the Cyprinidae. According to th a t author (’91, p. 574), a well developed myodome is found in most
of these fishes. In Cobitis, Misgurnus, Nemachilus and Acanthophthalmus, all of which have small
eyes, the myodome is said to have undergone retrogression. The weak muscles of the eye are said to
all arise in the postero-ventral corner of th e orbit, none of them entering th e cranial cavity. The
anterior edge of th e prootic is said to be formed of two lamellae one lying above the other, th e dorsal
one ending with a free anterior edge, while th e ventral one forms th e floor of th e cranial cavity.
The slit-like space between th e two lamellae, although it lies considerably posterior to th e hind end
of th e orbit, is said to unquestionably represent a myodome th a t has undergone reduction. Reference
to th e figures of Cobitis fossilis (pi. 29, fig. 12) shows th a t th e dorsal lamella must be a rudimentary
mesial process of th e prootic, and as such Sagemehl doubtless considered it, although he does n o t
definitely describe it as such. Anterior to this process, there is said (1. c., p. 549) to be a wide rhom-
boidal hypophysial fenestra. The nervus abducens, in those Cyprinidae th a t are said to have no
myodome, and of which Cobitis is one, is said n o t to perforate th e mesial process of th e prootic, b u t
to apparently issue by th e large optic fenestra. The condition of the myodome is thus here closely
similar to th a t in Lepidosteus; th a t posterior portion of th e canal th a t I have referred to as the
saccular space, existing in a reduced condition, well within the brain case, widely separated from
th e orbit, and in no direct relation whatever to any of th e eye-muscles. This condition is th u s the
reverse of th a t found in Dactylopterus, Gobius and Gadus, in which fishes this p a rt of th e myodome
has been crowded out of th e cranial cavity into th e orbit.
Ameiurus can now be considered, and of this fish I have one skull, prepared some years ago
and p a rtly disarticulated, and a few small alcoholic specimens. Of Ameiurus, McMurrich says:
„Below th e prootics, where they meet in the middle line below, and between them and th e anterior
portion of th e basioccipital above, and th e parasphenoid below, is a small cavity. This is th e almost
aborted rudiment of th e canal for th e orbital muscles, which is largely developed in many fishes,
b u t absent or rudimentary in Silurus, Ameiurus, Gadus, Lophius etc.“
This description of th e conditions in Ameiurus is n o t very satisfactory, and it is not explained,
either here or elsewhere in th e descriptions, th a t this so-called rudimentary myodome is widely
separated from the orbit and out of all relation to th e eye-muscles. Yet such is th e case. In th e
anterior three-fifths, approximately, of its length, th e ventral edge of th e prootic does n o t meet its
fellow of the opposite side, a wide hypophysial fenestra, closed ventrally by th e parasphenoid, being
left between the two bones. Posterior to this fenestra, th e ventral edges of th e prootics meet in th e
middle line, and th e two bones there form, on th e floor of th e cranial cavity, a prominent transverse
bolster which has closely th e position of the cross-canal of Lepidosteus; and it is certainly in this
bolster th a t McMurrich found th e small cavity th a t he considered as a rudimentary myodome. In
two of my small specimens there was no indication whatever of such a cavity; the bolster there being
completely filled with cartilage, b u t having, on its anterior surface, a slight median depression which
doubtless lodged the saccus vasculosus. In one other specimen, there was a median cavity extending
under, ra th e r th a n into th e bolster, while in a fourth specimen there was a smaller b u t similar cavity
on one side only of th e median line. The pitu ita ry body lies immediately anterior to th e bolster,
directly above th e hypophysial fenestra, and a vein th a t apparently drains the pitu ita ry region
begins here and runs directly forward along th e floor of th e cranial cavity, soon separating into two
parts, one on either side. The venous cross-comissure of Lepidosteus thus here seems replaced by
a median vein, this being a variation in detail b u t n o t in principle. The eye-muscles all have their
origins ventral to and between the optic and trigemino-facialis foramina, on the outer surface of
th a t median process of the parasphenoid th a t McMurrich considers as a basisphenoid completely
ankylosed with th e parasphenoid, b u t which Sagemehl (’91, p. 575) considers, in Silurus, as a simple
process of th e parasphenoid; none of the muscles entering th e cranial cavity. I t may furthermore
here be s tated th a t these muscles are said by Workman (’00) to be innervated as they are in Amia,
and hence as I have shown them to also be in Lepidosteus (present work), and th a t Ameiurus is the
only teleost in which this manner of innervation has, as yet, been described.
