338 M. Agassiz on Fossil Fishes. 



means of various dismemberments, separations, and combina- 

 tions, all the forms of the cranium were referred to the ver- 

 tebrae, by admitting that the number of pieces was invariably 

 fixed in every head, and that all the vertebrata, whatever 

 might be their organization in other respects, had in their 

 heads the same number of points of ossification. At a later 

 period, what was erroneous in this manner of regarding the 

 subject was detected ; but the idea of the vertebral composi- 

 tion of the head was still retained. It was admitted as a 

 general law, that the cranium was composed of three primi- 

 tive vertebrae, as the embryo is of three blastodermic leaflets ; 

 but that these vertebrae, like the leaflets, existed only ideally, 

 and that their presence, although easily demonstrated in cer- 

 tain cases, could only be slightly traced and with the greatest 

 difficulty in other instances. The notion thus laid down of 

 the virtual existence of cranial vertebrae did not encounter very 

 great opposition ; it could not be denied that there was a cer- 

 tain general resemblance between the osseous case of the brain 

 and the rachidian canal ; the occipital, in particular, had all the 

 characteristic features of a vertebra. But whenever an attempt 

 was made to push the analogy further, and to determine 

 rigorously the anterior vertebrse of the cranium, the observer 

 found himself arrested by insurmountable obstacles, and he 

 was obliged always to revert to the virtual existence. 



" In order to explain my idea clearly, let me have recourse 

 to an example. It is certain that organized bodies are some- 

 times endowed with virtual qualities, which, at a certain 

 period of the being's life, elude dissection, and all our means 

 of investigation. It is thus, that, at the moment of their 

 origin, the eggs of all animals have such a resemblance to 

 each other, that it would be impossible to distinguish, even by 

 the aid of the most powerful microscope, the ovarial egg of a 

 craw- fish for example, from that of true fish. And yet who 

 would deny that beings in every respect different from each 

 other exist in these eggs "? It is precisely because the difier- 

 ence manifests itself at a later period, in proportion as the 

 embryo develops itself, that we were authorized to conclude, 

 that, even from the earliest period, the eggs were different ; that 

 each had virtual qualities proper to itself, although they could 



