1857.] OWEN — LAOPHIS CROTALOIDES. 197 



from the back part of the under surface of the centrum, and a ridge 

 is continued from the fore-part of its base, gradually expanding as it 

 subsides, to the lower rim of the anterior articular cup of the cen- 

 trum, below which it makes no projection. 



With the small number of vertebrae of this fossil Serpent trans- 

 mitted for comparison, this character alone will not warrant a con- 

 clusion in favour of its affinities to the Crotalus, since they may all have 

 come from the anterior region of the trunk, where a similar hypapo- 

 physis is developed in the other genera of Serpents above mentioned. 



In point of size, the vertebrae from Salonica agree with the middle 

 trunk-vertebree of a Python tigris of 1 1 feet in length ; but they are 

 manifestly of a different genus and family of Serpents. 



Independently of the ridge continued forwards from the hypapo- 

 physis, there is a process, figs. 2 and 3, d', flattened behind, produced 

 downwards and forwards below and beyond the articular cup, from 

 the under part of the diapophysis, d, as in Crotalus, fig. 4. A small 

 and well-defined articular tubercle, d, projects from near the upper 

 part of the diapophysial surface, also as in Crotalus ; and the upper 

 part of the diapophysis is produced and bent outwards at d", beneath 

 the anterior zygapophysis, z, with which it is blended, and which 

 it seems to sustain, like a cantilever in roof-architecture ; the obtuse 

 point of the process extending outwards (fig. 3, d''), about a line 

 beyond the flat articular surface of the zygapophysis, z. A minute 

 tubercle beneath the outer end of that articular surface indicates the 

 homologous point in Python ; but the production of the pointed pro- 

 cess is not found in any of the vertebrae of the constricting Serpents. 

 It characterizes the vertebrae of the genera Coluber (fig. 7), Naja, 

 Crotalus, and Hydrus ; and the degree to which the process in 

 question is produced in the fossil Serpent is intermediate between 

 Coluber, where it is longest, and Crotalus, where it is less developed. 



The zygosphene, zs, is relatively broader in proportion to its depth 

 in the fossil Serpent (fig. 3) than in Python (fig. 6), and is slightly 

 excavated anteriorly, and without a median tubercle ; it more resem- 

 bles the form and proportions of that part in Coluber and Crotalus. 

 The posterior border of the neural arch, n, describes a gentle curve 

 convex backwards as it descends from the base of the neural spine to 

 the posterior zygapophysis, z\ The similarly-sized fossil vertebrae 

 of the Palceophis toliapicus offer a striking contrast with the Salonica 

 Ophidiolite in the configuration of this part, which is produced into 

 an angle (fig. 1, n) ; the posterior zygapophyses are more produced 

 outwards than in the anterior trunk-vertebrae of Python, resem- 

 bling those of the middle trunk-vertebrae of Python, which are 

 without the long hypapophysis. The base of the neural spine is 

 coextensive with the neural arch ; no well-defined part of the zygo- 

 sphene projects beyond it, as in Python, Naja, and Hydrus. In 

 this respect the Salonica Ophidiolite resembles Crotalus ; the neural 

 spine is more compressed than in Python, and its posterior border 

 slopes more backward ; but it is not sufficiently entire in any of the 

 specimens to permit this comparison to be pursued with advantage. 

 The process from above the concave part of the expanded articular 



p 2 



