40 PISCES. 



Numerous Palaeozoic teeth and fin-spines have been ascribed 

 to this order, but only on account of their close resemblance to 

 the corresponding hard parts of certain of its existing members. 

 Even the nearly complete example of the Carboniferous Sphena- 

 canthus in the British Museum, proves little more than that 

 the spines were placed in front of the two dorsal fins ; the 

 supporting cartilages of the fins cannot be satisfactorily ob- 

 served. There is, however, little doubt that many of the 

 Carboniferous and Permian Elasmobranchs were closely related 

 to the surviving Port Jackson Shark (Cestracion). 



Entire skeletons of undoubted Selachii are known first from 

 the Lower Lias (Palceospinax), and it is interesting to observe 

 how varied is the degree of development of the vertebral centra 

 even in genera most closely resembling each other in all re- 

 maining features. For example, the genus Hybodus (fig. 34, 

 p. 45) has a persistent notochord, without calcifications in the 

 notochordal sheath, and is known to range at least from the 

 Lower Lias to the Wealden. The Lower Liassic Palceospinax 

 scarcely differs from it except in the smoothness of the dorsal 

 fin-spines and shagreen; but its vertebral centra are always 

 conspicuous as constricted cylinders (the "cyclospondylic" type). 

 The Cretaceous Synechodus, again, is almost identical with 

 Palceospinax ; but its vertebral rings are strengthened by peri- 

 pheral calcifications which are chiefly disposed in longitudinal 

 ridges or radiating plates (the " asterospondylic " type). Among 

 Selachii, therefore, the degree of development of the vertebral 

 centra is shown to be of only secondary importance. The 

 manner of the specialization seqms to be of more fundamental 

 significance ; for it is commonly admitted that the sharks 

 (Selachoidei) and the rays (Batoidei) form two approximately 

 natural groups the one tending towards agility in swimming, 

 the other towards expertness in feeding on the bottom and 

 the fully-formed vertebral centra are strengthened in a different 

 way in these two groups. At the same time it must be remarked 

 that the classification is merely provisional; for there is ob- 

 viously much uncertainty as to the systematic position of those 

 forms with persistent notochord and simple vertebral rings, and 

 many of the extinct types are too imperfectly known as yet to 

 admit of satisfactory discussion. 



