Klobiodon rochei, O’Sullivan & Martill, 2018

O’Sullivan, Michael & Martill, David M., 2018, Pterosauria of the Great Oolite Group (Bathonian, Middle Jurassic) of Oxfordshire and Gloucestershire, England, Acta Palaeontologica Polonica 63 (4), pp. 617-644 : 624-635

publication ID	https://doi.org/ 10.4202/app.00490.2018
persistent identifier	https://treatment.plazi.org/id/03AA87D0-0762-3D40-6248-FE6AA6BD7D43
treatment provided by	Felipe
scientific name	Klobiodon rochei
status	sp. nov.

Treatment

Klobiodon rochei sp. nov.

Fig. 4 View Fig .

1878 Pterodactylus raptor ; Waterhouse 1878: 34.

1888 Rhamphocephalus depressirostris ; Lydekker 1888: 36.

1995 Rhamphocephalus depressirostris ( Lydekker, 1888) ; Benton and Spencer 1995: 144.

2012 Rhamphocephalus depressirostris ( Lydekker, 1888) ; Steel 2012: 1347.

ZooBank LSID: urn:lsid:zoobank.org:act:F79D2658-AF38-4F16-87C2-4BAC9F1021E5

Etymology: After comic book artist Nick Roche. Comic books are a medium where extinct animals are portrayed in an increasingly scientifically accurate manner, and Roche’s work in the late 2000s was one of the earlier examples in this renaissance in palaeoart.

Holotype: NHMUK PV OR 47991, a right lower jaw in right lateral view.

Type locality: Stonesfield, Oxfordshire, UK.

Type horizon: Stonesfield Slate Member, Taynton Limestone Formation, Bathonian, Middle Jurassic.

Material. — Holotype and OUM 28410 View Materials , an isolated mandibular symphysis from the type locality.

Diagnosis. —A rhamphorhynchine pterosaur possessing a unique combination of fang-like laniaries and short but robust medial teeth; medial teeth with a vertical height at least 1.3 times the width of the alveolar base; a posterior laniary 1.4–2.4 times the height of the first medial tooth; the longest laniaries are at least 1.5 times the depth of the dentary at its deepest point.

Description. —NHMUK PV OR 47991 ( Fig. 4 View Fig , Table 1) is a 140 mm long jaw preserved in left lateral view with a broken mandibular symphysis and posterior ramus. Using Rhamphorhynchus (NMHUK PV R 37002) and Dorygnathus (MBR 1920.16) as proxies ( Padian 2008a; Bonde and Leal 2015) NHMUK PV OR 47991 is estimated to be missing approximately 22% of the total jaw, giving it an estimated original length of ~ 180 mm. The mandible has a consistent depth of 13 mm posterior to the diverging rami at the mandibular symphysis. From the mandibular symphysis to the anterior break, the jaw depth increases to 17 mm. The ramus curves gently ventrally through an arc of approximately 165°. Mandibular specimen NHMUK PV OR 47991 preserves the second and third anterior teeth as well as the first two medial teeth. Two morphotypes are present; elongate recurved anterior laniaries and shorter, straighter medially positioned teeth. A concave margin on the dorsal surface anterior to the first preserved tooth is identified by Lydekker (1888) as the alveolus for the first tooth. The first preserved laniary is 26 mm long with an 8.6 mm diameter and is directed anteriorly 120° relative to the jaw line. It is elongate and strongly recurved throughout its length. The second preserved is 19 mm long with an 8 mm wide alveolus. It is more robust than its predecessor and less curved distally. Its mesial margin is broader and more sharply recurved relative to the first preserved tooth. It is angled at 110° relative to the jaw line. There is a marked size differentiation between the anterior laniaries and the medial teeth. Even assuming the first medial tooth is a replacement tooth due to its small size, such teeth are always at least 60% erupted ( Fastnacht 2008). Thus, the third preserved tooth is at least 1.4–2.4 times the height of the first medial tooth, giving the dental profile a stepped appearance ( Fig. 3 View Fig ). The first medial tooth is 7.9 mm tall with a 6.5 mm base. It is squatter and more triangular than the highly recurved laniaries. Like the posterior-most laniary it has a slightly recurved mesial margin and a straight, less expanded distal margin. The second medial tooth is 10.8 mm tall and with a 7.2 mm base. Despite the size difference, the two medial teeth are similarly shaped. Based on a comparison with the jaw of NHMUK PV R 37002, Klobiodon is estimated to have achieved an adult wingspan of ~ 2 m.

Besides holotype, only one other specimen can confidently be referred to Klobiodon . OUM J.28410 ( Fig. 9) is an isolated mandibular symphysis with three complete alveoli, 63 mm anteroposteriorly and 13 mm dorsoventrally at its deepest point. It is broken posterior to the third alveolus, around the midpoint of the incomplete fourth alveolus and the alveoli are approximately 8 mm wide mesiodistally. The first alveolus is directed dorsoanteriorly, whereas the second and third alveoli are more dorsally oriented. The anterior symphysis is developed into a large sweeping prow which makes up 28% of the preserved jaw. The large alveoli and the relatively deep jaw are similar to the arrangement of the holotype of Klobiodon rochei . However, Klobiodon rochei is primarily defined on dental characters whereas OUM J.28410 lacks teeth. Therefore, it can only tentatively be placed in the genus but is sufficiently similar in its overall morphology to be provisionally identified as Klobiodon cf. rochei .

