325 



the hindcrmost row. Few caninoid teeth in the frontpart of 

 the vomer, in one row, forming a A. A row of small teeth 

 on the palatines, with a series of distant canines, about 4 or 

 5 on each side. In the mandibles a band of about 4 rows of 

 small teeth in front, tapering laterally and with a series of 

 distant canines. About 10 — 12 rows of scales between eye and 

 hindborder of praeoperculum, 5 or ,6 rows of scales on oper- 

 culum. Linea lateralis running in a straight line, without any 

 abrupt curve or interruption. Dorsal beginning above origin 

 of pectorals and ending scarcely behind anal, its origin sepa- 

 rated by 17 — 20 scales from tip of snout. Origin of anal below 

 131'! dorsal ray. Pectorals about one eye-diameter shorter than 

 postorbital part of head. Ventrals originating behind origin of 

 dorsal, somewhat shorter than pectorals. Brownish above, 

 yellowish white below. Two dark parallel bands running from 

 the hindborder of the eye somewhat obliquely downwards to the 

 interoperculum. Four or 5 black ocelli, surrounded by a yellowish 

 ring and larger than the eye : the first partly on praeoper- 

 culum, partly on operculum, the following three at some dis- 

 tance of each other on the lateral line, the second, which may 

 be absent, above the anus, the third about below the middle 

 of the dorsal and the fourth below the last third of dorsal ; 

 the fifth in the middle of the base of the caudal fin. Dorsal, 

 anal and caudal with indistinct light and dark crossbars. Pec- 

 torals and ventrals more or less dusky, base of pectorals brown. 

 Length almost 400 mm. '). 



Habitat: Sumatra (Palembang !, Djambi !, Padang ?) ; Borneo 

 (Banjermassin, river Kapuas, Sebruang, Knapei, Matang, Danau 

 Sriang, Pulu Matjan). 



In rivers. 



l) O. spiritalis Fowler, from Padang, is intermediate iu many respects l>et\veen 

 O. pleurophthalnius and O. lucius. The number of scales in the L.l. (55), the 

 fm formula (D. 29, A. 28) do not give any definite clue to which of the two 

 species O. spiritalis belongs, although the low nunil)er of scales points to O. pleu- 

 rophthalnius. According to the figure the profile of the head is somewhat concave, 

 but not so much as in O. lucius. Tlie blotches on opercle and body correspond 

 fairly well with the ocelli of 0. plcurophthalmtts.^ but they miss the light rings 

 of the latter. The bars on the pectorals are again exactly what we see in speci- 

 mens of 0. lucius. 0. spiritalis may be a hybrid of the two species, although we 

 arc not aware that hybridisation is known in Ophiocephalus. 



