The oldrst recordcd anthracosaur amphibian Eoherpeton walsoni Panchen, from localities in the Vistan and Namurian of the Scottish Carboniferous, is described. It lacks a distinct otic notch, and rcconstruction of thc position of cranial ligaments and muscles suggest that a tympanum was never drvrloped. O n the basis of these observations it is concluded that arguments used to refute a relationship between anthracosaurs and amniotes, based on apparent differences in the morphology of their otic regions, are invalid. The order Anthracosauria is considered to be a monophyletic group characterized by the development of a sutural connection between the tabular and parietal boncs on a skull table which includes both the intertemporal and supratemporal as well as the tabular. I t is divided into two suborders: Seymouriamorpha characterized by a sutural Connection between the squamosal and intertemporal and the presence of a lateral otic tube, and Anthracosauroideae characterized by the absence of post-temporal fossae. The Discosauriscidae, Kotlassiidae and Seymouriidae are included in the Seymouriamorpha, and the Eoherpetontidae, Gephyrostegidac and Embolomeri comprisr the Anthracosauroideae. The Proterogyrinidae are ronsidcred thc most primitive family of embolomeres. A new phylogeny of Amphibia is presented. The old groupings of Labyrinthodontia, Lepospondyli and Lissamphibia are rejected and Amphibia is divided, on the basis of differences in morphology of the neurocranium and its relationship with the surrounding bones of the skull, into two major groups, the first including the Ichthyostcgalia, Tcmnospondyli, Microsauria and the extant Anura, Apoda and Urodela, while the Aistopoda, Nectridea, Palaeostegalia, Loxommatoidea, Anthracosauria and Diadectomorpha, together with the Amniota, form the second. Similarities in the dentition of the palaeostegalian CrasszpyrinuJ and loxommatoids suggest the two taxa are sister-groups, but the immediate relationships of anthracosaurs remain indeterminate and they here form an unresolved trichotomy with diadectomorphs and amniotes.

The end-Devonian to mid-Mississippian time interval has long been known for its depauperate palaeontological record, especially for tetrapods. This interval encapsulates the time of increasing terrestriality among tetrapods, but only two Tournaisian localities previously produced tetrapod fossils. Here we describe five new Tournaisian tetrapods (Perittodus apsconditus, Koilops herma, Ossirarus kierani, Diploradus austiumensis and Aytonerpeton microps) from two localities in their environmental context. A phylogenetic analysis retrieved three taxa as stem tetrapods, interspersed among Devonian and Carboniferous forms, and two as stem amphibians, suggesting a deep split among crown tetrapods. We also illustrate new tetrapod specimens from these and additional localities in the Scottish Borders region. The new taxa and specimens suggest that tetrapod diversification was well established by the Tournaisian. Sedimentary evidence indicates that the tetrapod fossils are usually associated with sandy siltstones overlying wetland palaeosols. Tetrapods were probably living on vegetated surfaces that were subsequently flooded. We show that atmospheric oxygen levels were stable across the Devonian/Carboniferous boundary, and did not inhibit the evolution of terrestriality. This wealth of tetrapods from Tournaisian localities highlights the potential for discoveries elsewhere.

The skull of Greererpeton burAemorani Romer, a temnospondyl amphibian from the Upper Mississippian at Greer, West Virginia is described. A detailed account of the stapes of a Mississippian amphibian is given for the fint time and its function is discussed. It is suggested that the stapes formed the principal element of support for the back of the braincase and resisted potential dislocation of the otico-occipital region from the skull roof during contraction of the hypaxial musculature.Creererpeton is included in the Colosteidae and an amended diagnosis of the family is given. Erpetosaurus differs from Coloskur. Creercrpcton and Pholidogaskr in the pattern of bones in the skull roof and palate, the dentition and the otic region and, consequently, it is removed from the Colosteidae. The Temnospondyli are considered to be a monophyletic group characterized by the development ofa connection between the dorsal portion of the occipital arch, the exoccipital bones, and the skull roof. The loxommatids are removed from the Temnospondyli as they retain the plesiomorphic condition of braincase attachment which relies exclusively on derivatives of the auditory capsules.On the basis of similarities in the structure of the braincase, palate and manus it is suggested that microsaun are the collateral descendants (sister group) of temnospondyls. This relationship may account for the large number of similarities in the three living groups of Amphibia: Anura are generally believed to have descended from temnospondyls, while the Urodela and Apoda are often considered to have descended from microsaun. These systematic conclusions endorse the recent suggestions that neither the Lepospondyli nor the Labyrinthodontia are natural groups, and both terms should be abandoned.

