TRIASSIC FISHES FROM SPITZBERGEN 
115 
median one on each side are also joined posteriorly to the most anterior postrostral 
plate. The centres of ossification are clearly discernible in all the bone-plates now described 
with the exception of the interrostrals. 
Nothing is preserved of the sculpture of the bones or of the sensory canals. 
* * 
Remarks. — At present it is impossible to decide with certainty whether this 
specimen belongs to any of the genera that have been described above or whether it 
represents a new one, but as, in spite of considerable deviations from the Axelia type, 
it still seems to be most closely connected with this, it is most probable that it will 
turn out to belong to some Mylacanthus or Scleracanthus species. 
Geological occurrence and locality. — Fish horizon, Mt Trident. 
Some remarks on previously described Coelacanthids. 
With the help of the experience gained from the material now described from the 
Spitzbergen Triassic it may be of use also to discuss briefly some questions of detail 
with regard to the previously described Coelacanthids. This seems to be all the more 
justified because I have had an opportunity to study the more important type specimens 
of Coelacanthids preserved in European museums. At the same time as I briefly mention 
in this connection some new observations I must, however, show to a considerable extent 
the mistakes in certain older views, and the following short exposition accordingly not 
only forms an addition to previously known facts but also partly takes the form of a 
criticism. 
The primordial neurocranium and its ossifications. — Heineke in 1907 (p. 8) 
described a basioccipital in Undina and the correctness of his interpretation of this bone 
seems to me obvious, partly from his own type specimen and partly from another 
specimen belonging to the British Museum (B. M. P. gi 3 o). Other occipital bones are 
not developed, as far as I could find out. It is true that in 1888 Reis (pp. 24—25) also 
described a supraoccipital and a paired exoccipital (lateral occipital) in Undina, but an 
investigation of the specimen on which he based his description shows, however, quite 
evidently that his supraoccipital is really identical with the basioccipital and his exoccipital 
with the posterior parts of the very much flattened prootico-opisthotic, seen from beneath. 
The detached fragments of bones that lie between the basioccipital and the prootico- 
opisthotics and which Reis takes to be vertebral arches, cannot be exactly defined (cf. Reis, 
1888, p. 25, PI. I, fig. 22). With regard to the bone or the part of that bone of Macro- 
poma that Huxley interpreted in 1866 (p. 36 ) as a possible «vertebra», which, in his 
opinion, corresponded to one of the «vertebrae» that are coalesced with the occipital 
region in the recent ganoids, it is impossible to pronounce any quite definite opinion. But 
it does not seem improbable that this bone, as Heineke (1907, p.8) has also pointed out, 
might be the basioccipital, which in this case has come loose from its original position. 
The prootico-opisthotic and the basisphenoid seem apparently, at least in the majority 
of Coelacanthids, to have been developed as in the Spitzbergen forms described above. 
Besides in these, the bones in question are partly or entirely preserved in Macropoma 
