†Protopsephurus liui Lu, 1994, from the Lower Cretaceous Yixian Formation of China, is both the oldest and the most primitive paddlefish (polyodontid) known. It was originally described based on a partial skeleton, a partial skull, and two partial trunk regions, leaving much of the species' osteology unknown. Here, based on the type specimens and many newly discovered well-preserved skeletons, we describe this species in detail.

†Protopsephurus is clearly a polyodontid based on several characters including: (1) the presence of stellate bones; (2) the presence of a series of long median dorsorostral and ventrorostral bones; (3) extreme elongation of the snout; (4) the presence of anterior, middle and posterior divisions of the fenestra longitudinalis; and (5) the presence of “microctenoid” scales (defined here). Within Polyodontidae, †Protopsephurus (which forms the monotypic series †Protopsephuri) lacks several characters derived for all other Polyodontidae (=Polyodonti). In contrast to †Protopsephuri, Polyodonti (including two fossil and two extant genera) have: (1) larger, more branched stellate bones; (2) a reduction of branchiostegals to a single element; (3) a sutural contact between posttemporal and dermosphenotic; (4) a longer anterior arm of the posttemporal; (5) a very large anterior division of the fenestra longitudinalis; (6) a loss of the pectoral fin spine; (7) the nasals reduced to little more than simple tubular ossifications; and (8) well-developed spine-like posterior projections of the subopercle. In addition, †Protopsephurus and †Paleopsephurus are apparently the only paddlefish species to have retained median elements at the rear of the skull roof (e.g., a median extrascapular bone).

The presence of a sturgeon-like pectoral spine in †Protopsephurus and some †peipiaosteids suggest that paddlefishes primitively had this structure and that the mere presence of a pectoral fin spine is not synapomorphic for Acipenseridae (i.e., the pectoral fin spine could be a derived feature of Acipenseristomi sensu Grande and Bemis, 1996). Strong development of the pectoral fin spine can still be used as a derived character of Acipenseridae because it is unique to that family among Acipenseriformes. Our phylogenetic analysis here does not support the monophyly of Acipenseristomi Grande and Bemis, 1996 (a group indicating Acipenseroidei to be more closely related to †Peipiaosteidae than to †Chondrosteidae). Further detailed studies of †Chondrosteidae are needed to more definitively resolve the three taxon interrelationship problem of Acipenseroidei, †Chondrosteidae, and †Peipiaosteidae. Paddlefishes remain one of the strongest indicators of historical biogeographic connection between East Asia and North America. Fossil and living members of the family date from the Early Cretaceous and Recent in China, and from the Late Cretaceous, Early Paleocene, Early Eocene and Recent in North America.

A new skull of Lesothosaurus has been found in the Upper Elliot Formation (Hettangian) of Masitise (Lesotho). It is nearly complete and provides new information, especially of the occipital region (horizontal paroccipital processes). The skull is not from a fully-grown individual. A specific determination is impossible and it is suggested that it should be referred to as Lesothosaurus sp. indet.

The parareptile Owenetta rubidgei is known from several skulls, all found in Upper Permian sediments of South Africa. The anatomy of a new species of Owenetta, O. kitchingorum, from the Early Triassic is described here on the basis of three nearly complete skeletons. This new species is distinguished from O. rubidgei on the basis of three dental and three cranial features. Postcranial features cannot be used in the diagnosis because the postcranial anatomy of O. rubidgei is unknown. The small parareptile Barasaurus from the Late Permian of Madagascar and Owenetta are united in the clade Owenettidae, as the sister-taxon of Procolophonidae. Controversies surrounding parareptilian phylogeny are discussed as they relate to Owenetta.

The poorly known genera Burnetia and Proburnetia, which are the only members of the therapsid family Burnetiidae, are redescribed and revised generic diagnoses are provided. Both genera are similar in having a prominent antorbital depression, a parasagittal ridge on the basicranial ramus of the pterygoid that extends from the transverse processes to the basisphenoid, a small temporal fenestra with no external area for adductor muscle attachment, and the same complement of bony protuberances on the skull roof including characteristic posteriorly projecting supratemporal “horns” formed by the squamosal and parietal bones. The morphology identified here is consistent with grouping the Burnetiidae in the Biarmosuchia, rather than in the Dinocephalia or Gorgonopsia as has been previously suggested.

Terminal fusions—fusions of skeletal elements thought to mark the cessation of significant growth—can be used to estimate the relative maturity of fossilized individuals. Little is known of how the absolute timing of terminal fusions relates to sexual and skeletal maturity in squamates. Examination of postnatal ontogenetic series of extant representatives of 14 crown squamate clades reveals that no terminal fusion universally coincides with the achievement of either sexually or skeletally mature size; however, certain fusions may serve as benchmarks of maturity within particular crown clades. Complete fusion of the braincase is a reliable benchmark for skeletal maturity in scleroglossans, but not iguanians. Complete fusion of long bone epiphyses will indicate that a squamate individual is within roughly 20% of maximum size, whereas complete fusion of the scapula and coracoid, the pelvis, and the astragalus and calcaneum can be taken as evidence only that an individual has achieved at least half of the maximum size of its species.

