1895] ^*^ * [Cope. 



Vertebrata. The structure of its temporal roof is essentially that of the 

 stegocephalous Batrachia, while the various postorbital bars of the 

 amnlote Vertebrata are explained by reference to the same part of its 

 structure. 



The relations of the opisthotic and paroccipital bones in the later orders 

 are apparently explained by their structure in the Cotylosauiia. I have 

 identified the element which intervenes between the exoccipital and the 

 supratemporal in the Lacertilia with the distal part of the opisthotico- 

 paroccipital element in the Testudinata, retaining for it the name of 

 paroccipital. Thinking that I have confounded this element with tlie 

 opisthotic, Baur differs with me,* and calls the element in question the 

 squamosal. My belief in the position taken was based on the character of 

 this region in the Pythonomorpha, where the paroccipital extends 

 proximad to the petrosal, and nearly or quite to the opisthotic, which it 

 does not do in the Lacertilia, a fact not sufficiently regarded by authors 

 generally. I assumed, in consequence of this structure in the Pythono- 

 morpha, that the single element in the Testudinata which extends from 

 the supraoccipital to the quadrate, really includes two elements, the 

 opisthotic proximally and the paroccipital distally. This view is confirmed 

 by the fact that the two elements in the Cotylosauria are distinct. I have 

 been able to locate the semicircular canals definitely in the genera 

 Empediasf and Chilonyx, and to fix the position of ihe opisthotic, petro- 

 sal and epiotic bones. In adults these elements are coo^sified, but they 

 appear to be separated by suture from the exoccipital and from the par- 

 occip.tal. The latter is a long bone and supports the quadrate exclusively, 

 the exoccipital being separated by a considerable interval. The form of 

 the quadrate is that characteristic of theromorous Reptilia and Batrachia. 

 It is distal in position and short, not reaching the cranial box by its proxi- 

 mal extremity, with which it is connected by the paroccipital, and by the 

 elements of the temporal roof. 



The palatal elements in this order are more or less in contact on the 

 middle line, and the pterygoids diverge abruptly from this point, and 

 return to the quadrate. The occipital condyle is single, and does not 

 include exoccipital elements (unknown in Elginia). 



Intercentra are present in Pariasauiidse, Diadectidse and Pariotichidae, 

 and they are wanting in Elginiidae. The hyposphen-hypantrum articu- 

 lation is present in the Diadectidae, but is wanting in the Elginiidae and 

 Pariasauridae. 



The scapular arch is best known in Pariotichidae, Pariasauridae:]: and 

 Diadectidae. In the two former there is a T-shaped sternum, over which 

 are applied the median extremities of the clavicles ; and there are a well- 

 developed coracoid and praecoracoid. In Diadectidae || (probably genus 

 Erapedias) the episternura is articulated by suture with the clavicles. 



* American Naturalist, 1895, p. 998. 



t Proceeds. Amer. Philos. Soc, 1885, p. 231. 



t Seeley, Philos. Trans. Roy. Soc. London, 1888, p. 89 ; 1892, p. 334. 



I Cope, Proceeds. Amer. Philos. Soc, 1883, p. 635. 



