taxonID	type	description	language	source
038287CBFFC2FFA7792751CEEB3DB54F.taxon	description	urn: lsid: zoobank. org: act: 39 A 734 F 4 - 1403 - 43 C 1 - 9 D 99 - 1 DB 3 F 1901478	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FFA7792751CEEB3DB54F.taxon	type_taxon	TYPE SPECIES. — Woznikella triradiata n. sp.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	description	(Figs 1 - 16) urn: lsid: zoobank. org: act: DACADFD 8 - C 682 - 4872 - 82 C 0 - 7 CB 34025868 D	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	materials_examined	HOLOTYPE. — ZPAL V. 34 / 1, partial skeleton.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	diagnosis	DIAGNOSIS. — A medium-sized stahleckeriid-related dicynodont with triradiate, Y-shaped branching pattern of premaxillary ridges, paired medial ridge on the mental surface of the dentary, slender scapula, wakly-angled scapular spine, well-demarcated acromion process projecting well beyond the border of the scapula, and coracoid foramen completely enclosed by the procoracoid, characterized by an autapomorphy: acromion directed anterodorsally rather than anteroventrally or anteriorly. Differs from non-kannemeyeriiform dicynodonts and Dinodontosaurus tener in presence of a distinct supinator process. Differs from most shansiodontids and stahleckeriids in the gracile scapular blade withsignificant terminal flaring of the scapula. Differs from more derived stahleckeriines in having a less pronounced supinator process. Differs from Ischigualastia jenseni Cox, 1962 and Jachaleria candelariensis Araujo & Gonzaga, 1980 in scapula not contributing to the coracoid foramen. Differs from Stahleckeria potens von Huene, 1935, Placerias hesternus Lucas, 1904, Eubrachiosaurus browni, and Zambiasaurus submersus Cox, 1969 in having a weakly developed scapular spine.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	etymology	DERIVATIO NOMINIS. — Woznikella in reference to the name of the type locality, Woźniki; triradiata in reference to the triradiate branching pattern of the premaxillary ridges. TYPE LOCALITY. — Woźniki, southern Poland.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	materials_examined	TYPE HORIZON. — The Patoka Member of the Grabowa Formation, Carnian (Sulej et al. 2011, see Szulc et al. 2015 b for the arguments for the Norian age of the locality; see Discussion for the rationale for its Carnian age). REFERRED SPECIMEN. — SMNS 91416, a fragmentary mandible from the Carnian of Markt Obernzenn (Bavaria, Germany), Stuttgart Formation (Shilfsandstein).	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	distribution	DISTRIBUTION. — Carnian (and? Norian) of Poland and Germany.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
038287CBFFC2FF8F79475229E85FB08A.taxon	description	DESCRIPTION Skull and mandible The preserved parts of the skull and mandible (Figs 1 - 6) lack pronounced rugosities, which may be either taxonomic or related to the young age of the individual. Based on the shape of the beak (both of the premaxilla and dentary) and the width of the preserved skull table elements, at least preorbitally the skull was apparently relatively narrow (Fig. 7). Unlike in, e. g., Kannemeyeria aganosteus Kammerer & Ordoñez, 2021, Kannemeyeria simocephalus, Kannemeyeria lophorhinus Renaut, Damiani, Yates & Hancox, 2003, Lystrosaurus spp. (with the possible exception of L. youngi), Rechnisaurus cristarhynchus Chowdhury, 1970, Sangusaurus edentatus Cox, 1969, Sangusaurus parringtonii Cruickshank, 1986 a, Shaanbeikannemeyeria xilougouensis Cheng, 1980, Shansiodon wangi Yeh, 1959, Shansiodon wuhsiangensis Yeh, 1959, Stahleckeria potens, Sungeodon kimkraemerae Maisch & Matzke, 2014, Tetragonias njalilus (von Huene, 1942), Ufudocyclops mukanelai, Vinceria andina Bonaparte, 1969, Wadiasaurus indicus Chowdhury, 1970, and Zambiasaurus submersus Cox, 1969 there is no conspicuous median ridge on the skull (Weithofer 1888; Broom 1899; Haughton 1915; Pearson 1924 a; Case 1934; Yuan & Young 1934 b; Young 1935; von Huene 1942; Camp 1956; Yeh 1959; Sun 1964; Cruickshank 1967; Chowdhury 1970; Crozier 1970; Cluver 1971; Kalandadze 1975; Cheng 1980; Cruickshank 1986 a; Bandyopadhyay 1989; Pickford 1995; Schwanke-Peruzzo & Araújo-Barbarena 1995; Maisch 2001; Renaut et al. 2003; Surkov et al. 2005; Vega-Dias et al. 2005; Morato 2006; Grine et al. 2006; Domnanovich & Marsicano 2012; Angielczyk et al. 2014; Maisch & Matzke 2014; Angielczyk et al. 2017; Kammerer et al. 2019; Kammerer & Ordoñez 2021). Note that the presence and degree of development of the median ridge is ontogeny-dependent at least in some dicynodont taxa (e. g., Kammerer & Ordoñez 2021). Comparison of overall proportions of the mandiblular bones shows that the mandible of Woznikella triradiata n. gen., n. sp. was proportionally significantly longer relative to its height than that of; e. g., Angonisaurus cruickshanki, Ischigualastia jenseni, Lystrosaurus spp., Myosaurus gracilis Haughton, 1917, Sangusaurus parringtonii, and Stahleckeria potens, and more reminiscent to that of, e. g., Kannemeyeria simocephalus, Sinokannemeyeria spp., or Wadiasaurus indicus (see, e. g., Cox & Li 1983; Pearson 1924 a; von Huene 1935, 1936; Young 1935; Broom 1937; Janensch 1952; Camp 1956; Sun 1963, 1964; Cox 1965; Cluver 1971, 1974; Keyser 1974; Cruickshank 1986 a; Bandyopadhyay 1988; Renaut 2000; Hancox et al. 2013; Angielczyk et al. 2017). Premaxilla The premaxilla (ZPAL V. 34 / 1 / 2; Fig. 1) is fused and well preserved. Only the right side and the tip of the left side of the dorsal process as well as a small fragment of the posteriormost tip are broken. The anterior part is skewed to the left, but it is unclear whether this deformation is taphonomical or if it occurred during the life of the animal. The anterior tip in dorsoventral aspect is blunt (squared off), rather than sharply pointed as in, e. g., Kannemeyeria lophorhinus, apparently Moghreberia nmachouensis Dutuit, 1980, Placerias hesternus, “ Putillosaurus sennikovi ” Surkov, 2005, and Ufudocyclops mukanelai (see Camp & Welles 1956; Cox 1965; Dutuit 1980, 1988; Renaut 2000; Surkov 2005; Angielczyk et al. 2014; Kammerer et al. 2019). Nonetheless, it appears to be narrower than in Dinodontosaurus brevirostris Cox, 1968, Dinodontosaurus tener (von Huene, 1935), Dolichuranus primaevus Keyser, 1973, Jachaleria candelariensis, Rechnisaurus cristarhynchus, Sinokannemeyeria sanchuanheensis Cheng, 1980, and Stahleckeria potens, in which the bone is about as wide as it is long in dorsal view and shows nearly no rostral tapering (Tupi Caldas 1936; Cox 1965, 1968; Chowdhury 1970; Keyser 1973; Araújo & Gonzaga 1980; Cheng 1980; Bandyopadhyay 1989; Maisch 2001; Vega-Dias & Schultz 2004; Vega-Dias et al. 2005; Morato 2006; Kammerer & Ordoñez 2021). Its edge and anterodorsal surface are slightly concave medially, so the beak is M-shaped in dorsoventral aspect, similar to Ischigualastia jenseni, Jachaleria candelariensis, Parakannemeyeria chengi Liu, 2004, Parakannemeyeria ningwuensis Sun, 1963, Parakannemeyeria youngi Sun, 1963, Rhadiodromus mariae Surkov, 2003, Uralokannemeyeria vjuschkovi Danilov, 1971, and Stahleckeria potens (see Sun 1963; Cox 1965; Danilov 1971; Araújo & Gonzaga 1980; Maisch 2001; Surkov 2003; Liu 2004; Vega-Dias & Schultz 2004; Kammerer & Ordoñez 2021). In contrast to, e. g., Dinodontosaurus brevirostris, Dinodontosaurus tener, Ischigualastia jenseni (pers. obs.: PVSJ 545), Jachaleria candelariensis, Lystrosaurus spp. (with the possible exception of L. georgi, L. hedini Young, 1935, and L. youngi), Parakannemeyeria ningwuensis, and Sinokannemeyeria yingchiaoensis Sun, 1963, there is no pronounced notch in the median part of the tomial edge (unless the tips of the beak of ZPAL V. 34 / 1 / 2 are damaged; Yuan & Young 1934 b; Young 1935; Sun 1963, 1964; Cluver 1971; Kalandadze 1975; Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004; Surkov 2005; Morato 2006; Grine et al. 2006; Kammerer & Ordoñez 2021). In lateral view (Fig. 1 C), accounting for deformation, the labial (tomial) edge of the premaxilla was apparently directed anteroventrally, roughly perpendicular to the anterodorsally facing surface of the nasal process. This differs from at least some specimens of Tetragonias njalilus, in which the labial edge of the beak was set at an obtuse angle to its anteroventrally facing surface and thus the tip of the beak was slightly recessed (Cruickshank 1967; Kammerer & Ordoñez 2021). The lateral surface of the anterior half is coarsely rugose, indicative of a keratinous rhamphotheca (Fig 1 B, D, E). The premaxilla forms the anterior edge of the naris. The anterior outline of the narial fossa is straight in the upper part and reaches very close to the ventral edge of the premaxilla, similar to, e. g., Acratophorus argentinensis (Bonaparte, 1965), Kannemeyeria simocephalus, Parakannemeyeria dolichocephala Sun, 1960, Parakannemeyeria ningwuensis, Parakannemeyeria youngi, “ Putillosaurus sennikovi ”, Rechnisaurus cristarhynchus, Repelinosaurus robustus Olivier, Battail, Bourquin, Rossignol, Steyer & Jalil, 2019, Rhadiodromus mariae, Rhinodicynodon gracile Kalanadze, 1970, Rabidosaurus cristatus Kalanadze, 1970, the South African Shansiodon sp., Sinokannemeyeria pearsoni Young, 1937, Sinokannemeyeria sanchuanheensis, Ufudocyclops mukanelai, Vinceria andina, Wadiasaurus indicus, and Xiyukannemeyeria brevirostris (Sun, 1978) (see Haughton 1915; Pearson 1924 a; Case 1934; Watson 1948; Sun 1960, 1963; Cruickshank 1965; Bonaparte 1966 a, 1967; Chowdhury 1970; Kalandadze 1970; Cheng 1980; Bandyopadhyay 1988, 1989; Renaut 2000; Renaut & Hancox 2001; Liu & Li 2003; Surkov 2003, 2005; Domnanovich & Marsicano 2012; Hancox et al. 2013; Kammerer et al. 2019; Olivier et al. 2019; Kammerer & Ordoñez 2021). The premaxilla differs dramatically in proportions from the anteroposteriorly short and vertically elongated premaxillae of Lystrosaurus spp. (Yuan & Young 1934 b; Young 1935; Sun 1964; e. g., Cluver 1971; Colbert 1974; Kalandadze 1975; Surkov et al. 2005; Grine et al. 2006). On the ventral (palatal) surface of the premaxilla, each of the anteriorly pronounced tips of the beak projects rearward to form a predominantly posteriorly and gently medially directed, rugose ridge (Fig. 1 C). Anteriorly, these ridges are separated by the median palatal groove, but merge at approximately 2 / 3 of the premaxilla’s length into a single, higher, posteriorly directed median (posterior) palatal ridge, resulting in a Y-shaped structure, similar to most Triassic dicynodonts with the exception of Kombuisia frerensis Hotton, 1974 and Myosaurus gracilis, which lack the anterior ridges (Cluver 1974; Hotton 1974; Fröbisch 2007). There are no lateral anterior ridges comparable to those of Kombuisia frerensis (see Hotton 1974). The anterior ridges in Woznikella triradiata n. gen., n. sp. are more widely separated than in, e. g., Jachaleria canderlariensis and Kannemeyeria lophorhinus (Crozier 1970; Araújo & Gonzaga 1980; Pickford 1995; Renaut 2000; Vega-Dias & Schultz 2004; Angielczyk et al. 2014). Unlike in, e. g., Acratophorus argentinensis, Angonisaurus cruickshanki Cox & Li, 1983, Counillonia superoculis Olivier, Battail, Bourquin, Rossignol, Steyer & Jalil, 2019, “ Cristonasus koltaeviensis ”, Dolichuranus primaevus, Kannemeyeria simocephalus, Kannemeyeria lophorhinus, “ Kannemeyeria ” latirostris Crozier, 1970, Jachaleria candelariensis, Parakannemeyeria dolichocephala, “ Putillosaurus sennikovi ”, Rechnisaurus cristarhynchus, Repelinosaurus robustus, Sangusaurus edentatus, Sangusaurus parringtonii, Sinokannemeyeria yingchiaoensis, Ufudocyclops mukanelai, Uralokannemeyeria vjuschkovi, and Zambiasaurus submersus, the ridges diverge anteriorly rather than being parallel (Pearson 1924 a; Case 1934; Toerien 1953; Sun 1960, 1963; Bonaparte 1966 a; Cox 1969; Chowdhury 1970; Crozier 1970; Danilov 1971; Keyser 1973; Araújo & Gonzaga 1980; Cox & Li 1983; Bandyopadhyay 1989; Pickford 1995; Schwanke-Peruzzo & Araújo-Barbarena 1995; Surkov 1999 a; Renaut 2000; Vega-Dias & Schultz 2004; Surkov 2005; Damiani et al. 2007; Hancox et al. 2013; Angielczyk et al. 2014, 2017; Kammerer et al. 2019; Olivier et al. 2019; Kammerer & Ordoñez 2021). They nearly reach