TRIASSIC FISHES FROM SPITZBERGEN 
33 
frontal and the dermosphenotic. And the following facts show that the Cbelacanthids 
also, with regard to the arrangement of the bone-plates on the dorsal side of the 
ethmoidal region, have in the main several features in common with the Rhipidistids. 
In Eusthenopteron (text fig. 57) we find, as I have been able to show after careful 
investigations of a well-preserved specimen in the museum here, the rostrals (R), 
interrostrals (Ir) 1 ) and the postrostrals (Ptr) with about the same conditions as in the 
above described Coelacanthids from the Triassic of Spitzbergen. On the other hand the 
antorbital (Ant) is an independent bone, situated in the normal way between the entrance to 
the orbit (orb) and the external nasal aperture (na), and the nasal bone (Na) is, as in 
Teleosts, a narrow bone, pierced by the supraorbital canal and 
situated on the dorsal side of the nasal capsule along the lateral 
margin of the postrostral. She supraorbitals (So) are probably 
represented by one long, narrow bone on the lateral side of 
each frontal (Fr). 
According to Watson and Day’s (1916) reconstructions 
Glyptopomus and Osteolepis show on the whole an arrangement 
very similar to Eusthenopteron and Coelacanthids in this respect. 
But what these two investigators have interpreted as the prefontal 
(Watson and Day’s text fig. 2) is presumably a posterior nasal ele¬ 
ment, and their nasal corresponds to my postrostral. Further, the 
lacrymal of their interpretation seems to represent both the real 
lacrymal and the antorbital of Eusthenopteron. In Osteolepis (cf. 
Watson and Day’s text figs. 5, 9), on the other hand, the nasal 
may be correctly determined, but probably the plate situated 
in front of it belongs to it as well. The postrostrals seem to be 
represented by at least two separate elements on each side of 
the median line, which, as we have seen, may also be the case 
in certain Coelacanthid forms. 
In Megalichthys, as far as one can judge from the figures 
given by Birks’ (1916), there are also probably conditions corres¬ 
ponding to what we have just described in the three preceding 
Rhipidistids. The nasal -f- ethmoid of Birks’ description (cf. his 
PI. 14, fig. 2) corresponds in any case at least to the postrostral, 
while, if his figures are correet, the rostrals probably form part of his premaxillaris. 
The nasal may be an independent bone situated laterally of the postrostral as in 
Eusthenopteron. Wellburn’s exposition of Megalichthys (1900) seems in many respects 
to be unreliable. 
By a new investigation of Dictyonosteus I was able to verify, as seen in text fig. 58, 
that this form too shows in all essentials the same conditions as the other fossil 
Crossopterygians here considered with regard to the mebrane bones on the dorsal side 
of the ethmoidal region. We can thus discover without difficulty the postrostrals (Ptr), an 
unpaired interrostral (Ir), the rosfrals (R), nasal (Na) and antorbitals (Ant). It is noteworthy, 
Text fig. 57. 
Eusthenopteron fordii 
Whiteaves. 
Cranial roof from a specimen 
belonging to the Palaeontological 
Institution Upsala. Vi- 
Ant, antorbital; Dsp/i, dermo¬ 
sphenotic; Ext, lateral extra- 
scapuiar plate; Ext. in, median 
extrascapular plate; Fr, frontal; 
Ir, interrostral; It, intertempo¬ 
ral; La, lacrymal; Mx, maxilla¬ 
ry ; Na, nasal; Pa, parietal; Pmx, 
premaxillary; Ptr, postrostral; 
R, rostrals; So, supraorbital; 
Stemp, supratemporal; na, nasal 
aperture; orb,, orbit. 
[ ) The interrostrale has probably been developed here as a paired bone. 
