242 
ERIK A : SON STENSIO 
The processus ascendentes of the parasphenoid in Birgeria cover from the lateral 
sides that part of the wall of the primordial cranium where the trigemino-facialis 
chamber and the myodome are situated. Perhaps they extended somewhat farther forward, 
and, if so was the case, an extracranial space would of course be added both to the myo¬ 
dome and the trigemino-facialis chamber, which otherwise seem to be intramural spaces. On 
the medial side of the process there is a small spine-like formation, whose position and 
direction suggest that in it we perhaps have the partial homologue of the alisphenoid 
pedicle of Amia, although this may preponderatingly have been of a membranous nature. 
With regard to the membrane bones on the cranial roof I have little that is new 
to add to what has already been shown by Traquair’s investigations, and in order not 
to repeat previously familiar facts with regard to the configuration of the different bones 
and their mutual relations I refer to Traquair’s excellent account of 1877 (1877 a, pp. 15 
—16) and Woodward’s of 1891 (1891b, pp. 423—527). The only fact that needs perhaps be 
pointed out is that the extrascapulars («supratemporals» Traquair) are in several cases re¬ 
presented by a single large plate on each side of the median line (Cheirolepis, Bc»~eosomus). 
According to what has been shown by my investigations of the membrane bones 
of the cranial roof in Crossopterygians (cf. pp. 62—66, 94—98, 102—105, i 33 —134, i 38 , 
text figs. 43, 51, 57, 58) it seems, however, as if a number of alterations ought to be made in 
the nomenclature used by Traquair and previous authors. TIiusTraquaWs squamosal should 
correctly be called the supratemporo-intertemporal, a term already introduced by me in my 
description above. His prefrontal ought to be termed antorbital (Allis 1905, p. 491), while his 
large ethmoid presumably comprises several different elements, among which there are 
especially the homologues of the nasals, postrostrals and interrostral elements of both sides 
of the Crossopterygians (cf. Goodrich 1909, p. 3 og and my description of the sensory canals 
of the cranium below). For the sake of brevity I have called it after its most important 
components the nasalo-postrostral. According to Traquair (1877 a, PI. I, fig. 8) no 
independent rostral elements seem to be present. They may as far as one can understand 
have fused with neighbouring bones, but as the position of the ethmoidal commissure 
between the infraorbital canals and the anterior portions of these canals are not known, 
it is impossible to decide with certainty how the fusions have taken place. It does not 
seem improbable, however, that a couple of the most medially situated rostral elements 
from part of the nasalo-postrostral. With regard to the homologues of the lateral rostral 
plates of the Crossopterygii, it seems as if in the Palaeoniscids, if Traquair’s view is 
correct, they form part either of the premaxillary or the antorbital of each side, most 
probably, however, the former. 
The labyrinth, as we have seen, has been principally situated in the prootico- 
opisthotic. I have clearly observed the cavities both for canalis semicircularis externus and 
canalis semicircularis posterior. Otholites are described in Palaeoniscus (Hennig 1915, 
p. 52—55), Aniblypterus (Fritsch 1895, p. 126; PI. 122, figs. 1, 3 , 4, 5) and Rliadinichthys 
(Traquair, 1910, p. 127). 
Visceral skeleton. 
The palato-quadrate seems as a rule to have been triangular in shape. In the 
anterior part it has a small autopalatine, in the posterior one the metapterygoid and 
