224 
PACIFIC SCIENCE, VoL VII, April, 1953 
fins, with 18-20 principal rays. Pectoral hns 
moderately low, below midaxis of body, with 
9-17 rays. Pelvic fins small, close together, 
behind middle of standard length, with 8-12 
rays; their position varies from far before dor- 
sal hn to behind dorsal base. Vertebrae 65- 
102, as far as known. No light organs present. 
Secondary sexual dimorphism not known. 
Specimens have been taken up to approxi- 
mately 20 inches in length. 
Sudis Rafinesque (1810) was the first genus 
of the barracudinas to be described, but it 
was indicated as being related to Sphyraena 
and placed in ’’Esocidi.” Cuvier described 
Paralepis in 1817 for two species mentioned 
by Risso (1810) and also related it to Sphy- 
raena. Apparently Bonaparte (1832-46) first 
elevated one of the genera to familial status 
when he placed ”Paralepidini” as a subfamily 
of Sphyraenidae (1832-41) and of Scopelidae 
(1846), using, ipso facto, Paralepis as the family 
type. He recognized both Sudis and Paralepis 
in this group. As far as I can find, all authors 
up to the time of Regan (1911) used Para- 
lepis as the type genus, with Gill (1874) the 
first author to taise the group to the status of 
a full family (Paralepididae). Dollo (1908) 
and many subsequent authors shortened the 
name to Paralepidae, but this is not proper 
form. Regan (1911) based the family name on 
Siidis because it is the oldest generic name. 
According to the Rules of the International 
Commission of Zoological Nomenclature as 
clarified by Opinions 133 and 141, however, 
this is not acceptable, because Paralepididae 
was proposed before Sudidae. Unless I have 
missed a reference earlier than Bonaparte, in 
which a family name was based on Sadis, 
Paralepididae is the proper name. It seems 
unlikely that there is an earlier reference, since 
Bonaparte was apparently the first ichthyolo- 
gist to use family endings. 
Relationships within the Family 
Most authors, except Ege, have examined 
only a few of the genera and usually only very 
immature material. As a result, few workers 
have attempted to analyze the major parale- 
pidid relationships. Most current authors have 
accepted Ege’s classification of the family 
which is based primarily on differences of 
postlarval development and recognizes only 
a few generic units. My investigations have 
primarily stressed the morphological aspects 
of as mature material as possible although I 
have utilized several complete ontogenetic 
series. As there is some important variation 
between stages of development, it is believed 
that a clearer picture of paralepidid classifica- 
tion can be obtained by utilizing more than 
one growth stage. 
Fig. 3. Top of head of a paralepidid (illustrated from 
a paratype of Lestidium Thunnorum n. sp.), showing the 
cranial ridges and the foramina in the premaxillaries. 
Although our present knowledge of these 
fishes is so inadequate as to make very exten- 
sive splitting seem premature, certain aggre- 
gates of species appear to be so widely sepa- 
rated from their nearest relatives on the evi- 
dence of prominent characters in stages older 
than postlarvae that their inclusion in Para- 
lepis seems much too conservative. The reten- 
tion of only three genera does not allow us to 
depict several important and easily recognized 
lines of evolution. In my own investigations, 
conclusions have been reached which involve 
a splitting of the genus Paralepis, a resurrec- 
tion of the previously suppressed genera Les- 
MAGNISUDIS 
PARALEPIS 
PROFUNDISUDIS 
NOTOLEPIS 
LESTIDIOPS 
LESTIDIUM 
LESTROLEPIS 
MACROPARALEPIS 
STEMONOSUDIS 
SUDIS 
Fig. 4. Diagram of the extent of relationships be- 
tween the genera of the family Paralepididae. 
