taxonID	type	description	language	source
603F87C7FFE8FFEF23EEFABCFA445740.taxon	type_taxon	TYPE SPECIES. — Saniwa ensidens Leidy, 1870 by original designation.	en	Augé, Marc Louis, Folie, Annelise, Smith, Richard, Phélizon, Alain, Gigase, Paul, Smith, Thierry (2022): Revision of the oldest varanid, Saniwa orsmaelensis Dollo, 1923, from the earliest Eocene of northwest Europe. Comptes Rendus Palevol 21 (25): 511-529, DOI: 10.5852/cr-palevol2022v21a25, URL: https://doi.org/10.5852/cr-palevol2022v21a25
603F87C7FFE8FFE6238BFA3BFDFB5480.taxon	materials_examined	LECTOTYPE. — IRSNB R 298, dorsal vertebra (Hoffstetter 1968: fig. 2; Estes 1983: fig. 48 D; Molnar 2004: fig. 2.11). TYPE LOCALITY. — Dormaal, Flemish Brabant, Belgium, earliest Eocene, MP 7. REFERRED MATERIAL. — Dormaal (MP 7). IRSNB: 2 incomplete left maxillae, IRSNB R 393, Vert- 8818 - 01; 2 incomplete left dentaries, IRSNB R 394; 3 fragmentary jaws, Vert- 20681 - 03, Vert- 20681 - 04, Vert- 20681 - 05; 7 dorsal vertebrae, IRSNB R 116, IRSNB R 395, Vert- 8737 - 04, Vert- 8793 - 05, Vert- 8818 - 03, Vert- 16786 - 13, Vert- 20681 - 02; 15 caudal vertebrae, IRSNB R 115, IRSNB R 396, IRSNB R 397, Vert- 8685 - 01, Vert- 8685 - 02, Vert- 8818 - 04, Vert- 8818 - 05, Vert- 16786 - 14, Vert- 16786 - 15, Vert- 16786 - 16, Vert- 20681 - 06; Vert- 20237 - 05, Vert- 20237 - 07, Vert- 20237 - 08, Vert- 20237 - 09; 19 dorsal vertebrae and 17 caudal vertebrae (unnumbered from collection P. Gigase); three anterior dorsal vertebrae, 23 dorsal vertebrae and 35 caudal vertebrae (unnumbered from collection R. Smith); one incomplete right humerus, IRSNB R 398; one femur, IRSNB R 69; one incomplete ilium?, Vert- 16786 - 11. Erquelinnes-Jeumont (MP 7), IRSNB: four dorsal vertebrae Vert- 6433 - 01, Vert- 6433 - 02, Vert- 6433 - 03, Vert- 6433 - 04. Le Quesnoy (MP 7): 15 dorsal vertebrae, MNHN. F. QNY 1.1152 ai, MNHN. F. QNY 2.1155 a-e, MNHN. F. QNY 2.1158; 14 caudal vertebrae, MNHN. F. QNY 1.1153 a-f, MNHN. F. QNY 2.1154, MNHN. F. QNY 2.1156 a-g) and tentatively 1 parietal, MNHN. F. QNY 2.1157.	en	Augé, Marc Louis, Folie, Annelise, Smith, Richard, Phélizon, Alain, Gigase, Paul, Smith, Thierry (2022): Revision of the oldest varanid, Saniwa orsmaelensis Dollo, 1923, from the earliest Eocene of northwest Europe. Comptes Rendus Palevol 21 (25): 511-529, DOI: 10.5852/cr-palevol2022v21a25, URL: https://doi.org/10.5852/cr-palevol2022v21a25
603F87C7FFE8FFE6238BFA3BFDFB5480.taxon	diagnosis	EMENDED DIAGNOSIS. — Differs from the genus Varanus in having a pseudozygosphene-pseudozygantrum complex on dorsal vertebrae and postzygapophyses aligned with the neural arch; differs from Varanus and Saniwa ensidens in having the supradental table of the maxilla that bears four teeth behind the last labial foramen; differs from Saniwa ensidens and S. edura by lateral margins of the parietal table not upturned; differs further from S. ensidens in having a parietal table narrower, parietal foramen surrounded by an ovoid-circular raised ridge, presence of a single pineal foramen (the parietal foramen); differs from Saniwa edura in having a rather smooth parietal table, and many closely spaced basal infoldings present on tooth bases.	en	Augé, Marc Louis, Folie, Annelise, Smith, Richard, Phélizon, Alain, Gigase, Paul, Smith, Thierry (2022): Revision of the oldest varanid, Saniwa orsmaelensis Dollo, 1923, from the earliest Eocene of northwest Europe. Comptes Rendus Palevol 21 (25): 511-529, DOI: 10.5852/cr-palevol2022v21a25, URL: https://doi.org/10.5852/cr-palevol2022v21a25
603F87C7FFE8FFE6238BFA3BFDFB5480.taxon	description	DESCRIPTION Skull and mandible Maxilla (Fig. 2). Specimen IRSNB R 393 (lot 8737 in Hecht & Hoffstetter 1962) is an incomplete right maxilla briefly described by Dollo (1923). The anterior part and nearly all the facial process are lacking. However, the posterior part of the bone is complete and indicates that it was slender. Hecht & Hoffstetter (1962: 10) have already carefully described this fossil but they did not provide figure. Although partially preserved, this maxilla shows relevant morphologic characters. In labial view, the external surface of the maxilla is unsculptured. Three labial foramina are aligned on the ventral edge of the maxilla, the last foramen (at the level of the fifth tooth position from the rear) is by far the largest. These foramina are located in a shallow, mesiodistally directed concavity. The ventral edge of the maxilla is slightly eroded anteriorly (mostly at the level of the seventh tooth