new or Utth-hnoivJi Fishes from Brazil. 11 



A single specimen from Jangada. 



This fish is evidently very closely allied to L. nudiventrisj 

 known from a single specimen from the llio San Francisco, 

 described by Cuvier and Valenciennes. It differs, however, 

 in having seven or eight shields on each side of the lower 

 surface, between the pectoral and ventral fins, instead of four. 



2'etragonopterus Moorii, sp. n. (PI. II. fig. 1.) 



Length of head 3| times in total length (without caudal), 

 depth of body 2|. IMaxillary toothless, extending to below 

 the centre of the eye ; diameter of eye ^ length of head, Ig- 

 length of snout, equal to interorbital width ; adipose eyelid 

 short. Dorsal I 10, originating above base of ventrals. 

 Anal II 28, originating a little behind the vertical of the 

 base of the dorsal. Pectorals reaching base of ventrals, 

 ventrals reaching origin of anal. Scales 37-38 ^ ; lateral line 

 complete. A black spot behind the shoulder ; a silvery 

 lateral stripe, turning to black on the tail and extending on 

 the caudal. 



Total length 75 milliui. 



Two specimens were collected by Mr. Moore on the 

 Chapala plateau. 



The nearest ally of this new species appears to be T. maxi- 

 mus, Stdr. {alosa, Gthr.), from the Peruvian Andes, which 

 differs in having the interorbital region wider. 



Beach YCHALCiNus, gen. no v. 



Intermediate between Tetraffonopterus, Cuv., and Luetlcem'a, 

 Stdr. Dentition as in tlie former, viz. two pramaxillary and 

 one mandibular row of tri- or quinquecuspid teeth ; body 

 elevated, with sharp ventral edge. Differing from both 

 in having a movable spine, directed forwards, in front of 

 the dorsal fin. 



In one of the three specimens (probably a male) this spine 

 is hammer-shaped, its free portion forming a longer anterior 

 and a shorter posterior branch, both of which are sharply 

 pointed ; in the two others (one of which I have ascertained 

 to be a female) the posterior process is wanting and the ante- 

 rior is more developed but not spinose, spoon- or saddle- 

 shaped, rounded at the end, concave below, and fitting into a 

 notcii in the back in front of the dorsal fin. The differences 

 in this curious arrangement will probably prove to be corre- 

 lative of the sexes. In Serrasalmo the first interneural bears 

 likewise a spine directed forwards, which is bicuspid behind 

 and scarcely movable. 



