138 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 51. 



and other characters. Further generic division of these two major 

 groups is based largely on dentition, amount of rostral projec- 

 tion, the presence or absence of serrations on the dorsal spine, and 

 other characters, which were formerly used in the primary division 

 of the subfamily. 



Most of the genera as here accepted seem to be well-defined groups 

 of doubtlessly related species, to which genera new species may be 

 referred with convenience and with little or no doubt. The value and 

 correctness of the classification as here adopted has been verified in 

 the study of several collections of Macrouroid fishes, which have not 

 been reported on. Future study may further subdivide some of the 

 larger genera into more compact groups, as the species vary within 

 wide limits. 



Macrourus herglax Lacepede (1800), the type-species of the genus 

 Macrourus^ is found to be congeneric with Coryphaenoldes rwpestris 

 Gunner (1765). Macrourus {=Macrurus, a changed spelling), there- 

 fore becomes a synonym of C oryijhaenokles^ and the family name 

 Macrouridae must also be discarded. The oldest and best known 

 genus is taken as the type of the family, which should apparently 

 stand as Coryphaenoididae. 



The most remarkable of the Coryphaenoididae collected on the ex- 

 pedition of the Albatross to Japan in 1906, is Squalogadus modi-flca- 

 tus, a new genus and species related to Macrouroides inflaticeps,'^ 

 from the Philippine Islands, but differing chiefly in the presence of 

 small ventral fins. On the evidence of this discovery, we do not 

 accept Macrouroides as the type of a distinct family. The single 

 dorsal fin of these two fishes is also shared by Lyconus,- which has 

 the anterior dorsal rays more or less elevated. Lyconus ^ appears to 

 be an ally of Bathygadus. 



The only constant character known, by which the Coryphaenoididae 

 can be separated from the Gadidae, is the absence of a caudal fin, the 

 dorsal and anal fins being continuous around the long, whip-like tail. 

 A pseudocaudal is frequently formed, probably after an injury to 

 the slender tip of the tail, but even in this case the vertical fins are 

 confluent around the tail. The posterior position of the ventrals is 

 not a constant character in the Coryphaenoididae. They are well in 

 advance of the pectorals in Squalogadus^ Cetonm-us^ and in several 

 species of Lionurus (most advanced in position in Lionurus gihher'^'^ 

 from the Hawaiian Islands). Regan ^ has noticed the close resem- 

 blance between Macruronus and the Gadidae. The Muraenolepididae 

 have the vertical fins confluent around the tail, but differ widely from 

 the Coryphaenoididae in the rounded caudal ; the restricted gill open- 



1 Smith and Radcliffe, Proc. U. S. Nat. Mus., vol. 43, 1913, p. 139, pi. 31, fig. 3. 

 - Giinther, Challenger Reports, vol. 23, Deep-Sea Fishes, 1887, p. 158. 



2 Regan, Ann. Mag. Nat. Hist., ser. 7, vol. 11, 1903, pp. 4G4 to 466. 



* Gilbert and Cramer, Proc. U. S. Nat. Mus., vol. 19, 1897, p. 426, pi. 44, fig. 2; Gilbert, 

 Bull. U. S. Fish Com.. 1903 (1905), pt. 2. p. 668, 