position from the rear). Medially, the supradental table (sensu Rage & Augé 2010) forms an obtuse ventral angle with the lateral wall, and no sulcus dentalis (dental gutter) is present. Eight tooth positions are preserved, the five most posterior teeth are almost complete and are posteriorly decreasing in size. The sub-pleurodont teeth are mediolaterally compressed, pointed, posteriorly recurved and their bases are enlarged. Mesio-distal keels very likely developed but are now eroded. All teeth have basoapical striae that betrayed the presence of plicidentine. The dorsal surface of the supradental table bears a large maxillary foramen at the level of the fifth tooth from the rear. Posteriorly there is a shallow, wide concavity without any furrow for the contact with the jugal. The posterior extremity of the maxilla is posterolaterally directed, rounded with a slight step behind the last tooth. Dentary (Fig. 3). Specimen IRSNB R 394 from Dormaal consists of the anterior part of an incomplete left dentary which is broken just behind the ninth tooth position. All the teeth are broken off and only two tooth bases (seventh and ninth) are visible. As preserved, this dentary is lightly built, elongated and shallow as in most extant Varanus species. The labial surface of the dentary is slightly convex and bears six irregularly spaced labial foramina, the first at the level of the boundary between the first and second tooth. Anteriorly, the dentary curves somewhat medially. In lingual view, the Meckelian groove is open; its anterior part is a shallow sulcus that faces only ventrally, curving to the mandibular symphysis. Behind the fourth tooth position it becomes posteriorly wider and opens ventrolingually. The mandibular septum is not observable because the posterior part of the dentary is broken. The medial rim of the subdental table (surface inclined in cross section and extending as far as the dental crest sensu Rage & Augé 2010) is crescentic, this concavity is widely present among anguimorph squamates (Estes 1964). Medial to the tooth bases, no subdental shelf or sulcus dentalis are present and a subrounded surface forms the upper margin of the Meckelian groove. On the poorly preserved tooth base of the seventh tooth, many closely spaced basal infoldings are present and these longitudinal striations are interpreted as an indication of the presence of plicidentine (Odermatt 1940; Bullet 1942; Caldwell 2003). This base is expanded and shows that the tooth attachment in this specimen is of the “ subpleurodont type ” or “ fully pleurodont ” as defined by Zaher & Rieppel (1999). Comparisons. All characters of the dentary and maxilla (overall morphology, subpleurodont implantation of the teeth, basal infoldings, plicidentine), caniniform morphology of the teeth) support platynotan, more precisely varanid relationships. Hecht & Hoffstetter (1962) expressed doubts about the referral to Saniwa and they did not exclude a possible referral to Palaeovaranus (previously Necrosaurus sensu Georgalis, 2017), e. g. by the presence of four teeth behind the maxilla foramen. However, at present the teeth of Palaeovaranus from Dormaal are known (from two undescribed dentaries) and their morphology differ significantly from that of Saniwa. Hecht & Hoffstetter (1962) pointed out the varanid features present in the maxilla IRSNB R 393: teeth sub-pleurodont, caniniform with development of plicidentine on their bases; enlargement of the last labial foramen such as in Varanus salvator (Dollo, 1923). However, Hecht & Hoffstetter (1962) also noted a character that distinguishes this maxilla from those of extant varanids (genus Varanus): the supradental table hosts four teeth behind the last labial foramen while in the genus Varanus only one or two teeth are present behind this foramen. In other words, the posterior part of this maxilla extended well below the orbit, in contrast with the maxilla of extant varanids that is excluded from the orbit. This plesiomorphic feature also appears in S. ensidens in which five teeth are present after the last labial foramen (Dollo 1923), and Gilmore (1922, 1928), Estes (1983: 184), Rieppel & Grande (2007) confirmed the presence of a long posterior (zygomatic) process of the maxilla in S. ensidens, which underlies the orbit with the last two or three teeth in a suborbital position, a plesiomorphic condition that is universally absent in Varanus, according to Rieppel & Grande (2007). Parietal (Fig. 4). Specimen MNHN. F. QNY 2.1157 from