132 BULLETIN or THE ESSEX INSTITUTE. 



ences in other respects. The absence of a palato-ptery- 

 goid cartilage is an especially noteworthy feature in this 

 connection which indicates the highly specialized nature 

 of this skull and renders it impossible to consider it a 

 very near approach to the ancestral Batrachian skull. 



The trabeculse are unlike both those of the Fishes and 

 those of the Batrachia. When their posterior ends are 

 compared with the chondrocranium of the former group 

 they seem rather to represent the supraorbital bands, 

 arising as they do from the antero-dorsal surface of the 

 otic capsules. But in passing forward, instead of curv- 

 ing up over the eye as supraorbital bands should do, they 

 curve downward and fuse at their anterior ends into an 

 ethmoid plate very much as occurs in the Batrachia. The 

 antorbital process arises from the dorsal margin of the 

 trabecula, a condition found in none of the other forms 

 studied ; and the ethmoid plate, instead of continuing 

 forward to form a floor beneath the nasal capsules, as is 

 the general method in both Fishes and Batrachia, curves 

 sharply upward at its anterior end and fuses with the dor- 

 sally situated posterior end of the nasal septum. The 

 possession of a remnant of the tegmen cranii is another 

 fish-like character. 



As was mentioned in the discussion of the chondro- 

 cranium of Necturus, there is some resemblance between 

 the nasal capsules of that form and those of Protopterus. 

 But, in view of the differences between the nasal capsules 

 of the various forms of Urodeles themselves and consid- 

 ering the many important points of difference in other 

 parts of the skull, it seems to me an entirely unwarrant- 

 able conclusion to assume any phyletic relationship be- 

 tween these two forms upon this account. 



Giinther has described the skull of Ceratodus as 

 consisting " of a completely closed inner cartilaginous 



