Phylogeny of the Tdeostomi. 343 



" has a dermal armour resembling that of the Antiarcha in 

 minute structure and a ventral plastron quite similar to that 

 of the latter. The lateral appendages, however, instead of 

 being complex and movable, are simple and fixed." Never- 

 theless Traquair* has given good reasons for regarding 

 Acanthaspis as a Coccosteid, and it would even seem that the 

 fixed spinous appendage may be diagnostic of that family. 



So tar, then, Coccosteus has been shown to resemble the 

 more generalized Crossopterygii in the arrangement of the 

 bones of the cranial roof, and reasons have been given for 

 regarding the Asterolepidas as closely related to the Cocco- 

 steidae "j". 



What, then, of the peculiar pectoral limb of the Astero- 

 lepidae? It has been sometimes assumed that this is not 

 homologous with the pectoral fin of other fishes, but evidence 

 in support of this assumption has not been forthcoming. 

 Bashford Dean even goes so far as to say that these ap- 

 pendages are now known to be the lateral head-angles [?of 

 Cephalaspis] produced and jointed for locomotion. This 

 extraordinary theory is evidently based on a complete mis- 

 conception as to the position of the Asterolepid limbs, and so 

 needs no discussion. Smith Woodward seems to think that 

 the fixed spinous appendage of Acanthaspis in some way 

 supports the view of the independent origin of the Asterolepid 

 pectoral, and I suppose therefore that he regards it as a stage 

 in the development of the latter. Personally, I am unable*to 

 imagine that a fixed spine could possibly give rise to a jointed 

 Arthropod-like limb with internal muscles. In fact, the 

 structure of such a limb, articulated to an anterior plate of 

 the body, in which latter is a large foramen, indicating that 

 tendons, blood-vessels, and nerves passed to the muscles of 

 the limb from the body, postulates for me an unarmed ancestor 

 with a muscular limb already developed. Just as the 

 similar limbs of the Crustacea are generally held to have been 



* Geol. Mag. (3) x. 1893, p. 148. 



f The reasons which have been given for regarding Coccosteids and 

 Asterolepids as not related are (1) the more vascular bone of the latter 

 (2) the presence of specialized paired tins in the former, and (3) the well- 

 ossified jaws of the Coccosteids. With regard to these, the resemblances 

 in the structure of the bony plates are very remarkable, and the differences 

 are evidently not well marked, or there could be no doubt as to the 

 position of a genus after the minute structure of the bone had been ascer- 

 tained. The Asterolepid pectoral is surely specialized enough and it is 

 purely gratuitous to assume its non-homology with that of oilier hshes 



As to the non-ossified lower jaw of the Asterolepids, instances are not 

 wanting m Chondrostei and Teleostei of degeneration of membrane-bones 

 or of the reversion of a bone to its primitive cartilaginous condition 



