taxonID	type	description	language	source
9E793A53FF97671EFF144A4BFBCFFD78.taxon	type_taxon	TYPE SPECIES. — Cadurceryx filholi Hoffstetter & Rage, 1972.	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF97671EFF144A4BFBCFFD78.taxon	diagnosis	EMENDED DIAGNOSIS. — Cadurceryx can be differentiated from all other snakes by the presence of additional apophyses throughout the trunk, cloacal, and caudal vertebrae, which form (at least partly) true articular joints linking neighbouring vertebrae (additional intervertebral articulations). These additional apophyses are: pterapophyses (present in all vertebrae), extensions of zygapophyses and neural spine, as well as extensions of pleurapophyses (present in caudal vertebrae). Moreover, there is a progressive increase of the complexity of succeeding vertebrae in the trunk portion of the column, from relatively simple anterior trunk vertebrae to moderately complex mid-trunk vertebrae, finally followed by highly complex vertebrae (“ Posterior trunk vertebrae ”). Cadurceryx is further diagnosed by the combination of the following features: centrum distinctly wider than long; presence of broad hypapophyses throughout the trunk and anterior cloacal vertebrae, which in posterior trunk vertebrae becomes even broader and possesses a distinct anteroventral corner forming approximately a right angle; presence of haemapophyses on more posterior caudal vertebrae; pterapophyses more prominent in mid- and posterior trunk vertebrae; pterapophyses extending forwards and backwards in trunk vertebrae, but extending almost exclusively forwards in caudal vertebrae; presence of a rather thick and occasionally anteriorly bifurcated neural spine in dorsal view; distinct interzygapophyseal connections present in trunk vertebrae but completely absent in cloacal and caudal vertebrae; in posterior trunk vertebrae, the posterior extensions of prezygapophyses and anterior extensions of postzygapophyses fuse with each other forming a thick bar; absence of paracotylar foramina.	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	description	(Figs 1 - 18)	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	materials_examined	TYPE MATERIAL. — Holotype. France • 1 specimen (a posterior trunk vertebra [not anterior cloacal vertebra as stated by Rage 1984 b]); Phosphorites du Quercy (imprecise locality); probably between middle and late Eocene; MNHN. F. QU 16301 (formerly MNHN Qu 301) (Hoffstetter & Rage 1972: fig. 6 a, pl. I. 1; Rage 1984 b: fig. 16 e; this paper, Fig. 1). TYPE LOCALITY AND AGE. — Similar to most other fossil specimens from the so called “ old collections ” of Quercy, there are no precise locality data for the holotype of Cadurceryx filholi, apart from the general information that it originates from the area of the Phosphorites du Quercy. However, this information is rather general, taking into consideration that the Phosphorites du Quercy include at least 170 fissure filling localities, distributed over a broad geographic area, encompassing large parts of the current Departments of Lot, Tarn-et-Garonne, Tarn, and Aveyron, all in the administrative region of Occitanie (Rage 2006; Sigé & Hugueney 2006; Georgalis et al. 2021 a, c; Pelissié et al. 2021). They also stratigraphically span over a considerable time period, from the early Eocene (MP 8 + 9) until the Early Miocene (MN 3); however, most of the respective fossiliferous localities range between the late middle Eocene (MP 16) and the late Oligocene (MP 28) (Rage 2006; Sigé & Hugueney 2006; Georgalis et al. 2021 a, c; Pelissié et al. 2021). NEW AND OTHER MATERIAL STUDIED HEREIN. — Lavergne (MP 16): two pairs of two fused mid-trunk vertebrae (SU. PAL. 2019.0.303.1 and SU. PAL. 2019.0.303.12), two anterior trunk vertebrae (SU. PAL. 2019.0.303.4 and SU. PAL. 2019.0.303.5), two mid-trunk vertebrae (SU. PAL. 2019.0.303.6 and SU. PAL. 2019.0.303.7), one posterior trunk vertebra (SU. PAL. 2019.0.303.8), two cloacal vertebrae (SU. PAL. 2019.0.303.9 and SU. PAL. 2019.0.303.10), nine caudal vertebrae (SU. PAL. 2019.0.303.11 - SU. PAL. 2019.0.303.19), two fused caudal vertebrae (SU. PAL. 2019.0.303.3), and 70 vertebrae (of which 13 trunk, 8 cloacal, and 49 caudal vertebrae; SU. PAL. 2019.0.303.20 - SU. PAL. 2019.0.303.89). Malpérié (MP 16): two caudal vertebrae (UM MAL 501 and UM MAL 503). La Bouffie (MP 17): 51 trunk vertebrae (UM BFI 3011 - UM BFI 3048, UM BFI 3082, UM BFI 3104, UM BFI 3117, UM BFI 3151 – UM BFI 3158, UM BFI 3160, and UM BFI 3161), one? cloacal (or anterior caudal) vertebra (UM BFI 3159), and seven caudal vertebrae (UM BFI 3049 - UM BFI 3051, UM BFI 3115, UM BFI 3116, UM BFI 3119, and UM BFI 3121). Rosières 2 (MP 17): one caudal vertebra (UM ROS 2 266). Sainte Néboule (MP 18): 10 caudal vertebrae (UM SNB 5135 - UM SNB 5142, UM SNB 5233, and UM SNB 5234). “ Old collections ” of Quercy (localities imprecisely known, middle or late Eocene): two trunk vertebrae (MNHN. F. QU 16304 and NHMW 2019 / 0064 / 0001). LOCALITIES AND AGE. — All localities of Lavergne, Malpérié, La Bouffie, Rosières 2, and Sainte Néboule are situated in the area of the Phosphorites du Quercy in southern France. Lavergne and Malpérié are the oldest among them, pertaining to the MP 16 zone of the middle Eocene, while Rosières 2 and Sainte Néboule are late Eocene, with Rosières 2 (MP 17) slightly older than Sainte Néboule (MP 18) (Biochrom’ 97 1997). As for the two vertebrae from the “ old collections ” of Quercy (MNHN. F. QU 16304 and NHMW 2019 / 0064 / 0001), similarly to the case of the holotype, they originate from imprecisely known localities and therefore their age could lie anywhen within middle to late Eocene.	