NO. 2UQ. JAPANESE MACROUROID FISHES— GILBERT AND HUBBS. 159 



The characters which have been used to separate the nominal 

 genera included in the synonomy vary to such an extent among the 

 different species that we are forced to regard them as of only specific 

 value. The strength of the serrations of the dorsal spine range, with 

 intermediate forms, from an obsolescent condition on the one hand, 

 as in N. abyssorum,^ to a well developed condition on the other 

 hand, as in N. clarki. 



This genus probably forms a natural group, but the dentition is 

 the only character by which it is constantly separated from Cory- 

 phaenoides. The premaxillary teeth are in a double series, instead of 

 a band, and the mandibular teeth are never in bands, as usual in 

 Coryphaenoides. The mouth is usually more nearly terminal, and 

 the physiognomy frequently approaches that of Bathygadus. 



6. NEMATONURUS LONGIFILIS (Gunther). 



Coryphaenoides longifllis Gunther, Ann. Mag. Nat. Hist., vol. 20, 1877, 



p. 439. 

 Macrurus longifllis GtJNTHEB, Challenger Reports, vol. 22, 1887, p. 151, pi. 35, 

 Moseleya longifllis Goode and Bean. Oceanic Ichthyology, 1895, p. 417. 

 Dolloa longifllis Jordan, American Naturalist, vol. 34, 1900, p. 897. 



This species, hitherto known only from the type, collected by the 

 Challenger south of Hondo, at a depth of 565 fathoms, was redis- 

 covered by the Albatross, which dredged three specimens, one at 

 station 4956 in 720 fathoms, and two at station 4980 in 507 fathoms, 

 both off the southeast coast of Hondo. 



The two specimens from station 4980 (the head 130 mm. long) are 

 in very poor condition. They differ in several respects, chiefly in the 

 weakly armed scales, and in the stronger pectoral, from the smaller 

 specimen, which was dredged at station 4956 (head 96 mm. long). 

 The larger specimens agree more closely than the smaller one with the 

 type description, as regards the character of the scales and the pec- 

 toral. This may be due to the fact that the type was a large specimen. 



Dorsal, II, 12 to 14 ; ventral, 8-9, 9, and 10 (in the three specimens) ; 

 pectoral, 17 or 18. 



The form of the body is well shown in Giinther's figure, except that 

 the artist probably rounded out the snout too much ; depth of body 

 about 1.6 in length of head ; width over pectoral bases less than half 

 depth. Head compressed, with subvertical sides, its width 2.4 to 2.7 

 in its depth; orbit suborbicular, the vertical diameter 0.9 the hori- 

 zontal diameter, which is contained 4.6 to 5.2 times in the head ; inter- 

 orbital width 4 to 4.6 in head ; length of snout about equal to inter- 

 orbital width ; width of suborbital 9 to 11 in head ; the free portion of 

 the barbel shorter than the anterior canines in the jaws; length of 



1 Gilbert, Proc. U. S. Nat. Mus., vol. 48, 1915, p. 374, pi. 21, flg. 23, from off Santa 

 Catallna Island. California. 



