Reassessment of the Triassic archosauriform : neither runner nor biped, but hopper.

The six known specimens of and casts made from their negative impressions were examined to reassess the osteological evidence that has been used to interpret s locomotion and phylogenetic relationships. It was found that the trunk was dorsoventrally compressed. The upper temporal fenestra was on the lateral surface of skull and two-thirds the size of the lower, the jaw joint posteriorly placed with short retroarticular process, and teeth short and subconical, but no evidence of external nares or antorbital fossae was found. The posterior trunk was covered with ~20 rows of closely spaced transversely elongate dorsal osteoderms. The coracoid was robust and elongate. The acetabulum was imperforate and the femoral head hemispherical and only weakly inturned such that the hip joint was unsuited to swinging in a parasagittal plane. The presence of four distal tarsals is confirmed. The marked disparity of tibial and fibular shaft diameters and of proximal tarsal dimensions indicates that the larger proximal tarsal is the astragalus and the significantly smaller tarsal is the calcaneum. The astragalus and calcaneum bear little resemblance to those of , and the prominent calcaneal tuber confirms that the ankle was crurotarsal. There is no evidence that preserved body and limb postures are unnatural, and most specimens are preserved in what is interpreted as a typical sprawling resting pose. A principal component analysis of skeletal measurements of and other vertebrates of known locomotor type found to plot with frogs, and that finding combined with skeletal morphology suggests was a sprawling quadrupedal hopper. Phylogenetic analyses found that was not an ornithodiran, but was either within the Doswelliidae or outside the clade consisting of the most recent common ancestor of the Erythrosuchidae and Archosauria and all its descendants.