INTRODUCTION

Cyamodontoid placodonts (Eureptilia, Sauropterygia), sometimes referred to as the 'reptilian rays' (cf. Naish 2004) and recently proposed to share a convergent lifestyle with modern sirenians (Diedrich 2009), possess a turtle-like armour. The cyamodontoid armour and the turtle shell evolved convergently and are only superficially similar. In contrast to turtle shells, the cyamodontoid dermal armour is not connected to the underlying postcranial endoskeleton (Gregory 1946; Peyer and Kuhn-Schnyder 1955; Westphal 1975, 1976; Pinna 1999; Rieppel 2002). Aside from the other species of Cyamodus, Cyamodus hildegardis Peyer 1931 is known from both cranial and postcranial material, mainly because of the exceptional preservation conditions of fossils in the UNESCO world heritage site of Monte San Giorgio, Ticino, southern Switzerland, from which the majority of specimens has been recovered (e.g., Peyer 1931, 1944; Kuhn-Schnyder 1964, 1979; Pinna 1980; Furrer 2003).

Nosotti and Pinna (1996) and Pinna (1999) proposed Cyamodus hildegardis to be a junior synonym of Cyamodus "laticeps" (Owen 1858), more closely related to the placochelyid Protenodontosaurus italicus Pinna 1990 than to Cyamodus material from the Germanic Triassic. According to Rieppel (2001), however, C. hildegardis forms a monophyletic group with Cyamodus kuhnschnyderi Nosotti and Pinna, 1993 and Cyamodus rostratus (Münster 1839) from the Germanic Triassic of Germany. Postcranial characters could not yet be included into analyses for the three species of Cyamodus, because remains are insufficiently known for the Germanic species (Rieppel 2001).

There are only three specimens of C. hildegardis in which the postcranium, especially the dermal armour, is preserved in articulation: two presumably adult specimens (holotype PIMUZ T4763; referred specimen PIMUZ T58) and a juvenile specimen (MSNM V458). For an overview of published specimens referred to C. hildegardis see Pinna (1992) and Rieppel (2002). Specimens PIMUZ T58 and MSNM V458, the latter first described by Pinna (1980), were previously figured in part by Westphal (1975, 1976) on the constructional aspects and development of the placodont armour, as well as in the comparative work on cyamodontoid armour by Rieppel (2002).

Skeletal and life reconstructions are substantial for understanding the anatomy of extinct vertebrates. Earlier life reconstructions of C. hildegardis, so far mainly published for museum exhibition purposes (e.g., Kuhn-Schnyder 1979; Furrer 2003; Figure 1), implied that the trunk of the animal was covered only by a single continuous main piece of dorsal armour (see also Diedrich 2009 for a recent skeletal reconstruction of C. rostratus). Pinna's (1992) schematic reconstruction based on both the PIMUZ and MSNM specimens on the other hand, already showed a separate shield covering the pelvic region and the base of the tail in C. hildegardis. This way, C. hildegardis was shown for the first time to have a similar body outline as Psephoderma alpinum for which such a bipartite armour structure is best known (Pinna and Nosotti 1989; Rieppel 2002). Pinna (1992) and later Rieppel (2000, 2002) pictured and discussed the various aspects of the postcranial armour of C. hildegardis. However, as is discussed below, several problems with their respective interpretations were encountered. The latest life reconstruction of C. hildegardis, an aquarelle painting by Fabio Fogliazza, was presented by Nosotti and Teruzzi (2008, figure 46), who underlined difficulties in reconstructing the general morphology of the armour. Although it is by far the best life reconstruction of the species and an artful piece of work in its own right which should not be belittled in any way, it is still deemed appropriate to point out that it is based mainly on earlier, partly erroneous, interpretations.

Herein, a re-examination and novel interpretation of the postcranial skeleton, especially the dermal armour, of C. hildegardis based on the two adult PIMUZ specimens is presented, resulting in a novel life reconstruction of the overall body shape of this placodont. On the other hand, matters of scalation patterns, the nature of the dermal armour or the functional role of individual armour plates and their intrinsic role in the general development of the armour are not engaged.