Our knowledge of the primitive and aberrant early tetrapodCrassigyrinus scoticuswas based on a partial skull roof and mandibles from the Lower Carboniferous of Gilmerton, Edinburgh, plus a skeleton lacking the hind limb and tail from the Namurian, basal Upper Carboniferous, of Cowdenbeath, Fife. New specimens from Cowdenbeath include the pelvic girdle, presacral and sacral rib and most of the hind limb. The ilium of the girdle had a firm articulation with the vertebral column via the sacral rib. The ischium, separated from the ilium by cartilage, was, like that found with the Cowdenbeath skeleton, ornamented as though a dermal bone. No pubis is preserved. The femur lacks an adductor crest, but has a strongly developed internal trochanter. Tibia and fibula are short stout bones, but axial torsion is present in the fibula rather than the tibia. This, and the structure of the femur, suggests a swimming rather than a walking limb. Probable metatarsals and phalanges are recorded. A skeletal reconstruction and a life restoration ofCrassigyrinusare presented in the light of its reconstructed anatomy and physiology.The ornamentation of the ischium ofCrassigyrinus, and that of the colosteidGreererpeton, suggests that the bone may be at least in part dermal. Its homology wth the pelvic fin basal scute of osteolepiform fishes is proposed and the homologies of pectoral and pelvic fins, and thus limbs, discussed.

SUMMARY I. The traditional view of the origin of tetrapod vertebrates is that they are descendants of fossil osteolepiform fish, of which Eusthenopteron is best known. In recent years both that conclusion and the methodology by which it has been reached have been challenged by practitioners of cladistic analysis. Particularly a recent review by Rosen et al. (1981) claims that Dipnoi (lungfish) are the sister‐group of the Tetrapoda, that Osteolepiformes is a non‐taxon and that Eusthenopteron is more distant from tetrapods than are Dipnoi, coelacanths and probably the fossil Porolepiformes. We attempt to refute all these concludions by use of the same cladistic technique. 2. We accept that all the above‐mentioned groups, together with some less well‐known taxa, can be united as Sarcopterygii by means of shared derived (apomorph) characters. We also agree that Porolepiformes and Actinistia (coelacanths) can be characterized as valid taxa. The primitive and enigmatic fossil fish Powichthys is accepted as representing the plesiomorph sister‐group of true porolepiforms. 3. Only two apomorph features, the course of the jaw adductor muscles and the position of incurrent and excurrent nostrils, appear to unite all the fish, living and fossil, currently regarded as Dipnoi. The characteristic tooth plates and the presence of petrodentine both exclude important primitive fossil forms. 4. Contrary to the opinion of Rosen et al., Osteolepiformes can be characterized — by the arrangement of bones forming the cheek plate, the presence of basal scutes to the fins and by the unjointed radials of the median fins. However, if these are true autapomorphies they exclude any osteolepiform from direct tetrapod ancestry. 5. Tetrapoda is a monophyletic group characterized by ten or more autapomorphies, including the bones of the cheek plate, a stapes and fenestra ovalis, and a series of characters of the appendicular skeleton. 6. Tetrapods have a true choana (internal nostril). We accept that the posterior (excurrent) nostril of Dipnoi is the homologue of the tetrapod choana. However, we assert that the posterior nostril of all bony fish is the homologue of the choana. This assertion would be refuted if any fish showed separate posterior nostril and choana. We reject the claim that this ‘three nostril condition’ occurred in porolepiforms and osteolepiforms. The evidence for a choana in porolepiforms is inadequate. Osteolepiforms had a true choana, characterized as in tetrapods by its relationship to the bones of the palate, but no third nostril. Dipnoans are not choanate. 7. Following cladistic practice, the relationship of the extant taxa is established first. Dipnoi are thus shown to be the living sister‐group of tetrapods, but only on ‘soft anatomy’ characters unavailable in fossils. Coelacanths are the living sister‐group of the taxon so formed. 8. The relationship of the fossil taxa to the extant sarcopterygians is then considered. The synapomorphy scheme proposed by Rosen et al. is discussed at length. Virtually all the characters t...