Leptocephalus acuticeps— O RTON 
189 
tive tissue between the right and left muscle 
layers, and each patch or spot consists of an 
oblique or more or less vertical cluster of small, 
contracted, and rather widely spaced melano- 
phores. On each specimen the first spot is at 
somites 25-26, the second at somites 50-52, and 
the third at somites 85-86. Each of the first two 
spots is close to a vertical blood vessel ( Fig. 4 ) . 
In addition to these three aggregations of mela- 
nophores, the median connective-tissue zone 
bears occasional very small, inconspicuous, sepa- 
rate melanophores scattered irregularly at wide 
intervals. Most of these cells are below the level 
of the notochord, but a very few are above it. 
The predominant pattern on the short tail ( Fig. 
3) consists of continuations of the supraspinal 
and middorsal rows. A midventral row, sym- 
metrical with the latter, is developing along the 
outer edges of the posterior anal-fin basals. There 
are occasional melanophores along the inner 
edges of the anal basals, and scattered melano- 
phores in the median connective-tissue zone 
(both above and below the midlateral axis), 
especially in the posterior half of the tail. The 
caudal fin is liberally dotted with very minute 
melanophores. The head is sparingly pigmented, 
except for the solidly dark eyes. There are a few 
tiny melanophores in the tip of each jaw, and 
a few along the lower-inner edge of the lower 
jaw. The heavy midventral band of pigment 
below the pericardium begins sufficiently far 
forward ( Fig. 2 ) that it might also be regarded 
as a part of the head pattern. 
COMPARISONS 
A. Leptocephalus acuticeps 
The specimen on which Regan based the 
larval name Leptocephalus acuticeps was col- 
lected in the South Atlantic (21° S, 37° 50' W) 
by the "Terra Nova” expedition. The data 
and illustrations in Regan (1916) and Bertin 
(1936), when combined, characterize L. acuti- 
ceps unusually completely, and show it to be 
very distinctive and more confidently identifiable 
than are most other described leptocephali. The 
given data on somite counts are even sufficiently 
complete to reveal several useful proportional 
characters that both authors had overlooked. In 
the following discussion of this specimen, the 
characters that pertain to somite numbers and 
positions are calculated from Regans original 
counts. Bertin omitted the first two somites, 
which appeared to be incomplete ventrally. My 
own counts on specimens examined include 
all discernible anterior somites regardless of 
