194 
PACIFIC SCIENCE, Vol. XVIII, April 1964 
dorsal fin), yet they retain an unmistakable 
structural similarity to these basic congrid larvae 
and their pigmentation is a simple variant of 
the Conger conger type of larval pattern. The 
heterocongrids probably should be considered 
genetically close to the typical congrids, no mat- 
ter how one may choose to juggle their nomen- 
clature. The nettastomid eels, which are still 
sometimes included in the Congridae (e.g., Gins- 
burg, 1951), exemplify the opposite extreme. 
Their known larvae differ strikingly from the 
typical congrid larvae in both morphology and 
pigmentation. The head and jaws are usually 
very elongated, the viscera are exceptionally 
short and complexly specialized, and the color 
patterns are unusual. These and other specializa- 
tions indicate that these larvae have evolved 
along distinctive lines and suggest that the 
nearest relationships of the nettastomids are to 
stocks that are remote from the congrids. Sev- 
eral groups of genera in addition to the net- 
tastomids seem far too discordant with the type 
genus, Conger, in both adult and larval charac- 
ters, to be retained within the same family. 
These include the dysommids, especially if the 
larva that Grassi ( 1913: 170, pi. 10, figs. 1, 5) 
assigned to Todarus brevirostre was correctly 
identified, and the muraenesocids. I agree with 
authors who have elevated each of these groups 
to family rank. It seems to me that certain other 
genera (e.g., Hoplunnis, Oxy conger, and Gaviali- 
ceps) that are sometimes placed in the Congri- 
dae should also be excluded, but their larvae are 
still unknown and their adult stages are too in- 
completely described to support effective discus- 
sion of affinities. Some of these forms may prove 
to be muraenesocids, when the limits of that 
family are better understood, but others (nota- 
bly, Gavialiceps toeniola Alcock) perhaps repre- 
sent phyletic lines that are distinctive enough 
to justify family rank. 
These examples help to establish criteria for 
the probable limits within which the larvae of 
true congrids have evolved, but a more precise 
understanding must await the specific identifica- 
tion of many more larvae, particularly in the less 
well-known genera that are of questionable 
status. Although present knowledge precludes a 
more authoritative discussion, this summary at 
least provides some basis for evaluating the pos- 
sible affinity of Leptocephalus acuticeps to the 
typical Congridae. 
Comparison of Leptocephalus acuticeps with 
congrid larvae. Neither D’ Ancona nor Bertin 
gave his reasons for considering Leptocephalus 
acuticeps to be a congrid, but they probably no- 
ticed the characters of long, unspecialized gut 
and short tail that it shares with the congrids 
and with certain other leptocephali. It is also 
likely that they noticed the partial similarity 
of its pigmentation to that of the Ariosoma 
balearica larva, for both of them had reported 
on larvae of the balearica group. However, com- 
parative study indicates that acuticeps does not 
belong here, even though the limited present 
knowledge reveals few absolute distinctions that 
firmly exclude it from the Congridae. There are 
some rather subtle morphological differences be- 
tween L. acuticeps and typical congrid larvae. 
In L. acuticeps the nasal capsule is conspicuously 
smaller, the eyeball lacks the white (or partly 
pigmented) supporting sheath that most but 
not all known congrid larvae have, and the 
tongue is fully adherent, in contrast to its usually 
free tip and edges in congrid larvae. The hy- 
purals are narrow and rather weakly chondrified, 
and there are only 3 or 4 caudal rays compared 
with the 6-10 generally reported for congrid 
eels. The somite counts of L. acuticeps exceed 
the vertebral counts of the better known con- 
grids, most of which fall between 130-160, but 
this is not an excluding character, for a few 
congrids are known to have counts that overlap 
or even exceed the known somite counts of L. 
acuticeps. For example, Asano (1962) listed 
vertebral counts of 173-181 for Congrina retro- 
tincta (Jordan and Snyder), and 203-206 for 
Uroconger lepturus (Richardson). The Scripps 
collection contains unidentified Indo-Pacific con- 
grid leptocephali with somite counts as high as 
230. The similarities in pigmentation between 
L. acuticeps and the larva of Ariosoma balearica 
include the very small size and dense spacing of 
the melanophores, and the presence of middorsal 
and anterior midventral surface rows and an 
internal row along the top of the gut. There are 
important differences in the rest of the pattern. 
L. acuticeps has a complete row of internal 
supraspinal melanophores and the distinctive 
three oblique internal spots, and lacks lateral 
