Raising my hand to proclaim a nesting error
Earlier (now trashed) I recovered Echinerpeton at the base of the Synapsida and Diapsida, but those elongate dorsal spines seemed odd at that node. Then I noticed that all other pelycosaurs had teeth only in front of the orbit. The skull is largely missing, so there’s no harm in shifting the jaws forward a bit. And suddenly Echinerpeton made more sense.
Echinerpeton intermedium (Reisz 1972), Late Carboniferous, 308 mya. Reisz (1972) tentatively classified Echinerpeton as an ophiacodontid in his initial description, and in 1986 he considered it an indeterminate “pelycosaur“. Benson (2012) could not nest Echinerpeton with certainty, perhaps because he used the wrong outgroups and mistakenly included caseasaurs because he followed tradition without the benefit of a large gamut reptile tree like we have here (Fig. 2).
Here Echinerpeton nests with Secodontosaurus. The snout was long because the last maxillary tooth was in front of the orbit. The maxilla was straight while the dentary was concave dorsally. Both were filled with long teeth.
The dorsal spines were long, but not as long as those of Secodontosaurus.The scapula was small and both the humerus and femur were short and slender. The ankle bones were round elements. Together these point to an aquatic, rather than a terrestrial niche. So Echinerpeton was a crocodile-like sphenacodont pelycosaur.
Benson RBJ 2012. Interrelationships of basal synapsids: Cranial and postcranial morphological partitions suggest different topologies. Journal of Systematic Palaeontology: 601-624.
Reisz R 1972. Pelycosaurian reptiles from the Middle Pennsylvanian of North America. Bulletin of the Museum of Comparative Zoology 144 (2): 27–62. online here