54 MEMOIRS OF TUE QUEENSLAND MUSEUM. 



reaching the margin at an extremely acute angle ; 7 to 10 basal radii, arranged fan- 

 ■\vise, the basal margin only weakly scalloped ; nucleus apicad of middle ; ctenoid 

 patch large, occupying the whole apical field, but the elements very small, the marginal 

 teeth extremely minute. The ctenoid structures are modified circuli, into which they 

 can be seen to pass ; they aj)pear to be, in fact, crowded segments of circuli, all much 

 longer than broad. 



These scales suggest those of Anabas, but differ as follows : — Basal radii less- 

 numerous, and not nearly so close together ; lateral circvili much finer and denser ; 

 no circuli meeting at an angle above the nucleus ; ctenoid patch much more extensive ; 

 apical teeth much smaller. The Anabas examined are A. munii from Egj'pt and 

 A. scandens from the Philippine Islands. 



There are many families in which the scales are of the same general type as 

 those of Toxotes, but in no case, so far as my material shows, is the resemblance 

 very close. 



SERRANIDiE. 

 Interpreting this family in a broad sense, the following may be included : — 



Radial system consisting of a more or less elongate ring in the nuclear region, from which proceed 

 strong radii, the basal ones many and long, the apical ones short ; circuli extremely fine 



Plesiops nigricans Rupp. 

 (Moreton Bay). 



Radii ordinary, all basal ; no nuclear ring . . . . . . . . . . . . . . 1. 



1. Scales broader than long ; basal radii 4 to 6 ; apical circuli transverse ; scales wholly cycloid 



Ambassis nalua Ham. Buch.. 

 (Cape Bowling Green). 



Scales longer than broad, with ctenoid elements, though not always with marginal teeth . . 2. 



2. Ctenoid patch large, marginal teeth numerous and well developed ; scales moderately broad . . 



Plectroplites ambiguus Rich. 

 (Condamine River). 



Ctenoid patch small, marginal teeth few or absent, scale narrower . . . . . . . . 3. 



3. Ctenoid (tooth or spine like) elements only developed in a patch just above the nucleus, the 



region beyond with broken remains of circuli, more or less V-shaped, conseciuently no 

 apical teeth whatever . . . . . . . . . . . . Epinephehis lanceolatus Forsk. 



(Moreton Bay). 



Ctenoid elements reaching the margin, where there are projecting spine -like teeth ; scales small 

 and very narrow . . . . . . . . . . . . Oligorus macquariensis C. & V. 



(Condamine River). 



It will be seen that at least three entirely different types of scale are represented 

 here. It must be stated, however, that some Plesiops scales (P. coralUcola, Eton Is.) 

 have the nucleus very near the apex, with long radii proceeding fan--wise basad from 

 it, no radial ring and no apical radii. In P. coralUcola scales, whether as just described 

 or with the radiaJ ring and apical as w^ell as basal radii, there are strong ctenoid elements 

 on the apical margin, with many sharp teeth. In P. nigricans the ctenoid features 

 are less distinct, largely concealed by the dark skin. According to Boulenger, 

 P. coralUcola is not to be separated from P. nigricans. 



Ambassis constitutes a very distinct group, Ambassinse or Ambassidse. Other 

 species of Epinephelus, as E. niveatus and E. megachir, have very strongly developed 

 apical teeth, so E. lanceolatus appears to be an aberrant species. Oligorus scales 

 closely resemble those of Ejnnephelus niveatus C. & V., from the Atlantic. 