A heretofore unknown teiid lizard, recovered from the Cedar Mountain Formation (Albian–Cenomanian boundary) of Emery County, Utah, is the oldest teiid that is represented by numerous specimens. This new taxon has a heterodont dentition with conical anterior teeth and transversely oriented bicuspid posterior teeth that are distinct from those of Peneteius (Late Cretaceous) and Teius and Dicrodon (Recent). It also shows ontogenetic variation in tooth shape, in which the posterior teeth become more massive and have more transversely expanded crowns, as the individual grows older. Further comparisons and analysis indicate (1) that North American Cretaceous Polyglyphanodontinae (characterized by teeth with transversely-oriented crests) achieved their unusual dental specializations independently from those of analogous modern taxa; (2) that their transverse tooth-cresting resulted from addition of a medial cusp (rather than crown rotation, as previously hypothesized and which apparently is the case for living taxa); and (3) that the species from the Cedar Mountain Formation represents a sister-taxon to other North American polyglyphanodontines Polyglyphanodon, Peneteius and presumably Paraglyphanodon and Dicothodon. The new taxon thus represents an example of the morphologically antecedent dental specializations that culminated in the strikingly specialized dentition of Polyglyphanodon sternbergi.

Recent discoveries from the Aptian–Albian Antlers (Oklahoma) and Cloverly (Montana, Wyoming) formations provide significant additions to the Early Cretaceous record of lizards in North America. The lizards from the Antlers Formation include two teiids (one named), an anguimorphan, and a series of fragmentary jaws with “paramacellodid”-like teeth. The lizards from the Cloverly Formation include a new species of Paramacellodus and many indeterminate jaw fragments. The apparent lack of shared lizard taxa between the two units calls into question their temporal equivalence, which is based on similarity of dinosaur taxa. Although apparently distinct, the lizard faunas of both units are very similar to that of the Late Jurassic Morrison Formation in being composed primarily of “paramacellodid” or “paramacellodid”-like taxa. It appears that there was a period of relative stasis in the evolution of lizards in North America between the Late Jurassic and Aptian–Albian, paralleling a similar trend in Europe between the Late Jurassic and the Barremian (or later). Reported lizards from the Aptian–Albian of Mongolia present a level of taxonomic diversity much different than that of the same time period in Oklahoma, Montana, and Wyoming, but more like that of the Late Cretaceous of North America, suggesting that taxonomic groups common to the Late Cretaceous and Tertiary occurred earlier in Asia than in North America.

The Mesozoic lizard fauna of Gondwana is virtually unknown. We report here on a lizard assemblage from the Upper Member of the Kota Formation of peninsular India, usually considered to be of Early–Middle Jurassic age. The dominant form, Bharatagama rebbanensis, gen. et sp. nov., has a predominantly acrodont dentition. Comparison with living and extinct taxa suggests that this new genus is a primitive acrodont iguanian distinct from the Cretaceous priscagamids. It predates known records of iguanian lizards by some 80 Ma, and provides evidence that iguanians had begun to diversify before the break-up of Pangea. A second fully pleurodont taxon is known from the same deposit. It is tentatively attributed to the Squamata but is too fragmentary for further determination.

The skull of the traversodontid cynodont Exaeretodon riograndensis, sp. nov. from the Santa Maria Formation, Rio Grande do Sul State, Brazil is described. The excellent preservation of one specimen reveals new cranial information about the genus. The presence of prootic crests in the extensive lateral flange, anterior to the fenestra ovalis, is a noteworthy basicranial trait, unique to this new taxon among non-mammalian cynodonts. These crests are interpreted as related to the origin of the posterior pterygoid muscle representing the first evidence, among non-mammalian cynodonts to mammals transition, of the attachment of this muscle onto the prootic bone. The new species is most similar to Exaeretodon frenguellii from the Ischigualasto Formation of Argentina. Small differences in the posterior projections of the jugal onto the squamosal and in the shape of prefrontals distinguish the two species. Additionally, the number of postcanines seems to be ontogenetically less variable in the Brazilian species. The postcanine number decreases from six plus one erupting in a small individual (preorbital length: 82.9 mm) to five plus one erupting in the largest specimen (pl. 121.5 mm).

A new species of Collocalia (Aves: Apodidae) is based on six fossils recovered from a Holocene rockshelter deposit on Mangaia, Cook Islands. The coracoid and carpometacarpus of the new species, C. manuoi, are larger and very different qualitatively from those of the extant C. sawtelli on Atiu, Cook Islands, the geographically nearest population of Collocalia. This is the first species of Collocalia to be described from fossils. The Cook Islands become the first Polynesian island group known to have had multiple species of Collocalia.