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	distribution	GEOGRAPHIC AND STRATIFRAPHIC DISTRIBUTION. — Middle and late Eocene of France. For a detailed account of all known occurrences of the taxon, see Discussion below.	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	diagnosis	EMENDED DIAGNOSIS. — As for the genus (see above).	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	description	DESCRIPTION The descriptions are based on the abundant vertebral sample from Lavergne. The vertebrae of Cadurceryx from Lavergne represent all major regions of the column (Figs 2; 3), except for anteriormost trunk (in particular atlas and axis) and possibly the last trunk vertebrae, i. e., those (or that) preceding immediately the cloacal region. All vertebrae, as characteristic of the so called constrictor snakes (i. e., booids and pythonoids), have centra distinctly wider than long. The vertebrae are relatively small, in a few largest trunk vertebrae the centrum length is 3.0 to 3.5 mm, centrum width 4.2 to 4.6 mm, whereas centrum length / neural arch width ratio 0.7 to 0.8. Apart from the vertebral structures found in all ophidians, virtually all vertebrae throughout the column possess additional apophyses providing (at least partly) articular joints between adjacent bones. These are: pterapophyses (present in all vertebrae), extensions of zygapophyses and neural spine, as well as extensions of pleurapophyses (in caudal vertebrae). The distinctive feature observed in the morphology of the trunk vertebrae of Cadurceryx is a progressive increase of the complexity of succeeding elements, from relatively simple anterior trunk vertebrae (see section “ Anterior trunk vertebrae ” below) to moderately complex mid-trunk vertebrae (“ Mid-trunk vertebrae ”), finally followed by highly complex vertebrae (“ Posterior trunk vertebrae ”). This unique morphological pattern (if we consider the morphology of additional apophyses) resembles closely the patterns characteristic of caudal (but not trunk!) portion of the column in both extant and extinct (the fossil genus Bransateryx Hoffstetter & Rage, 1972, at least) “ erycine ” snakes. The morphology of the post-trunk (i. e., cloacal and caudal) vertebrae, also provided with additional apophyses, is rather homogenous throughout the tail skeleton; in other words, there are no striking differences between the vertebrae coming from different parts of the cloacal / caudal portion of the column. Anterior trunk vertebrae (Fig. 4) We use this term to refer to a number (typically a few dozen in most alethinophidian snakes) of vertebrae located immediately behind the head and provided with distinct (but successively reduced in size) hypapophyses (taxa like elapids and natricids, provided with hypapophyses throughout the trunk portion of the column, are noteworthy exceptions). This trait is observed in 10 vertebrae from Lavergne, available for this study. The hypapophysis is subsquare in shape in lateral view and rather short; its posteroventral termination reaches the level of the condyle. Apart from the presence of the latter structure, the morphology of the anterior trunk vertebrae of Cadurceryx is relatively simple compared with those located more posteriorly. The anterior trunk vertebrae, however, possess small but distinct pterapophyses beginning atop the postzygapophyseal area of the neural arch and directed anteriorly. At first glance, these structures look like the postzygapophyseal wings observed in Eryx, rather than pterapophyses (cf. Szyndlar 1994: fig. 5 A), because they have the form of spurs projecting forwards but not backwards. However, in more posterior vertebrae located close to the anterior / mid-trunk transition (attributed to this region due to the presence of reduced and much different-shaped hypapophyses) the pterapophyses are produced into a short spur projecting backwards, as well (Fig. 5 A). The neural arch is moderately vaulted, but the difference between the vaulting of the arch of the anterior trunk vertebrae and of those located more posteriorly is not significant, unlike in most other snakes. The neural spine is as high as long in lateral view and not flattened dorsally. The subcentral ridges and subcentral grooves are prominent. The prezygapophyseal accessory processes are relatively short, strongly built, but are not expanded