Therapsid

Therapsids' temporal fenestrae were larger than those of the pelycosaurs. The jaws of some therapsids were more complex and powerful, and the teeth were differentiated into frontal incisors for nipping, great lateral canines for puncturing and tearing, and molars for shearing and chopping food.

Therapsid legs were positioned more vertically beneath their bodies than were the sprawling legs of reptiles and pelycosaurs. Also compared to these groups, the feet were more symmetrical, with the first and last toes short and the middle toes long, an indication that the foot's axis was placed parallel to that of the animal, not sprawling out sideways. This orientation would have given a more mammal-like gait than the lizard-like gait of the pelycosaurs.[2]

The physiology of therapsids is poorly understood. Most Permian therapsids had a pineal foramen, indicating that they had a parietal eye like many modern reptiles and amphibians. The parietal eye serves an important role in thermoregulation and the circadian rhythm of ectotherms, but is absent in modern mammals, which are endothermic.[3] Near the end of the Permian, dicynodonts, therocephalians, and cynodonts show parallel trends towards loss of the pineal foramen, and the foramen is completely absent in probainognathian cynodonts. Evidence from oxygen isotopes, which are correlated with body temperature, suggests that most Permian therapsids were ectotherms and that endothermy evolved convergently in dicynodonts and cynodonts near the end of the Permian.[4] In contrast, evidence from histology suggests that endothermy is shared across Therapsida,[5] whereas estimates of blood flow rate and lifespan in the mammaliaform Morganucodon suggest that even early mammaliaforms had reptile-like metabolic rates.[6] Evidence for respiratory turbinates, which have been hypothesized to be indicative of endothermy, was reported in the therocephalian Glanosuchus, but subsequent study showed that the apparent attachment sites for turbinates may simply be the result of distortion of the skull.[7]

The evolution of integument in therapsids is poorly known, and there are few fossils that provide direct evidence for the presence or absence of fur. The most basal synapsids with unambiguous direct evidence of fur are docodonts, which are mammaliaforms very closely related to crown-group mammals. Fossilized facial skin from the dinocephalian Estemmenosuchus has been described as showing that the skin was glandular and lacked both scales and hair.[8]

Coprolites containing what appear to be hairs have been found from the Permian.[9][10] Though the source of these hairs is not known with certainty, they may suggest that hair was present in at least some Permian therapsids.

The closure of the pineal foramen in probainognathian cynodonts may indicate a mutation in the regulatory gene Msx2, which is involved in both the closure of the skull roof and the maintenance of hair follicles in mice.[11] This suggests that hair may have first evolved in probainognathians, though it does not entirely rule out an earlier origin of fur.[11]

Whiskers probably evolved in probainognathian cynodonts.[11][12] Some studies had inferred an earlier origin for whiskers based on the presence of foramina on the snout of therocephalians and early cynodonts, but the arrangement of foramina in these taxa actually closely resembles lizards,[13] which would make the presence of mammal-like whiskers unlikely.[12]

Biarmosuchus, a biarmosuchian

Biarmosuchia is the most recently-recognized therapsid clade, first recognized as a distinct lineage by Hopson and Barghausen in 1986 and formally named by Sigogneau-Russell in 1989. Most biarmosuchians were previously classified as gorgonopsians. Biarmosuchia includes the distinctive Burnetiamorpha, but support for the monophyly of Biarmosuchia is relatively low. Many biarmosuchians are known for extensive cranial ornamentation.

Two genera of dinocephalians : Titanophoneus (an anteosaur) devouring a Ulemosaurus (a tapinocephalian)

Dinocephalia comprises two distinctive groups, the Anteosauria and Tapinocephalia.

Historically, carnivorous dinocephalians, including both anteosaurs and titanosuchids, were called titanosuchians and classified as members of Theriodontia, while the herbivorous Tapinocephalidae were classified as members of Anomodontia.

Lystrosaurus, a dicynodont anomodont

Anomodontia includes the dicynodonts, a clade of tusked, beaked herbivores, and the most diverse and long-lived clade of non-cynodont therapsids. Other members of Anomodontia include Suminia, which is thought to have been a climbing form.

Inostrancevia, a gorgonopsian

Gorgonopsia is an abundant but morphologically homogeneous group of saber-toothed predators.

Moschorhinus, a therocephalian

It has been suggested that Therocephalia might not be monophyletic, with some species more closely related to cynodonts than others.[24] However, most studies regard Therocephalia as monophyletic.

Trucidocynodon, a non-mammalian cynodont

Cynodonts are the most diverse and longest-lived of the therapsid groups, as Cynodontia includes mammals. Cynodonts are the only major therapsid clade to lack a Middle Permian fossil record, with the earliest-known cynodont being Charassognathus from the Wuchiapingian age of the Late Permian. Non-mammalian cynodonts include both carnivorous and herbivorous forms.