327 



those large perforations of the solum cavi praecerebralis that give rise 

 to the three-limbed rostral process, as Gegenbaur and Sbwertzoff 

 both maintain, seems to me wholly improbable; these latter perforations 

 probably being simply defects in the cartilage due to the presence of 

 the large number of ampullae found in this region in these fishes. 



In fishes other than selachians the fenestra praecerebralis and 

 cavum praecerebrale are both found in the Batoidei (Gegenbaur), 

 and it would seem evident that they must also both be found, some- 

 what modified, in the adult Cyclostome. In the adult ganoid and 

 teleost the fenestra praecerebralis is not found in any of the forms 

 that I am acquainted with, but remnants of the cavum praecerebrale 

 are found as the mesethmoidal fat cavities of Parker's (1873) and 

 my own (Allis 1903, 1910) descriptions of certain teleosts. In the 

 young of Lepidosteus the fenestra praecerebralis is apparently 

 represented in those two perforations, one on either side, of the roof 

 of the chondrocranium that lie anterior to the paraphyseal bar of 

 Veit's (1911) descriptions, and those perforations, although they can 

 not represent the entire anterior cranial fontanelle of Sagemehl's 

 descriptions of certain of the Characinidae and Cyprinidae, may 

 perhaps be included in it. 



The ectethmoidal chamber is represented in Pristis perrotteti in 

 some part of the canal that is said by Hoffmann (1. c. p. 251) to give 

 passage to the vena faciahs ventrahs, and it is probably also found in 

 similar form in others of the Batoidei. In ganoids and teleosts it has 

 apparently been either largely suppressed or incorporated in the orbit, 

 a small dorso-mesial portion alone persisting and, together with the 

 related orbito-nasal and basal communicating canals, being represented 

 in the fenestra orbito-nasalis of my descriptions of Amia (Allis, 1897); 

 a fenestra that becomes so large in certain teleosts that the nervus 

 olfactorius lies for a considerable distance exposed in the orbit. 



The nasal fontanelle is not found in ganoids, the subnasal plate 

 in these fishes forming a continuous floor beneath the nasal pit. In 

 teleosts this fontanelle would seem to find its homologue in the large 

 incisure that lies between the anterior portion of the body of the 

 ectethmoid and the lateral arm of that bone and that is so well re- 

 presented in Scorpaena (Allis, 1910). This incisure hes between the 

 anterior and posterior palatine articular surfaces, which latter surfaces 

 thus find their homologues in the ventro-lateral edges of the pre- and 

 post-fontanelle portions of the subnasal plate of Chlamydoselachus. 



