Dr. A. Gunther — Tertiary Fishes of Sumatra. 439 



series is 45 millim. long, and consists of twenty-three closely-set 

 conical slender teeth, of which the two foremost and the twelfth and 

 thirteenth are much stronger than the rest, 10 millim. long. The 

 greater portion of the maxillary was armed with a single series of 

 conical teeth of about the same size as the small posterior teeth of 

 the mandible. 



Common as Chirocentrus is on the coasts of the Indian Ocean and 

 Archipelago, I cannot without hesitation refer our fossil to this 

 genus, which, as far as T know, is a marine form. 



Notopterus primcevus. (PI. XIX.) 



Although no doubt whatever can be entertained as regards the 

 generic affinities of this fish, the specific characters unfortunately 

 can be given but very incompletely ; the single specimen before us 

 being mutilated in front and behind. It is evidently most closely 

 allied to the species now living in Sumatra. The lower part of the 

 praeoperculum and some ridges of the mandible are distinctly 

 serrated ; the abdominal portion of the vertebral column is short, 

 much shorter than the head, formed by about 13 short vertebra. 

 Forty-one vertebrae of the caudal portion are preserved. The 

 neural and hsemal spines are of moderate strength and length, 

 directed backwards. The interhsemals are about twice as numerous 

 as, and much longer than, the hsemals, generally two of them being 

 attached to one hasmal. The first hasmal and interhsemal are ex- 

 ceedingly strong, firmly wedged together, and forming the posterior 

 boundary of the abdominal cavity. Eibs well developed, extending 

 nearly to the lower part of the abdominal cavity. From their ex- 

 tremities are suspended the scleral bones, which terminate below in 

 a spine, the spines forming the external double serrated edge of the 

 abdomen in this genus. The anal rays are not directly articulated 

 to the interhsemals, but a series of shorter and thinner additional 

 bones is intercalated between the rays and spines, the number of 

 the additional bones being considerably increased above that of the 

 true interhsemals, and equal to that of the fin-rays. 



The dorsal fin is short and small, its position corresponding to the 

 fourteenth and fifteenth caudal vertebrae. Pectoral fin moderately, 

 ventral but little developed — as usual in this genus. Scales very 

 small. 



The specimen, in its present mutilated state, is 9J inches long, 

 and may have measured 11^ inches when perfect. 



II. — Fossils from Carbonaceous Shales. 



SexapsepTius. (PI. XVI. Figs. 2 & 3a, 6, c.) 



In the Tertiary coal-beds of Padang, conical or mammilliform 

 teeth, up to the size of a large pea, are of not uncommon occur- 

 rence ; their determination would have been a matter of great 

 difficulty ; but, fortunately, one or two slabs have been preserved, 

 showing these teeth in their serial connexion, and attached to a 

 bone, which, without any doubt, is the pharyngeal bone of a Cypri- 

 noid fish. Not knowing of any living Cyprinoid with teeth of similar 



