162 PROCEEDINGS OF THE NATIONAL MUSEUM. 



3. Left side, the scales in an area between pectoral, ventral, and 

 gill-opening; eight rows of scales extendmg obliquely forward at 

 a distance behind the pectoral equal to the length of the head ; this 

 group followed, after an interspace two-thirds length of head, by 

 an area about eight scale rows wide, and parallel to the first; a few 

 scales irregularly placed near these areas. Right side, no scales 

 with more than five ridges. 



4. Scales largely lost; a few with many ridges on the right side 

 below origin of the second dorsal. 



5. Left side, none. Right side, a few scales in oblique series below 

 anterior end of second dorsal ; four horizontal scale rows below mid- 

 dle of dorsal base. 



6. Left side, the scales in an area above the lateral line, below the 

 first dorsal base and the interdorsal space; about 30 continuous 

 oblique scale rows on middle of sides below lateral line. Right side, 

 scales largely lost, but a large area indicated behind pectoral. 



From these descriptions it is apparent that the many-ridged scales 

 occur in definite areas, usually on continuous scale rows, which may 

 extend obliquely forward, obliquely backward, or horizontally; and 

 it is further apparent that these areas are not similar on the two 

 sides of the fish. 



Ventral rays in Alaskan specimens: 



Rays 6 7 7-8 8 



Specimens 16 11 



8. NEMATONURUS BONA-NOX (Jordan and Thompson). i 



Con/ijhaenoides hona-nox Jordan and Thompson, Mem. Carnegie Mus., vol. 

 6. pt. 4, 1914, p. 305, pi. 38, fig. 1 and Icr. 



Two large specimens, overlooked by Jordan and Thompson, are 

 here reported on. One of them bears the label: " Enoshima- 22 

 II 05." 



This species differs from the description of Coryphaenoides alti- 

 plnnis,^ a Japanese species not obtained by the Alhatross, in the lower 

 dorsal spine, and especially in the more strongly armed scales, which 

 are excessively rough, usually with 7 to 9 strongly spinous ridges; 

 but a few of the scales in certain regions, as in N. pectoralis^ have 

 an increased number of spines (as many as 18). 



Premaxillary teeth biserial, rather blunt, and slightly arrow- 

 shaped, the inner series the smaller, rather irregular, becoming bi- 



1 Subsequent study has shown C. hona-nox to be identical with C. acrolepis (Bean) of 

 California and Bering Sea. The dentition of C. acrolepis is intermediate between that of 

 Coryphaenoides and that of Nematonurus. Nematonurus should be merged with Coryphae- 

 noides, hut may be retained as a subgenus; it should include C. acrolepis (Bean) and C. 

 lepturus (Gill and Townsend). 



2 Coryphaenoides altipinnis Giuither, Ann. Mag. Nat. Hist., ser. 4, vol. 20, 1877, p. 439. 

 Macrurus altipinnis Giinther, ChaUeniier Reports, vol. 22, Deep-Sea Fishes, 1887, p. 138, 



pi. 39, fig. A (in explanation of plate erroneously stated to be from the Japanese Sea). 



