[19] " THE OSTEOLOGY OF AMIA CALVA. 765 



that the formation of the odontoid process out of the body of the atlas 

 iu the Amniota only begins among the reptiles.^^ In higher fishes it is 

 very generally found that the anterior aspect of the first vertebra is not 

 excavated, but slightly convex. Now, though it seems to me to be im- 

 probable that the conditions in Amphibia can be traced directly to these 

 structures in fishes, yet here is a state of things that can be considered 

 parallel with that of the Amphibia. 



An explanation for the singular fact that in the higher fishes inde- 

 pendent vertebra are co-ossified with the occiput is not difficult to find, 

 and I believe the reason for this condition is to be found in the way 

 and method in which the parasphenoid makes its appearance. 



It has been fully and conclusively shown by Hertwig that teeth can 

 be discovered upon all the bones of the buccal cavity, which arise from 

 these osseous plates through sockets in their substance, and that the para- 

 sphenoid forms no exception to this rule, although teeth are found upon it 

 far more seldom than on the other bones of the mouth. If we now know 

 that the appearance of teeth in the Selachians is not confined to the 

 cavity of the mouth, but that tbey also extend upon the mucous mem- 

 brane of the fore-gut, as far as the gill slits, thus reaching far below the 

 anterior extremity of the vertebral column, then the supposition will not 

 be startling that the parasphenoid originally did not confine itself to 

 the basis cranii, but extended far behind it upon the vertebral column. 



In fact, we meet with the parasphenoid occupying this very position 

 in those fishes in which bone first begins to appear, in the cartilaginous 

 Ganoids, and in the Dipnoi. As already known, the parasphenoid of 

 Stohrs does not confine itself to the base of the true skull, but extends 

 backwards to be applied to the inferior surface of the centra of about 

 7 or 8 vertebrae. According to Wiederscheim this is the arrangement 

 in Polypterus, and Glinther tells us that it also occurs in Ceratodus, only 

 in these fishes the number of vertebrse covered by the parasphenoid is 

 fewer. This also must have been the state of things in the direct ances- 

 try of the existing bony Ganoids and Teleostei. Now, after the para- 

 sphenoid had ceased to be a tooth -bearing bone of the cavity of the 

 mouth, a curtailment from behind took place, and at the same time a 

 reduction in number and consolidation of the vertebrae resting upon this 

 bone, which was already firmly connected with the cranium, set in, to 

 replace the latter, a transformation the last traces of which can still be 

 seen in bony Ganoids and Teleosteans. 



The region of the labyrinth ^^ is bounded posteriorly by the foramen 



''I Gegeubaur, Grundzuge der vergl. Anaiomie, 2 Aufl., page 615. 



3' Labyrinth region, the term here used, applies more particularly to that space as 

 seen in the Teleostei and bony Ganoids, which, by the way, it does not entirely in- 

 clude, as the labyrinth in these fishes generally extends beyond the confines given ; 

 moreover, all the bones enumerated by me as belonging to the occipital region may, 

 under certain circumstances, serve for the inclosure of j)arts of this area. So I have 

 retained the term simply to avoid a new name. 