Le Quesnoy, France, is a well-preserved, unpaired bone, rectangular and comparatively elongated antero-posteriorly (Fig. 4). The extremities of the two anterolateral processes and, posteriorly, the two supratemporal processes are broken off. The dorsal surface of the parietal table bears some traces of weathering; otherwise it is fairly smooth except for weak striations developed near the parietal foramen. No osteodermal encrustations are present and the lateral margins of the parietal table are not upturned. The large parietal foramen opens anteriorly on the parietal table, it is not located on the fronto-parietal suture (Gilmore 1928; Estes 1983). The parietal foramen is surrounded by a raised ovoid to circular ridge. Anteriorly, the frontoparietal suture was not interdigitated except for some slight interdigitations near the midline of the frontal; inconspicuous anterolateral facets for the articulation with the frontal are present. The lateral margins of the parietal table are concave and form descending lateral walls that certainly received the pseudotemporalis superficialis muscle, as is also the case in Varanus. Posteriorly, the supratemporal processes are broken but their bases diverge posterolaterally. Thus, the posterior border of the parietal is broadly concave. On the ventral side, two weak crests separate the basis of the supratemporal processes from the main part of the parietal. The deep parietal fossa is located between those crests, well behind the parietal foramen. The ventral surface is rather smooth except for a strange circular depression in the parietal fossa. Because this depression does not open on the dorsal side of the parietal and is located well behind the parietal foramen, on a midpoint between the parietal foramen and the posterior border of the parietal, it could likely not represent a pineal foramen (sensu Smith et al. 2018), whereas the pineal foramen of S. ensidens is located just behind the parietal foramen (Smith et al. 2018: fig. 2 B, C). This depression is considered here to represent the parietal fossa. A similar structure is observed on a parietal of Pseudopus apodus (Pallas, 1775) (Klembara et al. 2017 a: fig. 17 D 12; b: fig. 6 B) and Ophisaurus manchenioi Blain & Bailon, 2019 (Blain & Bailon 2019: fig. 3 B).	en	Augé, Marc Louis, Folie, Annelise, Smith, Richard, Phélizon, Alain, Gigase, Paul, Smith, Thierry (2022): Revision of the oldest varanid, Saniwa orsmaelensis Dollo, 1923, from the earliest Eocene of northwest Europe. Comptes Rendus Palevol 21 (25): 511-529, DOI: 10.5852/cr-palevol2022v21a25, URL: https://doi.org/10.5852/cr-palevol2022v21a25
603F87C7FFE8FFE6238BFA3BFDFB5480.taxon	discussion	Comparisons. This parietal closely resembles those of Varanus and S. ensidens: temporal muscles originate from the dorsolateral surface of the parietal rather than from the ventral surface, leaving a wide amuscular parietal table; deep posterior parietal fossa; ventral ridges moderately developed; osteoderms reduced or absent; anterior position of the parietal foramen; general morphology of the bone (McDowell & Bogert 1954; Pregill et al. 1986). According to Evans (2008), the parietal foramen of Varanus is small, but Mertens (1942) noted significant ontogenetic variation of the relative size of the parietal foramen among the Varanus species. Also, the frontoparietal suture is likely simpler in Saniwa and more complex in Varanus (Conrad et al. 2008). Recently, Smith et al. (2018) described the pineal complex of the parietal of the S. ensidens specimens from the Bridger Formation in the Bridger Basin, Wyoming, with the presence of a widely open parietal foramen and a smaller pineal foramen. This complex is unfortunately not observable on specimen FMNH PR 2378 of S. ensidens from Green River (Rieppel & Grande 2007) or on specimen PTRM 5380 of S. edura (Smith, 2006). However, similarities can be noted between the parietal MNHN. F. QNY 2.1157 of S. orsmaelensis (Fig. 4) and that of the type specimen USNM 2185 of S. ensidens (Gilmore 1922: pl. 9, fig. 11). Indeed, anteriorly, the shape and location of the pineal foramen, the width of the parietal between the crests and the posterior border of the bone (concave anteriorly and forming a large notch) are very similar. The pineal foramen is described as being on the fronto-parietal suture. However, because the parietal of S. ensidens is not complete anteriorly, we cannot be sure of this, and specimen FMNH PR 2378 described by Rieppel & Grande (2007) confirms that