Leptocephalus ■ acuticeps— Orton 
195 
surface pigment. Few congrid larvae are known 
to have any trace of internal supraspinal pig- 
ment. The internal three-spot pattern of acuti- 
ceps has no known counterpart in congrids. 
Numerous differences in morphological charac- 
ters indicate that acuticeps is not related to 
Afiosoma baleatica . 
At present, the best evidence that supports the 
exclusion of L. acuticeps from the Congridae is 
its close resemblance to the definitely identified 
larvae of a different family, the Nemichthyidae. 
E. Nemichthys Larvae 
Characters of Nemichthys larvae. The dis- 
tinctive larvae of Nemichthys (family Nemich- 
thyidae) have been described and illustrated 
under several leptocephalus names, for their 
various size-groups and stages in metamorphosis 
have repeatedly been considered new kinds of 
larvae. Data on metamorphosis enabled Roule 
and Bertin (1929:61) and Beebe and Crane 
(1937:357) to assign all of these varying larvae 
to Nemichthys It is possible that the extensive 
described material may also include larvae of 
Nematoprora or Cercomitus, for these genera 
are closely related to Nemichthys and probably 
closely accord with it in larval characters. A 
complete review of the literature on the Ne- 
michthys group is not essential to the present 
paper, however, for the two references cited 
above give adequate surveys of the literature 
on the larvae up to 1937, and very little that is 
pertinent to this paper has been published since 
then. The discussion that follows is based both 
on the literature and on Nemichthys larvae 
from the eastern Pacific in the Scripps collection. 
Larvae of the Nemichthys group may attain 
a total length of at least 359 mm before meta- 
morphosis (Roule and Bertin, op. cit. ) , and 
they are therefore among the largest known 
leptocephali. Total length is a rather deceptive 
measure of their size, however, for they are also 
among the slenderest of the known leptocephali 
(except during their more conventionally pro- 
portioned earliest stages), and hence they look 
smaller than the length indicates. The general 
form of the well-grown Nemichthys larva is a 
long narrow ribbon that has a nearly uniform 
width along much of its length and ends in a 
very thin, pointed tip without a well-defined 
caudal fin. The slenderness of the larva becomes 
more exaggerated as the total length increases. 
The gut is a straight, simple tube that extends 
exceptionally far back, and the preanal somite 
count is therefore unusually high. The bulging 
cranium, strongly concave snout profile, and 
slender jaws give the head distinctive and some- 
what birdlike contours. The forms of Nemich- 
thys have the highest vertebral counts known 
in the eels, and both of the references cited 
above suggested that additional somites and 
vertebrae probably continue to form at the tail 
tip throughout life, in contrast to the definitive 
growth pattern known in other eels. The nar- 
rowness and dense spacing of the terminal seg- 
ments make a precise count difficult, especially 
on the smaller larvae. A 147-mm larva in the 
Scripps collection (SI062— 640— 26A) has 225 
preanal somites and about 50 postanal somites, 
making an approximate total of 275. Beebe and 
Crane (1937:363) reported that the total num- 
ber may reach 450 before metamorphosis, and 
the same authors (op. cit., p. 351) recorded a 
vertebral count of 660 in an adult Nemichthys . 
Roule and Bertin (1929:61) and Beebe and 
Crane (1937:357) had two kinds of nemich- 
thyid larvae, which they termed "A” and "B.” 
These probably represent different species, and 
possibly different genera, but the nomenclatural 
details need not be explored in this paper. 
Roule and Bertin found that the "Dana” col- 
lections contained 664 larvae of type A and 
only 26 of type B, and included a sufficient 
range of growth stages of each type to demon- 
strate that the two kinds do differ and are not 
themselves stages in a continuous series. The 
type B larva reaches a greater total length (to 
359 mm) than type A, and has a higher number 
of preanal somites (maximum known, 320). 
An internal row of very small melanophores 
along the top of the spinal cord and a similar 
row along the top of the gut both begin at 
about the 10th somite. The smaller type A larva 
( maximum length before metamorphosis, about 
253 mm) has fewer preanal somites (maximum 
known, 248), and its internal rows of minute 
supraspinal and supraintestinal melanophores 
begin farther back, at about the 25th somite. 
Type A has an important color-pattern charac- 
ter that the authors did not find in type B. This 
