ON CERTAIN CROSSOPTERYGIANS. 1245 



orbitotemporal region, apart from possible fontanelks, must have 

 been cartilaginous (cf. text-fig. 2). 



Each of the partes laterales has a somewhat concave posterior 

 edge, which faces postero-dorsa]]3^ This edge also is so developed 

 that in part it cannot have been in contact with cartilage, but 

 must have been free, forming the anterior boundary of a, great 

 foramen (*>, text-lig. 1). This foramen must, as comparisons 

 Avith Diplocercides make clear, be the trigeminus foramen, which 

 thus has been situated in the dorsal part of the synchondrosis 

 between the sphenoid and the ossifications in the labyrinth 

 regioti {cf. Stensiii, 1922 a). 



Along the posterior edge each pars lateralis has on its lateral 

 surface a I'ather broad but well-marked " wulst," which extends 

 from the upper edge of the pars lateralis downwards and 

 somewhat backwards to the upper end of the basipterygoicl 

 pi'ocess. This "wulst" {alsjjh, text-fig. 1; PI. I. figs. 1-4), 

 Avhich I have called the alisphenoid wulst (Sbensio, 1922 a, 

 pp. 180, 196-197 ; 1922 h), is perforated close to the dorsal end 

 bv a horizontally running canal {c.ophiJi.lat, text-fig. 1 ; PI. I. 

 figs. 1-4), which has its posterior opening on the posterior and 

 its anterior opening on the anterior side of the Avulst. At the 

 anterior edge of the ventral part of the same wulst, some distance 

 above the basipterygoid process, there is found the outer opening 

 of a caiial (V^, text-fig. 1 ; PI. I. figs. 1-4), leading backward, 

 inward, and somewhat mediall}^ to the cranial cavity. 



The division of the cranial cavity enclosed in the sphenoid is 

 compara,tively narrow throughout its extension (text-figs. 1,2; 

 PL I. fig. 3; cf. also Stensiti, 1918 &, pi. v. figs. 2. 3). While 

 its height decreases gently forward, its width, on the contrary, 

 seems rather to grow somewhat larger towards the anterior end. 

 Owing to the considerable heigiit of the pars basalis of the 

 sphenoid, the division of the cranial cavity in question, with the 

 exception of the large fossa hypophyseos, is situated totally in the 

 dorsal half of the neuroeranium, and it is evident that the 

 division next behind in the anterior half of the lab3"rinth region 

 also had a similar high position. 



From the shape of the division of the cranial cavity just 

 described, we can conclude with great probability that the 

 telencephalon was long ai:id rather high, and that the ventral 

 pa.rts of the diencephalon Avere strongly developed Thus the 

 anterior divisions of the brain of the fish must have i-esembled 

 the corresponding ones in the recent Ceratodus, at least Avith 

 regard to their general shape {cf. Bing & Burckhardt, 1905). 



The trigeminus branches proper, except the r. ophthalmicus 

 profundus and the lateralis branches accompanying these (or 

 perhaps their roots), have left the cranial caAdty through the 

 foramen already referred to, Avhich must be supjDosed to have 

 existed in the upper part of the synchondrosis between the 

 sphenoid and the bone or bones following next behind in the 

 labyrinth region. The position and probable size of this foramen 



