TRIASSIC FISHES FROM SPITZBERGEN 
I 2 I 
a caudal direction as well. The latter process runs horizontally near the base of the 
neurocranium and its posterior end joins the lateral side of the basioccipital when this is 
developed. 
As the corpora of the prootico-opisthotics are situated far apart, there is thus at 
the basis cranii between them an unossified interspace which in its continuation backwards 
is bounded on the lateral sides by their posterior processes, posteriorly by the basioccipital, 
when there is one, and anteriorly by the basisphenoid. In this interspace, which was at 
least partly filled with cartilage, the chorda dorsalis has extended. 
Behind the corpus of the prootico-opisthotic of either side the lateral wall of the 
labyrinth region was cartilaginous, apart from the small part consisting of the posterior 
process of the bone mentioned. Dorsally of the corpus of this bone on both sides of its 
dorsal process there have certainly been cartilaginous parts. The whole roof of the region 
has also consisted of cartilage, except for possible fontanels. 
The anterior and larger part of the orbitotemporel region, the part that forms the 
real interorbital wall, was unossified throughout and apparently cartilaginous; the posterior 
part of this region, on the other hand, is occupied to a rather considerable extent by 
an unpaired ossification, which can be called the basisphenoid. This bone also reaches, 
however, somewhat backwards into the most anterior part of the labyrinth region, but 
for practical reasons I prefer to describe it as it if was entirely within the orbitotemporal 
region. In extent it really corresponds fairly closely to the basisphenoid in Stegocephalians 
(Watson, 1912a; 1916; 1919; Huene, 19 i 3; etc.) and Reptiles. As comparisons with 
Dictyonosteus (Stensio, 1918 c) and Polypterus (Traquair, 1871) indicate, one is probably 
justified in considering it as the postero-ventral part of a large sphenoid, which originally 
extended forward to the ethmoidal region. While an additional component of this 
sphenoid, situated postero-dorsally and corresponding about to an alisphenoid is still 
present in Coelacanthids, although it is fused with a dermal element of the cranial roof, 
the homologue of the orbitosphenoid component of the sphenoid, on the other hand, is 
quite reduced. 
The corpus of the basisphenoid is situated behind the fossa hypophyseos and ought 
thus to belong to the chordal part of the cranium. Strong basipterygoid processes, 
situated high up, against which the metapterygoid of either side articulates, are developed, 
and besides these there issues from the corpus anteriorly on each side a lamella which 
bounds the posterior part of the fossa hypophyseos laterally and partly also ventrally, 
and, in addition, a paired process issues in a ventral direction and is attached to the 
parasphenoid. 
Between the last-mentioned processes from the basisphenoid corpus and the para¬ 
sphenoid there is found a more or less strongly developed, forward open canal, so situated 
that it may have had its exit through the supposed fenestration in the interorbital wall. 
Both on account of this position and because of the strong development of the eye 
bulbs in all Coelacanthids, this canal ought, it seems, to be interpreted as a myodome. 
In the ethmoidal region we only find one paired ossification, which, as far as can 
be judged, constitutes a preethmoid (septomaxillary in the same meaning as in Amici), 
while homologues of the exethmoids were not developed. Each preethmoid bone 
is situated in the ventral part of the region in question and extends so far medially 
Stensio, Triassic Fishes from Spitzbergen. 16 
