72 Sigurd Johnsen. 
p. 82, fig. 27) on some specimens taken at an intermediate depth 
of 300 fths. in the Bering Sea. The specimens were all of small 
size, total length of the type 34mm. (Copied here as text-fig. 1 
on p. 21). The coloration, which consists of distinct chromatophors, 
and the appearance of the ventral part of the pectoral fin, indicate 
the juvenile state of the specimen. GILBERT & BURKE summarize 
the specific characters of the genus in the following way: “A 
pelagic genus related to Paraliparis but differing from it and 
the other genera of the family in having 5 branchiostegals and 
the gill slit restricted to the front of the pectoral.” Now, one 
of the Paraliparis-species from this region, P. holomelas GILBERT 
(1895 p. 441) seems to have a gill-slit of the same size, being 
described as “extending from above opercular flap nearly to 
middle of base of pectorals”, the other species having the gill- 
slit restricted to the area above the pectoral’). P. holomelas 
is known from two specimens taken in the Bering Sea at 406 
and 1625 fathoms depth; they measured abt. 100 mm.; their 
coloration was black. The question then remains how much 
weight as a generic distinguishing mark should be laid upon 
Nectoliparis having only 5 branchiostegals. GARMAN (1892) in 
his monograph on the Discoboli gives 6 r. br. as typical for all 
three families of this order and regards other statements as very 
doubtful viz. 5 r. br. by Liparis Steineni, a little known form 
from South-Georgia, and 7 r. br. by Paraliparis bathybii, the 
only species of the genus where this character is observed (by 
COLLETT 1880 & 1905). Several of COLLETT’s specimens (from 
1905) are now in the Bg. M. I have in them only found 6 r. 
br. (V. text-fig. 3 on p. 23). As will be seen from the text-figs. 
2 & 3 interoperculum is stylifform and when the thick skin is 
not removed it may easily be counted as a seventh branchiostegal. 
I think this explains how COLLETT has arrived at the number seven 
1), JORDAN & EVERMANN (1898, II p. 2129) refer the American species 
of Paraliparis to three subgenera. P. holomelas is the only (American) 
representative of one subgenus, just because of the wide gill-slit. When, 
however, the authors retain for this subgenus the subgeneric name of 
Paraliparis proposed by COLLETT (1880), it should be remarked that the type 
of this genus and subgenus, Paraliparis (Paraliparis) bathybii, does not have 
this character. If therefore subgenera are to be used a new subgenus must 
be established for P. holomelas. 
R 
4 