laterally or posteriorly. Mid-trunk vertebrae (Fig. 5) In the vertebrae immediately succeeding the anterior trunk ones, the hypapophysis has a rather different shape: this structure is prominent, thick, and moderately broad in ventral view, while in lateral view, this is relatively anteroposteriorly long. Behind the cotyle rim, there is a shallow (and rather indistinct) depression in the keel (Fig. 5 F). The subcentral grooves, on both sides of the hypapophysis, are deep and accompanied by prominent subcentral ridges. In most (but not all) vertebrae, the ridges terminate sharply in the form of distinct spurs before reaching the condyle, the feature well seen in lateral view (Fig. 5 E). These vertebrae are distinctly higher than long in lateral view and distinctly broader than long in ventral view. The neural arch is moderately depressed. The neural canal is significantly narrower than the suborbicular cotyle. The condyle is also rounded or slightly depressed. The paradiapophyses are relatively small, without distinct subdivision into para- and diapophyseal portions. The subcentral and lateral foramina are small but distinct (the latter, usually covered by post-prezygapophyseal structures, are hardly seen in lateral view). The neural spine, approximately as high as long, has its anterior margin slightly overhanging and the posterior margin usually strongly overhanging. The upper margin of the neural spine can be slightly depressed and sometimes slightly bifurcated anteriorly and posteriorly, in dorsal view; the bifurcations do not overlap in neighbouring vertebrae (Fig. 5 A, C). The pterapophyses, well developed on both sides of the base of the neural spine, are elongate and protrude distinctly both anteriorly and posteriorly. As seen in the pairs of fused vertebrae, the anterior projections of pterapophyses of the more posteriorly located vertebra externally overlap the posterior projections of pterapophyses of the preceding vertebra (Fig. 5 C, D). The prezygapophyses have prominent extensions: prezygapophyseal accessory processes and posterior extensions of prezygapophyses. The prezygapophyseal accessory processes, developed beneath the elongated prezygapophyseal articular facets in the form of long laminae, project forwards and backwards. The postzygapophyses, provided with subsquareshaped articular facets, are distinctly separated off the pterapophyses located above; anteriorly, the postzygapophyses are produced into more or less prominent anterior extensions. The posterior extensions of prezygapophyses and anterior extensions of postzygapophyses are approaching to each other but are not fused (Fig. 5 E). Posterior trunk vertebrae (Fig. 6) Some vertebral elements of the posterior trunk vertebrae (pterapophyses, diapophyses, neural canal, cotyles and condyles, zygosphenes, and zygantra) do not differ significantly from those of the mid-trunk vertebrae. Other elements, however, are apparently overgrown and / or significantly enlarged. The main difference between the vertebrae described below and those described in the preceding section is progressing complexity of the vertebral structures, observed in particular (but not only) in the development of additional apophyses. At first glance, the most striking feature differentiating the mid-trunk and posterior trunk vertebrae is, in the latter vertebrae, the significant enlargement of the posterior extensions of prezygapophyses and anterior extensions of postzygapophyses that fuse with each other forming a thick bar, especially well visible in lateral views (Fig. 6 A). Hence, each bar, resulting from the fusion of the zygapophyseal extensions, is distinctly separated from the main body of the vertebra by a large foramen. In rare cases, the fusion may be incomplete on one side of a vertebra (Fig. 6 J). The neural spine is of similar height as in mid-trunk vertebrae but thicker (in anterior and posterior views), more flattened dorsally, and with more distinct anterior and posterior bifurcations. In lateral view, the hypapophysis is more prominent and relatively higher (i. e., extending further downward) than in mid-trunk vertebrae and provided with more prominent (deeper) depression behind the cotyle rim (Fig. 6 A, F). A characteristic feature of the hypapophysis in this portion of the vertebral column is its distinct anteroventral corner that forms approximately a right angle. In a few vertebrae, there exist distinct spurs originating from the centrum wall above the postero-dorsal part of the paradiapophyses and projecting posteriorly (“ une apophyse incline vers le bas et l’arrière ” of Hoffstetter & Rage 1972). These structures, although quite prominent, can be obscured by overgrown posterior parts of the prezygapophyseal accessory processes and hence hardly visible in lateral view (Fig. 6 A); instead, they are well seen, immediately behind the paradiapophyses, in ventral view (Fig. 6 E). The posterior trunk vertebrae are more depressed than those from the preceding regions of the