TRIASSIC FISHES FROM SPITZBERGEN 
249 
portion of the musculus coraco-arcualis. By analogy me may thus be justified in 
concluding that a similar portion of this muscle occurs in Acrorhabdus. 
It is also noteworthy that between the anterior portion of the musculus coraco- 
arcualis. of the right and left sides in Acrorhabdus there was a well developed urohyal 
(basibranchiostegal), as in Teleosts and Coelacanthids. 
I have nothing important to add about the delevopment of the pectoral fins. 
The pelvic girdle is known so far only in Birgeria, where it consists of a fairly 
large bone, expanded in the postero-distal part. This bone shows no indications at all 
of division into segments, as in the sturgeons. It probably comprises not only a 
component homologous to the pelvis of the Cladodontids and Pleuracanthids but also 
a component that in these forms is represented by the proximal segments of the anterior 
endoskeletal radials. 
Behind the pelvic plate and following immediately upon it there must have been 
either a single long metapterygeal element or a series of several small ones, which were 
cartilaginous. The endoskeletal radials, which were probably about 16 in number, arti¬ 
culated against the pelvic plate and the metapterygeal elements just mentioned. Only 
the 4—5 most anterior ones, however, articulated against the pelvic plate. The all 
consist of a long proximal and a short distal segment, the former of which is always 
ossified, while the latter only shows ossification in some of the middle endoskeletal radials. 
A row of ossified proximal segments of endoskeletal radials is known in addition in 
Coccolepis (Woodward, 1895a, 1898a; Traquair, 1911b) and Elpisopholis (Woodward 1908a). 
The lepidotrichia of the ventral fins are generally jointed right to their bases. An 
exception to this is seen, however, in Birgeria, where at least a large number of the 
anterior ones are unjointed in their proximal parts. 
Squamation. 
As we now know, the Palaeoniscids show considerable variation with regard to 
the development of the squamation. Thus in Chryphiolepis (Traqair, 1907, p. 104—106) 
and Coccolepis (Traquair, ignb* p. 11—15) the scales are cycloid apart from those 
covering the dorsal lobe of the caudal fin, and in Birgeria, Psilichthys, Elpisopholis and 
Phanerosteon the scales are quite reduced except for the dorsal lobe of the caudal fin, 
where they are constantly retained. 
Main lateral canal of the body. 
In Birgeria we find the lateral line of the body enclosed in small tubular bones 
lying in a continuous longitudinal row. In Elpisopholis (Woodward, 1908 a, p. 20) it is 
situated in a row of <-shaped small shields. On the upper lobe of the caudal fin it 
passes off along the ventral margin of the axis, i. e. close to the bases of the lepido¬ 
trichia, as in the sturgeons. 
Some general remarks. 
As we see, the Palaeoniscids include forms that have attained rather different 
degrees of specialization. Such genera as, e. g. Acrorhabdus and Boreosomus much 
Stensio, Triassic Fishes from Spitzbergen. 32 
