When first described,
was considered an enigma because of the incomplete preservation of the holotype material; however, autapomorphic features of the skeleton allow for unambiguous referral of material to this taxon (Hunt et al. 2002;
Polycyn et al. 2002;
Small and Downs 2002;
Parker and Irmis 2005;
Hunt et al. 2005;
Irmis 2005). Newly collected material from Petrified Forest National Park described here suggests that
is a non-archosaur archosauriform, a finding that is further supported if the Ghost Ranch material is referable to this taxon (Small and Downs 2002;
Hunt et al. 2002). Evidence provided by
Small and Downs (2002) for the inclusion of
within Archosauriformes include thecodont tooth implantation and an ossified laterosphenoid; however, these authors also mention the possible presence of a thyroid fenestra in the pelvis, which is a plesiomorphic character found in some
basal archosauromorphs and squamates (Dilkes 1998).
Comparisons between the type material from the Blue Mesa Member of the Chinle Formation and the referred material from the Petrified Forest Member of the Chinle Formation reveal very few morphological differences. In the holotype material (PEFO 2427) the internal tuberosity of the humerus is pointed, similar to one of the Petrified Forest Member specimens (PEFO 33978), whereas in another specimen (PEFO 34035) from the Petrified Forest Member the internal tuberosity is gently rounded. The sacral centra of PEFO 2427 have double ventral keels, whereas in PEFO 34035 it is a single, broad keel. However, because of the fragmentary preservation of this material and the small sample size, these differences do not clearly represent unambiguous autapomorphies; therefore all of the referred material is assigned to the type species,
. What remains to be determined is if any differences exist between the Ghost Ranch and Stinking Springs specimens. If one or both of these specimens show no distinct differences with the poorly preserved holotype material, but are significantly different from each other and/or the Petrified Forest Member material described here, then the holotype material must be considered non-diagnostic and
would be a nomen dubium. This would require new taxonomic assignments for all
of the referred material (Parker and Irmis 2005).