TRIASSIC FISHES FROM SPITZBERGEN 
l3i 
on both the dorsal and the ventral side with a simultaneous reduction of the unpaired 
fins. Like Reis (1892) and Dollo (1895), he considers the diphycercy in the recent 
Ceratodus to be secondary, and he finds evidence for this view of his in the fossil forms, 
in Ctenodus among others. 
While in the primitive Dipnoi and the Crossopterygii the caudal fin was of a 
heterocercal shape, vve find, as he also points out, in the somewhat more spezialized 
representatives of these fishes a number of types in which it is rather symmetrically 
developed. Among the Crossopterygii he pays the closest attention to Eusthenopteron in 
this respect, and he considers, like Dollo (1895), that the caudal fin of the Coelacanthids 
is developed from a type of this sort. He then continues as follows (pp. 55—56): 
«Die Schwanzflosse der Coelacanthidae ist derselben der Rhizodontidae sehr ahnlich, 
wir konnen sie nur als ein weiteres Umbildungsprodukt in derselben Richtung ansehen. 
— Ausbildung einer vollkommenen Symmetric und progressives Wachstum der Dorn- 
fortsatze, welche auch dorsal eine Abgliederung erworben haben, zeichnen das Skelett 
dieser Flosse aus. — Was die Bedeutung der Pinselflosse anbelangi, so scheint es mir, 
dafi sie vollkommen dem mittleren Schwanzlappen des Eusthenopteron entspricht und 
dafi folglich die sekundare Symmetric genau in derselben Weise ausgebildet wurde. » 
The diphycercal shape of the caudal fin in the Coelacanthids should thus be 
secondary and the endoskeletal supporting elements connected with the axial skeleton 
are to be considered as distal segments of the neural and haemal spines. 
Air-bladder. — With regard to the air-bladder I have only to remark that it 
has been impossible so far to trace it in any of the Triassic forms from Spitzbergen. 
Squamation. — The most important characters of the scales are well known 
from a number of earlier works and as I have nothing of importance to add here I 
shall simply refer to these (Huxley, 1866; Reis, 1888, 1892; Zittel, 1887; Williamson, 1849; 
Woodward, 1891b; 1909). 
Coelacanthids compared with a number of other fishes and 
certain tetrapods. 
Coelacanthids and other Crossopterygians. 
Coelacanthids and Rhipidistids. — In the Rhipidistids the ossifications of 
the primordial neurocranium are very seldom preserved. It is really only in Dictyonosteus 
(Stensio, 1918c) and Megalichthys (Young, 1866; Cope, i 883 ; Watson, 1912a; Birks, 1916) 
and to some extent in Dendrodus and Cricodus (Rohon, 1889) that we know a number 
of them more or less incompletely. 1 ) 
In the occipital region in Megalichthys, according to Young’s description (1866, 
p. 605), only a basioccipital seems to ba developed in the same way as in Coelacanthids. 
z ) Pander (i860, pp. 15—16) described fragments of substitution bones from the anterior portions of the 
head of Osteolepis. 
The primordial neurocranium of Rhizodopsis is said by Watson and Day, 1916 (pp. i3—14) to have 
consisted of cartilage. I do not feel convinced, however, that this is really the case, but consider it more probable 
that the «cartilage* was bone of a similar structure as in Dictyonosteus. 
17* 
