The relationships of placoderm fishes

Aspects of placoderm morphology are considered in the light of current competing hypotheses of placoderm relationship. It is suggested that the macromeric dermal skeleton in the skull roof, cheek and shoulder girdle resulted from a single morphogenetic change, and that lack of correspondence with the osteichthyan bone pattern indicates independent development. The presence of a high scapular process in some placoderms shows that resemblances within the group to the reduced scapulocoracoid of osteichthyans is a parallelism, whilst its different relation to the prepectoral dermal spinal element in placoderms and acanthodians shows that these spines are not homologous. The pelvic claspers of ptyctodontids were carried on a cartilaginous basipterygium, and as far as is known differ from corresponding structures in chondrichthyans only in their macromeric rather than micromeric dermal investment. The placoderm eyestalk has similar morphological relations to that of elasmobranchs, and is either a synapomorphy of these groups or a gnathostome symplesiomorphy. Extrinsic eye muscle arrangement in placoderms may have been primitive in the anterior insertion of the internal rectus, and posterior insertion of the superior oblique. The persistence of an ethmo-otic fissure is primitive for gnathostomes, although the precerebral fontanelle of chondrichthyans may be a remnant of this feature. In a reinterpretation of the rhenanid Jagorina the ‘antorbital cartilage’ and ‘palatoquadrate’ of earlier descriptions are determined as the palatoquadrate and meckelian cartilage respectively. The placoderm epihyal retained its connection with the ceratohyal, and the separate opercular cartilage supporting the dermal operculum against the braincase is a placoderm autapomorphy. Character analysis with reference to four competing schemes of placoderm interrelationships suggests that the suborbital plate, stenobasal fin, parasphenoid, eyestalk and opercular cartilage were primitively present, but that dermal laminae supporting the braincase and toothplates developed within the group. The existence of a dental lamina and serial tooth replacement is uncertain. It is concluded that placoderms are either the sister-group of all other gnathostomes, or of chondrichthyans alone. A sister-group relationship with osteichthyans is rejected.