the pineal foramen is well enclosed in the parietal. The two bones differ by the extension of the parietal processes, which are more laterally expended on S. ensidens than on S. orsmaelensis. Vertebrae Dorsal vertebrae (Fig. 5). Vertebrae are procoelous with roughly triangular, tapering centra, condyles and cotyles oval, laterally expanded and with a flange. The articular surface of the condylar ball is directed dorsally, and the condylar surface is scarcely apparent in ventral view. This situation appears reversed for the cotyle. In dorsal view, the neural arch is slanting anteriorly with the presence of anterior pars tectiformis. A weak midline keel runs along the neural arch, but it does not reach the anterior margin of the vertebra. Its posterior end is flared and markedly taller than the anterior end, forming the neural spine when preserved. Hence the neural spine is, at best, moderately antero-posteriorly elongated. The prezygapopyseal facets form an angle of about 40 degrees with the horizontal plane. The anterior aspect of the neural arch presents a pair of rudimentary inclined elevations. Clearly, they do not possess articular facets and they cannot be considered real zygosphenes. As Albino & Brizuela (2014) recognized, there is no zygosphenal articulation in Anguidae and more generally in extant Anguimorpha. This structure is not homologous to the zygosphene and it is called pseudozygosphene, according to Hoffstetter & Gasc (1969) (for additional comments on zygosphenes see Albino & Brizuela 2014). Pseudozygantrum is only variably present at the rear of the vertebra. In lateral view, the synapophyseal facets are approximately vertical and they extend the entire height of the centrum. In ventral view, a precondylar constriction is visible at the rear of the centrum. This constriction is less developed than on the dorsal vertebrae of S. ensidens (see Gilmore 1928). The ventral aspect of the centrum is smooth (without ridges) and slightly convex transversely. Pregill et al. (1986) noted that a marked precondylar constriction of the vertebrae is an autapomorphic feature of the Varanus lineage because this constriction is very weak or absent in Lanthanotus Steindachner, 1877 and Heloderma Wiegmann, 1829 and absent in Palaeovaranidae (Rage, 1978). Caudal vertebrae (Fig. 6). The centrum is rather elongated and narrow, there is no trace of autotomic septum and the cotyle and condyle are, as on the dorsal vertebrae, oval (dorsoventrally flattened). In lateral view, the vertebrae only preserve the dorsoventrally flattened bases of the transverse processes (one per side) that are attached to the anterior fourth of the centrum. In dorsal view, the prezygapophyseal facets are oval and project beyond the lateral margins of the neural arch. Between them, there is no evidence of a pseudozygosphene. No striae are present on the neural arch. The neural spine of a caudal vertebra from Dormaal (specimen IRSNB R 397, Fig. 6 H) is preserved. It is moderately high and rather antero-posteriorly extended but it does not reach the anterior margin of the vertebra. Its anterior margin is somewhat concave and its posterior margin slants posterodorsally. In ventral view, a pair of short articular protuberances located on the posterior third of the ventral surface of the centrum indicates the point of articulation of the chevron bones. A ridge runs anteriorly from the base of each protuberances and a median concavity occupies the space between the two ridges. As in S. ensidens, the precondylar constriction is less developed than on dorsal vertebrae. Comparisons. These vertebrae present anguimorph characters such as a smooth ventral surface of the centrum, synapophyses nearly vertical and hemicylindrical, and two keels on the ventral face of the centrum of the caudal vertebrae (Augé & Guevel 2018). Moreover, the general morphology of the vertebrae resembles those of extant varanids by the following characters: presence of a single pair of postero-laterally oriented transverse processes placed in the anterior part of the centrum and absence of an autotomic plane, orientation of the cotyle and condyle are of characteristic varanid form (e. g. posterior retraction of the ventral rim of the cotyle) and the well developed chevron facets just anterior to the posterior margin of the centrum on caudal vertebrae (Etheridge 1967; Augé & Guevel 2018). This last character would be a synapomorphy of the Varanidae (Augé & Guevel 2018). Additional similarities to varanid taxa