TRIASSIC FISHES FROM SPITZBERGEN 
65 
It cannot be proper to retain the earlier name squamosal for such a compound 
bone as that now under consideration in Coelacanthids, as the so-called squamosal or 
dermopterotic in the Actinopterygii may according to our present knowledge correspond 
to the supratemporal and intertemporal of the Rhipidistia fused into one bone. The name 
supratemporo-extrascapular proposed by me thus seems to be the only one justifiable. 
I shall have further reason to return to the question of the homologues of the supra¬ 
temporo-extrascapular later .on in this work. 
Fusions of a kind similar to those now described between the membrane bones 
of the posterior part of the cranial roof of Coelacanthids have also taken place in the 
anterior part. Thus the «frontal» bone according to the terminology of earlier palaeonto¬ 
logists seems first and foremost to comprise the dermosphenotic- (dermal postfrontal), 
which is normally found in the Rhipidistia as an independant bone, though among them 
it may also sometimes fuse with the frontal, as is found to be the case in Rliizodopsis 
(Allis, 1919b, p. 76). The course of the sensory canal shews plainly, as can be 
seen from the description of Axelia robusta given later, that in Coelacanthids the homo- 
logue of the dermosphenotic of the Rhipidistia forms the postero-lateral corner of the 
so-called « frontal*. It is also plain that this so-called «frontal» has become fused 
ventrally with a part of cartilage bone, the above-mentioned alisphenoid, which belongs 
to the postero-dorsal part of the orbitotemporal region. The compound bone which 
has thus been formed through the fusion of the three bones mentioned, I shall call 
in the continuation, for the sake of shortness, the fronto-dermosphenotic, although it 
ought properly to be called the fronto-dermosphenotico-alisphenoid. 
In W. sinuosa the two fronto-dermosphenotic bones (Fr. dsph, text figs. 19, 20, 21, 
25; PI. 4, figs. 1, 2; PI. 5, figs. 1, 2; PI. 8, fig. 1), are long and separated from each 
other by a straight median suture. At their posterior end they are considerably broader 
than the parieto-intertemporal bones; they are narrowest at their middle part, where 
they are of about the same width as the last-mentioned bones. They are situated for 
the most part over the orbitotemporal region and even stretch forward to some extent 
over the ethmoidal region, but reach no farther backwards than over the anterior part 
of the basisphenoid, where they meet the parieto-intertemporal bones in a straight 
transverse suture, which, to be more exact, lies over the anterior end of the basi- 
pterygoid processes (text figs. 19, 20). The centre of ossification in each fronto-dermo¬ 
sphenotic bone (text fig. 21) lies very far back and rather nearer to the lateral border 
than to the medial one. From this centre on the ventral side of the bone, in a ventro- 
caudal and somewhat medial direction, there issues a process (Alsph, text fig. 20; PI. 4, 
figs. 1, 2; PI. 5, fig. 1; PI. 8, fig. 1), which meets the upper anterior corner of the 
lamella al of the basisphenoid. The topographical position as well as the structure of 
the process in question seem to me to indicate that it is the homologue of that part 
of the sphenoid of Dictyonosteus which corresponds to the alisphenoid in the higher 
Actinopterygii. Further proofs of the correctness of this view will be given later in the 
description of Axelia robusta. 
The anterior greater part of the lateral margin of the fronto-dermosphenotic is 
concave. In front of the somewhat projecting postero-lateral corner, which, as I have 
demonstrated, represents the dermosphenotic, there are placed along the lateral margin 
S tensio, Triassic Fishes from Spitzbergen. 9 
