[5] REVIEW OF THE SCLENID. 347 
a. Dorsal spines well separated, the first dorsal spine * attached to the third or fourth 
interneural, not more than two* of 
the spine-bearing interneurals being 
placed between the same pair of ver- 
tebra; soft rays of dorsal fin 17 to 32 
(37 to 40 in Lonchurus, 45 to 50 in Sci- 
cenoides) ; occipital crest not greatly 
b. Lower pharyngeals separate. elevated. 
c. Lower jaw without barbels. 
d. Caudal fin moderately scaly, its distal portion usually more or less naked, 
the scales not numerous enough to 
give athickened appearance to the fin. 
e. Teeth well developed, permanent in both jaws. 
Jj. Lower pharyngeals rather narrow; their teeth conic and mostly sharp ; 
none of them molar; outer teeth of 
upper jaw more or less enlarged. 
g. Gill-rakers comparatively long and slender; mouth more or less ob- 
lique, anal fin usually (but not always) 
inserted posteriorly; preorbital usually 
narrow, flat ; edge of snout above up- 
per jaw with the pores and slits little 
conspicuous or obsolete. 
h. Preopercle without bony teeth or serrations, its membranaceous mar- 
gin entire, crenulate or ciliate (two or 
three slender spinules present in Col- 
lichthys); teeth of lower jaw in few 
4 series. 
i. Skull excessively cavernous, soft and spongy to the touch, the inter- 
orbital space very broad; eye very 
small; mouth large, oblique; preoper- 
cle witha broad membranaceus border, 
which is striated and fringed; scales 
small; spinous dorsal short and weak; 
anal spines weak ; caudal fin pointed. 
os ‘Pseudobranchix wanting; air-bladder with a lateral horn-like 
process on each side, this dividing into 
many branches in the skin of a perito- 
neal membrane; both jaws with small, | 
unequal, canine-like teeth, those of the 
upper jaw in the outer, of the lower jaw 
in the inner series ; forehead very con- 
vex; soft dorsal aS long, of 27 to 50 
rays; anal fin small; 3; pectoral fin long ; 
gill-rakers (X + 14) slender but rather 
short; lower jaw included; ‘ verte- 
bree 14 -+ 10” (Bleeker) ; ‘‘ vertebrie 12 
+12” (Cur. § Val.). (Asiatic species. ) 
SCLENOIDES. + 
* These characters (which separate the rest of the Scienine from Lques) have been 
verified in part of the genera only, and the statement of them may need some modi- 
fication when the entire group is considered. The genus Lonchurus especially should 
be examined in this regard. 
t Scienoides Blyth, Journ. Asiat. Sci. Beng., 29, 1861; type Otolithus biauritus Can- 
tor. The characters here given are drawn from Sciwnoides pama. This genus seems 
nearest to Nebris, but it shows several resemblances to Lonchurus, «If it really has 
vertebre 14+ 10, as stated by Bleeker, it should be placed among the Ofolithine. 