are the anteriorly depressed neural arch (pars tectiformis), presence of a distinct precondylar constriction, and absence of a median ventral ridge on the dorsal vertebrae. These two last features also occur in the agamid genus Uromastyx Merrem, 1820 (Hoffstetter & Gasc 1969; Rage & Augé 2015). However, the caudal vertebrae of Uromastyx (present in the European Paleogene, Oligocene epoch) as in most agamids are characterized by circular cotyle and condyle, devoid of articular surface for the chevron bones, and lateral transverse processes are attached to the mid-length of the centrum. The main characters that distinguish Saniwa from Varanus are the presence of a pseudozygosphene-pseudozygantrum complex on the dorsal vertebrae (e. g. Dollo 1923; Rieppel & Grande 2007) and the anterior border of the vertebrae (including the prezygapophyses) which forms a line instead of a central notch on Varanus vertebrae (Gilmore 1928). Moreover, the precondylar constriction is less marked than in Varanus. Several authors also indicate that the neural spine on the dorsal and caudal vertebrae of Saniwa, which is anteroposteriorly well developed and lower than in Varanus, is a diagnostic character of Saniwa (Dollo 1923; Estes 1983; Rieppel & Grande 2007). However, other authors also indicate that the shape, size and extension of the neural spine are changing both along the vertebral column and with ontogeny (Gilmore 1928; Rieppel & Grande 2007). This character has therefore to be considered with caution. Limbs Humerus (Fig. 7). Specimen IRSNB R 398 is the distal end of a right humerus; the ulnar and radial condyles are lacking (Fig. 7). The entepicondyle is incomplete but it seems well developed and it is set off from the posterior margin of the diaphysis of the humerus. The entepicondylar foramen is absent as in S. ensidens. The ectepicondyle is broken but it is prolonged distally by a short ascending crest (ectepicondylar crest) which is pierced by an ectepicondylar foramen. In ventral view, a large triangular depression is present just proximal to the condyles and a small foramen opens on the proximal rim of the depression. Comparisons. This bone is tentatively referred to S. orsmaelensis based on size and varanid features. The presence of an ectepicondylar crest is consistent with referral of the humerus to varanids and more specifically to Saniwa (Lécuru 1969). In addition, the large triangular depression above the condyles, pierced by a small foramen that opens on the proximal rim of the depression, is characteristic of varanid squamates (Smith 2009). According to Lécuru (1968) and Smith (2009) this foramen is observed in nearly all limbed squamates, but it is quite reduced or even absent in Varanus. Moreover, the absence of an olecranon scar on the dorsal side is also considered a varanid feature (Hecht & Hoffstetter 1962). A potentially conflicting character is the moderately laterally projected entepicondyle. Rieppel & Grande (2007) and Gilmore (1922, 1928) noted that the entepicondyle is more distinctly developed in S. ensidens than in Varanus but it is worth noting that the entepicondyle of the specimen from Dormaal is incomplete. Two other large anguimorphans, Palaeovaranus and Placosaurus Gervais, 1848, are also present at Dormaal (Sullivan et al. 2012), and the morphology of their limbs is as yet virtually unknown. Hence the referral of this humerus to Saniwa cannot be accepted without reservations. As a last point, it could be worth noting that Hoffstetter left a manuscript note with the specimen indicating that he refers this humerus to S. orsmaelensis. Femur (Fig. 8). The right femur IRSNB R 69 is 28 mm long and is nearly complete (Fig. 8). It lacks only the proximal main condyle (articulating with the acetabulum of the pelvis) and the distolateral condyle. It is long, slender and sigmoidal. Proximally, the greater trochanter is present and well developed. A scar is likely present along the shaft just below this latter trochanter. A shallow groove is present dorsally between the trochanter and the base of the proximal main condyle. The ventral side is smooth. Distally, a small rounded medial condyle is preserved. The lateral condyle is broken and preserves only its base, which suggests that it was oval. Dorsally, a deep groove between the two condyles extends perpendicularly to the shaft up to the diaphysis. Ventrally, the extremity of the lateral condyle extends and forms the base of a rounded crest. The epiphysis is therefore triangular in