274 PALEONTOLOGY. 



of reptiles and mammals, in having the component elements 

 of many parts of their skeleton permanently separable from 

 each other ; a circumstance which renders the study of their 

 osteology, especially that of the skull, a matter of much 

 difficulty. 



Many of the numerous bones which the student meets 

 with in the cranium of fishes, are detached elementary parts 

 of the cranial vertebra), which are united to their neighbouring 

 elements in the cranial vertebrae of reptiles and quadrupeds. 



The bodies of the spinal vertebrae are formed of concentric 

 layers, which are broadest at the circumference, and are 

 narrowest at the centre. They are therefore doubly con- 

 cave, and are readily identified in a fossil state by this 

 character alone. 



The skin of fishes is covered with scales, which contain 

 about forty-six per cent, of the 'phosphate of lime ; so that 

 scales in reality resemble bone in composition. In this 

 respect they differ essentially from the horny covering of 

 reptiles. In consequence of the scales of fishes being found 

 preserved when all other parts of their structure have- 

 disappeared, a new arrangement of this class has been 

 made, which is based on the organisation of their dermal 

 covering. 



The teeth of fishes present great varieties, as regards their 

 number, form, and structure, so that no general description 

 can possibly convey a correct idea of their dentition ; and as 

 our limits prevent us from entering into details, we must 

 direct the student to the best sources of information. Pro- 

 fessor Owen's "Lectures on Comparative Anatomy," vol. ii. 

 lect. 9, contains a valuable summary on the subject, and 

 more elaborate details will be found in his great work on 

 Odontography. Professor Eymer Jones's article, Pisces, 

 in the " Cyclopaedia of Anatomy," ought likewise to be 

 consulted. But the student would do well to procure very 

 thin vertical and transverse sections of the teeth of the 

 different orders and families of fishes, mounted in Canada 

 balsam, which he ought to study attentively with the micro- 

 scope. Figures and descriptions are very good, but the 

 natural objects are far better ; and these preparations will at 

 the same time form good standards for the comparison of fossil 

 specimens. The Cuvierian classification of fishes, although 



