TRIASSIC FISHES FROM SPITZBERGEN 
275 
The antorbital in Amia, apart from its anterior end, which we have just dealt 
with, seems to be homologues with at least an anterior part of the antorbital of Palaeo- 
niscids, Platysomids and Catopterids. 
The nasal in Amia is pierced, as we know (Allis 1889, PI. XLI), in a longitudinal 
direction by the most anterior part of the supraorbital canal and, in spite of its consider¬ 
able width and extension in a medial direction, probably corresponds as a whole only 
to the nasal elements in the primitive Crossopterygians (cf. p. 98, 102 —105, 114, i 33 —135 
above). In Palaeoniscids, Platysomids and Catopterids, as we have seen, all the homo- 
logues of the nasal and postrostral elements of the Crossopterygians are probably fused 
into a single large plate, which also seems to include the interrostrals and sometimes 
one or more rostral elements (Perleidus). 
In order to understand in Amia what has become of the homologues of the post- 
rostrals of the primitive Crossopterygians we must first examine the conditions in 
Polypterus. 
In this last-mentioned fish (Traquair, 1871; Allis, 1900 a, fig. 2, pp. 435—437; 1900 b, 
p.262; 1905, pp. 417—420; 1917, pp. hi, 112) we find an unpaired so-called ethmoid, 
which is long and relatively narrow. Backwards it extends between the nasal elements 
of the right and left sides to the anterior end of the frontals and its posterior longer 
part is covered from the sides to a rather considerable extend by the nasal elements. 
In its anterior end, which is somewhat expanded, it is pierced by the ethmoidal commissure 
between the two infraorbital sensory canals. It thus corresponds (cf. Allis, 1900 b, p. 262) 
with its anterior part to a median or a couple of medial rostral elements, while its 
posterior parts probably include the homologues of the inter- and postrostrals. 
I may now be supposed that in Amia the homologues of the post- and interrostral 
elements gradually became overgrown with nasal elements in the same way as has 
begun in Polypterus, at the same time becoming fused with premaxillaries anteriorly. 
This view seems, at least as far as can be now seen, to be far more probable than the 
one that Allis (1898, pp. 433 —436; 450—456; 1909 a, p. 28) tried to maintain, namely 
that the posterior processes from the maxillaries of Amia should be the homologues of 
the articulatory processes of Teleosts (cf. Allis 1919 a). 
Palaeoniscids, Platysomids and Catopterids compared with Lepidosteus. 
With regard to the primordial neurocranium I have nothing to add to what I 
have said in my description of Birgeria mougeoti (cf. pp. 152—180 above). Here I shall 
confine myself to some remarks about the bones on the dorsal side of the ethmoidal 
region. The homologue of the middle rostral of Colobodus is represented in Lepidosteus, 
as in Amia, by the bone that was formerly called the ethmoid (Allis, 1905, pp. 407—417), 
and the homologue of the paired lateral rostral of Colobodus may have fused in Lepidosteus, 
in the same way as in Amia, with the antorbital, representing a considerable part of 
that bone. 
The specimens of Lepidosteus investigated by me (belonging both to L. platystomus 
and to a smaller long-snouted species) show that, as Allis pointed out in 1905, 
pp. 4I 5—4x6, the so-called ethmo-nasal of each side is connected anteriorly under the 
nasals uninterruptedly with the premaxillary on its side and that, in addition, its 
35* 