column. This difference is most apparent in the most posterior trunk vertebrae (usually termed “ posteriormost trunk ” or “ last trunk ”, i. e., the last and possibly penultimate vertebrae), described in the following paragraphs. The available material from Lavergne contains four such vertebrae that no doubt originate from the very end of the trunk portion of the column (Fig. 6 H, I, J-O). The most characteristic feature distinguishing them from more anterior trunk vertebrae is the increasing height (i. e., downward projection) of their hypapophyses (Fig. 6 I-K); these can be termed “ cloacal hypapophyses ”. These hypapophyses are even more prominent than preceding vertebrae and their anteroventral corner is even more distinct and straighter. Another important feature of these vertebrae is an apparent reduction in size of the posterior projections of pterapophyses; interestingly, the same morphological pattern is observed in the anterior trunk vertebrae (see below) as well as those from the cloacal and caudal regions (see above). Concluding, most vertebral elements observed in the available posteriormost trunk and anteriormost cloacal vertebrae display very similar (unless almost identical) morphological patterns (the presence of lymphapophyses in cloacal vertebrae is an obvious exception). Unfortunately, there exists one significant deviation from this “ smooth ” trunk-to-cloacal transition, namely the presence of interzygapophyseal connections in the former vs. their complete disappearance in the latter (in fact, in all following vertebrae until the tail tip). This “ sharp transition ” may indicate that Cadurceryx could actually possess (an) additional trunk vertebra (e) immediately preceding the cloacal region, which is unfortunately absent in the available fossil material. Cloacal vertebrae (Fig. 7) The material from Lavergne makes it possible to reconstruct roughly the cloacal and caudal portions of the vertebral column. Interestingly, the structure of virtually all vertebrae coming from this region of the column is rather homogenous, from the cloaca to the tail tip. This situation is in contrast to that observed in the extant genus Eryx and the extinct Bransateryx, where the simply built cloacal and anterior caudal vertebrae are followed by more and more complex ones, with the highest rate of complexity appearing in posteriormost caudal vertebrae. As in other snakes, the cloacal vertebrae of Cadurceryx clearly differ from the caudal vertebrae by having (remnants of) lymphapophyses instead of pleurapophyses. Besides, they display quite complex morphology in the upper portion of vertebral bodies and in this aspect they do not differ substantially from the following caudal vertebrae. The available material from Lavergne consists of 10 cloacal vertebrae. Two of them, provided with short hypapophyses (resembling those of the last trunk vertebrae), are considered anterior cloacal vertebrae. The remaining eight cloacal vertebrae are provided with (remnants of) haemapophyses, hence, they are considered to be located more posteriorly in the cloacal portion of the column. In most examined snakes, in particular in snakes with complex vertebrae, the cloacal vertebrae morphologically closely resemble the posteriormost trunk vertebrae (obviously, except for the subcentral structures). Unfortunately, in the case of Cadurceryx from Lavergne, vertebrae that could be interpreted with full certainty as posteriormost trunk ones have not been found (see below). Their possible presence could shed light on the enigmatic transition between the “ trunk-like ” and “ cloacal / caudal ” patterns of complexity. Caudal vertebrae (Figs 8 - 10) Sixty caudal vertebrae have been picked off the available Lavergne material. Anteriormost caudal vertebrae possess a hypapophysis, while, as characteristic of most snakes, succeeding caudal vertebrae are provided with paired haemapophyses; unfortunately, these latter structures are broken in virtually all available caudal vertebrae. As mentioned in the preceding paragraphs, the vertebrae coming from the tail region display rather homogenous morphology. However, protruding vertebral elements (in particular pterapophyses, as well as extending parts of prezygapophyses and pleurapophyses) are less strongly developed in the vertebrae interpreted as located more anteriorly (Fig. 9) than those located more posteriorly (Fig. 10). Compared with trunk vertebrae, most caudal vertebrae are relatively much higher than long in lateral view, as well as distinctly broader than long in ventral view. The neural arch is depressed. The neural canal is relatively very small (except for vertebrae pertaining to juvenile individuals, e. g. Fig. 10 L). The neural spine is moderately small and short in lateral view, but relatively thick in anterior / posterior views and sometimes depressed dorsally. The postzygapophyses are simply built. With one exception only (Fig. 10 A), the postzygapophyses are devoid of any anterior extensions. Although the