the anterolateral corners of the beak, unlike, e. g., in Acratophorus argentinensis, Dinodontosaurus spp., Jachaleria candelariensis, Lystrosaurus spp., Rechnisaurus cristarhynchus, Shaanbeikannemeyeria xilougouensis, Stahleckeria potens, Vinceria andina, and Wadiasaurus indicus, in which the snout stretches out further laterally (von Huene 1935; Toerien 1953; Cox 1965, 1968; Bonaparte 1966 a; Chowdhury 1970; Cluver 1971; Cheng 1980; Araújo & Gonzaga 1980; Bandyopadhyay 1988, 1989; Renaut 2000; Renaut & Hancox 2001; Vega-Dias & Schultz 2004; Morato 2006; Domnanovich & Marsicano 2012; Abdala et al. 2013; Kammerer & Ordoñez 2021). The three ridges are continuous with each other and the point of their branching is well demarcated and ventrally convex, in contrast to Jachaleria candelariensis, Kannemeyeria lophorhinus, Rabidosaurus cristatus (pers. obs.), and Rhinodicynodon gracile, in which the posterior (median) ridge is flanked by the anterior ones but not connected to them (Crozier 1970; Araújo & Gonzaga 1980; Pickford 1995; Renaut 2000; Vega-Dias & Schultz 2004; Angielczyk et al. 2014; Kammerer & Ordoñez 2021) or, e. g., Acratophorus argentinensis, Angonisaurus cruickshanki, Counillonia superoculis, Dolichuranus primaevus, “ Kannemeyeria ” latirostris, Lystrosaurus spp., Parakannemeyeria dolichocephala, Placerias hesternus, Rechnisaurus cristarhynchus, Repelinosaurus robustus, Sangusaurus parringtonii, Shaanbeikannemeyeria xilougouensis, the South African Shansiodon sp., Sinokannemeyeria baidaoyuensis Liu, 2015, Ufudocyclops mukanelai, Vinceria andina, Wadiasaurus indicus, and Zambiasarus submersus, in which the ridges are separate and the branching point is indistinct (Camp 1956; Camp & Welles 1956; Sun 1960; Cox 1965, 1969; Crozier 1970; Chowdhury 1970; Cluver 1971; Keyser 1973; Cheng 1980; Cox & Li 1983; Bandyopadhyay 1988, 1989; Renaut 2000; Renaut & Hancox 2001; Damiani et al. 2007; Domnanovich & Marsicano 2012; Hancox et al. 2013; Angielczyk et al. 2014, 2017; Liu 2015; Kammerer et al. 2019; Olivier et al. 2019; Kammerer & Ordoñez 2021). In Kannemeyeria simocephalus, both of the latter morphologies can be observed, but the lateral and median palatal ridges never seem to form a well-defined connection (Pearson 1924 a; Case 1934; Toerien 1953; Renaut 2000). This character is usually poorly presented in published figures, but apparently the only Triassic dicynodonts with such a well-defined triradiate branching pattern of the premaxillary ridges comparable to that of Woznikella triradiata n. gen., n. sp. Are Dinodontosaurus brevirostris and Dinodontosaurus tener (see Cox 1965, 1968; Morato 2006; Kammerer & Ordoñez 2021), Ischigualastia jenseni (see Kammerer & Ordoñez 2021; pers. obs.: PVSJ 545), Kannemeyeria aganosteus (see Kammerer & Ordoñez 2021), Rhadiodromus mariae (see Surkov 2003), Tetragonias njalilus (von Huene 1942; Cruickshank 1967; Hancox et al. 2013), Uralokannemeyeria vjuschkovi (see Danilov 1971), and the specimens illustrated by Camp (1956) as a juvenile Kannemeyeria “ vanhoepeni ” and by Cruickshank (1965) as a juvenile Kannemeyeria “ latifrons ” Broom 1899. In the latter two cases this is based only on the interpretative drawings – this morphology is not clearly described, and no photographs of these specimens were published, so the accuracy of the drawings cannot be verified. In Ufudocyclops mukanelai, the anterior and posterior ridges do not form a connection, but the surface between them is slightly convex (Hancox et al. 2013; Kammerer et al. 2019). As preserved, the anteriormost parts of these ridges are slightly deeper than the lateral labial edge, and thus visible in lateral view, in contrast to, e. g., Placerias hesternus (see Camp & Welles 1956; Cox 1965). The area for the connection with the palatines and vomers is very poorly preserved, so it is not possible to establish the anterior extent of those bones. The dorsal surface of the palatal plate is convex and a single longitudinal ridge spans along its entire midline (Fig. 1 A). Large parts of the surfaces on either side of this ridge probably served as facets for the maxillae. These facets are outlined with gentle, anterodorsally aligned, posterodorsally convex grooves, suggesting that the maxillae nearly reached anteriorly the front of the narial fossa, spanning along approximately posterior two thirds of the premaxillae length (Fig. 1 B, D). This is relatively further than in, e. g., “ Nasoplanites danilovi ”, Placerias hesternus, and “ Putillosaurus sennikovi ”, but comparable to Kannemeyeria simocephalus (see Pearson 1924 a; Camp & Welles 1956; Surkov 1999 a; Renaut 2000; Surkov 2005). Both premaxillae are damaged in a way that obscures the connections with the septomaxillae. Nasal The right nasal (ZPAL V. 34 / 1 / 42; Fig. 2 A-D) is preserved almost completely. It consists of two parts, lateral and dorsal. The bone is slightly deformed due to compaction, which is clearly visible in the anteromedial part, which is concave dorsally (not corresponding with the alignment of the preserved part of the dorsal process of the premaxilla, thus unlikely to represent nasal depressions; compare to, e. g., Chowdhury 1970; Bandyopadhyay 1989; Kammerer & Ordoñez 2021). The deformation also likely affected the angle between the lateral and dorsal plates (Fig. 2 B). Unlike in, e. g., “ Planitorostris pechoriensis ” (see Surkov 1999 b), the lateral part has a coarsely rugose surface (although the development of the rugosity is mild in comparison with many Kannemeyeriiform taxa) with numerous large vascular openings. It suggests that Woznikella triradiata n. gen., n. sp. had either a sensitive tip of the snout or keratinous covering in that area (e. g., Keyser & Cruickshank 1979; Surkov 2003, 2006; Morato et al. 2005; Morato 2006; Benoit et al. 2018). The lateral surface of the nasal is roughly even with the lateral edge of the dorsal table and is vertical and mostly flat, without prominent lateral depression (Fig. 2 D), unlike in, e. g., some specimens of Dinodontosaurus brevirostris and Dinodontosaurus tener, Rhadiodromus klimovi (Efremov, 1938), Rhadiodromus mariae, Rhinodicynodon gracile, Sinokannemeyeria yingchiaoensis, and Xiyukannemeyeria brevirostris (Sun 1963, 1978; Cox 1965, 1968; Vjuschkov 1969; Kalandadze 1970; Surkov 2003; Morato 2006). This may imply the absence of a postnarial depression (although the lack of maxilla makes it uncertain). The ventral edge of the lateral part is sinusoidal. Two ventral processes, which formed the sutural connection with the lacrimal, have distinct ridges on the lateral surfaces. A shallow depression, which is probably the area for the suture with the dorsal process of the premaxilla, is visible on the dorsal surface (Fig. 2 A). The edge of this sutural field is directed posteromedially and it spans for about half the length of the dorsal plate of the nasal. Aside from that, there are no pronounced grooves comparable with those described in, e. g., “ Calleonasus furvus ” or “ Elatosaurus facetus ” (see Kalandadze & Sennikov 1985; Surkov 1999 a). The contact with the septomaxilla was apparently very small, only at the edge of the naris, due to most of the ventral edge of the nasal being occupied by the lacrimal. The contact with the prefrontal is slightly broken and difficult to interpret, but at least part of the facet for the prefrontal is preserved in the posterior third of the bone. As preserved, the medial portions of the nasals and frontals are separated by a diamond-shaped empty space, which was filled either by a short, pointy median anterior process of the frontal, or by a supernumerary internasal bone (Fig. 7 A, B), as in Jachaleria spp. or some individuals of Lystrosaurus spp. (Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004; Jasinoski et al. 2014; Kammerer & Ordoñez 2021). The nasal roofed the naris, and its anterior edge formed the thick posterodprsal edge of the nostril (Fig. 2 B, C), similar to, e. g., Dolichuranus primaevus, Acratophorus argentinensis, Kannemeyeria lophorhinus, Kannemeyeria simocephalus, Parakannemeyeria dolichocephala, Parakannemeyeria ningwuensis, Rabidosaurus cristatus, the South African Shansiodon sp., Sinokannemeyeria pearsoni, Sinokannemeyeria yingchiaoensis, Sungeodon kimkraemerae, Ufudocyclops mukanelai, Uralokannemeyeria vjuschkovi, and Vinceria andina (Pearson 1924 a; Case 1934; Sun 1960, 1963; Bonaparte 1966 a, 1967, 1969; Kalandadze 1970; Danilov 1971; Keyser 1973; Pickford 1995; Renaut 2000; Maisch 2001; Renaut & Hancox 2001; Damiani et al. 2007; Domnanovich & Marsicano 2012; Hancox et al. 2013; Maisch & Matzke 2014; Angielczyk et al. 2014; Kammerer et al. 2019; Kammerer & Ordoñez 2021). Lacrimal The right lacrimal (ZPAL V. 34 / 1 / 80; Fig. 2 E-H )) is almost complete, only a small triangular fragment is missing in the middle of the posterior edge. The bone is subtriangular, approximately as high as it is long, and thus unusually equidimensional compared to other dicynodonts. The lateral surface is flat and the sutural surfaces are prominent (Fig. 2 E). The ventral edge is the most massive, with a triangular surface for the contact with the maxilla and, possibly the jugal and / or palatine (Fig. 2 H). Unlike, e. g., Placerias hesternus, there is no conspicuous jugal process (Camp & Welles 1956). The anteroventral orbit margin is formed by a thin lappet of bone. The suture for the prefrontal is inverted V-shaped. Internally, a ridge is visible in the posterior part of the bone (Fig. 2 F). This ridge is most distinctive in the ventral part, where it floors and encases anteriorly a pocket-like lacrimal pit, which opens posteriorly as a large lacrimal foramen (Fig. 2 G). The ridge continues faintly along the posterior edge towards the dorsal part of the bone. The anterior opening of the lacrimal duct is located on the internal surface, in the anterior part of the bone, and is directed anterodorsally. The specimen articulates well with the preserved nasal (ZPAL V. 34 / 1 / 42) and based on that articulation it appears that the lacrimal nearly reached the naris, preventing or nearly preventing the contact between the nasal and the maxilla – the very posterior edge of the naris was formed by another bone, but it is currently impossible to establish whether this was maxilla or septomaxilla (Fig. 7). The anterior extent of the lacrimal, which is usually covered by the more superficial bones, is unknown in many other Triassic dicynodonts, but it is similar to that of Woznikella triradiata n. gen., n. sp. at least in Ischigualastia jenseni, Jachaleria candelariensis, Kannemeyeria simocephalus, Parakannemeyeria dolichocephala, Parakannemeyeria shenmuensis Cheng, 1980, Parakannemeyeria youngi, Placerias hesternus, Rhadiodromus mariae, Shaanbeikannemeyeria xilougouensis, possibly Shansiodon wangi, the South African Shansiodon sp., Stahleckeria potens, Sinokannemeyeria baidaoyuensis, Sinokannemeyeria pearsoni, Sinokannemeyeria sanchuanheensis, Sungeodon kimkraemerae, and Vinceria andina, in which the lacrimal reaches or nearly reaches the nostril (Pearson 1924 a; von Huene 1935; Camp 1956; Camp & Welles 1956; Yeh 1959; Sun 1960, 1963; Cox 1965; Bonaparte 1969; Cheng 1980; Araújo & Gonzaga 1980; Hancox 1998; Renaut 2000; Renaut & Hancox 2001; Surkov 2003; Vega-Dias & Schultz 2004; Domnanovich & Marsicano 2012; Abdala et al. 2013; Hancox et al. 2013; Maisch & Matzke 2014; Liu 2015; Kammerer & Ordoñez 2021). In contrast to Kannemeyeria simocephalus, Parakannemeyeria youngi (and, likely, the other species of that genus), Placerias hesternus, and Stahleckeria potens, however, except for a narrow margin along the dorsoanterior and anterior border, which shows characteristics of a squamous suture, the lacrimal in Woznikella triradiata n. gen., n. sp. was almost entirely exposed externally (Pearson 1924 a; von Huene 1935; Camp 1956; Sun 1963; Renaut 2000; Maisch 2001; Abdala et al. 2013; Kammerer & Ordoñez 2021). In Dinodontosaurus brevirostris and Dinodontosaurus tener the morphology of the lacrimal is variable (Cox 1965, 1968; Morato 2006). In the specimens MCZ 1628, MCZ 3454, DGM 309, and the “ Chanaria platyceps ” holotype (UNLaR 14) the lacrimal has a large exposure, unlike in the specimens MCZ 1670 