56 MR. B. P. ALLIS ON POSTORBITAL ARTICULATION OF 



upward to the positions the corresponding bones of the adult 

 actually occupy, and that part of ea,ch anterior process of the 

 median basisphenoid on which the base of the related alisphenoid 

 rests would correspond to the outer end of the processus basiptery- 

 goideus of Amia and Lacerta, the remainder of the anterior 

 process apparently corresponding to that little lamellar process that, 

 in Amia, forms the lateral wall of the basal portion of the orbital 

 opening of the myodome and is traversed by the foramen for the 

 efferent pseudobranchial artery. There would have been, internal 

 to the alisphenoid, a space traversed by the vena jugularis and 

 the ramus ophthalmicus profundus, the mesial boundary of this 

 space being formed by cartilage that represents the basi sphenoidal 

 leg of the alisphenoid. This leg of the alisphenoid would have 

 been continuous, posteriorly, with the mesial wall of a trigemino- 

 facialis chamber, that wall being either of cartilage or of 

 membrane, and the base of the leg of one side of the head would 

 have been connected with that of its fellow of the opposite side 

 by the preclinoid wall. The optic chiasma must have lain upon 

 the dorsal surface of the latter wall, and the lobi inferiores in a 

 depression posterior to it, on the dorsal surface of the membrane 

 (more or less ossified) that covers the dorso-anteriorly presented 

 ventral surface of the cartilago acrochordalis. A median opening 

 in this membrane would lead into a large pituitary sac, which 

 would extend ventrally into the space enclosed between the 

 ventral processes of the basisphenoid, these lattei' processes thus 

 corresponding to those antero-ventral processes of the parachordals 

 of recent fishes that bound laterally the pituitary fossa. The 

 basisphenoid of these fishes would thus apparently correspond to 

 a bone formed by the fused prootics of recent fishes. 



There are thus several features in the cr-anial anatomy of these 

 fishes that indicate that the so-called dorsal process of the basi- 

 sphenoid corresponds to some part of the postorbital process of 

 recent fishes, and not to the basipterygoid process, the most 

 important being that the process gives articulation to a processus 

 metapteiygoideus metapterygoidei and not to a processus basalis, 

 that it has no supporting relations to the alisphenoid, and that 

 the truncus maxillo-mandibularis trigemini quite certainly ran 

 outward anterior to it, between it and the alisphenoid. The 

 basisphenoid of these fishes is such a peculiar bone that all com- 

 parisons with recent fishes are more or less tentative, and there 

 is one further feature in their cranial anatomy that seems equally 

 peculiar. In all i^ecent bony fishes (Orossopterygii, Holostei, 

 Teleostei) it is, so far as I know, an invariable rule that the dorsal 

 section of the hyomandibular latero-sensory canal runs upward 

 toward, or actually to, a postotic portion of the main infraorbital 

 latero-sensory canal, while in the Plagiostomi this canal turns 

 forward and falls into the infraorbital portion of the main infra- 

 orbital canal at the ventro-posterior corner ©f the orbit. Now, 

 whether correlated to the course of this canal or not, it is a fact 

 that, in all the former fishes the hyomandibula articulates with 



