116 Mr. C. T. Regan on the 



Subfamily SxuoMATEiNiE, 



■with sacciform processes extending back from the last branchial arch, 

 14-15 abdominal and 17-21 caudal vertebrfp, non-protractile premaxilla- 

 ries, ventral fins rarely persistent. 



Genus 1. Stromatetis ; with ample branchial apertures. 

 Subgenus a. Stromateus, no pehic spine. 



„ b. Peprilus, a trenchant pelvic spine. 



„ c. Poronohis, a trenchant pelvic spine and a row of vertical 



slits above the lateral line. 

 „ d. Apolectus, the lateral line keeled posteriorly. 

 Genus 2. Stromateoichs, with restricted branchial apertures. 

 Subgenus a. Stromateoides, dorsal and anal spines trenchant. 



„ b, Chondroplites, dorsal and anal spines concealed, sub- 

 cartilaginons. 

 Genus 3. Pscnopsis, with persistent perfect ventrals. 



In 1896 Steinclaclmer * proposed to unite tiiis last genus 

 (type Psenes anomalus, Schleg.) with his genus Farapsetius 

 (type Parapsettus panome7)s{s), which he had previously con- 

 sidered as allied to Psettus, but now placed in the Stromateidte. 

 I have examined Parapsettus jxinamensis, and it appears by no 

 means closely allied to Psenopsis, although superficially it 

 bears a great resemblance to Stromateoides. Dr. Steindachner 

 has been misled by the presence in the oesophagus of numerous 

 conical papillae, which have a similar appearance to the 

 toothed processes o£ Stromateus, but are soft and fleshy, 

 and not confined to the sides of the oesophagus. Similar 

 papilla3 are present in Platax, EjJuppus, and Chcetodipterus, 

 which, together with Parapsettus, constitute the family 

 Ephippidge, as defined by Messrs. Jordan and Evermann. 

 Psettus is not closely related to Parapsettus, and belongs to 

 the Scorpididre. 



Messrs. Goode and Bean (' Oceanic Ichtliyology,' p. 215) 

 unite the genera Icosteus and Icichthys with the Stromateida?. 

 The affinities of these rare fishes are obscure, but Icosteus, 

 which I have examined, certainly does not possess teeth in 

 the oesophagus. 



Gill, in 1893 (Nat. Ac. Sci. vi., Mem. 5, p. 115), made a 

 list of the fishes of New Zealand, in which the Stromateidos 

 are represented by a single si)ecies, Neptotichthys violaceus^ 

 llutton. Eutton t places this fish in the Carangidee., and no 

 reason is given by Gill for his alteration. On investigation I 

 find it to be a Scorpis, closely allied to Scorpis c/nletisis, Guich. 

 In the course of attempting to find definite characters by 

 which the so-called Scombriform fishes could be diagnosed, I 

 have discovered that the Stromateidai must be enlarged by 

 the addition of the following genera: — Nomeus, Cuh'ceps, 



* Steindachner, Ann. Hofnnis. Wien, xi. p. 211 (1806) 

 t Trans. ^. Z IlsI. v. p. 2(J1, pis. Mii. c^ xxii. p. 278. 



