TRIASSIC FISHES FROM SPITZBERGEN 
l § 5 
Hbr,_ 
sinus into an anterior and a posterior short truncate process. Whether that of branchial 
arch IV was ossified or consisted of cartilage is a question Shat cannot be answered 
from the present material. 
At least in the second and probably also in the first pair of branchial arches the 
pharyngobranchials were divided into infra- and suprapharyngobranchial, all of which, 
except the infrapharyngobranchial II ( Ifbr 2 , text fig. 65; PI. 23 , fig. 2), were probably 
cartilaginous. Infrapharyngobranchial II consists of a relatively large bone, shaped 
somewhat like in hour-glass, which had a carti¬ 
laginous epiphyse at either end. The pharyngo¬ 
branchials of the third pair of branchial arches, 
and presumably also of the fourth pair as well, 
do not seem, at least if we judge from the dorsal 
end of epibranchial III, to have been divided into 
infra- and suprapharyngobranchials, and it is also 
probable that they had no ossification. 
The gill filaments have had in their interior 
ossified or calcified supporting gill-rays, some of 
which are still preserved. As far as can be seen, 
the gill-rays are not comb-shaped. 
The ventral ends of the branchial arches Cbr 2 -- 
were joined in the usual way by a row of copulae. 
The material present does not show the conditions 
of this row in front of the first gill-arch pair. 
Behind this pair and between it and pair IV it 
was strong. In this part it was also strikingly 
well ossified, as we find there no fewer than three 
well developed ossifications ( Coj — Co s , text fig. 66; 
PI. 23 , fig. 1). The most anterior of these, which 
is at the same time the shortest, is situated bet¬ 
ween the first and second pairs of branchial 
arches, the following one between pairs II and III, 
and finally the most posterior one, which is also 
the longest, between pairs III and IV. All the 
ossifications were separated from each other by 
short cartilaginous interspaces and caudally of the last one there was also certainly 
cartilage. It is of course impossible to decide the relations of these ossifications to the 
segmentation of the row of copulae 1 ). 
The investigations of Traquair (1877 a, p. 22), Fritsch (1895, vol. 3 , p. 79) and others 
have already shown that the branchial skeleton in the Palaeoniscids was well ossified, 
*) As Gaupp (1904, pp. 910, 914) has pointed out, our knowledge of the copular skeleton of the Ganoids 
and the Teleosts is very incomplete. Moreover he name copula or basibranchial has been used as a term both for 
each of the cartilaginous segments of the copular row and for each of the ossifications that appear in this, even if 
two or more of these ossifications corresponded to one cartilaginous segment. It is easy to understand that such 
a method of procedure has caused a great deal of confusion. In this work I have consequently followed Gaupp and 
avoided calling the copular ossifications copulae. 
Stensio, Triassic Fishes from Spitsbergen. 24 
Text fig. 66. Birgeria mougeoti Agassiz. 
Parts of the visceral skeleton from the ventral side. 
After P. 171. »/ 4 . 
Cbr t -Cbr 4 , cefatobranchials; Co r Co 3 , ossifications iu 
the copulae; Ecpt, ectopterygoid; Hbr t -Hbr 3 , hypo- 
branchials; Mco, mixicoronoid; pr. int, process 
possibly representing a fastening for ligamentnm 
interarcuale ventrale. 
