Since the 19th century, the Campanian and Maastrichtian continental deposits of southern France have yielded numerous dinosaur remains [Le Loeuff, 1991; 1998; Buffetaut et al., 1997; Laurent et al., 1991; Allain and Suberbiola, 2003]. The ornithopod remains that have not been referred to the hadrosaurids have been systematically attributed to Rhabdodon [Buffetaut and Le Loeuff, 1991; Buffetaut et al., 1996; Garcia et al., 1999; Pincemaille-Quillévéré, 2002]. This genus, initially named by Matheron  after its discovery in the lower Maastrichtian of La Nerthe (Bouches-du-Rhône), belongs to the Euornithopoda [sensu Sereno, 1999]. Rhabdodon represents the most common element of the dinosaur assemblages from the late Cretaceous of southern France [e.g. Allain and Suberbiola, 2003]. Nevertheless, since the localities have only provided some fragmentary material [Pincemaille-Quillévéré, 2002], the global morphology of this dinosaur and its phylogenetic placement within the euornithopods are still debated. The cranial morphology of Rhabdodon is particularly poorly understood due to the rarity of cranial remains preserved in the localities of southern France [Matheron, 1869; Garcia et al., 1999; Buffetaut et al., 1999; Pincemaille-Quillévéré, 2002]. Buffetaut et al.  first mentioned the discovery of a braincase (M4) referred to Rhabdodon, at Massecaps, a locality close to the village of Cruzy (Hérault, France). More recently, a new braincase (MN25) has been discovered at Montplô Nord, another locality close to Cruzy (specimens M4 and MN25 are conserved in the Museum of Cruzy). Both these localities have revealed a diverse and abundant vertebrate fauna suggesting a late Campanian to early Maastrichtian age [Buffetaut et al., 1999].
These braincases are described here in an attempt to detect potential autapomorphic characters in Rhabdodon, and compared to a more complete braincase of Tenontosaurus, an euornithopod from the Lower Cretaceous of North America, considered as the sister group of Rhabdodon [Weishampel et al., 1998; 2003; Garcia et al., 1999; Pincemaille-Quillévéré, 2002], in order to determine the potential differences and synapomorphies between the occiputs of the two genera. Finally, the braincases from Cruzy are compared to those of the other euornithopods described in the literature.
Specimen M4 (figs. 1–4) is incomplete but exceptionally well preserved. This braincase belongs to a juvenile individual, as shown by the numerous visible suture lines between the different cranial elements. Specimen MN25 (fig. 5) is badly deformed and attributable to an adult individual. Until now, all the ornithopods from the Upper Cretaceous of southern France have been referred either to hadrosaurs or to Rhabdodon. The Hadrosauridae show a low nuchal crest and their exoccipitals meet and form a bar on the dorsal border of the foramen magnum, excluding the supraoccipital from this border. Specimens M4 and MN25 do not present any nuchal crest and the supraoccipital participates in the dorsal border of the foramen magnum. Both braincases M4 and MN25 are therefore attributable to Rhabdodon.
Specimens M4 and MN25 have been compared to the occiput of a juvenile Tenontosaurus tilletti (fig. 6 : MCZ 4205, conserved in the Museum of Comparative Zoology, Harvard University). This reveals that Tenontosaurus and Rhabdodon share numerous characters : (1) the exoccipitals form the lateral borders of the foramen magnum, its ventral border being occupied by the basioccipital; (2) the occipital condyle is partly constituted by the exoccipitals, and in the same proportions; (3) the supraoccipital is rostrally oriented; (4) the suture line located between the prootic and the laterosphenoid shows the same outline; (5) the cresta prootica starts within the paroccipital process and extends onto the opisthotic; (6) the cresta prootica is transversal and non-horizontal; (7) the distribution of the cranial nerves is homologuous along the lateral surface of the braincase. Nevertheless, the braincase of Tenontosaurus differs from that of Rhabdodon in several significant respects : (1) the exoccipitals are dorsally connected, excluding the supraoccipital from the dorsal border of the foramen magnum; (2) two small dorsal humps are present at the level of the suture of the exoccipitals; (3) the supraoccipital is excluded from the dorsal border of the foramen magnum, which gives it a triangular shape; (4) the paroccipital processes are short, laterally flattened, and wing-shaped, and are more mediodorsally oriented than in Rhabdodon; (5) the cresta prootica follows a concave line and ends up on the prootic, at the level of the opening of the trigeminal nerve; (6) the external curve of the laterosphenoids is stronger; (7) the suture between the basioccipital and the opisthotic is very clear. The first of these unshared characters suggests that Rhabdodon belongs to Norman’s  ‘hypsilophodontoid’ clade and Tenontosaurus to the more evolved ‘iguanodontoid’ clade. The fusion of the exoccipitals on the dorsal border of the foramen magnum, together with other cranial adaptations, may have reduced the stress caused by a more elaborate mastication. Rhabdodon appears to have had a more primitive type of mastication. The strip formed by the reunion of the exoccipitals is less expanded dorsoventrally in Tenontosaurus tilletti than in the ‘iguanodontoid’ and ‘hadrosauroid’ clades. Tenontosaurus may therefore represent an intermediate group between the ‘hypsilophodontoid’ and ‘iguanodontoid’ clades.