and MCZ 1687 of the same species. However, in MCZ 1628 it is subrectangular and still significantly separated from the nostril, which is located in the anterior third of the narial fossa, whereas in MCZ 3454, DGM 309, and UNLaR 14 it reaches further rostrally (Cox 1965, 1968). In Dinodontosaurus brevirostris MCZ 3454 and likely UNLaR 14, as well as in Dinodontosaurus tener DGM 309, Parakannemeyeria youngi, Sinokannemeyeria baidaoyuensis, and Sinokannemeyeria pearsoni the lacrimal meets the septomaxilla anteriorly, separating the maxilla from the nasal, but the septomaxilla seems to cover about third of this distance (Sun 1963; Cox 1968; Liu 2015), thus more than in Woznikella triradiata n. gen., n. sp. In Dinodontosaurus tener specimen DGM 530 R the separation is incomplete and the lacrimal is relatively small, probably due to relatively short preorbital length of that skull (Cox 1968). Variable lacrimals, generally large but not separating the nasals from the maxillae, were observed in the series of specimens described by Morato (2006). Some age-related variation was also described in Kannemeyeria simocephalus by Renaut (2000), with larger individuals having proportionally smaller exposure of the lacrimal, but in no case was the anterior part of the lacrimal externally exposed in vivo in that species. In Jachaleria candelariensis the lacrimal, as exposed laterally, is much lower dorsoventrally and more strap-like (Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004). The nasal is also completely or almost completely separated from the maxilla by the lacrimal and septomaxilla in Lystrosaurus spp. and the South African Shansiodon sp., but due to the shortening of the preorbital part of the skull, the lacrimal remains relatively short (e. g., Sun 1964; Cluver 1971; Li 1988; Hancox et al. 2013). Frontal The right frontal (ZPAL V. 34 / 1 / 1; Fig. 3) is almost complete, only the top of the dorsal (preparietal) process, a small fragment of the medial edge, and the anterior tip are broken. The frontal plate is nearly as wide as long and its anterior edge is bowed. The frontal pair thus was fan-shaped in dorsal view and apparently lacked prominent and narrow lateral processes or conspicuous embayments for the nasal or prefrontals, similar to, e. g., Dolichuranus primaevus, Lystrosaurus spp. (with the exception of L. curvatus (Owen, 1876) and L. hedini), the South African Shansiodon sp., Sinokannemeyeria sanchuanheensis, and Tetragonias njalilus, but different from, e. g., Acratophorus argentinensis, Dinodontosaurus spp., Elephantosaurus jachimovitschi Vjuschkov, 1969, Ischigualastia jenseni, Jachaleria candelariensis, “ Kannemeyeria ” latirostris, Moghreberia nmachouensis, Parakannemeyeria dolichocephala, Parakannemeyeria youngi, “ Parvobestiola bashkiriensis ”, Placerias hesternus, Rechnisaurus cristarhynchus, Rhadiodromus mariae, Rhinodicynodon gracile, Shaanbeikannemeyeria xilougouensis, Sinokannemeyeria yingchiaoensis, Stahleckeria potens, Ufudocyclops mukanelai, and Vinceria andina (von Huene 1935, 1942; Young 1935; Camp & Welles 1956; Sun 1960, 1963; Cox 1965, 1968; Bonaparte 1966 a; Cruickshank 1967; Vjuschkov 1969; Crozier 1970; Kalandadze 1970; Chowdhury 1970; Cluver 1971; Cheng 1980; Araújo & Gonzaga 1980; Dutuit 1988; Bandyopadhyay 1989; Lucas & Harris 1996; Surkov 1999 a, 2003; Maisch 2001; Vega-Dias & Schultz 2004; Vega-Dias et al. 2005; Morato 2006; Damiani et al. 2007; Domnanovich & Marsicano 2012; Abdala et al. 2013; Hancox et al. 2013; Kammerer 2018; Kammerer et al. 2019; Kammerer & Ordoñez 2021). The anterior median process is broken but, considering that the posterior edge of the nasal appears to be complete, it likely either was relatively short and ended in a point, or a supernumerary internasal bone was present, as in Jachaleria spp. and some individuals of Lystrosaurus spp. (Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004; Jasinoski et al. 2014; Kammerer & Ordoñez 2021). It must be noted, however, that the layout of the anterior frontal suture exhibits some variability, even bilaterally within a single individual (e. g., Camp 1956; Cox 1965; Renaut 2000; Kammerer & Ordoñez 2021). The frontal is proportionally much wider than in the narrow-roofed skulls of Counillonia superoculis, Myosaurus gracilis, Kombuisia frerensis, Kombuisia antarctica Fröbisch, Angielczyk & Sidor, 2010, and Repelinosaurus robustus (see Haughton 1917; Cluver 1974; Hotton 1974; Hammer & Cosgriff 1981; Fröbisch 2007; Olivier et al. 2019). It is much longer than the anteroposteriorly shortened (at least, as exposed externally) frontal of Sangusaurus parringtonii (see Angielczyk et al. 2017; Kammerer & Ordoñez 2021). The most distinctive feature of the frontal is the strongly elevated dorsal process similar to that of Kannemeyeria simocephalus (e. g., Weithofer 1888; Broom 1937; Watson 1948; Kammerer & Ordoñez 2021). On the anterior and lateral surfaces of this process, the area for sutural contact with the postorbital is visible. This area continues throughout the posterodorsal edge of the bone, reaching its lateral edge, where it ends in a point. The lateral edge of the frontal contributed to the margin of the orbit, and in dorsoventral aspect is slightly oblique to the long axis of the skull (turned anteromedially). The contribution is restricted, similar as in, e. g., Dolichuranus primaevus and Rechnisaurus cristarhynchus, but larger than in at least some specimens of Dinodontosaurus spp., Ischigualastia jenseni, “ Kannemeyeria ” latirostris, Stahleckeria potens, and Uralokannemeyeria vjuschkovi, and smaller than in most specimens of Kannemeyeria simocephalus, Kombuisia spp., and Myosaurus gracilis (see Haughton 1917; Pearson 1924 a; Case 1934; von Huene 1935; Camp 1956; Cox 1965, 1968; Crozier 1970; Danilov 1971; Cluver 1974; Hammer & Cosgriff 1981; Bandyopadhyay 1989; Renaut 2000; Maisch 2001; Vega-Dias et al. 2005; Morato 2006; Damiani et al. 2007; Fröbisch 2007; Fröbisch et al. 2010; Abdala et al. 2013; Kammerer & Ordoñez 2021). This differs from Jachaleria candelariensis, in which the frontal is excluded from the edge of the orbit by the prefrontal and postorbital (Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004). In anterior view, the part of the skull roof formed by the frontal has its medial part noticeably more elevated dorsally than the lateral one, so this part of the skull was convex, similar to, e. g., Ischigualastia jenseni (see Cox 1965; Kammerer & Ordoñez 2021). The pineal foramen is located in the posteromedial corner of the bone. In that aspect, Woznikella triradiata n. gen., n. sp. differs from Kombuisia frerensis, in which the pineal foramen is absent (Hotton 1974; Fröbisch 2007). The area of the suture with the left frontal is damaged, but it appears that the preparietal bone was absent. The internal side is also poorly preserved, and a large fragment of sediment was left unprepared to avoid destroying the specimen, obscuring most of the internal structures. Still, a deep fissure is visible in the center of the internal surface of the bone. Dentary The dentaries (ZPAL V. 34 / 1 / 3; Fig. 4) are fused together at the symphysis. Each dentary is a massive yet narrow bone with a dorsally projected and dorsolaterally pointed anterodorsal tip, triangular in cross-section. Between the tips there is a narrow notch (Fig. 4 D), like in Angonisaurus cruickshanki and Kannemeyeria simocephalus (see Pearson 1924 a; Cox & Li 1983; Renaut 2000; Hancox et al. 2013). In lateral view the dentary is gently hooked (Fig. 4 D), similarly to that of Angonisaurus cruickshanki, Dolichuranus primaevus, Kannemeyeria simocephalus, Kannemeyeria lophorhinus, “ Kannemeyeria ” latirostris, Kombuisia frerensis, Kombuisia antarctica, Lystrosaurus spp., Myosaurus gracilis, Parakannemeyeria dolichocephala, Rhinodicynodon gracile, Sangusaurus parringtonii, Tetragonias njalilus, Wadiasaurus indicus, and Xiyukannemeyeria brevirostris (see Broom 1923, 1937; Pearson 1924 a; von Huene 1942; Sun 1960; Cruickshank 1967, 1986 a; Crozier 1970; Kalandadze 1970; Cluver 1971, 1974; Keyser 1973; Hotton 1974; Hammer & Cosgriff 1981; Cox & Li 1983; Bandyopadhyay 1988; Renaut 2000; Liu & Li 2003; Damiani et al. 2007; Fröbisch 2007; Fröbisch et al. 2010; Hancox et al. 2013; Angielczyk et al. 2014, 2017; Kammerer 2018) and unlike the non-hookeed or minimally hooked mandibles of, e. g., Dinodontosaurus s pp., Ischigualastia jenseni, Parakannemeyeria ningwuensis, Sinokannemeyeria yingchiaoensis, Stahleckeria potens, and Vinceria andina (see von Huene 1935; Romer & Price 1944; Camp 1956; Sun 1963; Cox 1965, 1968; Lucas 2002; Domnanovich & Marsicano 2012; Abdala et al. 2013; Kammerer 2018; Kammerer & Ordoñez 2021; Escobar et al. 2023). Unlike in Dinodontosaurus brevirostris, the labial (tomial) edge is continuous with the posterior part of the bone and directed either dorsally or (in the anteriormost part) posterodorsally, but never predominantly anterodorsally (Cox 1968; Kammerer & Ordoñez 2021; Escobar et al. 2023). The anterior portion of the lateral dentary surface exhibits a coarsely rugose texture indicating the presence of a rhamphotheca-covered beak. The rugosities extend ventrally, towards the mentum, and posteriorly. The posteroventral limit of the rugose area is concave. From the rostral corner of that concavity, rostrodorsally, crossing the rugose surface towards the dorsal edge of the dentaries, paired, anastomosing vascular grooves are visible. Just anterior to them, shallow, gently expressed lateral grooves similar to those of Dolichuranus primaevus, Kannemeyeria simocephalus, Kannemeyeria lophorhinus, Parakannemeyeria ningwuensis, Sangusaurus parringtonii, and Sinokannemeyeria yingchiaoensis span from the lateral surface of each of the tips of the beak posteroventrally towards the ventral edge of the bone (Camp 1956; Sun 1963; Cruickshank 1986 a; Renaut 2000; Damiani et al. 2007; Angielczyk et al. 2017). Like in Kannemeyeria simocephalus, Kannemeyeria lophorhinus, and Sangusaurus parringtonii, the grooves do not follow the ventral curvature of the dentary, but are straight (Camp 1956; Renaut 2000; Angielczyk et al. 2017). Paired, more defined grooves coupled with paired ridges are also present laterally on the dentaries of at least the South African Lystrosaurus spp. (Cluver 1971). The lateral dentary shelf (Cluver 1971) is not developed, but the lateral surfaces of the posterior part of the dentary above the mandibular fenestra show a laterally symmetrical, rounded, indistinct broadening at the level of the split of the dorsal (tomial), medial (lingual), and lateral (labial) laminae of the dentary ramus, at the roof of the Meckelian canal. This differs significantly from the exceptionally pronounced dentary shelf of Pentasaurus goggai Kammerer, 2018 (see Kammerer 2018). The symphysis in dorsal view is relatively long and narrow but the mandible broadens ventrally (Fig. 4 A), similar to the mandibles of, e. g., Kannemeyeria simocephalus, Sinokannemeyeria yingchiaoensis, and Tetragonias njalilus (see Pearson 1924 a; Sun 1963; Cruickshank 1967). It bears a single wide, deep groove in the junction between the left and right rami, as is typical for dicynodonts. The dorsal surface of each ramus shows two parallel longitudinal ridges separated by a marked groove (posterior dentary sulcus posteriorly and dentary table anteriorly; see discussion in Angielczyk & Rubidge 2013) along its length, with the lingual one higher than the labial around the midlength, and diminishing gradually posteriorly and sharply anteriorly. The anterior end of each lingual ridge is associated with a rounded, mediolaterally compressed peg. Anterior to that peg each lingual ridge continues rostrally, but becomes less prominent and gently diverges laterally, making the median trough slightly wider frontally and eventually ending in the lateral point of the beak. This morphology is different than in Dinodontosaurus tener as figured by von Huene (1935) and later by Lucas & Harris (1996), in which the symphyseal part is shorter, the lingual ridges nearly meet anteriorly, and the terminal part of the mandible appears wider and more rounded, but (aside from rostral broadening of the median groove) it