittatus 
  Snyder 
  (1907, 
  p. 
  101), 
  Stanford 
  

   University 
  Natural 
  History 
  Museum 
  No. 
  20156, 
  collected 
  by 
  Jordan 
  

   and 
  Snyder 
  at 
  Wakanoura, 
  Japan, 
  is 
  without 
  question 
  U. 
  tragula. 
  

   The 
  following 
  counts 
  for 
  this 
  specimen 
  almost 
  all 
  fall 
  on 
  the 
  modes 
  of 
  

   the 
  frequency 
  distributions 
  given 
  above 
  for 
  tragula: 
  pectoral 
  fin 
  

   rays 
  28, 
  gillrakers 
  2, 
  4 
  + 
  1 
  + 
  11, 
  5 
  totaling 
  23, 
  barbel 
  length 
  in 
  head 
  

   length 
  58 
  percent, 
  vertical 
  scale 
  rows 
  29. 
  Other 
  important 
  characters 
  

   as 
  the 
  silvery 
  peritoneum, 
  dark 
  lateral 
  stripe 
  on 
  body, 
  dark 
  dorsal 
  

   saddles, 
  and 
  oblique 
  bars 
  on 
  caudal 
  fin 
  are 
  also 
  identical 
  to 
  those 
  of 
  

   tragula. 
  Snyder 
  gave 
  a 
  "length" 
  of 
  175 
  mm. 
  in 
  contrast 
  to 
  my 
  

   measure 
  of 
  228 
  mm. 
  (standard 
  length) 
  for 
  the 
  above 
  specimen. 
  

   This 
  discrepancy 
  is 
  not 
  explainable. 
  

  

  