KISHKr^ OK HAWAIIAN ISLANDS. ^)'i)'^ 



Toetli oiv jaws and palatiiu's hi very narrow bands; Ijroad bands investing tho mesoptervjioids; 

 head of vomer hearing on <'ach side a strongly protrnding lolie, devoid of teetli, tlieconeave interspace 

 hearing niinnte asperities; gill-rakers long anil slender, 4 t H in niindieron onti'rareh; preo|iercnlar 

 margin vertically placeil and evenly ronnded. 



Inseilion of lirst dorsal ray slightly behind the vertical from base of ventrals, the last ilorsal ray 

 vertically over the first anal ray; adipos(! dorsal oversecond postero-anal jjliotophorc, well in advance 

 iif last anal ray; pectorals very long and jioiiited, their tips reaching vent and extending beyond tips 

 of ventrals. 



Scales thin, with entire edges, and cadiicons, (jccasional scales oidy lieing preserved in oiu' material. 



Arrangement of photophores: Agreeing with M. /■<'/»/i((/v//rt in the ol.>tnse-angled triangle formed 

 by the .S snpra-aiials, in the presence of 2 posterolaterals, in the wide separation of the '2 candals, 

 and in the n umal [visition of the snpra-ventral. A niinnte preocular present on lower anterior 

 orbital rim, similar to other photophores, but ninch smaller; first and second, and second and third 

 pairs of thoracic photophores more widely se|iarated than the others, the third and fourth jiairs nearer 

 together than the others; ujiper jiectoral s|iot immediately below third scale of lateral line; snpra- 

 ventral on a level with midille pectoral and middle snpra-anal .spots, ami constantly a little above ante- 

 rior snpra-anal; upper supra-anal, upper posterolateral ami upper caudal spots in contact with lateral 

 line at ISth, 26th, and 40th .scales, respectively; lower ]iosteroIateral in advance of njiper, and very 

 sliglilly in advance of last antero-anal. 



The number of photophores is as follows: I'reoculars 1; niandilinlars .'!; o]ierculars 2, the lower 

 not concealed; pectorals M, the niiildle one innnediately below the base of lowest pectoral ray, the 

 lower one on a line joining the middle ]iectoral and the lower opercular spots; thoracics o, those of the 

 last jiair more widely separated; ventrals 4, the third and fourtli pairs le.ss widely spaced than the 

 others; anals 6 to 8+6 to S (IS to 1.5); caudals 2; supraventrals 1; supra-anals 8; posterolaterals 2. 



The anals form the only variable groups. In the 17 specimens (giving 34 cases) of the collection, 

 we find them distriliuted as follows: Anlero-anals 6 in 14 cases, 7 in 18 cases, 8 in 2 ca.ses; postero- 

 anals 6 in 2 cases, 7 in 23 ca.ses, 8 in 9 cases. In both groups then, the mode is 7, the anterior group 

 varying princijially toward 6, the posterior group toward 8. The totals of the 2 groups, in the 34 cases, 

 are as follows: 13 in 7 cases, 14 in 25 cases, 15 in 2 cases. In the type there is a w-ell-develo|ied luminous 

 luitch on the lower side of the caudal iieduncle, and an incipient patch on the dorsal siirt'ace. 



< ieneral color dark, with bluish rettections from oiiercles and .scales on sides; caudal with a ilusky 

 patch at base of each lobi>; other tins translucent. 



Spec'imens were taken at the following stations: Nos. 3878, south of l.anai; .3ii27. west of Niiliau; 

 3I)S0, south of Oahii; and 4145, west of N'iihau; all at the surface. 



Named for Dr. .\ugust Ilraner, the authoi- of a most useful review of the genus Miii-liiphum. 



Dasyscopelus spinosus (Steindachner). 



Forty-nine specimens, the longest 6!t mm., were taken at the surface about the Hawaiian Islands, 

 and in the vicinity of Laysan Island. They agree in all essential features wdth the excellent dcscrip- 

 ti(jns and figures cited below. The jiectoral fins are a little longer and more jiointed than previously 

 represented, reaching with their tips a little beyond the ventrals, but not quite to vent. The scales of 

 the lateraKline are little longer than the neighboring scales. They are almost wholly concealed by 

 the overlapping series above and below, the .scales of which are arranged vertically with reference 

 to them. 



Dorsal with 13 or 14 rays, anal 111; 40 scales in lateral line, in a cross-series, excluding the mid- 

 dorsal and the mid-ventral rows. 



Variation in number and disiiosilion of pliolophiiies is conlined to the anal series. The total num- 

 ber of ihe.se varies from 13 to 15 (abnormally 12 in the anterior group on one side of one siiecimen), 

 the prevailing number being 14 (7 I 7). The follow ing coniliinations occur; 7 | 7 and 7 -f 7 in 26 speci- 

 mens; 7-7 and 7 + 6 in 3; 7 ; 7 and 6 ; 7 in 1 ; 7 j- 7 and 7 - 5 in 1; 7 — 6 and 7 }- 6 in 6; 7 ; 8 

 and 7 + 8 in 1 ; 7 + 8 and 7 4 7 in 1 ; 7 : 8 and 8 -' 7 in 1 ; 7 • 8 and 6 — 7 in 1 ; 8 - 6 and 8 • (> in 5; 

 8^6 and 8 + 7 in 2; 8 -p 7 and 8 : 7 in 1. From this it ajipears that the predominating combinations 

 are 7 + 7, 7 + 6, and 8^-6. In the anterior group, 8 occurs in 17 cases, 7 in 79, and 6 in 2; in the 

 posterior group, 8 occui-s in 5 cases, 7 in 65, 6 in 27, and 5 in 1. Thus, while 7 is the jirevailing 

 number in eacn group, the anterior group varies most frequently toward 8, the posterior group in tlie 



