26 o 
ERIK A: SON STENSIO 
All the membrane bones described here are ornamented with rather course and 
flat ganoine tubercles, in most cases closely arranged (PI. 33 ; PL 35, fig. 1). 
Visceral skeleton. 
The quadrate is well ossified and consits of cancellous bone substance as in the 
Palaeoniscids. No details are known as to its extension and shape. Nothing is preserved 
of the metapterygoid and the autopalatine. 
The ecto- and entopterygoid both seem to have been present, but are incomplete. 
The maxillary (Mr, text fig. 81; PI. 33 ; PI. 33 ; PI. 34, figs. 1, 2; PI. 35, fig. 1) is of 
the Palaeoniscid type, i. e. its suborbital part is low, while its postorbital part forms a- 
large, rather high plate on the cheek. The postorbital part seems to have been relatively 
longer than in Colobodus (Stolley 1920; Andersson 1916a). 
The mandible is robust and, as far as can be seen, had a straight labial margin 
without any coronoid process. Among its bones only the large high dentalo-splenial can 
be clearly distinguished, and this is pierced by the mandibular canal as 
in the Palaeoniscids and other Actinopterygians, so that it probably 
comprises equivalents not only to the dental of the primitive Crosso- 
pterygii but also to a number of their splenial. elements, as I have 
indicated by the name I have chosen for it. 
The hyomandibular (text fig. 82) is an angular bone as in the 
Palaeoniscids. The angle is situated at about the same height as the 
upper part of the operculum and at this place there issues on the 
posterior margin a short and wide processus opercularis (pr. op), so 
that the musculature of the opercular apparatus must have shown 
somewhat analogous conditions to those in Amia and the Teleosts 
(cf. pp. 2x3—214, 226 above). The dorsal part of the bone is considerably 
wider than the ventral one. Both the dorsal and the ventral ends and 
the processus opercularis have been furnished with cartilagionous epi¬ 
physes. A canal for truncus hyoideomandibularis facialis (can. tr.fac) can be distinctly 
observed 1 ). 
It is evident that the hyomandibular was oblique in the same way as in the Palaeo¬ 
niscids, although not so much as in the majority of these. The corresponding bone 
certainly occupied a similar position in Colobodus and Dollopterus too. 
The preoperculum ( Po , text fig. 81; PI. 33 ; PI. 34, figs. 1, 2; PI. 35* fig. 1) is a large 
and wide triangular plate of about the same type as in Perleidus altolepis (cf. text 
fig. 78). Its long posterior margin, which meets the operculum (Op) and the sub¬ 
operculum (Sop), is convex, its anterior margin concave, and its forward and downward 
pointing margin, which is in contact with the postero-dorsal margin of the maxillary, 
is also concave. The anterior corner forms a little point, which projects in a characteristic 
way forward and downward for a short distance along the antero-dorsal margin of the 
maxillary. The ventral corner is pointed, the dorsal one, on the other hand, probably 
T ) It is of interest to note that the hyomandibular in P. woodwardi greatly resembles in shape that of 
Colobodus bassani (cf. Andersson, 1916a, Hm, PI. II). In this latter species too it is pierced by a canal for truncus 
hyoideomandibularis facialis. '•* 
Text fig. 82. 
Perleidus wood¬ 
wardi nom. nov. 
Left hyomandibular 
from the lateral side. 
After P. iS 5 . Vi- 
Can. tr.fac, canal for 
thetruncus hyoideo¬ 
mandibularis facia¬ 
lis; pr. op, processus 