is generally similar to that of Kannemeyeria simocephalus and Tetragonias njalilus (see Pearson 1924 a; Cruickshank 1967; Renaut 2000). The lingual ridge in the latter, however, is concave dorsally, lower than the labial, and the anterior pegs appear slightly larger, so they are visible laterally (Cruickshank 1967). Anteroposteriorly short lingual ridges taller than the labial ridges are also present in Angonisaurus cruickshanki (“ rugosities ” of Cox & Li 1983; Hancox et al. 2013). In Stahleckeria potens the lingual ridge is also taller than the labial (von Huene 1935; Camp 1956; Abdala et al. 2013; Kammerer 2018). The morphology of Woznikella triradiata n. gen., n. sp. is also different than in Placerias hesternus (Camp & Welles 1956; Kammerer 2018; pers. obs.) in which the labial ridge is consistently taller than the lingual and straight. In the latter species the rostral ends of the lingual ridges are visible in lateral view (Camp & Welles 1956), but in Woznikella triradiata n. gen., n. sp. they are obscured by the very high dorsal labial edges of the rami. The anatomy of the scoop-shaped dorsal surfaces of the symphyseal areas of Kombuisia frerensis and Myosaurus gracilis is much simpler and the median groove in these species is wider (Cluver 1974; Hotton 1974; Hammer & Cosgriff 1981; Fröbisch 2007). In dorsoventral aspect, the dentaries have roughly parallel lateral edges in their anterior part and strongly diverge posteriorly. They are much more slender than the dentaries of Dolichuranus primaevus, Kombuisia spp., Lystrosaurus spp., Myosaurus gracilis, Pentasaurus goggai, Shaanbeikannemeyeria xilougouensis, Shansiodon wangi, Shansiodon wuhsiangensis, Sinokannemeyeria baidaoyuensis, Vinceria andina, and Xiyukannemeyeria brevirostris (see Young 1935; Yeh 1959; Cluver 1971, 1974; Keyser 1973; Hotton 1974; Li 1980; Hammer & Cosgriff 1981; Liu & Li 2003; Damiani et al. 2007; Fröbisch 2007; Fröbisch et al. 2010; Domnanovich & Marsicano 2012; Liu 2015; Kammerer 2018). The ventral part is complete only in the right ramus and the mentum is strongly bowed, more so than in, e. g., Dinodontosaurus brevirostris or Shaanbeikannemeyeria xilougouensis (Cox 1968; Kammerer & Ordoñez 2021; Liu 2022; Escobar et al. 2023). It is, however, strikingly different from the top-heavy, rostrally vertical dentaries of Sungeodon kimkraemerae and Vinceria andina, in which the rostral and the posteroventral surfaces are set at an acute angle (Domnanovich & Marsicano 2012; Maisch & Matzke 2014). The lateral bone lamella of the dentary in Woznikella triradiata n. gen., n. sp. becomes thinner ventrally, forming a large surface for a suture with the splenial (not preserved) and angular. On the anteroventral (mental) surface of the dentaries two sharply defined, parallel, longitudinal grooves are separated by a wide ridge with a narrow but distinct third groove running medially along the symphysis (Fig. 4 B, D). A ridge-like sagittal structure is also present in Kannemeyeria simocephalus, Parakannemeyeria ningwuensis, Pentasaurus goggai, Placerias hesternus, Sinokannemeyeria sanchuanheensis, Sungeodon kimkraemerae, Tetragonias njalilus, and Wadiasaurus indicus (Camp & Welles 1956; Sun 1963; Cruickshank 1967; Cheng 1980; Bandyopadhyay 1988; Renaut 2000; Maisch & Matzke 2014; Kammerer 2018). In Tetragonias njalilus, however, it projects its own tip from the anterior edge of the beak (Cruickshank 1967). The tips of the beak bear narrow, longitudinal grooves close to their lateral edges. The Meckelian canal exposed on the ventral surface of the dorsal part of the disarticulated dentary takes the form of a longitudinal groove which is widest anteriorly and narrows posteriorly (Fig. 4 B). This groove ends in a point in contrast to Placerias hesternus, in which it forms a vertical edge (Camp & Welles 1956 and pers. obs.). In most dicynodonts at least part of this groove accommodates an anterior process of the surangular. The posteriormost part of the dentary is formed by a thin blade, which contacted the surangular. Its ventral edge contributed to the mandibular fenestra. Below that, a plate-like process for the angular was present, but only its base is preserved on the right side of ZPAL V. 34 / 1 / 3. Given the imprint on the left angular (ZPAL V. 34 / 1 / 81), this process was relatively large, about one fourth to one third the length of the dentary, very different than, e. g., the unusually small process of Dolichuranus primaevus (see Damiani et al. 2007). Unlike Myosaurus gracilis and Tetragonias njalilus, the dentary was apparently level with the more posterior parts of the mandible, not set at an angle (von Huene 1942; Cruickshank 1967; Cluver 1974). ZPAL V. 34 / 1 / 3 is notably similar to SMNS 91416, the partial mandible from the Stuttgart Formation in Bavaria (Germany) described by Schoch (2012). The similarities include: general shape in lateral view (hooked, with bowed mental surface); presence of a median notch; presence and shape of anterodorsal tips; rugosity of the surface; shape of the ramphotheca, with distinctly concave posteroventral edge; position and shape of a lateral groove; position and shape of lateral vascular grooves; presence of a symphyseal ridge on the rostroventral (mental) surface of the mandible flanked by lateral grooves and with a sagittal groove (less distinct than in ZPAL V. 34 / 1 / 3, possibly due to difference in ontogenetic age); presence of what appears to be a lingual ridge higher than the labial edge. Based on the published photographs, SMNS 91416 appears to be wider than ZPAL V. 34 / 1 / 3, but this, at least to some extent, is caused by the foreshortening due to the perspective and slightly different scale of the panels captioned as the dorsal and ventral view. Because of the morphological similarity, geographical and temporal proximity, and absence of other dicynodonts in the European Carnian, we therefore tentatively refer SMNS 91416 to Woznikella triradiata n. gen., n. sp. Angular Only the posterior portion of the left angular (ZPAL V. 34 / 1 / 81) is preserved, and its posterodorsal edge is incomplete. The right angular (ZPAL V. 34 / 1 / 5; Fig. 5) is almost complete. The ventral and dorsal edges are deeply concave in lateromedial aspect. The reflected lamina is very well preserved in ZPAL V. 34 / 1 / 5, even though it is a thin plate. Its medial surface is slightly concave, whereas the lateral surface is convex. Medial to the reflected lamina is the main body of the angular, and between the two there is a deep notch, the anterior limit of which is exposed in lateral view. In lateromedial aspect, the main body protrudes dorsally above the reflected lamina (Fig. 4 A, D). Unfortunately, the dorsal edge of the angular is broken in both specimens so its exact height remains unknown – the probable outline of the articular surface on the labial side of the surangular suggests that approximately 5 mm are missing. On the medial surface of the main body a long, shallow groove is visible (Fig. 4 D). Its edges are pronounced and mostly parallel in its posterior part, and anteriorly it becomes very wide. It is gently sinuous, with its anterior end turned gently dorsally, the middle part approximately parallel to the long axis of the bone, and the posterior end downturned. In the posterior and middle part, it is situated closer to the dorsal than to the ventral edge of the bone, whereas anteriorly its ventral edge comes closer to the ventral edge of the bone while the dorsal edge diminishes, leaving the groove open dorsally. An additional short ridge is present above the middle part of the groove, close to the half of the angular. These structures probably form the area of connection with the surangular. A large, clearly demarcated, depressed and mostly flat area for the suture with the dentary is visible on the lateral surface in the anterior part of the angular (Fig. 4 A, B). Its posterior end is located around the mid-length of the angular, its dorsal edge quickly merges with the dorsal edge of the angular, and its ventral edge is gently bowed, concave dorsally, and reaches the ventral edge of the angular close to its anterior tip. Surangular The right surangular is fused with the articular and prearticular (ZPAL V. 34 / 1 / 4; Fig. 6). As preserved, its dorsal edge forms a triangular coronoid eminence. The boundary of that structure is damaged, but what appears to be remnants of its natural edge suggest that the general shape is not be significantly altered. The surangular is similar to the homologous bone of, e. g., Angonisaurus cruickshanki and Lisowicia bojani (see Cox & Li 1983; Sulej & Niedźwiedzki 2019). The surangular has a completely flat lingual surface. A distinct ridge on the labial surface begins in front of the coronoid eminence and posteriorly it continues onto the labial surface of the articular. A narrow ridge and a groove beneath it are close to the ventral edge of the labial side, but the anteroventral part of the bone is broken. This structure was probably the area of contact with the angular. Prearticular The right prearticular is fused with the articular and surangular (ZPAL V. 34 / 1 / 4; Fig. 6). It is a thin blade of roughly rectangular shape. Its dorsal edge is curved, so the highest point is in the middle of the bone, in front of the coronoid eminence of the surangular. Articular The right articular is fused with the surangular and prearticular (ZPAL V. 34 / 1 / 4; Fig. 6) and is broadened posterodorsally into a triple-edged condyle. The central ridge is the longest and highest, the inner ridge is the shortest. The retroarticular (digastric) process is very short, as in, e. g., Ischigulastia jenseni, most Shaanbeikannemeyeria xilougouensis, and Wadiasaurus indicus, but it is longer than in Vinceria andina (see Cox 1965; Bandyopadhyay 1988; Domnanovich & Marsicano 2012; Liu 2022). Note, however, that this character shows some variability, even bilaterally within a single individual (Camp 1956). The fourth, labialmost ridge, which is located outside of the condyle, begins on the surangular like in Stachleckeria potens and ends on the retroarticular (digastric) process like in Kannemeyeria simocephalus (see Camp 1956). Vertebral column All centra and neural arches are preserved separately (Figs 8; 9 A-M), with the exception of ZPAL V. 34 / 1 / 12, a posterior dorsal vertebra (Fig. 8 N-Q). Although the elements were mostly found scattered, their morphology and documented location within the association allow the establishment of their approximate relative positions within the sequence. Neural arch of the atlas Both halves of the neural arch of the atlas are preserved. The left part of the atlas neural arch is complete (ZPAL V. 34 / 1 / 27; Fig. 8 A-F). In lateral view, it is composed of an almost flat lateral process and an elongated posterodorsal process. The posterodorsal process is slightly asymmetric and located entirely to the left of the midline of the neural canal, most likely acting functionally as a postzygapophysis. The postzygapophyseal surface is almost oval with nearly straight medial edge. Above the postzygapophysis, a distinct ridge is present on the dorsomedial side of the process. At the base of the posterodorsal process, a large prezygapophysis is very distinct with a deep notch in its posterior margin. On the anterior portion (main body) there is a flat facet for the exoccipital and a concave, rounded facet for the centrum ventromedially. The lateral process is directed predominantly posterodorsally, its long axis nearly parallel to the posterodorsal process. It is very high and its lateral surface forms a large, oval, flat area, with a gentle but clear, laterally directed lip along its dorsal edge. The overall morphology is nearly identical to that of the neural arch of the atlas of Kannemeyeria simocephalus (see Pearson 1924 b; Govender 2005; Govender et al. 2008) but more elongated posterodorsally than in Wadiasarus indicus (see Bandyopadhyay 1988). The right part of the atlas neural arch (ZPAL V. 34 / 1 / 69) is preserved in the same bone accumulation but has a slightly different state of preservation and shape due to deformation. It is broken into at least three pieces, slightly distorted, and