TRIASSIC FISHES FROM SPITZBERGEN 
2i3 
bones. We thus have here a dorsally open fossa, which undoubtedly has been developed 
in connection with the insertion of the musculus adductor mandibulae. 
The bone now described must, as can easily be seen, represent a considerable part 
of the Meckelian cartilage and nothing like it is known so far in any Teleostome fish. 
It is evident, however, that the bone in question has such an extension that it must 
presumably comprise not only the homologues of the articular and the autangular but 
also that of the mentomandibular. One might possibly suspect that the autocoronal may 
also forms part of it. 
Of the membrane bones of the mandible the dentalo-splenial, angular and supra- 
angular (particular* Traquair, 1877 a, p. 19) are incompletely preserved. The dentalo- 
splenial has, as already mentioned, along its dorsal margin a medial tooth-bearing 
lamella that covers the long substitution bone in the Meckelian cartilage. The supra- 
angular was as usual a small bone forming together with the dentalo-spenial the lateral 
wall of the fossa for the insertion the musculus adductor mandibula described just 
above. No further details are known as to the boundaries or shape of the three bones 
in question. 
The hyomandibular is a large and powerful bone whose shape is seen in text 
fig. 70 (Hm). Its middle part, which is relatively thin, consists of compact bone. The 
dorsal and ventral parts, both of which are rather thick, have consisted of cartilage 
interiorly, and this cartilage has formed a continuation of the cartilaginous epiphyse which 
has certainly been present at either end of the bone. The bone substance is radiate, 
and a distinct centre of ossification is discernible, situated in the middle part of the 
bone. From the posterior margin of this part there issues a long processus opercularis 
(pr . op) whose distal end was evidently covered with cartilage. As we shall see, a 
similar process is also found in the Palaeoniscid genus Acrorhabdus described below. 
The occurrence of this process shows beyond all doubt that with regard to the 
development of the musculature of the gill-cover both Boreosomus and Acrorhabdus have 
been specialized in the same direction as the higher Ganoids and Teleosts. A musculus 
dilatator operculi would consequently have been developed from the musculus constrictor 
superficialis I, at the same time as musculus constrictor superficialis II was probably 
weakened and subdivided into more sharply localized independent muscles (cf. Vetter, 
1878; Ruge, 1897; Sagemehl, 1885, p. 62; Luther, 1909, igi 3 ). 
This state of affairs is especially noteworthy as the hyomandibular in Palaeoniscids 
which have hitherto been known, has shown more primitive conditions. According to 
Traquair’s exposition (1877a, p. 17, p. 3 g; PI.I, fig. 3 ; PI.II, fig. 2) the processus opercularis 
is thus not developed or at any rate not distinct in the genera Palaeoniscus, Nematoptychius, 
Oxygnatlius, Elonichthys, Cycloptychius and Amblypterus, to which may also be added 
Gyrolepis, which I have had an opportunity to investigate myself. Finally by Traquair’s 
investigations (1879, p. 366 ) it was also shown that a distinct processus opercularis was 
not present either in Cheirodus among the Platysomids. 
The morphological importance of these facts was obviously quite clear to Traquair, 
as he says with regard to the relation between Polyodon and Palaeoniscus (1877a, p. 3 g): 
«In neither is the operculum attached to the hyomandibular by a joint, but secured 
in its place only by skin and muscles.» In other words, no musculus dilatator operculi 
