THE SHORE EISIH<:S. ;!()«) 



Teuthis triostegus (Liimne). 



Chaetoilon triosd'gus Linne, Syst. Nat. ed. 10, 1758, p. 274. 



Acanlhurus triontedus GtiNTHER, Cat., 1861, 3, p. 327. D,\Y, FLshcs of Inilia, I.S70, p. 204, pi. 48, fig. 2. 



Ihjiiiliis triodtcyu^ JoHU.K^ & .Se.\le, Bull. U. S, Bui-. Fish.. l<IO(i, 25, \>. 3.54. 



Nos. 05872, Is inches long from Fakarava, and 05924, M. C. Z. 29480, 

 three specimens, li to 2§ inches long from Makemo, each in Paumotu Islands. 

 Nos. 08927, 2^ inches long from Jaluit, and 08910, M. C. Z. 29483, li inches long 

 from Arhno Atoll, each in Marshall Islands. 



Jordan and Scale say: "this seems like sandwichensis, but lacks one cross- 

 band and is very pale; only 4 bands on sides." Regarding sandvicensis they say, 

 "This is the most abundant species of the genus about Hawaii, where it replaces 

 the closely allied H. triostegus. The differences, although slight, are constant." 

 We have compared our specimens with a specimen of sandvicensis from 

 Hawaii, and find the number of cross bands the same, the only difference being 

 the somewhat darker color of the Hawaiian specimen and the continuation of 

 the shoulder band downward and backward below the upper base of the pec- 

 toral, which is well shown in the figure given by .lordan & Evermann (Bull. 

 U. S. Fish. Comm. 23, jit. l.fig. 172). In our specimens the band does not extend 

 below the pectoral, which character is well shown in Day's figure of T. triostegus. 



Another specimen. No. 09072 (larval form) Is inches long from .Vrhno Atoll, 

 Marshall Islands. 



In color this specimen differs but slightly from Teuthia elcguns Gannan, 

 Deep Sea Fishes, p. 70, ])!. 50, fig. 2, the only differences being in the faint V- 

 shaped mark on caudal jieduncle, fainter dots at the base of the anal, none at 

 base of the dorsal and a much darker end to the caudal. It differs in other re- 

 spects as follows: the spines at the ends of alternate ridges at the bases of dorsal 

 and anal are arranged in groujjs of 2-4 instead of 3-(3; the second dorsal spine 

 is equal to the distance from the posterior edge of the pupil to the tip of the snout, 

 instead of equal to the eye and snout; and it has no denticles on the anterior 

 edge, but merely presents a rough and irregular surface, where T. elegans is said 

 to have about nine denticles, it is distinctly grooved on each side, with the 

 appearance of two close-set spines, the i)osterioi' portion of which seems to be 

 finely serrate on the sides; and the margin of the spinous dorsal instead of hav- 



