76 
PACIFIC SCIENCE, Vol. VIII, January, 1954 
isthmus pushed up against the roof of the 
mouth, giving the head a flattened appear- 
ance as compared with D. (rwers-andersomP) . 
I believe this apparent difference is due to the 
manner of preservation and that the head 
shapes of the two species are actually very 
similar. 
The differences in counts and measure- 
ments of the two Hawaiian species of Dipla- 
canthopoma are given in Table 1. Among these 
the most marked discrepancies are the some- 
what larger head and eye, the longer ventral 
filaments, and the more numerous predorsal 
scales in D. (sp.?). However, more significant 
structural differences are also present. In D. 
(sp.?) the cleithrum just above the pectoral 
base projects backward as a sharp point which 
can easily be felt through the skin; in D. 
{rivers- under soni?) there is no such point. In 
D. (sp.?) the basal part of the pectoral fin 
rays are scaled; in D. {rivers -andersoni?) there 
are no scales on the base of the pectoral rays, 
though a fleshy sheath extends on to the fin. 
Again, in D. (sp.?) there is a patch of teeth 
at the base of the third gill arch on each side; 
in D. {rivers-andersoni?) there are no such 
patches. Finally, in the latter species the 
lower pharyngeals appear to be much more 
widely separated than in D. (sp.?). Aside from 
these differences, the description of D. {rivers- 
andersoni?), given above, will apply equally 
well to D. (sp.?). 
D. (sp.?) appears to differ from D. hrachy- 
soma in about the same way that D. {rivers- 
andersoni?) does, but I cannot with any 
certainty distinguish it from any of the other 
six described species in the genus. Indeed, 
it is quite possible that this specimen, and 
not the one I have provisionally called D. 
rivers-andersoni, represents that species. 
Genus Cataetyx Gunther 
Type species: Cataetyx messieri Gunther 
(1887: 104), by monotypy. 
The genus Cataetyx has been interpreted 
very broadly by Norman (1939: 83) as in- 
cluding those brotulids with the following 
characters : 
Head low, more or less depressed. Opercles 
and cheeks scaled. Interorbital region and 
preopercles without spines. Snout and lower 
jaw without barbels. Vomer and palatines 
toothed. No canine teeth. Two to five devel- 
oped gill rakers on the first arch. Body scaled, 
not very tapering posteriorly. Lateral line pres- 
ent, double, inconspicuous. Caudal united 
with dorsal and anal, dorsal with more than 
90 rays, anal with more than 70. Tips of 
cleithra firmly united. Pelvics each of a simple ' 
filament originating behind eye. Flap-like ap- | 
pendages near pylorus. i 
Within the genus Cataetyx as thus defined, i 
the species C. messieri, C. rubrirostris, the spe- | 
cies described below, and perhaps C. laticeps ' 
(of which the teeth are said only to be "'in j 
viliform bands,” Koefoed, 1927: 137) would ; 
seem to form a closely related group. ^ These | 
may be differentiated from other species in- | 
eluded in the genus by Norman (1939: 90) j 
by having the head scaled to forward of the ! 
eyes and some enlarged teeth on the sides of | 
the mandibles and on the vomer and pala- I 
tines. Whether the genus should be restricted | 
to such species, I do not have the material to | 
determine. Under any circumstances, how- | 
ever, Pteridium alleni Byrne does not appear j 
to belong in Cataetyx, where it was placed by i 
Norman {loc. citi). j 
Cataetyx hawatiensis n. sp. | 
Tables 1, 2; Figs. Ig, 2c, 3c j| 
Holotype: U. S. N. M. 162715, a male 222 
mm. in standard length, collected off the 
Mauna Loa lava flow, Hawaii, by Gosline, 
Hayes, Keen, and Ellis, June 6, 1950. 
Body tapering to a rounded point poste- ;| 
riorly, covered with nonembedded, overlap- |i 
. ! 
Another species which may belong in this group jl, 
but which is impossible to place from the description ! 
is Oculospinus brevis Koefoed (1927: 138). Koefoed’s I 
new genus {Oculospinus) and both of his new species i 
have been omitted from Norman’s "Synopsis” (1939: 
79-92). ! 