distal view. Comparisons. The femur was first mentioned by Dollo (1923), who indicated that it presents varanid characters and can be attributed to S. orsmaelensis. However, Hecht & Hoffstetter (1962) indicated that its systematic attribution should be uncertain because it presents affinities with palaeovaranids. They also indicated that this bone could be a juvenile without epiphyses. None of the authors figured the specimens. The femur from Dormaal, is very similar to that of Varanus niloticus (comparison to specimen MRAC 92 - 048 - R- 0038 from Togo). The specimen from Dormaal is about half the size of the recent one but presents the same sigmoid curve and the same structures. The lateral distal condyle is more distally extended. Some additional crests are developed on the recent specimen in lateral view at the level of the epiphyses. However, this could be due to erosion of the specimen. Given these similarities, we consider the femur IRSNB R 69 part of the material attributed to S. orsmaelensis. ATTRIBUTION OF THE MATERIAL TO SANIWA ORSMAELENSIS The bones described hereall come from the locality of Dormaal and are grouped together based on their individual morphology, their comparable size and proportion. Consequently, they are all attributed to a single taxon presenting clear anguimorph characters. Hecht & Hoffstetter (1962) indicated that three large anguimorphs are present in Dormaal: Placosaurus (Anguidae; Sullivan et al. 2012), Palaeovaranus (previously Necrosaurus, Palaeovaranidae), and Saniwa (Varanidae). Each of the bones described here presents varanid characteristics: subpleurodont and caniniform teeth presenting plicidentine, a large last (posterior) foramen on the maxilla, a single pair of transverse processes placed in the anterior part of the vertebral centrum, absence of an autotomic plane, cotyle visible ventrally, anteriorly depressed neural arch (pars tectiformis), presence of a distinct precondylar constriction, absence of a median ventral ridge on dorsal vertebra, well developed chevron facets anterior to the posterior margin of the centrum on caudal vertebrae, presence of a small ventral foramen above the humeral condyle, absence of an olecranon scar on the dorsal face of the humerus (Dollo 1923; Hecht & Hoffstetter 1962; Etheridge 1967; Smith 2009). Referral of the specimens described here to the genus Saniwa is based on their general similarity to the type species S. ensidens: presence of four to five teeth behind the last labial foramen, pseudozygosphene-pseudozygantrum complex on the dorsal vertebrae, thin anterior end of the dentaries (Gilmore 1922; Dollo 1923; Hecht & Hoffstetter 1962). The similarities have already been mentioned several times in the literature (e. g. Hecht & Hoffstetter 1962; Rieppel & Grande 2007) and some authors (e. g. Estes 1983) even indicated that the two species could be synonymous with the Dormaal specimens representing just a brief incursion of S. ensidens in Europe. However, some differences have to be noted: the last (posterior) foramen of the maxilla is larger than on S. ensidens (Rieppel & Grande 2007: fig. 4 contra Fig. 2), presence of four teeth after the large last foramen of the maxilla instead of five teeth on S. ensidens (Dollo 1923). These differences are weak, especially because the variability is unknown and therefore it would not be parsimonious to consider the Dormaal taxon a new genus. This is why we refer them here to the species Saniwa orsmaelensis. Dorsal and caudal vertebrae found at Le Quesnoy present the same morphology as the specimens from Dormaal (including the presence of a pseudozygosphene-pseudozygantrum complex, see Figs 5; 6). Additionally, the parietal also presents varanid characters such as the presence of two lateral crests instead of a single sagittal crest (Palaeovaranus cayluxi Zittel, 1887 [Augé 2005: fig. 188 a]) or a narrow extension of the table (“ Necrosaurus eucarinatus ” Kuhn, 1940 [Augé 2005: figs 194, 195]) posteriorly (Rage 1978; Rage & Augé 2010). Because no other varanid taxa have been found at Le Quesnoy, we also attribute this parietal and the vertebrae to S. orsmaelensis. In addition to these arguments, it has to be noted that Dormaal and Le Quesnoy are located in the same geographic area corresponding to the southern North Sea Basin (Northwestern Europe, Fig. 1) and sedimentary conditions are nearly similar: the Dormaal Sands are a series of fluviatile beds, typical of the lower part of the Tienen Formation which is very similar