prezygapophyses are developed into prominent posterior extensions, the latter structures are never fused with postzygapophyses, unlike in the posterior trunk vertebrae. The prominent pterapophyses are maintained in the entire post-trunk portion of the column, however, in overwhelming majority of the post-trunk vertebrae they are extending exclusively forwards and not backwards, therefore unlike as in the trunk vertebrae of Cadurceryx itself as well as caudal vertebrae of Eryx and Bransateryx, having pterapophyses provided with distinct both anterior and posterior projections. Despite the lack of the backward projections, the pterapophyses belonging to more posterior caudal vertebrae of Cadurceryx may have overlapped those of their anterior neighbours. Surprisingly, as seen in two fused caudal vertebrae (Fig. 8 C), the (left) pterapophysis belonging to the posterior vertebra overlaps the (left) pterapophysis of the anterior vertebra from inside and not outside, as characteristic of the trunk vertebrae (see below). Notably, in a few caudal vertebrae of Cadurceryx, there exist minute protuberances above the postzygapophyseal region that may mimic vestigial rear projections of pterapophyses (Fig. 9 H), whereas one caudal vertebra from Lavergne (Fig. 9 L-M) has pterapophyses distinctly separated off the postzygapophyses and provided with distinct posterior projections. The pleurapophyses (although badly preserved in most caudal vertebrae) were no doubt strongly built structures, stretching latero-ventrally. The pleurapophyses are provided with prominent anterior and posterior extensions, from above distinctly separated by a deep furrow from the prezygapophyses and their posterior extensions. As seen in two fused caudal vertebrae (Fig. 8 B), the anterior extension of pleurapophysis is wedged between the posterior extension of prezygapophysis (from above) and the posterior extension of pleurapophysis of the preceding vertebra (from below). This condition does not differ from that observed in the recent genus Eryx, while in the Oligo-Miocene genus Bransateryx the anterior extension of pleurapophysis is articulated from above by the posterior extension of pleurapophysis of the preceding vertebra (Szyndlar 1994).	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	materials_examined	ADDITIONAL MATERIAL OF CADURCERYX FILHOLI FROM THE PHOSPHORITES DU QUERCY Besides the abundant vertebral material from Lavergne described above, which pertains to practically almost all portions of the column, we present here additional vertebrae of Cadurceryx filholi from other localities in the area of the Phosphorites du Quercy. These include caudal vertebrae from Malpérié (MP 16; Fig. 11 A-J) and Rosières 2 (MP 17; Fig. 11 K-M) that were previously described and figured byHoffstetter & Rage (1972), plus several undescribed trunk and caudal vertebrae from La Bouffie (MP 17; Figs 12 - 16), and a few caudal vertebrae from Sainte Néboule (MP 18; Fig. 17). Moreover, we present also two further trunk vertebrae from the “ old collections ” of Quercy: one vertebra (MNHN. F. QU 16304; Fig. 18 A-F), originally described and figured by Hoffstetter & Rage (1972: fig. 6 B), and one previously undescribed vertebra from the collections of NHMW (NHMW 2019 / 0064 / 0001; Fig. 18 G-L). All these vertebrae match well with the morphology and intracolumnar variation observed in the above documented Lavergne sample. We therefore assign all of them to Cadurceryx filholi. This resemblance further prompts us to consider that a single species of Cadurceryx, i. e., the type species Cadurceryx filholi, was present in all these studied localities and perhaps, in general in France during the middle and late Eocene. It is further worth noting that the snake sample from Sainte Néboule, although it comprised vertebrae of several other snake taxa (GLG, personal observation), yielded only caudal vertebrae of Cadurceryx but not a single trunk one (see also Discussion below).	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
9E793A53FF976703FC954F2DFDDFF875.taxon	discussion	A CRITICAL APPROACH OF HOFFSTETTER & RAGE’S (1972) ORIGINAL DESCRIPTION AND OTHER EARLY DOCUMENTATIONS OF CADURCERYX The genus Cadurceryx, with the type (and then only known) species Cadurceryx filholi, was originally described by Hoffstetter & Rage (1972), based on remains (exclusively isolated vertebrae) coming from several Eocene sites of France. The holotype vertebra (MNHN. F. QU 16301), originating from an unidentified locality in the Phosphorites du Quercy (“ anciennes collections ”), was defined by Hoffstetter & Rage (1972) as “ une vertèbre complexe à tubercules costaux ”, owing to its complex morphology, as originating from the posterior portion of the column; subsequently Rage (1984 b) treated the holotype as a probably anterior cloacal vertebra. The astonishing statement of Rage (1984 b) that the holotype vertebra of Cadurceryx filholi may have come from the cloacal region of the column seems apparently to be