120 
ERIK A : SON STENSIQ 
Of the Jurassic forms Cpccodenna gigas cannot, as Reis (1900, p. 191) has pointed 
out, be taken any longer as a Coccodenna species. According to the conclusions I have 
been able to draw, it certainly represents a new genus. 
The organization of the Coelacanthids. Summary. 
The primordial neurocranium and its ossifications. — Looked at as a 
whole the primordial neurocranium of the Coelacanthids is high and often relatively 
narrow. Its basis is straight from the snout and backwards at least as far as the 
anterior part of the labyrinth region, while its roof, on the other hand, is developed 
with an angle that is in most cases' prominent, as the posterior part is horizontal and 
the anterior part slopes more or less forward. The angle between the two parts is 
situated above the posterior part of the orbitotemporal region. 
In most cases the ethmoidal regio.n is broad and moderately long and, seen from 
above, bluntly rounded. In profile, on the other hand, it is either relatively high and 
blunt anteriorly ( Coelacanthiis, Axelia) or else relatively lower and more pointed ( Macro- 
poma, Undina, Wimania). 
The orbitotemporal region is characterized in most cases by a considerable length 
and is also always high. It is only in certain forms of the Axelia type that it seems to 
have been more distinctly shortened. It was narrowest (thinnest) at the ventro-medial 
part, at which farthest posteriorly beneath the fossa hypophyseos it probably had a 
fenestration immediately above the parasphenoid. While in such forms as Coelacanthus, 
Sassenia, Axelia its ventral parts seem to have been comparatively wide, in others 
(e. g. Undina and Macropoma ), on the contrary, these parts must have been considerably 
narrower. 
Posteriorly the orbitotemporal region has passed without any sharp boundary into 
the labyrinth region. This latter region forms the broadest part of the neurocranium, 
is well developed in other respects, and seems to have had a fairly normal length. On 
its dorsal side a paired muscle fossa («fossa temporalis*) was apparently developed in 
the posterior half. 
The occipital region is very incompletely known but was most probably short. 
It is true that a large number of ossifications appear in the primordial neurocranium, 
but the main part of these undoubtedly consisted of cartilage in the forms nearer known. 
In Undina (Heineke, 1907) and possibly in Macropoma (Huxley, 1866; Heineke, 1907) 
too we only know one basioccipital in the occipital region. In these forms no other 
ossifications seem to exist in this region. At present it is impossible to ascertain whether 
the other Coelacanthids also showed the same conditions or whether there were also 
forms that had a quite cartilaginous occipital region. In all the Coelacanthids the chorda 
dorsalis extended into the occipital region and possibly still farther forward. 
The lateral walls of the labyrinth region are formed to a rather considerable 
extent by a paired prootico-opisthotic, which seems to have surrounded an anterior 
part of the labyrinth. The main part of the bone is really situated in the anterior, 
ventral part of the labyrinth region and from this main part, which can conveniently be 
called the corpus, there issues a strong process dorsally and, as a rule, a long one in 
