NOTES ON FISHKS <>F HAWAII. 213 



est above corner of mouth; maxillary extending to a point below anterior edge of pupil; lower jaw 

 projecting slightly beyond upper; jaws each with a row of enlarged canines, behind which is .1 band 

 of villiform teeth, among which are a few slightly enlarged ones; narrow bands of minute teeth on 

 vomer and palatines; tongue smooth; gillrakers 3 10, long, slender, an<l pointed; dorsal inserted 

 directly above tip of opercular spine, the spines slender, stiff and pungent, the lirst spine not quite 

 half the height of second; longest 1 third to fifth) Bpine2.7 in head; no notch between spinous and soft 

 dorsal, rays about equal to spines in length; anal not quite so high as dorsal, thi eavier and 



shorter; lirst spine slightly more than one-third length of the third, the latter .;."i in head; ventral 

 pointed, extending slightly beyond vent; dorsal, anal, and ventral without scales; pectoral falciform, 



extending to a point abo 1 rigin of anal, its base with minute scales; ram la 1 deeply forked, the rays 



with scales (the rays being broken, their length ran nol be determined). Scales finely ctenoid; a row 

 of enlarge^ 1 soft scales extending from upper edge < if opercle to nape; a row of similar scales below eye; 

 cheeks and opercles with scales; maxillary, preorbital, snout, and interorbital naked; preopercle 

 smooth, the edge serrated; opercle with a flat, pointed spine at the upper angle; lateral line curved 

 much like the contour of back. .No distinctive color-markings in an alcoholic specimen, except 

 numerous obscure mottlings and minute spot- of ;i deeper color, which may have been greenish on a 

 red background, and a subdued dark dash on opercle. The appearance of tic specimen, extended 

 stomach, etc.. indicates that it came from deep water. 



Type, no. 577M I'. S. National Museum, about II inches long, collected by Jordan and F.ver- 

 mann at Honolulu, It ml, and by some oversight not distinguished at tin' time from Rooseveltia 1. milium',. 



15. Rooseveltia brighami (Scale). DHkiki; Kalikali. 



Scrranusbrigliami 8q£le, osional Papers Bishop Museum, vol. 1, no. t. 1901, :. Honolulu. 



Apettui brighami, Jenkins, Hull. 1'. s. Fish C6mm., vol. xxii, 1902 [Sept 28, 1903 . 152 Honolulu. Snyder, ibid. (Jan. 19, 



L90i), 527; Honolulu. Jordan & Evermann, Bull. r. s. Pish Comm., vol. win. 190 S3 1-1. xvi: 



Honolulu; Kailun. 



This species, having the top of the head without the elevated crests 1 haracteristic of butianus and 

 Apsilus, should not have been referred to N rranus nor to Apsih 



16. Chaetopterus dubius Gflnther. 



(Aprion microdon Steindachner. 

 (Bowersia ulaula Jordan and Evermann,) 



A specimen l"i inches long, from the market at Honolulu, seems to differ in no way from a Japanese 

 example collected at Fukaura. 



Head 3.4 in length; depth 3.6; snout 3.8; interorbital 3.3; preorbital 10.5; maxillary 2.9; dorsal 

 x, 11; anal tit, 8; scales in lateral scries 7<t; between lateral line and anal I I. Basihyal with tin oval 

 patch of villiform teeth on its posterior half. 



This specimen corresponds evidently to Bowersia ulaula of Jordan and Evermann; and the num- 

 ber of settles in the lateral line in Aprion microdon, as described by Steindachner, shows that his species 

 is ft dubius rather than Bowersia violescens. This species is known as opakapdka in the markets, not 

 being distinguished from the equally common Bowersia violescens by the fishermen. Ulaula, very red, 

 is the name of Etelis eturus and Etelis marshi. 



17. Bowersia violescens Jordan & Evermann. 



(Apsilus microdon .Ionian A: Evermann, young; not Aprion microdon Steindachner.) 



The small specimens described by Jordan and Evermann as Aprion microdon are the young of 

 HiwYrsia n,,lrs,;ns. We find no teeth on the tongue in any of these and none in a cotj pe of Bowersia 

 violescens. The genus Bowersia is therefore to be separated from Clmiopterus by the absence of these 

 teeth, not by their presence, as originally stated. The teeth are present in Chsetopterus dubius. The 

 absence of lingual teeth, therefore, the smaller mouth, and especially the larger scales, separate 

 Bowersia violescens from Chsetopterus dubius. 



is. Etelis evurus Jordan & Kvermann. 



This species, with its congener, Etelis marshi, is common in the winter markets, and both are 

 known as ulaula (very red). The two species differ widely from each other and a study of the skele- 

 tons will probably place them in different genera. 



