Ordovician placoderm, Eriptychius, reconstructed in lateral view

Dearden et al 2023 provided
the first µCT images of the Ordovician placoderm, Eriptychius, which they published in dorsal and ventral views (Fig 1). Earlier those data were assembled using for a Bauplan a wide, flat, jawless placoderm, Arctolepis (Fig 1). That seemed to be a good idea at the time.

Today a lateral view of Eriptychius is presented
using DGS methods and a Shenacanthus Bauplan Fig 1). This Early Silurian taxon had a relatively narrow skull and thoracic shield that appears to be a better match for Eriptychius.

Since Eriptychius is broken and crushed into many small parts,
assembly is a best guess (Fig 1) based on similar parts found in better articulated more or less closely related specimens.

Earlier than the Early Silurian
there were already two clades of placoderms: those with a flattened skull, like Early Silurian Xiushanosteus (Fig 1), and those with a more plesiomorphic skull, represented here by Shenacanthus – now related to Eriptychius (Fig 1).

This morphological split in placoderms is the current subject of study
in the present effort to fully resolve the fish subset of the LRT. Flatter skulls on bottom-dwelling placoderms are clearly derived. This is based on the many eel-like ancestral taxa, like lamprey and hagfish. A problem arises because the weak-to-unknown jaw elements of the disc-like bottom dwellers, like Xiushanosteus (Fig 1) make them appear to be more primitive. Reversals, like convergence, make scoring more difficult.

Jaws were invented more than once.
Teeth were invented more than once.
Fins were invented more than once and lost more than once.

Thank you for your patience
and understanding as I wade through such problems in my quest for complete resolution.

References
Dearden RP et al 2023. The oldest three-dimensionally preserved vertebrate neurocranium. Nature https://doi.org/10.1038/s41586-023-06538-y

wiki/Xiushanosteus – not yet posted
wiki/Eriptychiida
wiki/Shenacanthus – not yet posted