the surface of the bone is reddish and in a bad condition. Both processes, lateral and posterodorsal, are eroded. The lateral process is misshapen and only the base of the dorsal process is preserved. Both modes of preservation are congruent with the state of the rest of the skeleton (see the Material section). Cervical vertebrae Five cervical vertebral centra (ZPAL V. 34 / 1 / 66, ZPAL V. 34 / 1 / 69, ZPAL V. 34 / 1 / 70, ZPAL V. 34 / 1 / 76, and ZPAL V. 34 / 1 / 77; Fig. 8 K-P) and three cervical neural arches (ZPAL V. 34 / 1 / 14, and ZPAL V. 34 / 1 / 30; Fig. 8 G-J) are preserved and are mostly complete (ZPAL V. 34 / 1 / 14 has its neural process deformed and broken, ZPAL V. 34 / 1 / 70 is incomplete laterally on its left side, ZPAL V. 34 / 1 / 76 has its posteroventral part broken off). Based on their morphology, the centra can be arranged in a following sequence (anterior to posterior): ZPAL V. 34 / 1 / 66, ZPAL V. 34 / 1 / 70, ZPAL V. 34 / 1 / 69, ZPAL V. 34 / 1 / 77, ZPAL V. 34 / 1 / 76. The neural arches cannot be attributed to any particular centrum, but ZPAL V. 34 / 1 / 14 is larger and has a longer neural process than ZPAL V. 34 / 1 / 30, indicating that it is more posterior. The centra are wide and rounded, subpentagonal in their anteroposterior aspect. Anterior and posterior surfaces are concave (weakly amphicoelous) with pronounced notochordal pits in their centers. The anterior centra are anteroposteriorly short, but gradually increase in length posteriorly, so ZPAL V. 34 / 1 / 76 is about double the length of ZPAL V. 34 / 1 / 66. The two anteriormost centra (ZPAL V. 34 / 1 / 66 and ZPAL V. 34 / 1 / 70) have small, knob-like parapophyses located around the mid-height of the centrum (Fig. 8 K-M), but the parapophyses of the more posterior centra increase in size and extend further dorsally, in ZPAL V. 34 / 1 / 77 and ZPAL V. 34 / 1 / 76 reaching around half of the centrum’s length and nearly reaching the articular surfaces for the neural arch pedicles (Fig. 8 N-P; compare with Camp & Welles 1956; Sun 1963; Bandyopadhyay 1988; Surkov 1998 a; Vega-Dias & Schultz 2004). The dorsal surfaces are wide, and the surfaces for articulation with the neural arches change from wider than long in the anterior cervical vertebral centra to almost rectangular in the posterior, and gently slope ventrolaterally. The centra show little to no keeling ventrally. Ventrally and laterally, particularly above the parapophyses, the cortices of the three anteriormost preserved centra are perforated by vascular canals. The neural arch ZPAL V. 34 / 1 / 30 has a very low and anteroposteriorly short neural process, with an additional articulation surface on its anterior edge (Fig. 8 G-J). The apex of the neural process is subtriangular in cross-section, with sharpened anterior and posterolateral edges, and the posterior surface is concave. In anteroposterior aspect, the process is slightly expanded laterally near the apex (Fig. 8 G, I). In lateral view, the anterior edge of the process is gently concave and slanted posterodorsally, with the posterior edge subvertical (Fig. 8 H). There is a narrow, dorsoventrally aligned rugose band on the posterior surface, likely an attachment point for muscles or ligaments (Fig. 8 I). The neural process is much higher, although plastically deformed and dorsally incomplete, in ZPAL V. 34 / 1 / 14. It is, furthermore, shifted posteriorly, nearly completely beyond the posterior limit of the transverse processes. The zygapophyses are nearly horizontal. The prezygapophyses are oval, and the postzygapophyses are ovoid (long axes directed posterolaterally) and concave ventrally. In lateral view, the dorsal (non-articular) surfaces of the postzygapophyses of ZPAL V. 34 / 1 / 30 are continuous with and parallel to the dorsal (articular) surfaces of the prezygapophyses, and they gently slope anteroventrally. A distinctive vertical ridge is present above each postzygapophysis. In ZPAL V. 34 / 1 / 14, the articular surfaces of both the pre- and postzygapophyses are roughly level with each other, but aligned more ventromedially, and the postzygapophyses are more elongate than in ZPAL V. 34 / 1 / 30, with their long axes directed more posteriorly. In both specimens, the prezygapophyses and postzygapophyses are separated mesially along their entire length (the prezygapophyses wider than the postzygapophyses), but in ZPAL V. 34 / 1 / 14 the latter are closer to each other than in ZPAL V. 34 / 1 / 30. The diapophyses are present beneath the prezygapophyses and are relatively small dorsoventrally, but the transverse processes extend horizontally just above the neurocentral suture, with their ventrolateral surfaces convex (compare with Young 1937; Sun 1963; Cheng 1980; Bandyopadhyay 1988; Vega-Dias & Schultz 2004; Liu 2015). Dorsal vertebrae There are six isolated neural arches (ZPAL V. 34 / 1 / 13, ZPAL V. 34 / 1 / 31, ZPAL V. 34 / 1 / 33, ZPAL V. 34 / 1 / 34, ZPAL V. 34 / 1 / 35, and ZPAL V. 34 / 1 / 36; Fig. 9 A-J), four isolated vertebral centra (ZPAL V. 34 / 1 / 11, ZPAL V. 34 / 1 / 67, ZPAL V. 34 / 1 / 68, and ZPAL V. 34 / 1 / 71; Fig. 9 K-M), and a single mostly complete vertebra (ZPAL V. 34 / 1 / 12; Fig. 8 N-Q) that can be classified as belonging to the dorsal series. Based on recovery documentation, out of the isolated elements, the centrum ZPAL V. 34 / 1 / 71 and the neural arch ZPAL V. 34 / 1 / 36 were found in near articulation, and ZPAL V. 34 / 1 / 34 was found right behind them. The centrum ZPAL V. 34 / 1 / 67 was found below the neural arch ZPAL V. 34 / 1 / 31. The shapes and sizes of the neurocentral articulation facets as well as the state of preservation suggest that the centra ZPAL V. 34 / 1 / 68 and ZPAL V. 34 / 1 / 11 (Fig. 9 K-M) may belong to the neural arches ZPAL V. 34 / 1 / 34 (Fig. 9 K-M) and ZPAL V. 34 / 1 / 33 (Fig. 9 G-J), respectively, and that they come from neighboring segments. Based on those data and gradual changes of morphology, the elements may be hypothesized to form the following relative sequence (anterior to posterior): ZPAL V. 34 / 1 / 31, ZPAL V. 34 / 1 / 67, ZPAL V. 34 / 1 / 35, ZPAL V. 34 / 1 / 71 + ZPAL V. 34 / 1 / 36, ZPAL V. 34 / 1 / 68 + ZPAL V. 34 / 1 / 34, ZPAL V. 34 / 1 / 11 + ZPAL V. 34 / 1 / 33, ZPAL V. 34 / 1 / 13, ZPAL V. 34 / 1 / 12. However, this sequence should be treated with caution, particularly when it comes to the vertebral centra, due to the lesser number of morphological details observable on them. The anterior part of the dorsal vertebral column is represented by a single neural arch, ZPAL V. 34 / 1 / 31 (Fig. 9 A-D). Its anterior position within the dorsal section of the vertebral column is indicated by its extensive, wide synapophyses spanning from the tips of the transverse processes, ventromedially towards the neurocentral suture, and apparently continuing onto the centrum (compare with Pearson 1924 b; Young 1937; Sun 1963; Cox 1965; Cruickshank 1967; Bandyopadhyay 1988). The transverse processes are wing-like, massive, and similar to those of the cervical neural arches, but gently raised laterally. Their ventrolateral surfaces are nearly straight, and their edges are angular. The dorsal edges of the transverse processes face anterodorsally and form distinct lateromedial ridges anteriorly and posterodorsally. The synapophyses are wider dorsally than ventrally and skewed posterodorsally in lateral view. In contrast to the cervical neural arches, the prezygapophyses of ZPAL V. 34 / 1 / 31 have subrectangular outlines and are aligned dorsolaterally rather than subhorizontally; the angle between their anterior edges is approximately 90 ° in the medial part, but their lateral parts gently turn dorsally. The postzygapophyses are ovoid, shorter than in the cervical neural arch ZPAL V. 34 / 1 / 14, and set at a more acute angle than the prezygapophyses (about 60 °). Contralaterally, both the pre- and postzygapophyses are separated only by a narrow notch. The neural process is high and anteroposteriorly long with a rounded and gently anteroposteriorly expanded dorsal end. It is directed gently posteriorly. The neural process apex is fusiform in cross-section and only slightly expanded laterally. At the base of the neural process there is a distinct vertical ridge anteriorly, reaching the notch between the prezygapophyses, and a vertically expanded, rhomboid depression posteriorly, limited by the postzygapophyses ventrolaterally and by low, rounded ridges dorsolaterally. The anterior edge of the base of the neural process reaches anteriorly past the level of the anterior edge of the synapophyses. ZPAL V. 34 / 1 / 35 represents a slightly more posterior neural arch (Fig. 9 E, F). In contrast to ZPAL V. 34 / 1 / 31, the dorsolaterally directed transverse process in this specimen is free from the synapophysis and is plate-like. Only a low ridge continues from the dorsal end of the synapophysis onto the posteroventral surface of the transverse process. The dorsal extent of the synapophyses is below the dorsolateral edges of the prezygapophyses, which is lower than in ZPAL V. 34 / 1 / 31 but higher than in succeeding vertebrae. It is well defined in anteroposterior aspect and clearly separated from the transverse process, as in the remaining, more posterior neural arches. The dorsal part of the synapophysis is rounded and anteroposteriorly wider than the ventral part. The prezygapophyses are set at a more acute angle than in ZPAL V. 34 / 1 / 31 (about 60 °), the postzygapophyses are set at a greater angle (about 80 °) and the dorsal process is directed more posterodorsally. The morphology is reminiscent of that figured and described by Pearson (1924 b), Govender (2005), and Govender et al. (2008) for the mid-dorsal vertebrae of Kannemeyeria simocephalus. The succeeding neural arches show gradual changes in morphology: 1) the synapophyses recede farther ventrally compared to ZPAL V. 34 / 1 / 35 and develop a conspicuous dorsal tip, thus attaining a subtriangular shape with gently bowed anterodorsal and posterodorsal edges; 2) the prezygapophyses and postzygapophyses are ovoid and become more horizontal; 3) the dorsal processes become inclined even more posterodorsally and their bases shift more posteriorly relative to the anterior extent of the synapophyses; 4) the apices of the dorsal processes expand more laterally, particularly in the anterior part, so they become teardrop-shaped in cross-section, and develop subtle medial grooves anteriorly; and 5) the depressions in the posterior bases of the neural processes become deeper and more conspicuous. The posteriormost preserved vertebra, ZPAL V. 34 / 1 / 12, presents the most extreme morphology (Fig. 9 N-Q). The synapophyses in this specimen are particularly low but anteroposteriorly wide, the prezygapophyses are short but the postzygapophyseal part is extended posteriorly due to the posterior shift of the neural process. The outline of the articular surface of the postzygapophyses, nonetheless, takes up only the posterior half of that area. The transverse processes are more horizontal, both in the lateral and anteroposterior aspect. The posterodorsal alignment of the neural process is particularly pronounced. In lateral view, the process presents subtle sigmoidal curvature, with the midsection directed more posteriorly than the base and the apex. The base of the process presents a strong ridge, which unlike in ZPAL V. 34 / 1 / 31 does not extend to the notch between the prezygapophyses, but ends at the level of the base of the transverse processes. The apex of the neural process is expanded also in the middle part, thus presenting a pentagonal cross-section. The groove in its anterior face is subtle, but more conspicuous than in the preceding neural arches. The dorsal vertebral centra are weakly amphicoelous with distinct notochordal pits in the anterior and posterior facets, and are ovoid in anterioposterior aspect. They are comparable in length to the posterior cervical vertebral centra but slightly narrower dorsally, so the facets for the neural arches are longer than wide. Some of them (particularly ZPAL V. 34 / 1 / 67 and ZPAL V. 34 / 1 / 12) bear a rounded ventral keel. The central parts of synapophyses are located in the anterodorsal