Ameiurus thus presents conditions similar to b u t even more extreme th an those in Lepidosteus,
the cross-canal of the la tte r fish here being a solid bolster, due, quite undoubtedly, to th e invading
growth of the surrounding cartilage. Yet, th e condition in Ameiurus is said (McMurrich) to represent
an aborted condition of th e teleostean myodome, while th a t in Lepidosteus is said (Sagemehl) to
represent a condition from which th a t myodome was developed! Wishing to determine which of these
two directly opposed assumptions is the correct one, I have consulted the geological record, and,
while th a t record certainly does n o t give a definite solution, it has suggested what seems to me such
a solution; for it has led me to ascribe a totally different motive to th e origin of the myodome than
th a t proposed by Sagemehl, which, as is well known, is th e simple seeking of a better point of origin
b y one or more of th e muscles of th e eyeball.
According to Zittel (’87—’90), whom I have selected because of the convenient tables given,
representatives of the families Siluridae and Ginglymodi, in the la tte r of which the living Lepidostei
are placed, are found in th e Eocene formations, and no earlier; and in all th e living representatives
of these two families th a t have been examined in this connection, there is no functional myodome.
Representatives of th e family Halecomorphi, in which Amia, with a well developed myodome, is
placed, are found considerably earlier, in the Jurassic deposits. Certain other families of th e Lepi-
dosteidae occur earlier th a n any of the Amiadae, th e Stylodontidae being found in th e D yas (Permian)
deposits; but, as two Jurassic representatives of th is la tte r family, Lepidotus and Dapedius, are
said by Woodward (’93) to have a „basicranial canal“ — th a t is a myodome -— it is evident th a t
Lepidosteus can not descend from them, if its myodome represents a primary condition.
The earliest teleosts are said by Zittel to be the Hoplopleuridae and Clupeidae, both of which
are said to be found in th e Triassic, considerably earlier th an the families Siluridae and Ginglymodi,
and as early, even, as any representatives of th e order Lepidosteidae excepting only th e Stylodontidae.
Of these two families of teleosts the Hoplopleuridae are extinct, while the Clupeidae are well represented
by living species, several of which, Clupea, Elops, Albula and Megalops, are known to have a well-
developed myodome. In Clupea harengus, which I have-examined for this purposei the prootic bridge
is thick and has a bevelled anterior edge which slopes antero-ventrally, the whole bridge much
resembling th a t of Gadus excepting th a t it lies in a horizontal instead of a strongly inclined position.
Several other families of teleosts existed a t the beginning of the Cretaceous period, and hence
earlier th an th e Ginglymodi and Siluridae. These families are th e Saurocephalidae (extinct), Strato-
dontidae (extinct), Salmonidae and Scopelidae, all placed by Zittel in his order Physostomi; the
Labridae and Chromidae of his order Pharyngognathi; and the Berycidae, Sparidae, Xiphiidae,
Carangidae, Cataphracti, Gobiidae and Aulostomi of his order Acanthopteri. In living representatives
o f these several families, Parker shows a myodome in Salmo salar, of the Salmonidae; in
Dactylopterus and Peristedion, of the living Cataphracti, I have described it in the present work:
in Gobius of the Gobiidae, a myodome similar to th a t in Dactylopterus is found, as also already
Zoologica. Heft 57. I 26