M. Agassiz on Fossil Fishes. 339 



not be discovered by our senses. If, on the contrary, any 

 one should find two eggs perfectly alike, and should observe 

 two beings perfectly identical issue from them, he would 

 greatly err if he ascribed to these eggs different virtual quali- 

 ties. It is, therefore, necessary, in order to be in a condition 

 to suppose that virtual properties peculiar to it are concealed 

 in an animal, that these properties should manifest themselves 

 once in some phase or other of its development. Now, ap- 

 plying this principle to the theory of cranial vertebrae, we 

 should say that if these vertebra) virtually exist in the adult, 

 they must needs shew themselves in reality, at a certain period 

 of development. If, on the contrary, they are found neither 

 in the embryo nor the adult, I am of opinion that we are en- 

 titled likewise to dispute their virtual existence. 



" Here, however, an objection may be made to me, drawn 

 from the physiological value of the vertebrae, the function of 

 which, as is well known, is, on the one hand, to furnish a 

 solid support to the muscular contractions which determine the 

 movements of the trunk, and, on the other, to protect the 

 centres of the ner\^ous system, by forming a more or less solid 

 case completely around them. The bodies of the vertebrae are 

 particularly destined to the first of these offices, the neura- 

 pophyses to the second. What can be more natural than to 

 admit, from the consideration of this, that, in the head, the 

 bodies of the vertebrae diminish in proportion as the moving 

 function becomes lost, while the neurapophyses are consider- 

 ably developed for protecting the brain, the volume of which 

 is very considerable, when compared with that of the spinal 

 marrow % Have we not an example of this feet in the verte- 

 brae of the tail, where the neurapophyses become completely 

 obliterated, and a simple cylindrical body alone remains i 

 Now, may it not be the case, that, in the head, the bodies of 

 the vertebrae have disappeared ; and that, in consequence, there 

 is a prolongation of the cord only as far as the moving"] func- 

 tions of the vertebrae extend % There is some truth in this 

 argument, and it would be difficult to refute it a priori. But 

 it loses all its force the moment that we enter upon a de- 

 tailed examination of the bones of the head. Thus, what 

 would we call, according to this hypothesis, the principal sphe- 