an unintentional mistake (although certainly difficult to be explained how the mistake was made)! Jean-Claude Rage, who on many occasions in the 1990 s and early 2000 s was discussing with ZS about Cadurceryx, never actually supported the above opinion; what is more, the only figure of a trunk vertebra of Cadurceryx (UM [formerly USTL] LAV 1275), morphologically not differing from the holotype, published by Rage (2013: fig. 3 h, i) was described in the caption as just “ trunk vertebra ”. In any case, as we demonstrated in the descriptions above, the morphology of the holotype specimen matches that of a posterior trunk vertebra. Eight other trunk vertebrae, also belonging to the old collections of the Phosphorites du Quercy, were referred to C. filholi, as well (Hoffstetter & Rage 1972). Of them, four vertebrae display the complex morphology (MNHN. F. QU 16304; Hoffstetter & Rage 1972: fig. 6 B; this paper, Fig. 17 A-F), similarly to the holotype, whereas four others (MNHN. F. QU 16306; Hoffstetter & Rage 1972: fig. 6 C) are simply built, devoid of any additional apophyses (“ vertèbres dorsales typiques ”). Hoffstetter & Rage (1972) attributed the latter vertebrae to the anterior trunk portion of the column of C. filholi, justifying that decision with their similarity to trunk vertebrae of the extant “ erycine ” snakes (“ présentent le style des Erycinae ”), in which they envisaged Cadurceryx to pertain. Hoffstetter & Rage (1972) noticed that, at first glance, the complexity pattern of the trunk vertebrae of Cadurceryx closely resembles (but not in all aspects) that characteristic of caudal vertebrae of the living erycids and charinaids. Apart from the remains belonging to the “ old collections ” from the Phosphorites du Quercy, Hoffstetter & Rage (1972) described also a number of vertebrae from precisely dated Eocene localities in France. These were identified as either Cadurceryx cf. filholi or Cadurceryx sp., and originated from the middle Eocene (MP 16) of Robiac and the late Eocene (MP 17) Quercy localities of Malpérié and Les Pradigues. Among them, Hoffstetter & Rage (1972) found a few complex caudal (or cloacal) vertebrae (“ vertèbres caudales typiques ”) (UM MAL 501 and UM MAL 503; Hoffstetter & Rage 1972: 95, fig. 6 D, pl. I fig. 4; this paper, Fig. 11 A-J). They noticed that these vertebrae are less complex than the preceding trunk vertebrae (“ vertèbres complexes à tubercules costaux ”) and identified them as belonging to Cadurceryx sp. (and not C. filholi). At least one vertebra from Malpérié (UM MAL 502), described briefly and illustrated byHoffstetter & Rage (1972: pl. I, fig. 3), with reduced protruding elements accompanying the zygapophyses and provided with a distinct (and small anteroposteriorly) hypapophysis, apparently pertained (in the light of our present study; see above) to the anterior trunk portion of the column. Hoffstetter & Rage (1972), however, referred it as a “ vertèbre complexe à tubercules costaux ”, therefore as coming from the posterior trunk portion of the column, whereas its hypapophysis (short but distinct) was unexpectedly considered a haemal keel (“ carène hémale ”). This (rather strange) judgement may have resulted from the absolute (?) conviction of these authors that all anterior trunk vertebrae of Cadurceryx were simply built, therefore the vertebra in question should have originated from another region of the column. In the section devoted to the fossils from Malpérié, Hoffstetter & Rage (1972: 95) mentioned the presence of vertebrae provided with cloacal hypapophyses (“ hypapophyse « cloacale » ou « précloacale » ”), the feature characteristic of posteriormost trunk vertebrae, i. e., those located immediately before cloacal vertebrae. It is not certain, based on Hoffstetter & Rage’s (1972) text, whether these vertebrae belonged to the fossils coming from Malpérié or another locality; we have been unable to trace these vertebrae. The caudal vertebrae from Malpérié (originally referred to Cadurceryx sp.) attracted a special attention of Hoffstetter & Rage (1972: 95), who observed (considering this fact as curious) that they are a little less complex (“ un peu moins complexes ”) than the preceding (i. e., trunk) ones. These authors also noticed that the arrangement of additional apophyses surmounting the caudal vertebrae differs from that observed in the trunk vertebrae, in particular “ les ptérapophyses sont confondues avec la lame qui surmonte les postzygapophyses [= anterior extensions of postzygapophyses] ” (Hoffstetter & Rage 1972: 95). It is difficult, however, to guess from the above statement which of the either two structures is lacking. In our opinion, the anterior extensions of postzygapophyses are the elements missing in the caudal (and also cloacal) vertebrae, whereas the prominent pterapophyses are maintained in the entire post-trunk portion of the column. This explains the absence of interzygapophyseal fusions in post-trunk vertebrae (see below). Subsequently to this 1972 