Nanolestes drescherae, new genus and species, a stem-lineage representative of Zatheria, is represented by 48 isolated teeth, a dentary with p2 only, and an anterior dentary fragment with p3–p5 from the Kimmeridgian of the Guimarota coal mine, Portugal. Lower molars of N. drescherae have an enlarged, unicuspid talonid and a small additional cuspule on the cristid obliqua not referable to any of the standard talonid cusps. No trace of an incipient talonid basin is present. Tooth formula is I?/4, C?/1, P?/5, M?/5; all lower teeth except for incisors and m5 are double-rooted. The dentary has a well developed angular process, an internal groove extending beyond m3, and a mandibular foramen in an anterior position. Upper molars have a comparatively large stylocone; cusp “C” is present, stylar cusps are well developed, and additional cusps on the paracrista are present. The former attribution of these specimens to the peramurids cannot be corroborated. The “Porto Pinheiro Molar” from the latest Jurassic/earliest Cretaceous of Porto Pinheiro (or Dinheiro) is referred to Nanolestes as N. krusati, new species. There are no autapomorphies for Arguimuridae Dashzeveg, 1994. The teeth preserved in the holotype of Arguitherium cromptoni Daszheveg, 1994 are p3–p5 and not p4–m1 as assumed in the original description. The teeth described as upper molars of the symmetrodont Thereuodon Sigogneau-Russell, 1989 are posterior upper deciduous premolars (DP?3–5) of holotherians (probably zatherians).

The Pisco Formation (Miocene–Pliocene) of Peru contains a diverse marine vertebrate fauna. Among the more unusual members of the fauna is the semi-aquatic to aquatic nothrothere sloth, Thalassocnus. Continued fieldwork indicates that, in addition to the late Miocene type species, Thalassocnus natans, the genus is also represented by two additional species in higher horizons of the Pisco Formation. The skull morphology of T. natans is described along with that of Thalassocnus littoralis, sp. nov. and Thalassocnus carolomartini, sp. nov. from the earliest and late early to early late Pliocene, respectively. Many aspects of the cranium of the oldest species of Thalassocnus, T. natans, compare well to those of terrestrial nothrotheres. The terminal species, T. carolomartini, has the most derived features of the skull, absent in all other nothrotheres, that are related to feeding in a marine environment.

Clarkforkian late Paleocene freshwater limestone from the Clarks Fork Basin, Wyoming, has yielded four specimens of the plesiadapiform paromomyid, Acidomomys hebeticus gen. et sp. nov. A. hebeticus has a strong metaconid on p4, a small double-rooted p3, and a very diminutive i2. Presence of p3 and i2 is primitive, and their retention in A. hebeticus suggests that this taxon is part of a previously unknown lineage that diverged from other paromomyids by the early or middle Paleocene.

Acidomomys hebeticus is represented by virtually complete upper and lower dentitions, all from less than fully mature individuals, with teeth at one of three stages of development. The dental eruption sequence differs from that of other plesiadapiforms in having p3 erupt before p4 (p3 erupts after p4 in microsyopids, and p3 and p4 erupt simultaneously in plesiadapids). A. hebeticus differs from Plesiadapis and primitive primates, in which the premolars erupt after m3 is in place, in having p3/P3 erupt before or at the same time as that of m3/M3. It is similar to Plesiadapis in having p4 erupt after the molars are in place, a condition considered primitive in primates. While these differences in dental eruption sequence between representatives of Paromomyidae, Microsyopidae, and Plesiadapidae could be phylogenetically significant, they may instead reflect adaptive differences in dental morphology, facial architecture, or life history.

A fossil cranium from Rusinga Island, Kenya, is described as a new species of free-tailed bat, Tadarida rusingae (Chiroptera, Molossidae). The sediments where the skull was found are dated by Potassium–Argon assay at about 17.5–18.0 million years before present (early Miocene). Comparisons with fossil molossids and all the extant species of Tadarida show that T. rusingae is distinguished by its large size, well developed posterior sagittal crest, deeply domed palate, and several features of the upper molars. This skull represents one of the most complete fossil molossid bats known.

Fieldwork conducted since 1981 has greatly increased the sample of proboscidean fossils from Wadi Moghara, Egypt. The Moghara proboscidean assemblage is taxonomically more diverse than previously suspected, comprising four taxa: Gomphotherium angustidens libycum, Afrochoerodon kisumuensis, cf. Archaeobelodon, and Zygolophodon aegyptensis, sp. nov. Biochronological analysis of the proboscideans supports previous findings based on the remainder of the fauna that the age of Moghara is early Miocene, approximately 18–17 Ma. The composition of the Moghara proboscidean assemblage suggests complex biogeographic distribution patterns of proboscideans throughout Eurasia and Afro–Arabia during the early Miocene. Moghara and other pene-contemporaneous Afro–Arabian sites were apparently characterized by a relatively high degree of mammalian species-level endemism.

