TRIASSIC FISHES FROM SPITZBERGEN 
191 
As we know the endoskeletal radials of the dorsal and anal fins are generally divided 
into three segments (cf, Schmalhaussen 1912, igi 3 a, igi 3 b), it now seems probable to 
me that the proximal ossifications in the five posterior endoskeletal radials of the anal 
fin in B. mougeoti were situated in the proximal segments, while the distal ossifications 
(Ms) belonged to the middle segments. The distal segments have been small and carti¬ 
laginous, as in the Actinopterygii in general. Thus we have in the posterior part of the 
anal fin both ossified axonosts and baseosts, to use Cope’s (1887, p. 1016) and Woodward’s 
(1891b) terminology. 
The long ossifications in the 16 most anterior endoskeletal radials of the anal fin 
must either have been situated entirely in the proximal segments, which would then be 
very much lengthened at the expense of the middle ones, or have arisen by a fusion 
of the ossifications of the middle and proximal segments in each radial. If the former 
alternative is correct both the middle and the distal segments would of course have 
been small and cartilaginous. 
In certain primitive Palaeoniscids, e. g. Elpisopholis (Woodward, 1908 a, p. 20), both 
axonosts and baseosts are present, though the baseosts are short. On the other hand in 
Coccolepis, which is geologically the youngest representative of the Palaeoniscids, baseosts 
are absent, at least as independent bones (Woodward, 1895a, pp. 4—8; Traquair, 19x1b, 
pp. 1 3 —14). The modification of the endoskeleton of the unpaired fins thus seems in 
the Palaeoniscids to go in such a direction that the baseosts disappear either by being 
fused with the axonosts or being reduced. Birgeria mougeoti shows in this respect an 
intermediate stage between Elpisopholis and Coccolepis. 
Another circumstance with regard to the endoskeletal radials in the anal fin of 
B. mougeoti ought also to be noted in this connection. It is evident that the endoskeletal 
radials correspond to some extent to the number of the haemal spines only in the 
anterior part of the anal fin. In the middle part of the fin we find two and in its 
posterior part two or three inner radials to each haemal spine. 
Similar conditions have already been shown by Traquair for the dorsal fin of 
Phanerosteon (1911a, p. 167) and seem also to occur in Palaeoniscids in general, both in 
dorsal and anal fins ( Psilichthys, Hall, 1900, p. 147). 
The dorsal fin (D, PI. 21, fig. 4; PI. 22, fig. 2) in Birgeria mougeoti was situated 
rather far caudally and had a large posterior part opposite the anal fin. It is very 
incompletely preserved, so that nothing is to be seen with regard to its shape. It is 
evident, however, that it was large and had a considerable number of lepidotrichs, 
which showed in all respects the same conditions as those of the anal fin. The endskeleton 
consists, as is usual in Palaeoniscids, of axonosts and baseosts (Traquair, 1877 a, p. 24; 
Woodward, 1891b, p.427; 1898 a, p. 84; 1908 a, p. 20), which seem, as in Phanerosteon 
(Traquair, 1911a, p. 167, fig. 9) to have been of about the same length. This would 
apparently mean that the proximal and middle segments of the endoskeletal radials were 
equally strongly developed, a condition that, according to Schmalhaussen’s investigations 
(1912—xgi 3 ), is to be considered as rather primitive. As in the anal fin, the number of 
the endoskeletal radials was greater than that of the neural spines. I; ) 
l ) A fragment either of a dorsal or an anal fin, probably belonging to B. mougeoti, has a certain interest, 
as it shows that endoskeletal radials were also present in front of the fins in question (cf. TRAQUAIR, 1877 a, p. 24). 