corners of the centra. As in the case of their neural parts, they become smaller posteriorly. Most of them present convex posteroventral edges, similar to those in the posterior cervical vertebrae, with the exception of ZPAL V. 34 / 1 / 12, in which this edge is straight and the synapophysis has a rhomboid outline. The observed morphological anteroposterior progression is consistent with the changes observed previously in other Kannemeyeriiformes (compare with Pearson 1924 b; Young 1937; Sun 1963; Cruickshank 1967; Bandyopadhyay 1988). It is furthermore consistent with the shape and size of the ribs and their articular regions (see below). According to Vega-Dias et al. (2004) the vertebrae of Triassic dicynodonts have no diagnostic value at the family or generic level. However, it must be noted that the relative scarcity of published data on dicynodont vertebrae (in many cases, only selected vertebrae were figured and their precise location within the column was not determined), has made meaningful comparisons difficult, and variation in these elements may be more extensive than currently recognized. In contrast to Jachaleria candelariensis, the dorsal vertebral centra of Woznikella triradiata n. gen., n. sp. are significantly longer and the transverse processes are more pronounced (Araújo & Gonzaga 1980; Vega-Dias & Schultz 2004; Martinelli et al. 2020). They are also narrower relative to height, particularly in the ventral region, than in, e. g., Dinodontosaurus tener, Ischigualastia jenseni (at least for part of the dorsal vertebral column), Jachaleria colorata Bonaparte, 1970, Kannemeyeria simocephalus, Lisowicia bojani, Pentasaurus goggai, Rhinodicynodon gracile, Sangusaurus parringtonii, Sinokannemeyeria pearsoni, and Stahleckeria potens (see von Huene 1935; Young 1937; Romer & Price 1944; Cox 1965; Araújo & Gonzaga 1980; Surkov 1998 a; Vega-Dias & Schultz 2004; Morato 2006; Angielczyk et al. 2017; Kammerer 2018; Sulej & Niedźwiedzki 2019; Martinelli et al. 2020), being more in line with, e. g., Parakannemeyeria shenmuensis, Placerias hesternus, Shansiodon wangi, Sinokannemeyeria yingchiaoensis, Wadiasaurus indicus, or Zambiasaurus submersus (e. g., Camp & Welles 1956; Yeh 1959; Sun 1963; Cox 1969; Cheng 1980; Bandyopadhyay 1988). The shape of the centra, however, aside from taphonomic factors, may potentially be impacted by ontogeny (e. g., von Huene 1935). If the assignment of the neural arches to the vertebral centra is correct, the posterior ends of their pedicels could project posteriorly in the mid-dorsal vertebrae, forming protuberances consistent with the morphology seen in at least some other Kannemeyeriiformes (von Huene 1935; Araújo & Gonzaga 1980; Bandyopadhyay 1988; Vega-Dias & Schultz 2004; Martinelli et al. 2020). Ribs Ribs are represented by some almost complete specimens and many fragments of varying sizes. Among them several morphotypes may be distinguished, which generally agree with the morphologies presented for Jachaleria candelariensis by Vega-Dias & Schultz (2004), Kannemeyeria simocephalus by Pearson (1924 b), and Sinokannemeyeria yingchiaoensis by Sun (1963), and correspond well with the articulation facets observed on the vertebrae (see above). ZPAL V. 34 / 1 / 26 is a very short (14 cm along the parietal edge), anteroposteriorly flattened rib with a distinct curvature and conspicuous decrease of the visceroparietal width along its length (Fig. 10 I, J, Q). The proximal part in anteroposterior aspect is gently flared, slightly sloped anteromedially in proximal view, and exceptionally flat (29 × 8 mm). Unlike the remaining ribs, it does not form a defined articular surface and its edge is uneven, so it is likely that this rib formed a sutural connection with the vertebra rather than a movable articulation. As preserved, the specimen terminates in a blunt 4 × 6 mm large tip. Possibly this is a break, but it seems unlikely that the rib continued much further in life. Given its different morphology, this specimen is tentatively identified as a cervical rib. Ribs with anteroposteriorly narrow and visceroparietally wide shafts are common and attain the largest sizes, possibly coming from the widest, anterior section of the ribcage. The best-preserved specimen in this category is ZPAL V. 34 / 1 / 82 (Fig. 10 A, B, L). It is a very long rib with a complete head and tubercle, but with a broken distal end. The rib head is set on a short neck and is relatively small compared to the expanded tubercle, the visceroparietal size of which nearly equals the visceroparietal width of the shaft. This amounts to the proximalmost part of the rib being almost two times the size of the shaft visceroparietal width. The articular surface for the diapophysis forms an inverted pear-shape in proximal view and is much larger than the oval surface for the parapophysis. Both facets are separated by a constriction giving the proximal part of the rib an inverted figure 8 shape in articular view. The posterior edges of the articular surfaces protrude in a ridge-like fashion, as does the dorsal edge of the facet for the diapophysis. The specimen is broken in several places and its distal part is crushed and distorted, but it appears to originally have a relatively straight shaft with nearly parallel edges. The specimen ZPAL V. 34 / 1 / 16 presents overall the same morphology but the edges of the articular facets also form conspicuous ridges anteriorly (Fig. 10 K). The shaft of this specimen is gently sinuous in visceroparietal aspect, but it is not clear whether this waviness was present in vivo, or if it was acquired postmortem because of distortion. In the specimen ZPAL V. 34 / 1 / 83, which represents the same morphotype, the curvature of the shaft is slightly more pronounced, likely as a result of a smaller circumference of the rib cage at its level. In ZPAL V. 34 / 1 / 15, the head and tubercle merge in a single, elongated subvertical articular surface (Fig. 10 M). More posterior, visceroparietally wide ribs with a smaller single, vertical or nearly vertical articular surface lying in the same plane as the shaft and a curved proximal region (ZPAL V. 34 / 1 / 20, ZPAL V. 34 / 1 / 21, ZPAL V. 34 / 1 / 85) can reach similar visceroparietal width as the previous type, but were likely shorter, considering their curvature. In ZPAL V. 34 / 1 / 20 (Fig. 10 C, D, N) and ZPAL V. 34 / 1 / 85, the articular surface is sinuous in anteroposterior aspect, convex in its ventral part, and recessed in the dorsal part. In ZPAL V. 34 / 1 / 20 it is oblique, directed ventromedially in anteroposterior aspect, and dorsally projected into a small process, whereas in ZPAL V. 34 / 1 / 85 the dorsal part is nearly horizontal in anteroposterior aspect. In both specimens, the articular surfaces are slightly sloped anteroventrally in proximal view. The dorsal and ventral parts are continuous but in proximal view there is a very subtle constriction in the middle of the articular surface, giving it a peanut-shaped outline and suggesting that this area is formed by merging of the head and tubercle. In ZPAL V. 34 / 1 / 21, the articular surface in proximal view is vertical and convex, inverted comma-shaped, with the thinner end directed posterodorsally and the anterodorsal edge convex. The specimen is visceroparietally narrower than ZPAL V. 34 / 1 / 20 and ZPAL V. 34 / 1 / 85. In all specimens of this type, the shaft is anteroposteriorly flattened in the proximal part, the anterior and posterior surfaces are gently concave just distal to the curve, and more distally the shaft becomes oval in cross-section and visceroparietally narrower. The distal (ventral) ends of the ribs belonging to the previous two morphotypes are completely preserved in ZPAL V. 34 / 1 / 18 and ZPAL V. 34 / 1 / 20. Based on these specimens, it may be established that the oval cross-section, which is present around the midlength of the ribs, becomes flattened again and increases in visceroparietal width distally. The ribs terminate with swollen and rounded ends, ovoid in distal view. Strongly curved ribs with narrow and long shafts likely come from the posteriormost part of the rib cage. This kind of rib is represented by the specimens ZPAL V. 34 / 1 / 17 (Fig. 10 G, H, P) and ZPAL V. 34 / 1 / 23. The articular surface is only preserved in ZPAL V. 34 / 1 / 17. It is horizontal, single, and elongated (the anteroposterior width about twice that of the shaft) with a straight ventral edge and a convex dorsal edge. In dorsovental aspect, the anteriormost point of the articular surface is protruding. The neck expands dorsally towards a smooth dorsal curve, which constitutes the visceroparietally widest part of the shaft. The shaft just distal to the curve becomes subtriangular in cross-section, with a rounded visceral and posterior edge, an angular anterodorsal edge, and a concave posterovisceral surface, forming along the dorsolateral part of the rib a small, posteriorly directed lappet of bone. More distally, the rib becomes oval in cross-section, with the longer axis directed visceroparietally. Scapula The nearly complete, right scapula (ZPAL V. 34 / 1 / 7; Fig. 11 A-D) only lacks some minor fragments of its ventral and posterior edge and the posterodorsal tip of the blade. It is roughly hourglass-shaped in lateromedial aspect, with the anterior edge more concave than the posterior, and the minimal breadth just below the midlength, above the acromion (Fig. 11 A, B). It is gently bowed and twisted in anteroposterior view, with its dorsal and ventral ends turned medially and deflected anteromedially (Fig. 11 C, D). The scapular spine is damaged but seems to be only weakly developed and the prespinal surface is flat (Fig. 11 C) – in this aspect it differs from dicynodonts such as Eubrachiosaurus browni, Placerias hesternus, Stahleckeria potens, the indeterminate stahleckeriines (MCZ 3459 and CRILAR-Pv 82) from the Chañares Formation, Rhinodicynodon gracile, Tetragonias njalilus, or even Zambiasaurus submersus, which have a more pronounced spine (von Huene 1935; Romer & Price 1944; Camp & Welles 1956; Cruickshank 1967; Cox 1968, 1969; Surkov 1998 a; Kammerer et al. 2013; Escobar et al. 2021). The acromion is conspicuous, mediolaterally flattened, hooked medially, protrudes well past the anterior margin of the scapula, and has a blunt, thickened tip forming medially a small tubercle (Fig. 11 B) similar to that of Stahleckeria potens and Dinodontosaurus sp. (“ Dicynodon turpior ”) figured by von Huene (1935). The presence of a well-developed acromion differentiates Woznikella triradiata n. gen., n. sp. from cf. Dolichuranus primaevus and Ischigualastia jenseni, which lack this process (Cox 1965; Govender & Yates 2009; Escobar et al. 2021). Nonetheless, the acromion is not as large, relative to the rest of the scapula, as in, e. g., Acratophorus argentinensis, Dinodontosaurus tener, Dinodontosaurus sp. (“ Dicynodon turpior ”), Lystrosaurus spp., Rhinodicynodon gracile, or Shansiodon wangi (see Broom 1908 a; von Huene 1935; Young 1935; Yeh 1959; Cox 1965; Bonaparte 1966 a; DeFauw 1986; Surkov 1998 a; Morato 2006; Ray 2006; Kammerer et al. 2013; Escobar et al. 2021). Unlike in most Triassic dicynodonts (e. g., Pearson 1924 b; von Huene 1935; Broom 1937; Romer & Price 1944; Yeh 1959; Sun 1960, 1963; Bonaparte 1966 a; Cruickshank 1967; Colbert 1974; Araújo & Gonzaga 1980; Cox & Li 1983; Bandyopadhyay 1988; Lucas & Harris 1996; Surkov 1998 a; Govender 2005; Surkov et al. 2005; Ray 2006; Govender et al. 2008; Kammerer et al. 2013; Escobar et al. 2021), it is directed anterodorsally rather than anteriorly, laterally, or antero- or ventrolaterally. The acromion had been broken off during recovery or preparation, causing a wide crack laterally, but its medial and ventral surfaces indicate that the pieces were well fitted together, and that no misalignment occurred. Other than the direction of the acromion, the scapula mostly resembles the narrow and gently flared “ kannemeyeriid-like ” (sensu Kammerer et al. 2013; note that the morphologies are not restricted to the clades after which they are named) scapulae of, e. g., Kannemeyeria simocephalus, Parakannemeyeria youngi, Rhinodicynodon gracile, Sinokannemeyeria yingchiaoensis, Wadiasaurus