Jordan and Starks: Fishes from Ceylon. 453 



to the straight portion, but run obliquely upward above the curved 

 portion. No median pore at symphysis, but two on each side, as in 

 Bola OS sea. 



loi. Bola ossea (Day). 



No. 8107a, C. M. Two small specimens seem to be referable to 

 this species. No median pore at symphysis of mandible; two pores 

 on each side. The inner row of mandibular teeth are enlarged (Day 

 places the species in "Johnius" a group without enlarged teeth; his 

 specimen was, however, very much larger than ours). Edge of pre- 

 opercle crenulated, somewhat spiny at the angle. The scales run 

 slightly obliquely both above and below the lateral line. Apparently 

 scales covered at least the basal half of the soft dorsal membrane, 

 though in the specimens at hand they only remain on a few places. 



Genus Johnius Bloch. 



We accept the generic name Johnius for those species of this com- 

 plicated group, in which the teeth of the mandible are even and the 

 second anal spine moderate. Scicena {Corvina) has the lower teeth 

 uniform and the second anal spine very large. 



104. Johnius carutta Bloch. 



No. 8io8a-&, C. M. Three specimens with the lateral line fol- 

 lowing a light streak, and otherwise agreeing well with Day's descrip- 

 tion. The inner mandibular teeth are not enlarged. In specimens 

 six inches long the edge of the pre-opercle is strongly crenulated. In 

 a specimen nine inches long it is scarcely crenulated, but without long 

 teeth. The membrane of the soft dorsal is inconspicuously scaled 

 nearly or quite to its outer edge. The series of scales run rather 

 obliquely upward both above and below the lateral line. 



103. Johnius maculatus (Bloch & Schneider). 



No. 8109a, C. M. One specimen of this well-marked species. 

 Pre-opercle rather strongly denticulated at its angle. Mandible with 

 a band of villiform teeth, uniform in size. Soft dorsal entirely naked. 

 Series of scales running obliquely upward both above and below the 

 lateral line, though at the lateral line where they cross it their con- 

 tinuity is somewhat broken. Posteriorly, where the lateral line be- 

 comes straight, the lines of scales also become straight and parallel to 

 it, rather than crossing it. 



ANN. CARN. MUS., XI, 30, OCT. 31, I917. 



