A good look at the base of the Therapsida (Fig. 1) reveals several interesting convergences (taxa that look alike superficially). Prior studies used shared traits between Dimetrodon in the Sphenacodontia and Eotitanosuchus in the Therapsida to support a link between sphenacodonts and basal therapsids, but that is not supported in the large reptile tree. Previously unnoticed, the sphenacodont, Haptodus, bears a superficial resemblance to a basal therapsid, Stenocybus, which may be a juvenile of a longer snouted form.
Dimetrodon and Eotitanosuchus
The dorsally convex skull in Dimetrodon and Eotitanosuchus (Fig. 1) appears to be a shared trait, but it is convergent in the large reptile tree. Both share a deeply convex maxilla and similar proportions in the orbit vs. rostrum. Both share a posteriorly sloping cranium, broken by elevated lateral temporal fenestrae in Eotitanosuchus. The jawline was shallowest beneath the orbit in both taxa.
Haptodus and Stenocybus
The short rostrum/large orbit skull shapes of Haptodus and Stenocybus (Fig. 1) were arrived at via convergence, as are the relatively short canines and convex rostrum. A larger suite of traits and several intervening taxa separate these two. Now, Stenocybus was never allied with Haptodus in the literature. Stenocybus was previously allied with anteosaurid dinocephalians (stepped incisors a key trait), but here it nests more parsimoniously with hipposaurids.
The strength of the large reptile tree (lots of taxa, lots of traits) separates these convergent taxa. In a smaller study they would have nested closer together.
As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.
Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.