indicus, the unnamed stahleckeriine (MCZ 3459) from the Chañares Formation, the morphotype B stahleckeriid of Govender (2005), and possibly Eubrachiosaurus browni in its general shape and proportions (Watson 1917; Pearson 1924 b; Broom 1937; Sun 1963; Cox 1968; Bandyopadhyay 1988; Surkov 1998 a; Govender 2005; Ray 2006; Govender et al. 2008; Kammerer et al. 2013). Govender (2005) and Govender et al. (2008) noted that Kannemeyeria simocephalus had relatively wide scapular blade and that the narrow-bladed scapulae previously attributed to that species by Pearson (1924 b) differ significantly, resembling more Govender’s (2005) morphotypeB stahleckeriid, the scapula of which, however, also fits better the “ kannemeyeriid-like type ” sensu Kammerer et al. (2013). ZPAL V. 34 / 1 / 7 differs from the typical “ shansiodontid- ” and “ stahleckeriid-like ” (sensu Kammerer et al. 2013) scapulae of Acratophorus argentinensis, Dinodontosaurus tener, Dinodontosaurus sp. (“ Dicynodon turpior ”), Ischigualastia jenseni, Jachaleria candelariensis, cf. Kannemeyeria lophorhinus, Shansiodon wangi, Tetragonias njalilus, and Stahleckeria potens, which are less gracile, have a more expanded scapular blade, and less expanded point of contact with the coracoids (von Huene 1935; Romer & Price 1944; Yeh 1959; Cox 1965; Bonaparte 1966 a; Cruickshank 1967; Araújo & Gonzaga 1980; Lucas 2002; Vega-Dias & Schultz 2004; Vega-Dias et al. 2005; Morato 2006; Govender & Yates 2009). The scapulae attributed to Lystrosaurus spp. show some variability, but generally seem to be stouter as well (e. g., Huxley 1865; Broom 1908 a; Young 1935; Colbert 1974; DeFauw 1986; Ray 2006). The scapula of Zambiasaurus submersus is similar in its slenderness, but less flared at both ends (Cox 1969; Govender 2005; Kammerer et al. 2013), whereas the scapula of Placerias hesternus is less constricted (Camp & Welles 1956; Kammerer et al. 2013). A lesser degree of flaring around the glenoid area also differentiates it from the scapula of Angonisaurus cruickshanki (Cox & Li 1983; Govender 2005). The scapula of Lisowicia bojani is less gracile as well, and has a significantly more expanded ventral part (Sulej & Niedźwiedzki 2019). The scapula of Woznikella triradiata n. gen., n. sp. is thickest at the glenoid, the articular surface of which in the scapular part is gently concave, rhomboid with rounded corners and set at an angle of ~ 100 ° relative to the anteroventral facet for the coracoids. This articular surface is approximately half the length of the gently convex facet for the coracoids. The attachment site for the triceps muscle is present as a rugose field on the posterolateral edge of the scapula above the glenoid (Fig. 11 A, D). There is no trace of a m. teres major scar (e. g., Vega-Dias & Schultz 2004), but it might have occupied the unpreserved posterior tip of the scapula. Along most of the visceral surface of the scapula, a gentle but clear striation is present but there is no tubercle at the base of the acromion like that described by Govender (2005) and Govender et al. (2008) in Kannemeyeria simocephalus or Tetragonias njalilus. Govender (2005) described a fossa near the same area in her morphotype B stahleckeriid, and a gentle depression seems to be also present at the anteromedially surface of ZPAL V. 34 / 1 / 7, right above the acromion, but it is smaller and it is difficult to establish whether the two are homologous. Procoracoid The right procoracoid (ZPAL V. 34 / 1 / 107; Fig. 11 E, F) is preserved almost in its entirety but with its anterior end damaged. The facet for the scapula lacks sutural characteristics and the ventral edge is rounded and porous, but it is unclear whether this state is caused by weathering or if the bone was still finished in cartilage at the time of the animal’s death. The posterior suture with the coracoid is nearly straight and closed – it is recognizable as a raised, rugose ridge on both the visceral and external surface of the specimen. The procoracoid foramen is fully enclosed by the procoracoid, although it is very close to the contact with the scapula, especially on the visceral side. As a result, its walls are directed slightly posteroventrally. The procoracoid is very gently sigmoid in anterior view, with its ventral part thin and faintly directed medially. Its external surface is otherwise nearly flat. Viscerally, however, there is a pronounced thickening forming a rounded shelf below the procoracoid foramen and in the posterodorsal part of the bone, behind the foramen and close to the glenoid. There is no indication that the procoracoid itself contributed to the articular surface of the glenoid, unlike, e. g., Tetragonias njalilus (see Cruickshank 1967). Coracoid Only a minute part of the anterior edge of the right coracoid is preserved connected by a closed, nearly straight suture to the right procoracoid (ZPAL V. 34 / 1 / 107; Fig. 11 E, F). Nothing can be said about that bone other than that the preserved part thickens dorsally, towards the glenoid area. Clavicle Both clavicles are preserved (right ZPAL V. 34 / 1 / 74, left ZPAL V. 34 / 1 / 75; Fig. 12) but the medial ends of both are broken near the area for contact with the interclavicle. Each clavicle is twisted; the lateral part is vertical while the medial end is horizontal. The lateral part is gently bent posteriorly. The lateral end with the area of contact with the acromion process of scapula is well preserved and well developed. Its anterior surface is convex, whereas the posterior one is concave in ventral part and its dorsal part forms a massive ridge along the dorsal edge of the lateral half of the bone. The ventral edge of the lateral end forms a ridge along the posterior margin of the rest of the clavicle. Unlike in Sinokannemeyeria yingchiaoensis, the clavicles of Woznikella triradiata n. gen., n. sp. lack finger-like processes at the scapular ends (Sun 1963). This may be taxonomic or may result from their incomplete ossification due to young ontogenetic age in the individual. Humerus The right humerus (ZPAL V. 34 / 1 / 6; Fig. 13) is preserved in two parts, lacks the anterior portion of the proximal end and the midshaft, and most of the ventral surfaces of the preserved fragments are heavily damaged. The articular surface of the proximal end is slightly raised and subtriangular in dorsal aspect, but still rather inconspicuous (Fig. 13 A, C, F), at least in part due to taphonomic factors, but possibly also due to immature ontogenetic age of the individual – the well-preserved fragments of its surface are porous, indicative of a well-developed cartilaginous finish. As in Kannemeyeria simocephalus, its distal corner terminates in a gentle ridge (Govender 2005; Govender et al. 2008). Similarly, the medial process (insertion of m. subcoracoscapularis) is short, lacks any rounded tip, lies nearly in line with the posterior (medial) surface of the shaft, and takes form of a relatively short, raised lip, continuous along the proximal and posterior edge of the proximal humeral end, rather than a distinct projection. In that respect, it is similar to, e. g., cf. Dolichuranus primaevus, Lisowicia bojani, Parakannemeyeria shenmuensis, Parakannemeyeria youngi, Sinokannemeyeria yingchiaoensis, the morphotype B stahleckeriid of Govender (2005), possibly Eubrachiosaurus browni and – to some extent – Angonisaurus cruickshanki, Kannemeyeria simocephalus, and Sinokannemeyeria pearsoni, but different from, e. g., Dinodontosaurus sp. (“ Dicynodon turpior ”), Dinodontosaurus tener, at least some representatives of Lystrosaurus and Placerias hesternus, Shansiodon wangi, Stahleckeria potens, Wadiasaurus indicus, Xiyukannemeyeria brevirostris, and Zambiasaurus submersus (see Williston 1904; Watson 1917; Pearson 1924 b; von Huene 1935; Young 1937; Camp & Welles 1956; Yeh 1959; Sun 1963, 1978; Cox 1965, 1969; Cheng 1980; Bandyopadhyay 1988; Lucas & Harris 1996; Lucas 2002; Govender 2005; Ray 2006; Dzik et al. 2008 a; Govender et al. 2008; Govender & Yates 2009; Kammerer et al. 2013; Sulej & Niedźwiedzki 2019). A similar morphology is also present in the specimen MCZ 3118 of Ischigualastia jenseni as pictured by Cox (1965), but the specimen PVL 3807 has a more pronounced rounded medial process (Kammerer et al. 2013). The proximal end thickens in that area, but the significant damage of the ventral surface obscures details of the morphology. The deltopectoral crest is not preserved, so its exact size and shape cannot be determined, but based on the curvature of the proximal edge of the proximal end, at least proximally it did not reach much further anteriorly (laterally) from the proximal condyle than the medial process reaches posteriorly (medially). The bicipital fossa in the preserved part seems to be relatively shallow and gently concave, but the ventral surface is significantly damaged (Fig. 13 B, F). The attachment of m. latissimus dorsi is inconspicuous, taking form of a rugose, longitudinally extended area (Fig. 13 C). The distal end is relatively well preserved in dorsal aspect, presenting a marked smaller entepi- and larger ectepicondyle with rugosities and porous distal surfaces indicative of a cartilaginous cap (Fig. 13 A, B, D, E, G). The distal end is slightly narrower than the proximal end. Its outline in this aspect is triangular and its anterior (lateral) and posterior (medial) expansion is relatively gentle and symmetrical, similar to that in Kannemeyeria simocephalus, some specimens of Lystrosaurus spp. (except L. georgi, L. hedini, ‘ Lystrosaurus weidenreichi ’ Young, 1939, and some specimens from South Africa and Antarctica; see Young 1935, 1939; Colbert 1974; DeFauw 1986; Surkov et al. 2005), and Zambiasaurus submersus, but unlike, e. g., cf. Acratophorus argentinensis, Dolichuranus primaevus, Dinodontosaurus tener, Dinodontosaurus sp. (“ Dicynodon turpior ”), Eubrachiosaurus browni, Ischigualastia jenseni, Parakannemeyeria dolichocephala, Parakannemeyeria ningwuensis, Pentasaurus goggai, some (but apparently not all – see Camp & Welles 1956) specimens of Placerias hesternus, Rhinodicynodon gracile, Sangusaurus parringtonii, Shansiodon wangi, Sinokannemeyeria baidaoyuensis, Sinokannemeyeria pearsoni, Sinokannemeyeria sanchuanheensis, Sinokannemeyeria yingchiaoensis, Stahleckeria potens, Wadiasaurus indicus, or Xiyukannemeyeria brevirostris (Williston 1904; Watson 1917; Pearson 1924 b; von Huene 1935; Young 1937; Camp 1956; Yeh 1959; Sun 1960, 1963, 1978; Tripathi & Puri 1961; Cox 1965, 1969; Bonaparte 1966 a, 1967; Colbert 1974; Cheng 1980; Bandyopadhyay 1988; Lucas & Harris 1996; Lucas 1998, 2002; Surkov 1998 a; Surkov et al. 2005; Vega-Dias et al. 2005; Morato 2006; Ray 2006; Govender et al. 2008; Govender & Yates 2009; Kammerer et al. 2013; Angielczyk et al. 2014, 2017; Liu 2015; Kammerer 2018; Kammerer & Ordoñez 2021). It should be noted, however, that the proportions, symmetry, and shape of the distal humeral end are in Kannemeyeriiformes subject to change during ontogeny and intraspecific variability (Cox 1969; Kammerer et al. 2013; Angielczyk et al. 2014). The supinator process is easily noticeable, proximodistally elongated and rugose, but rather low (Fig. 13 B, E). This morphology differs from non-kannemeyeriiform dicynodonts and Dinodontosaurus tener which lack the supinator process (Morato 2006; Kammerer 2018), but also stahleckeriines (Eubrachiosaurus browni, Ischigualastia jenseni, Stahleckeria potens) and Lisowicia bojani, which have a very pronounced, tab-like supinator process (Williston 1904; Cox 1965; Vega-Dias et al. 2005; Dzik et al. 2008 a; Kammerer et al. 2013; Kammerer 2018; Sulej & Niedźwiedzki 2019). In that respect, it resembles Pentasaurus goggai and Zambiasaurus submersus the most (Cox 1969; Angielczyk et al. 2014; Kammerer 2018). The intercondylar groove is well-defined both distally and dorsally. The olecranon fossa is well expressed distally but fades towards the base of the distal end, its broadly concave and its anterior (lateral) and posterior (medial) edges are rounded. The trochlea is readily distinguishable but relatively low and restricted in extent. In ventral aspect, most of the surface except for the anterior (lateral) portion is