266 Mr. G. A. Boulenger on the 
Teeth on vomer and palatines .............. Parapercis, Blky.* 
ANIAIN ONCE OMW S26 600000000000 5000006 Neopercis, Stdr. 
B. Dorsal fins perfectly distinct. 
Teeth on vomer and palatines; head armed.... Centropereis, Ogilby.* 
Teeth on yomer and palatines; opercle with 
a dat ispine po. shinee hee oe cer enioneonts Pseuduphritis, Casteln.* 
Teeth on vomer only; a preeorbital spine...... Acanthaphritis, Gthr. 
Palate toothiless™ .f5sccricmikas owl eccyee baelnee eaenee Eleginops, Gill.’ 
6. Body naked; habit eottoid ...... Bovichthys, C. & V. 
B. A single dorsal ; scales extremely small ; 
snout spatulate ; palate toothless. 
Lateral lines two; opercular spines.......... Gerlachia, Dollo.* 
Lateral line single ; opercular spines ........ Bathydraco, Gthr. 
Lateral line single; no opercular spines ...... Racovitzia, Dollo.* 
II. Gillmembranes broadly united to the 
isthmus; habit cottoid; body naked ; 
headeATMEM sryek Lives couyeveiactecnss iste Harpagifer, Rich. 
The genera marked with an asterisk are only known from 
the external characters, and so long as the bones at the base 
of the pectoral fin have not been examined their correct 
systematic position remains uncertain, as some may belong to 
the Leptoseopidee. Pagetodus, Rich., rests on an insufficient 
description and figure. 
In Percophis, which has always been regarded as allied to 
Parapercis, the pectoral rays and the pterygials are in the 
same condition as in 7’rachinus, and a subocular lamina is 
likewise present; but the scapular fenestra is situated en- 
tirely in the scapula (fig. B, p. 264). This genus constitutes 
the family Percophiide. 
Bembrops, Stdr. (Hypsicometes, Goode), which resembles 
some of the Nototheniide, and the freshwater genus Chimar- 
rhichthys, Haast, differ from the preceding only in the absence 
of the subocular lamina. With them I would also associate 
Leptoscopus, which, in spite of the great external resemblance 
it bears to Uranoscopus, differs from it not only in the much 
more elongate caudal region, but in the distinet pterygials, 
the sessile ribs, and the feebly developed parapophyses. 
The name of this family, which leads to the Batrachide, 
should be Leptoscopide. 
The Uranoscopide, thus reduced to Uranoscopus, Anema, 
aud Cathetostoma*, are characterized by the much reduced 
pterygials, fused with the scapula and the coracoid, the 
scapular fenestra in the scapula, the parapophyses strongly 
1 Percis, Bl. Schn., nec Scopoli. 
2 Aphritis, Cuv., nec Latr,; Cottoperca, Stdr. 
3 Eleginus, Ouv., nec Fischer. 
4 Dactyloscopus does not differ from Clinus in its pectoral arch, I 
therefore follow Dr. Gunther in placing it with the Blenniidee. 
