12 ILUNOIS BIOLOGICAL MONOGRAPHS [12 



that condition than to the condition found in the chondrocranium of Amiurus. 

 The epiphysial bar, which is found again among the Teleosts in Gasterosteus ^ ^ 

 (Swinnerton 1902), the Characinidae (Sagemehl 1885), the Cyprinidae (Sage- 

 mehl 1891) and other Siluridae (Pollard 1895), is the striking feature of a com- 

 parison between these two forms, as Polypterus is specialized in so many other 

 features, that it differs from even the majority of the ganoids and has been 

 mentioned as representing the ancestral stage of the StegocephaH. 



In Gymnarchus niloticus (Assheton, 1907) the early chondrocranium of 

 which has been very briefly described from Budgett's collections and notes, 

 there is a long sht-like foramen in the wall of the orbit jvist anterior to the otic 

 capsule and a smaller anterior foramen separated from this for the passage of 

 the opticus. There is also a transverse bar similar to the epiphysial bar of 

 Amiurus and Polypterus, connected however with the anterior ethmoidal roof 

 by a median longitudinal bar of cartilage, which divides the anterior fontanelle 

 into two parts. This may be the representative of a group in which the reduc- 

 tion from the soHd cartilaginous roof of the lower ganoids is taking place, a 

 continuation of which would produce the condition found in Polypterus and 

 Amiurus. 



In Salmo salar (Parker 1873; Gaupp 1906) a cartilaginous bar extends on 

 each side of the cranium above the orbit, from the anterior end of the otic 

 capsule to the posterior margin of the tegmen cranii, which extends back 

 as far as the middle of the orbit. Posteriorly, between this bar and the otic 

 capsiile, the branches of the facialis and the jugular vein pass through separate 

 foramina in the cartilage. The bar itself (called by Gaupp the taenia mar- 

 ginahs) extends anteriorly above the trigeminus and does not enclose any of 

 its branches. Here, as in Amiurus, the large fenestra left in the wall between 

 this cartilage and the trabeculae cranii is closed by membrane, through which 

 the nerves issuing from the cranium pass. In Amiurus, as we have seen, both 

 the fifth and seventh nerves issue through this membrane and with them the 

 external carotid. In Sahno, as in Acanthias, these two nerves are separated 

 from each other by a bar of cartilage, thus showing a nearer degree of relation- 

 ship to the Selachian than to the Siluroid condition. The closer resemblance 

 of Salmo to Acanthias is again evidenced by the presence of a well-developed 

 tegmen cranii, not found in Amiurus. 



In Ceratodus and the Urodeles the alisphenoid cartilage is fused ventrally v 

 with the trabecula during the first stages of development and later unites with 

 the ethmoid and otic regions. As in the Ganoids and Teleosts, its anterior ^ 

 end and the trabeoila enclose the optic nerve. Since the posterior ends of 

 these cartilages enclose the trigeminus and the faciaHs, there is no question 

 as to their homology with the alisphenoid and trabecular cartilages of these 

 groups. Gaupp (1906) calls the former the crista trabeculae, thus adding to 

 the confusion of names. 