destroyed. Most of the capitulum and the edges of the entepicondylar foramen are not preserved. Distal to the supinator process, on the ventral surface of the bone, a single, oval puncture is present, likely a bite mark (Fig. 13 B). The incompleteness of the bone makes it impossible to establish the exact inclination (angle) between the proximal and distal end, but it seems to have been relatively low. Ulna The probable left ulna (ZPAL V. 34 / 1 / 9; Fig. 14) is mostly complete (despite the lack of the olecranon process), but its proximal end is severely damaged and distorted, making identification difficult. As preserved, the proximal end is wide, roughly triangular in anteroposterior (dorsoventral) view (Fig. 14 A, B). In anterior (dorsal) aspect, a clearly marked radial notch and a gently bowed outline of what seems to be the sigmoid notch are visible (Fig. 14 A). Although the lack of olecranon may be an effect of damage or immaturity of the individual, the whole edge of the sigmoid process is sharp, unlike the very rounded edges in the specimens of the (juvenile or worn) holotype of Shaanbeikannnemeyeria “ buerdongia ” figured by Li (1980) or juvenile Stahleckeria potens figured by Lucas (2002) and Vega-Dias et al. (2005). It particularly differs from the ulnae of, e. g., Ischigualastia jenseni, Lisowicia bojani, Parakannemeyeria youngi, Placerias hesternus, Shansiodon wangi, Sinokannemeyeria yingchiaoensis, Stahleckeria potens, Wadiasaurus indicus, and the unidentified kannemeyeriid from the Baidaoyu locality (IVPP V 19365) which have nearly vertical sigmoid notches (von Huene 1935; Romer & Price 1944; Camp & Welles 1956; Yeh 1959; Sun 1963; Cox 1965; Bandyopadhyay 1988; Lucas 1993 a; Ray 2006; Liu 2015; Sulej & Niedźwiedzki 2019). Despite the damage, the curvature of the bone is clearly visible, as the proximal end is gently turned and expanded medially. The distal end of the bone is slightly expanded, fusiform in distal view, ends bluntly and is rather featureless (Fig. 14 C-G), in a manner resembling, e. g., juvenile Dinodontosaurus tener, Dolichuranus primaevus, Jachaleria candelariensis, Kannemeyeria simocephalus, Lystrosaurus spp., Parakannemeyeria youngi, and Zambiasaurus submersus (see Young 1935; Sun 1963; Cox 1969; Colbert 1974; Vega-Dias & Schultz 2004; Govender 2005; Morato 2006; Govender et al. 2008; Govender & Yates 2009). Unlike, e. g., Dinodontosaurus tener as figured by Cox (1965), it lacks a pronounced medial process for contact with the radius. Right above the distal end, in anterior (dorsal) aspect there is a gentle depression, likely accommodating the distal end of the radius, but possibly exaggerated by crushing (Fig. 14 C-E). Despite the crushing, the ulna is significantly more gracile than in the juvenile Dinodontosaurus tener figured by Morato (2006), Lisowicia bojani figured by Sulej & Niedźwiedzki (2019), and in Stahleckeria potens as figured by Lucas (2002) and Vega-Dias et al. (2005). The diagnostic value of this character is unclear, however, as the ulna of the latter was described by Vega-Dias et al. (2005) as diagenetically swollen. Radius The proximal end of the right radius (ZPAL V. 34 / 1 / 10. Fig. 15) is crushed, and the proximal part of the shaft and the distal end are missing as well. Despite crushing, the proximal articular surface is subovoid with the lateral portion gently directed ventrally (posteriorly) in a comma-shaped fashion (Fig. 15 E). The edge of this portion is gently raised and rounded, surrounding a subtle depression. Medially, this depression is limited by a comparably gentle convexity, comprising most of the articular surface. The raised edges of the lateral portion do not merge smoothly with the convexity, but instead they end medially as low but noticeable bumps, separated from the convexity by shallow fossae. This separation may, however, be an artifact of preservation or preparation, or an effect of incomplete ossification. In the dorsal (anterior) and ventral (posterior) view the edges of the proximal articular surface are sinuous, with the lateral part of the edge higher than the medial part in the anterior (dorsal) view and the medial part of the edge being higher than the lateral part in the posterior (ventral) view. Due to this sinuousness, the articular convexity is exposed in the anterior (dorsal) view (Fig. 15 A). This shape may be impacted by crushing, but other dicynodonts (e. g., Jachaleria candelariensis) show a similar morphology (e. g., Vega-Dias & Schultz 2004). The shaft is relatively slender and as preserved the bone shows only moderate expansion at each end. Distally, the specimen is slightly flattened anteroposteriorly (dorsoventrally) and ends in a sediment-filled concavity, either as a result of taphonomical processes or incomplete ossification due to immaturity of the individual (Fig. 15 J). There is no evidence of pronounced distal expansions, such as in Stahleckeria potens or Dinodontosaurus tener radii (Romer & Price 1944; Lucas 2002; Vega-Dias et al. 2005; Morato 2006). Along the anterolateral edge of the bone an elongated muscle scar is present, directed slightly proximomedially. The damage and incompleteness of the specimen preclude observation of further meaningful features. In general outline, the bone resembles the radius of Dinodontosaurus sp. (“ Dicynodon turpior ”), Dinodontosaurus tener, at least some species of Lystrosaurus, Shaanbeikannemeyeria xilougouensis, Sinokannemeyeria baidaoyuensis, Sinokannemeyeria yingchiaoensis, Xiyukannemeyeria brevirostris, as well as Parakannemeyeria sp., Jachaleria candelariensis, or Wadiasaurus indicus, but is not as robust as the latter three (von Huene 1935; Young 1935; Sun 1963, 1978; Cox 1965; Colbert 1974; Li 1980; Bandyopadhyay 1988; Lucas & Harris 1996; Vega-Dias & Schultz 2004; Ray 2006; Liu 2015; Liu 2022). It differs from the radii of, e. g., Ischigualastria jenseni, Stahleckeria potens, and Wadiasaurus indicus, which are much less gracile (von Huene 1935; Romer & Price 1944; Cox 1965; Lucas 2002; Vega-Dias et al. 2005; Ray 2006). Femur The left femur (ZPAL V. 34 / 1 / 8; Fig. 16) is preserved in three pieces: the proximal end (Fig. 16 A-E), part of the shaft (Fig. 16 FI), and the medial condyle (Fig. 16 J-M). The articular surface of the proximal head is nearly circular in proximal view (Fig. 16 B) and gently turned anteromedially (Fig. 16 A), although still directed mostly proximally. Unlike, e. g., Ischigualastia jenseni, Placerias hesternus, Rhinodicynodon gracile, Shansiodon wangi, or Sinokannemeyeria pearsoni, there is no clear neck separating it from the rest of the proximal end (Young 1937; Camp & Welles 1956; Yeh 1959; Cox 1965; Surkov 1998 a; Kammerer et al. 2013). Still, the proximal end is more geometrically complex than in most Permian dicynodonts and Lystrosaurus spp. (e. g., Young 1935; Colbert 1974; Yuhe 1983; DeFauw 1986; Surkov et al. 2005; Ray 2006). The greater trochanter is roughly half the size of the articular head, both in the anteroposterior (Fig. 16 A, D) and in the proximal aspect (Fig. 16 E), but it is clearly demarcated, unlike in e. g., Dinodontosaurus tener, Dolichuranus primaevus, Kannemeyeria simocephalus, the specimen figured by Pearson (1924 b) as K. simocephalus (but not attributable to that species according to Govender 2005 and Govender & Yates 2009), Shaanbeikannemeyeria xilougouensis, or the unnamed stahleckeriids from the Manda Beds (NMT RB 463) and the Pekin Formation (NCSM 21719) (Pearson 1924 b; Govender 2005; Morato 2006; Govender & Yates 2009; Green 2012; Kammerer et al. 2013, 2017; Liu 2022). In proximal aspect it is positioned approximately at the level of the middle of the articular head (Fig. 16 E), unlike, e. g., Dinodontosaurus tener and Sangusaurus parringtonii, in which the greater trochanter is aligned with the posterior limit of the articular head (Morato 2006; Angielczyk et al. 2017), but similar to Lisowicia bojani, Tetragonias njalilus, the unnamed Manda Beds specimens (Cruickshank 1967; Kammerer et al. 2017; Sulej & Niedźwiedzki 2019), and Placerias hesternus (pers. obs.), although in the latter four the major trochanter is less bulbous. The outline of the proximal end in anteroposterior aspect resembles Tetragonias njalilus specimen UMCZ T 754 figured by Kammerer et al. (2013; but much less so the specimen figured by Cruickshank 1967), and the unnamed Manda Beds stahleckeriid (NMT RB 463) both in anteroposterior and mediolateral aspects (Kammerer et al. 2017). Laterally, the greater trochanter projects a tubercle, narrower than its main body and limited in proximodistal span (Fig. 16 C). The third trochanter is not preserved and there is no indication of a pronounced trochanteric crest spanning between the greater and third trochanter, at least in the proximal part of the bone. Unlike most dicynodonts, the mediolateral diameter of the bone starts to clearly decrease just below the tip of the trochanter major (Fig. 16 N-I ′). This is unlike, e. g., Dolichuranus primaevus, Dinodontosaurus tener, Dinodontosaurus sp. (“ Dicynodon turpior ”), Lisowicia bojani, at least some species of Lystrosaurus, Parakannemeyeria youngi, Rhinodicynodon gracile, Sangusaurus parringtonii, Shaanbeikannemeyeria xilougouensis, Shansiodon wangi, Sinokannemeyeria pearsoni, Stahleckeria potens, Wadiasaurus indicus, the unnamed Denwa Formation specimens, the unnamed Pekin Formation specimen (NCSM 21719), the unnamed stahleckeriid from the Santa Rosa Formation (NMMNH P- 13001), and the morphotype B stahleckeriid of Govender (2005) but similar to, e. g., young Kannemeyeria simocephalus and possibly Parakannemeyeria dolichocephala (see von Huene 1935; Young 1935, 1937; Yeh 1959; Sun 1963; Cox 1965; Li 1980; Yuhe 1983; Cruickshank 1986 b; DeFauw 1986; Bandyopadhyay 1988; Lucas & Harris 1996; Surkov 1998 a; Bandyopadhyay & Sengupta 1999; Govender 2005; Morato 2006; Ray 2006; Dzik et al. 2008 a; Govender et al. 2008; Govender & Yates 2009; Green 2012; Kammerer et al. 2013; Angielczyk et al. 2017; Sulej & Niedźwiedzki 2019). A clear trochanteric crest is present in the Ischigualastia jenseni specimen figured by Cox (1965), but in the specimen PVL 3807 the morphology is more reminiscent of ZPAL V. 34 / 1 / 8 (Kammerer et al. 2013). Some dicynodonts, such as Acratophorus argentinensis, Placerias hesternus (pers. obs.), and the Manda Beds stahleckeriid (NMT RB 463), exhibit a trochanteric crest that originates proximally wide and inconspicuous in height, but becomes thinner and / or more pronounced distally (Bonaparte 1966 a, 1967, Kammerer et al. 2017), whereas Tetragonias njalilus has a more distinct separation of the major and third trochanter (Fröbisch 2006; Kammerer et al. 2017); such morphologies cannot be ruled out for ZPAL V. 34 / 1 / 8 due to its incompleteness. Juvenile Zambiasaurus submersus has a rather shallow trochanteric crest, but the greater trochanter is apparently narrower and less bulbous than in ZPAL V. 34 / 1 / 8 – the morphology in adults is unfortunately unknown (Cox 1969). The base of the proximal end is flattened anteroposteriorly, as is the preserved part of the shaft (Fig. 16 F-I). The mediolateral diameter of these parts is similar, unlike in Lystrosaurus spp., Sangusaurus parringtonii, Shansiodon wangi, or Wadiasaurus indicus in which the femur exhibits a clear constriction along the diaphysis (Young 1935; Yeh 1959; Yuhe 1983; Cruickshank 1986 b; DeFauw 1986; Bandyopadhyay 1988; Ray 2006; Angielczyk et al. 2017). The medial condyle of the distal end (Fig. 16 J-M) terminates at a relatively sharp angle in the anterior aspect and is relatively narrow and elongated anteroposteriorly in ventral view (Fig. 16 M), similar as in the unnamed Manda Beds stahleckeriid (Kammerer et al. 2017). Its surface is rugose. The preserved fragment indicates a clear separation of the condyles both in the anteroposterior and in the distal aspect.	en	Szczygielski, Tomasz, Sulej, Tomasz (2023): Woznikella triradiata n. gen., n. sp. - a new kannemeyeriiform dicynodont from the Late Triassic of northern Pangea and the global distribution of Triassic dicynodonts. Comptes Rendus Palevol 22 (16): 279-406, DOI: 10.5852/cr-palevol2023v22a16, URL: http://dx.doi.org/10.5852/cr-palevol2023v22a16
