TRIASSIC FISHES FROM SPITZBERGEN 
opisthotic or the basisphenoid. — In 1888 (pp. 46, 55—57) Reis took the epihyal to be 
the metapterygoid. 
In all the Coelacanthids in question the urohyal seems to have resembled that of 
Wimania and Sassenia. In the majority of Jurassic and Cretaceous forms, however, it is 
considerably narrower than in the two Spitzbergen genera just mentioned and in the 
Palaeozoic Coelacantluis species. I have been unable to come to any conclusion as to what is 
really represented by the cruriform bone that Huxley (1866, p. 19, PI. IV, fig. 1) has described 
and figured in a specimen*) of Coelacantluis lepturus Agassiz; On the other hand the 
ossified mass which Davis (1884, p. 430; PL 47, fig. 1) compares in a specimen (J 3 . M. 
P. J7S4) of C. tingleyensis Davis with the above-mentioned cruriform of C. lepturus consists 
exclusively of ossifications belonging to the primordial neurocranium, which have been 
strongly compressed. Immediately to the left of the anterior end on the urohyal I can 
thus identify the lower surface of the prootico-opisthotic, and farther forward between 
the two jugular plates there is a part representing the compressed basisphenoid. 
The sensory canals of the head. —- The sensory canals of the head are, as 
we known, developed to a different degree in different Coelacanthid forms and the 
pores too may show considerable variations both in size and number. In Coelacantluis 
Undina and Macropoma the pores often seem to be rather small and numerous. Thus, 
for instance, on the supraangulo-angulare of Coelacantluis elegans Newberry I have found 
no less than twelve and on the same bone of Undina gulo the number is probably still 
greater. 
In Macropoma speciosum Reis has taken the sensory canal in the supratemporo- 
extrascapular and the parieto-intertemporal as a fossa temporalis (Reis 1888, p. 62, 
PI. IV, fig. 2). ' 
Some remarks on the different genera of the Coelacanthids. 
The present state of our knowledge of the Coelacanthids seems to indicate that 
these fishes show a rather greater variability than we have been inclined to think, 
and also that their generic characters as a rule are somewhat more strongly marked 
than was thought to be the case by a number of authors (Lutken, .1868, p: 61; Newberry, 
1888, p. 71; Reis, 1900, p. 191). In my opinion it is therefore necessary to maintain the 
majority of the species described earlier, in spite of the fact that in a number of cases 
we cannot yet fully define them. On the whole I agree on this question most closely 
with Woodward’s view, as put forward in his «Catalogue» of 1891. 
Although I do not think it improbable that with more extensive knowledge of the 
Carboniferous and Permian Coelacanthids it may be necessary to divide them into two 
genera, as Reis (1888) has suggested, still a present a division of this sort cannot be 
conveniently carried out, as our knowledge of the differences between them is very 
incomplete. On the other hand, even under present conditions it seems to be necessary 
to exclude at least certain of the Triassic species from the genus Coelacantluis, as for 
instance C. lunzensis Reis (cfi Schlosser, 1918, p. 87). 
1 I have not had an opportunity to study this specimen. 