to the channel deposits of the Sparnacian facies of the Paris Basin (Smith & Smith 1996; Steurbaut et al. 1999). According to the paleobotanical association, the Dormaal paleoriver system was bordered by warm-temperate to humid subtropical forests (Fairon-Demaret & Smith 2002). The Le Quesnoy locality exposes a succession of fluvio-lacustrine deposits, typical of the Sparnacian facies. They are also indicative of a paleoriver system bordered by wet, subtropical forests with similar fauna and flora (Nel et al. 1999, 2004; Jacques & De Franceschi 2005). Actually, if we compare the “ lizards ” (non-ophidian squamates) assemblages from Dormaal and Le Quesnoy, the species-poor locality (Le Quesnoy) is a subset of the assemblage of Dormaal (all “ lizard ” species and genera present at Le Quesnoy are shared with the assemblage of Dormaal). This pattern points to the similarity between the “ lizard ” assemblages from Dormaal and Le Quesnoy. Thus, their differences in diversity and abundance mainly reflect differences in quantity and exploitation of the fossiliferous substratum. As Hecht & Hoffstetter (1962) recognized, the cranial elements of Saniwa are difficult to separate from those of a purported Varanoidea, the genus Palaeovaranus which is present at Dormaal. In contrast, the vertebrae attributed to Palaeovaranus or Saniwa are readily distinguishable: the centrum of the dorsal vertebrae of Palaeovaranus bears no precondylar constriction; in Palaeovaranus, articulations with the chevron bones of the caudal vertebrae are in contact with the condyle on the ventral side of the centrum and the transverse processes are inserted more posteriorly in the centrum than those of Saniwa (Rage 1978: fig. 5; Rage & Augé 2015: fig. 3). Comparisons to skull elements of Palaeovaranus from the Eocene of the Phosphorites du Quercy (two species are currently described, P. cayluxi and P. eucarinatus [Estes 1983; Augé 2005]) allow further distinctions between Saniwa and Palaeovaranus. Despite a similar overall morphology of the dentary of S. orsmaelensis and that of Palaeovaranus (e. g. P. cayluxi, Augé 2005: figs 185, 186), numerous closely spaced basal infoldings are present on the tooth base of Saniwa whereas there are few, widely spaced striations in Palaeovaranus. The parietal of Palaeovaranus presents prominent dermal incrustations on the parietal table, which are often represented by oval, keeled osteoderms characteristic of the genus. In contrast, the parietal of Saniwa is devoid of dermal incrustation. A large parietal fossa is present in Palaeovaranus, whereas it is less developed in Saniwa. The ventral crests on the ventral surface of the parietal are absent or extremely reduced in Palaeovaranus. Thus, although the skull elements of Saniwa superficially resemble those of Palaeovaranus, the two taxa are always distinguishable by several features. Beside this, on the caudal vertebrae, the transverse processes are close to the condyle, which is not the case for the Palaeovaranus vertebrae (Hecht & Hoffstetter 1962), and the condyle and cotyle are oval in Saniwa and more rounded in Palaeovaranus (Estes 1983). Although the vertebrae do not bring new information to help the characterisation of S. orsmaelensis, it is not the case with the parietal, which brings additional data. Indeed, the parietal of S. orsmaelensis differs from that of S. ensidens by the oval shape of the parietal foramen (instead of a round foramen [Rieppel & Grande 2007]), by less laterally extended lateral processes (instead of well laterally extended processes [Smith et al. 2018]), by the parietal fossa which is more ventrally placed (instead of being posteriorly placed [Smith et al. 2018; Fig. 4]) and maybe the absence of a pineal foramen (presence of the foramen in specimens YPM VP 1074 and 613 from the Bridger Formation of Grizzly Buttes in Uinta County [Smith et al. 2018]). It resembles S. edura by the shape and size of the parietal foramen (Smith et al. 2018) but differs from this species by having a less ornamented surface. These differences are again too weak to identify a new genus and reinforce the validity of the species S. orsmaelensis.	en	Augé, Marc Louis, Folie, Annelise, Smith, Richard, Phélizon, Alain, Gigase, Paul, Smith, Thierry (2022): Revision of the oldest varanid, Saniwa orsmaelensis Dollo, 1923, from the earliest Eocene of northwest Europe. Comptes Rendus Palevol 21 (25): 511-529, DOI: 10.5852/cr-palevol2022v21a25, URL: https://doi.org/10.5852/cr-palevol2022v21a25