original description, Cadurceryx was mentioned (but not described in detail) from several Eocene localities in the Phosphorites du Quercy (see also Discussion below). In the literature, there exist illustrations of three previously undescribed vertebrae. Of them, two caudal vertebrae were illustrated: Cadurceryx cf. filholi from Le Bretou (MP 16) (UM BRT 1317; Rage 1988: fig. 17) and Cadurceryx filholi from Rosières 2 (MP 17) (UM ROS 2 266; Rage 1984 b: fig. 16 D; this paper, Fig. 11 K-M). As revealed by our current study, these two vertebrae rather “ deviate ” from the morphology “ typical ” of Cadurceryx filholi; it is unclear if this deviation could reflect some species level distinction, but we tentatively treat it here as some kind of intraspecific variation. The third illustrated specimen (shown in anterior and dorsal views) is a trunk vertebra from Lavergne (UM [formerly USTL] LAV 1275; Rage 2013: fig. 3 h, i). Finally, Szyndlar & Rage (2003), based on unpublished observations, briefly reported that additional apophyses in Cadurceryx were not restricted to the posterior trunk portion of the column but that they were also present on more anteriorly located trunk vertebrae. As a further note, based on the herein documentation that Cadurceryx possessed complex structures and additional apophyses throughout its vertebral column, from its beginning to the end of the tail, it is evident that the trunk vertebrae devoid of additional apophyses that were presented in the original Hoffstetter & Rage (1972) publication (“ vertèbres dorsales typiques ” of Hoffstetter & Rage 1972; e. g. specimens MNHN. F. QU 16306; see Hoffstetter & Rage 1972: fig. 6 C) cannot belong to this taxon and instead they are here reidentified as belonging to some other snake (provisionally referred here as Constrictores indet.). STATUS OF THE ENGLISH MATERIAL REFERRED TO CADURCERYX The first snake remains identified as Cadurceryx outside of France were reported from the Eocene of England. The oldest fossil, a single “ Cadurceryx - like caudal vertebra ”, was reported from the middle Eocene locality of Creechbarrow (MP 16) by Hooker (1986), with this record being reproduced as Cadurceryx sp. by Milner (1986) and cf. Cadurceryx sp. indet. by Benton & Spencer (1995). Unfortunately, these were all simple mentions and no description or figure or that material was published. The presence of other material of Cadurceryx sp. was mentioned also from the late Eocene of Hordle Cliff (MP 17) by Milner (1986); also in this case, no information about the above fossil (s) was provided. Nevertheless, other excavations in Hordle Cliff, starting at the 1990 s, yielded new fossil ophidian remains, which were eventually housed in the collections of MSUVP. Of them, five caudal vertebrae from the collections of MSUVP were described as the new species Cadurceryx pearchi by Holman et al. (2006). Interestingly, the above paper did not include any mention about possible presence or absence of trunk vertebrae in the examined MSUVP material, while four other snake papers that were also based on the Hordle Cliff material from the MSUVP collections (Holman 1993, 1996; Holman & Harrison 1998 a, b), provided no evidence for the existence of trunk vertebrae of Cadurceryx. Accordingly, there is no evidence whether trunk vertebrae were actually absent in this MSUVP collection or they (if present) were ignored by the above authors. Anyway, possible absence of trunk vertebrae in the situation when caudals are present, seems astonishing (but this is anyway not the only such case for this taxon; see also Discussion below). Apparently, in order to compare the newly erected species Cadurceryx pearchi with the type species Cadurceryx filholi, Holman et al. (2006) used only one illustration of a caudal vertebra referred to the latter species, published in the snake volume of Handbuch der Paläoherpetologie by Rage (1984 b: fig. 16 D; this paper, Fig. 11 K-M). The choice of this vertebra (UM ROS 2 266; originating from Rosières 2) seems unfortunate, because its morphology, especially in the anterior aspect, notably deviates from that characteristic of most caudal vertebrae of Cadurceryx. However, the reason for which Holman et al. (2006) chose just this vertebra seems quite clear: it was (and so far still is!) the only illustrated caudal vertebra identified to the species level as Cadurceryx filholi. At those times, illustrations of three other caudal vertebrae referred to Cadurceryx were available in the literature (two from Malpérié: UM MAL 501 and UM MAL 503; Hoffstetter & Rage 1972: fig. 6 D; pl. I, fig. 4; this paper, Figs 11 A-D, E-H; and one from Le Bretou: UM BRT 1317; Rage 1988: fig. 17), but they were ignored byHolman et al. (2006), probably because they were identified either only to the genus level for the material of Malpérié (Cadurceryx sp. of Hoffstetter & Rage 1972) or were not confidently identified anyway to the species level (Cadurceryx cf. filholi of Rage 1988). Ironically, the two