Continued excavation at the type locality of Ambulocetus natans led to the recovery of a majority of the axial skeleton of the holotype of this early Eocene cetacean, including both innominates, the sacrum, and most of the thoracic cage and thoracolumbar vertebral column. Additional appendicular, caudal, and cranial materials were also recovered, resulting in a specimen that is now approximately 80 percent complete. This new material allows refined interpretations of its functional morphology. Ambulocetus has a longer thoracolumbar column than that reported for later remingtonocetid and protocetid genera, suggesting that previous estimates of spinal length derived from models of mesonychid ancestry may be inaccurate. Ambulocetus also possesses a co-ossified ecto–mesocuneiform, a character found in some early and middle Eocene artiodactyls, but not mesonychids. New postcranial material provides further evidence of a systemic shift to aquatic locomotion.

Canis dirus preserved in the late Pleistocene Rancho La Brea tar seep deposits display a remarkably high incidence of teeth broken in life as compared with modern species. This might reflect greater carcass utilization in the Pleistocene relative to the present. Here we compare tooth fracture frequencies in C. dirus between two localities at Rancho La Brea, Pit 13 (about 15 ka) and Pit 61/67 (about 12 ka). Results indicate that Pit 13 C. dirus exhibited heavier tooth wear and fractured their teeth about three times more often than those of the younger Pit 61/67. This might reflect a difference in diet between the two periods or a greater preservation of older individuals (which are more likely to have broken teeth) in Pit 13. To test for differences in the mortality profiles of the two pits, we estimated individual age C. dirus from pulp cavity dimensions of lower canine teeth, as measured from radiographs of preserved mandibles. We first verified the validity of the approach using a sample of extant Canis latrans that had been previously aged using cementum counts. Pulp cavity analysis of the C. dirus indicated no significant difference between pits in the age structure of the preserved populations. Consequently, it appears that feeding behavior of C. dirus differed in the latest Pleistocene in that bone consumption declined. Pulp cavity analysis has great potential for further studies of population parameters, such as mortality profiles, of extinct carnivorans. It is easy, non-destructive, and provides a continuum of age values rather than discrete categories.

The boundary between Hemphillian and Blancan North American Land Mammal Ages has been elusive and equivocal. Our work identifies that boundary at about 4.9 to 5.0 Ma, based on paleomagnetic and radioisotopic dating of strata producing Blancan and Hemphillian mammals in Meadow and Spring Valleys of Lincoln County, eastern Nevada, and in the Pine Nut Mountains of western Nevada. Magnetostratigraphic study of the Panaca Formation in Lincoln County yielded a composite section with six magnetozones. The Healdsburg Tephra was identified near the top of the section in Meadow Valley, thereby placing the Lincoln County composite section relative to the Geomagnetic Polarity Time Scale (GPTS). The arvicolid rodent Mimomys, an indicator of the Blancan LMA, appears near the base of magnetic chron C3n.3r, about 4.98 Ma, in the Panaca Formation. Both Blancan and Hemphillian mammals have been reported from a stratigraphic sequence near Buckeye Creek in the Pine Nut Mountains of western Nevada. Magnetostratigraphic study of this area produced a thick (280 m) section with six magnetozones. A pumice layer near the top of magnetozone D yielded an ⁴⁰Ar/³⁹Ar date of 4.96 Ma, thereby correlating magnetozone D with Chron C3n.4n (Thvera subchron) of the GPTS. A Hemphillian rhino is recorded below the pumice layer and a Blancan bear is recorded above it. Correlation of the Lincoln County and Pine Nut Mountain sections in Nevada is consistent with placement of the Hemphillian/Blancan boundary in chron C3n.3r or the top of chron C3n.4n of the GPTS, about 4.9 to 5.0 Ma.

New studies based on enamel microstructure, as well as morphometric and morphologic comparative analyses have led to the revision of the Siwalik leporids described to date. A dp3 has been placed in Pliosiwalagus whitei, and a p3 previously assigned to Pliosiwalagus whitei has been placed in Pliosiwalagus sp. A specimen with p3p4, previously assigned to Caprolagus sivalensis has been reassigned to the genus Pliosiwalagus. A generalized phylogeny of Old World leporids has been suggested. Enamel microstructure studies of various leporid genera reveal difference in distribution of radial and irregular enamel within the thin and thick enamel of p3s and p4s.

Morphological differences among the p3s of fossil and extant leporid genera of Asia, Europe and Africa have been quantified using “Thin Plate Spline” program. Results from this analysis support the interpretations based on morphological similarities.