TRIASSIC FISHES FROM SPITZBERGEN 
77 
the hypophysis was in connection with the mucous membrane of the mouth cavity by a 
foramen in the parasphenoid (cf. Lehn 1918, figfs. 9—12). In addition the ontogenetic 
development of the hypophysis also of course supports the idea that certain forms it may 
remain in more or less direct contact with the parasphenoid, or be partly surrounded 
by that bone. Nor do I consider that Day's interpretation of the anterior process as a 
basipterygoid process can be maintained either. For a process with about the same 
position as this one in the Palaeoniscids also occurs in Lepidosteus, where it undoubtedly 
belongs to the parasphenoid, in spite of the fact that, together with an antero-ventral 
part of the prootic, it is developed with a facet for articulation with the metapterygoid 
(Allis, 1909 a, p. 184; Veit, 1907, p. 186). It thus seems presumable that at least certain 
Palaeoniscids showed in this respect about the same conditions as Lepidosteus. Others, 
on the contrary, such as, e. g., B. mougeoti seem to have been without any articulation 
between the palatoquadrate and the neurocranium. 
The trigemino-facialis chamber and the myodome. — I have already 
had an opportunity above (p. 166) to describe a paired, large, forward open cavity, 
situated behind the orbit. In connection with this I have also noted that this cavity 
presumably corresponds to a great extent both to the trigemino-facialis chamber and 
the myodome. Now that we have got to know the relation between the primordial 
neurocranium and the parasphenoid it may be convenient to look more closely at this 
postorbital cavity. 
It has extended rather considerably rostro-caudally, reaching from the anterior part 
of the labyrinth region forward to about the boundary between the posterior and the 
middle third of the sphenoid. Its most posterior part, which falls within the labyrinth 
region, was probably surrounded on all sides by cartilage. Its anterior and incomparably 
much larger part, which belongs to the orbitotemporal region, is bounded in its present 
state of preservation entirely by bone (text fig. 64 B). The basal part, bsph, of the 
sphenoid and the basal parachordal plate (pc) form together the medial wall, arid the 
parasphenoid with its processus ascendens forms the lateral (ventro-lateral) one. The 
dorsal (dorso-medial) wall consists of the pars lateralis of the sphenoid, and the ventral 
wall in the anterior part of the basipterygoid process (bp) and in the posterior one of 
the parasphenoid. It is, however, rather probable that a lamella of cartilage may have 
continued forward from the labyrinth region both along the lateral wall of the cavity 
and along its ventral wall to the posterior margin of the basipterygoid process, 
especially if we may judge from Saurichthys, as we shall see below (Part II). If this 
were the case the parasphenoid would in reality not at all or perhaps only slightly 
(farthest forward on the lateral side) have formed part of the boundary of the post¬ 
orbital cavity. 
The postorbital cavity is low and wide in shape. The width in greatest anteriorly 
and decreases continuously posteriorly. In cross section the anterior part of the cavity 
(niyi ; trig. fac. ch, text fig. 64 B) has an irregularly quadrangular shape, while its posterior 
part is more oval. 
At least from a functional point of view the postorbital cavity can be devided 
into a ventro-medial and a dorso-lateral half, the former of which functioned as a myodome > 
the latter presumably as a trigemino-facialis chamber (cf. Allis 1897 a, pp. 492—506; 
Stensid, Triassic Fishes from Spitzbergen. 23 