vertebrae from Malpérié (as we can state today) display clearly the morphology characteristic of the species C. filholi, unlike the caudal one from Rosières 2. Hence, the diagnosis of Cadurceryx pearchi seems inaccurate, because it reflects exclusively the differences between the single vertebra from Rosières 2 and the English holotype (MSUVP 2061; Holman et al. 2006: figs 2, 3), stating that: 1) C. pearchi has the cotyle larger than the neural canal, while in C. filholi vice versa; and 2) the prezygapophyses of C. pearchi are produced into knoblike tubercles, absent in C. filholi. However, the relatively broad neural canal in the Rosières fossil (feature 1) apparently results from the juvenile age of the snake (e. g. Georgalis & Scheyer 2019; Szyndlar & Georgalis 2023). Regarding the second diagnostic trait of C. filholi (feature 2), namely the “ anterior ends of prezygapophyses […] produced into knoblike tubercles ”, it is very difficult to guess what the latter term means based on Holman’s et al. (2006: 58) text and figures. The only reasonable explanation is that the authors used the term “ knoblike tubercle ” as a synonym of “ prezygapophyseal accessory process ”. The prezygapophyseal accessory processes (prominent in most cases), however, occur in caudal vertebrae of C. filholi; no doubt, their absence in the vertebra from Rosières results again from the juvenile age of the snake. To close the comments on the paper by Holman et al. (2006), we should add that one of the paratypes of Cadurceryx pearchi (MSUVP 2059) was listed as “ Caudal [sic] vertebrae of Pterygoboa sp. indet. (MSUVP 2059) ” on their caption of the drawing of the specimen (Holman et al. 2006: 57, fig. 3 i-l), although the same (single) vertebra was described as “ Caudal [sic] vertebrae of Cadurceryx pearchi sp. nov. paratype (MSUVP 2059) ” on their caption of the photograph of the specimen (Holman et al. 2006: 56, fig. 2 g-i). This is most probably a misspelling, as we see no reason why Holman et al. (2006) regarded this particular specimen as belonging to Pterygoboa. In any case, there is so far no evidence that the genus Pterygoboa Holman, 1976, a bizarre “ erycine ” from the Oligo-Miocene of North America was ever present in England or Europe in general; besides, the vertebral morphology of Pterygoboa (that is also characterized by the presence of additional apophyses) is drastically different from Cadurceryx (Holman 1976, 1998; Mead & Schubert 2013). Here we present a few caudal vertebrae from the late Eocene (MP 17 A) of Christchurch Bay at Hordle Cliff (Mammal Bed, Totland Bay Member), England, from the collections of NHMUK (Figs 19 - 21). These specimens are apparently the same ones mentioned in the faunal list of Milner (1986). Interestingly, in this NHMUK sample, among numerous vertebrae of several other snake taxa from Hordle Cliff, there were only a few Cadurceryx - like caudal vertebrae available and not a single trunk vertebra. Nevertheless, the vertebrae still seem to pertain to different regions of the caudal vertebral column, as it can be attested by the differences regarding the anteroposterior length and dorsoventral height across the specimens (Figs 19 - 21). In any case, the new English caudal vertebrae bear practically no important differences from the known caudal vertebrae of Cadurceryx filholi from the Phosphorites du Quercy. These being said, to summarize, the English form from Hordle Cliff seems to could potentially belong to Cadurceryx, although the absence of any trunk vertebra renders such taxonomic identification as tentative (at max). Nevertheless, based on the diagnosis of Holman et al. (2006), coupled with the observation of the NHMUK caudal vertebrae, a species level distinction of the English Cadurceryx from the French species Cadurceryx filholi (although possible) cannot be demonstrated with certainty. The holotype of Cadurceryx pearchi and all known specimens from England are all incomplete caudal vertebrae and the diagnosis proposed by Holman et al. (2006) refers to variable features. Accordingly, Cadurceryx pearchi is considered to be a nomen dubium, with the material potentially pertaining to an indeterminate species of Cadurceryx (i. e.,? Cadurceryx sp.). Only new and more complete material from the Eocene of England, ideally comprising also trunk vertebrae, will eventually show whether indeed a distinct species of this genus was inhabiting England or whether they are conspecific with the French species Cadurceryx filholi or even whether these “ puzzling ” caudal vertebrae from Hordle pertain instead to a (different) taxon (potentially an erycid or a charinaid).	en	Szyndlar, Zbigniew, Georgalis, Georgios L. (2025): When snake vertebrae go to an extreme - revision, vertebral morphology, and intracolumnar variation of the enigmatic snake Cadurceryx Hoffstetter & Rage, 1972, from the Eocene of Europe. Comptes Rendus Palevol 24 (29): 587-620, DOI: 10.5852/cr-palevol2025v24a29, URL: https://doi.org/10.5852/cr-palevol2025v24a29
