A giant brachyopoid temnospondyl from the Upper Triassic or Lower Jurassic of Lesotho J. SÉBASTIEN STEYER 1 and ROSS DAMIANI 2 Key words. – Stegocephalian, Triassic-Jurassic, Southern Africa, Gigantism. Abstract. – A jaw fragment of a giant temnospondyl from the Upper Triassic or Lower Jurassic of Lesotho (southern Africa), initially regarded as a Triassic mastodonsaurid because of its size, is redescribed in detail and considered to be a member of the Brachyopoidea (Brachyopidae + Chigutisauridae sensu Warren and Marsicano [2000]) based on its dental morphology, presence of a well-developed ectopterygoid tusk, and the concavity of the ventral margin of the skull in lateral view. Recognition of the specimen as a brachyopid, rather than as a chigutisaurid, is of palaeobiogeogra- phical significance in representing one of the youngest known brachyopids from Gondwana. The Lesotho specimen is also of palaeobiological interest in that an estimate of its overall size indicates that it represents one of the largest am- phibians sensu lato ever known. Un Temnospondyle brachyopoïde géant du Trias supérieur-Jurassique inférieur ? du Lésotho Mots clés. – Stégocéphale, Trias-Jurassique, Afrique australe, Gigantisme Résumé. – Un fragment de mâchoire d’un stégocéphale géant du Trias supérieur – Jurassique inférieur ? du Lésotho (Afrique australe), autrefois rapidement attribué à un capitosaure triasique à cause de sa taille, est ici redécrit en détail et considéré comme appartenant à un brachyopoïde d’après notamment la morphologie dentaire, la présence d’un croc ectoptérygoïdien bien développé, et la concavité du bord ventral du crâne en vue latérale. Cet unique temnospondyle du Lesotho correspond à un des amphibiens sensu lato les plus grands connus au monde. Outre les implications paléobio- géographiques que pose la présence d’un représentant du groupe des brachyopoïdes dans le Trias supérieur – Jurassique inférieur ? du Lésotho (peut-être le plus jeune brachyopide du Gondwana), ce spécimen soulève d’intéressantes ques- tions paléobiologiques sur le gigantisme des amphibiens en tant que tétrapodes non-amniotiques. INTRODUCTION In 1970, B. Battail, P. Ellenberger and L. Ginsburg discov- ered a skull fragment of a “grand Stégocéphale” from the uppermost part of the lower “Red Beds” (i.e. uppermost Lower Elliot Formation) at Alwinskop, near Quthing, south- western Lesotho. This fragment was later briefly described by Dutuit and Ginsburg [1982] as a mastodonsaurid, pri- marily on the basis of its giant size and presumed Triassic age. However, they considered the specimen to differ from other mastodonsaurids in the absence of a sensory sulcus, its curvature in lateral view, the position of the interpterygoid vacuity and choana relative to the maxilla, and the small size of the choana. Dutuit and Ginsburg [1982] concluded by referring the specimen to the “Mastodonsauridae, genus, species novae”. Until recently, the Mastodonsauridae was usually restricted to Mastodonsaurus and Heptasaurus from the Triassic of Eu- rope [Schoch, 1999], but has recently been redefined by Damiani [2001] to encompass all taxa previously included in the Capitosauridae. The Lesotho specimen therefore rep- resented the only known member of the Mastodonsauridae sensu stricto outside of Europe, and remains the only known temnospondyl from Lesotho. In this paper, we provide a detailed redescription of the Lesotho temnospondyl following additional preparation of the specimen. This permits a reassessment of its taxonomic affinities which, in turn, provides new paleobiogeographical and palaeobiological information. STRATIGRAPHY According to Dutuit and Ginsburg [1982], the specimen co- mes from the uppermost strata of the Lower “Red Beds”, in the “A7” zone of Ellenberger [1970]. The Elliot Formation (formerly “Red Beds”) of the “Stormberg Group” (Karoo Basin) of southern Africa [South African Committee for Stratigraphy, 1980] is a typical non-marine, fluvio-lacus- trine redbed sequence that straddles the Triassic-Jurassic boundary. It is stratigraphically subdivided into two (lower and upper) [Knoll, 2004, 2005] or three (lower, middle and upper) units [Kitching and Raath, 1984 ; Smith and Kit- ching, 1997]. Bull. Soc. géol. Fr., 2005, t. 176, n o 3, pp. 243-248 Bull. Soc. géol. Fr., 2005, n o 3 1. Bâtiment de Paléontologie, Département Histoire de la Terre, UMR 5143 du CNRS, Muséum national d’Histoire naturelle, 8 rue Buffon, F-75005 Paris, France, email: steyer@mnhn.fr. 2. Bernard Price Institute for Palaeontological Research, School of Geosciences, University of the Witwatersrand, Private Bag 3, Wits 2050, Johannes- burg, South Africa, email: rossano@mail.com. Manuscrit déposé le 23 août 2004 ; accepté après révision le 12 janvier 2004.