92 ERIK A: SON STENSIO 
limited to the ventral processes and a part of their dorsal extension. In addition to 
these processes the bone has, as in W. sinuosa, basipterygoid processes and anterior 
lamellae, all of which are joined to a powerful corpus. 
The corpus forms, in the same way as I have described in W. sinuosa, a triangular 
plate with its rounded point ventrally and its base dorsally. From side to side, however, 
it is wider and in an antero-caudal direction it is thicker than in the species just mentioned 
and it also takes up a more vertical position than in that species. On the corpus can 
be distinguished one anterior, one posterior and two lateral surfaces. Of these the dorsal 
one (text fig. 40; PL 14, fig. 3 ) has almost a rectangular shape, with its anterior and 
posterior margins long and the lateral ones short. All the margins are concave, the 
posterior ones least, the lateral ones most. The postero-lateral 
corners are comparatively slightly pronounced; the basipterygoid 
processes (e) issue in the antero-lateral corners. The median part 
of the dorsal surface is somewhat deepened in the longitudinal 
direction of the animal. The two lateral surfaces (text fig. 3 g, 
PI. 1 3 ) are relatively wide and low and pass into those of the 
anterior lamella and ventral processes without any boundary. 
The posterior surface is large and has the shape of a triangle 
with the well rounded point ventrally. It faces not only back¬ 
wards but at the same time somewhat ventrally as well. The 
anterior surface (text fig. 41; PI. 14, fig. 5) is freely visible only 
on a very small portion of its dorso-median part. Otherwise the 
anterior lamellae al issue with a broad base from its lateral and 
ventral parts. Finally it is noteworthy that the surface in 
question also faces somewhat dorsally. 
The two basipterygoid processes (e text figs. 3 g, 40, 41; 
Pis. 11—1 3 ; PI. 14, figs. 2, 3 , 5) agree fairly well in shape with 
the corresponding ones in W. sinuosa, but are stouter. They 
project from the antero-lateral parts of the dorsal surface of 
the basisphenoid corpus in an antero-latero-dorsal direction, so that they diverge 
from each other considerably more than in the Wimania species mentioned. In other 
respects it is worthy of special mention with regard to this that in the species in question 
they issue strikingly low on the lateral wall of the neurocranium (text fig. 3 g), in 
which respect they thus somewhat resemble these of Dictyonosteus. 
The ventral processes (v) of the basisphenoid are, seen from the lateral side 
(text fig. 3 g; PI. i 3 ), short and broad. Seen from in front (text fig. 41; PI. 14, fig. 5), 
they appear to be far apart and do not pass straight downwards but somewhat laterally 
as well, and in addition appear comparatively thick also in a transverse section. The 
ventral ends of these processes are attached to the parasphenoid. 
Finally, with regard to the lamellae ( al ) of the basisphenoid these are distinguished, 
first, by a rather considerable thickness, especially in their posterior parts attached to 
the corpus (text fig. 41; PI. 14, fig. 5). Their ventral parts bend towards and meet 
each other in the median line. Their dorsal parts are posteriorly connected with the 
antero-medial border of the basipterygoid processes and so form a strong support for 
Text fig. 41. Axelia robusta. 
Basisphenoid in anterior aspect. 
From P. ig 5 . 3 /„. 
Bsph, basisphenoid; Psph, para¬ 
sphenoid (cross-section); al, for¬ 
ward pointing lamella from the 
basisphenoid corpus; e, basi¬ 
pterygoid process; m, the entran¬ 
ce of the supposed myodome; r, 
longitudinal ridge on the medial 
side of the forward pointing 
lamellae; s, unpaired . vertical • 
furrow leading down from the 
fossa hypophyseos to the myo¬ 
dome: ventral process from the 
basisphenoid corpus; V,e, outer 
opening of the canal for n. oph¬ 
thalmicus profundus; V t i, inner- 
opening of the canal for n. oph¬ 
thalmicus profundus. 
