THE OSTEOLOGY OP AMIURUS CATUS. 307 



anterior ray of the pectoral fin also points to the lapse of a consider- 

 able period of time, during which small successive changes have been 

 wrought, and the extent of the modifications of the dorsal fin for the 

 same purpose point to the same conclusion. Other evidences of a 

 similar nature are to be seen in the absence of any neural arches 

 corresponding with the haemal processes which support the rays of 

 the caudal fin, and in the complete abortion of the radialia of the 

 ventral fin. 



All these latter points are, however, subordinate to the first in 

 determining the relative position of Arniurus. Since the course of 

 development, as is shown both by the ontological history of any 

 form, and by the study of the various vertebrate groups, leads from 

 a purely cai'tilaginoua to a purely osseous skeleton, the amount of 

 cartilage present in the skeleton of any fish is in indirect relation to 

 the extent of its development. This character is necessarily less 

 subject to the modification of external conditions than other parts, 

 so that even though certain of these may undergo great specialization, 

 yet if a considerable amount of cartilage be present in the skeleton, 

 the form under consideration must be considered as standing com- 

 paratively low in the group. The Lophobranchiates, for instance, 

 have undergone modifications, even more striking than those of 

 Arniurus, but since the relative amount of cartilage in the skull is 

 greater, and the parts modified may all be readily influenced by the 

 -conditions of existence, the members of this group must be placed 

 lower among the Teleosts than Arniurus. 



In conclusion, a few words concerning the process of ossification. 

 From what has already been said in this paper, it will be seen that 

 what may be termed several modes of ossification are present. We 

 have, in the first place, the deposition of the bone in general connec- 

 tive tissue, forming certain of the ' Deckknochen,' and the bones 

 around the mucous canals ; we have, secondly, cement-bone, as in the 

 premaxillse and dentary ; and we have, thirdly, perichondral bone, as 

 in the prootic, palatine, etc. It has also been shown that all these 

 forms of bone formation pass into one another perfectly, no dividing 

 line marking the termination of one form and the commencement of 

 another. Not only, however, do they thus pass into one another, 

 but they also replace each other. This is very evident in the case of 

 the frontal, maxillae, vomer, parasphenoid and mucous canal bones. 

 At one time these bones were probably formed by the union of the 