Remarks. —Basal pterosaur mandibles are highly variable between taxa ( Fig. 10 View Fig ) and often taxonomically informative. The Triassic pterosaurs Austriadactylus Dalla Vecchia, Wild, Hopf, and Reitner, 2002 and Preondactylus Wild, 1984 have similar dental morphologies with tightly packed,

O’SULLIVAN AND MARTILL—BATHONIAN PTEROSAURS FROM ENGLAND 625

10 mm

sub-triangular serrated teeth ( Dalla Vecchia et al. 2002; Dalla Vecchia 2003). Eudimorphodon Zambelli, 1973 , Carniadactylus Dalla Vecchia, 2009 , and Caviramus Fröbisch and Fröbisch, 2006 share similar complex dentitions, each with tightly spaced heterodont teeth and relatively enlarged anterior laniaries ( Wild 1984; Stecher 2008; Dalla Vecchia 2009). Dimorphodon Buckland, 1829 from the lowermost Jurassic of southern England caniform teeth which are tightly packed in the anterior half of the jaws, becoming more widely spaced with the gaps reducing posteriorly ( Buckland 1829; Padian 1984a).

Campylognathoides Strand, 1928 from the Toarcian of southern Germany has several anterior enlarged caniform teeth comparable in size to the anterior laniaries seen in several of the more basal pterosaurs ( Plieninger 1894). Scaphognathines (sensu Lü et al. 2012) are well known for their jaws whose depth can be as much as 1.4 times the length of the longest dentary tooth crown ( Goldfuss 1831; Wellnhofer 1991; Cheng et al. 2012). The tip of the jaw possesses a short prow, less prominent than that of rhamphorhynchines ( Wellnhofer 1975; Cheng et al. 2012). The teeth are slightly taller medially, giving the dentition a slightly arched profile ( Wellnhofer 1978; Cheng et al. 2012; Bennett 2014) and widely spaced. The curvature and robustness of the teeth in scaphognathines is somewhat variable between taxa and their position in the jaw ( Carpenter et al. 2003; Cheng et al. 2012; Bennett 2014).

Rhamphorhynchines (sensu Lü et al. 2012) possess large, procumbent fang-like laniaries ( Wellnhofer 1975, 1978; He et al. 1983; Padian 2008b; Hone et al. 2012; Lü et al. 2012) that can mesh together to form a “fish-grab” ( Wellnhofer 1991; Kellner and Tomida 2000; Unwin 2003), a cage-like structure at the anterior rostrum associated which some authors associate with an at least partially piscivorous diet e.g., Wellnhofer 1991). The mandibular symphysis in rhamphorhynchines develops into a hooked anterior prow that may vary in length ontogenetically ( Wellnhofer 1975, 1978). The Toarcian Dorygnathus has three enlarged recurved laniaries, which become less procumbent posteriorly. Directly behind the posterior most laniary there is a marked step in tooth height, with the next tooth being 16–33% the height of the last laniary. Posteriorly the successive teeth are of a similar height, are more erect and have a high triangular outline in lateral aspect. The Chinese rhamphorhynchine Angustinaripterus He, Yang, and Su, 1983 has teeth of relatively equal height with the anterior third becoming more strongly procumbent. The Late Jurassic rhamphorhynchines Bellubrunnus Hone, Tischlinger, Frey, and Röper, 2012 from the Late Jurassic of Solnhofen and Qinglongopterus Lü, Unwin, Zhao, Gao, and Shen, 2012 from the Oxfordian of China have a similar dentition to Rhamphorhynchus with the teeth being slightly procumbent and fang-like, and the anteromesial teeth being the most anteriorly inclined. The teeth show a similar profile to those of as Scaphognathus Goldfuss, 1831 with the “arch” created by the elongate third tooth. The Middle Jurassic scaphognathine Jianchangnathus robustus Cheng, Wang, Jiang, and Kellner, 2012 possesses a similar dental arrangement to Scaphognathus but with much larger, more anteriorly oriented laniaries, deep jaw and convex prow. Bennett (2014) argued that these features did not generically separate the two taxa and referred J. robustus to Scaphognathus robustus .

Wukongopteridae (sensu Wang et al. 2009) have numerous slightly recurved, well-spaced teeth ( Wang et al. 2009; Lü et al. 2011; Martill and Etches 2012). Ctenochasmatoids sensu Unwin 2003; also Pterodactylidae sensu Pereda-Suberbiola et al. 2012 ) possess a wide variety of dental morphologies which fall into two broad categories; well-spaced and slender, or tightly packed and slender ( Wellnhofer 1991). Dsungaripteroids have evenly spaced robust teeth situated in a straight jaw with the more derived dsungaripterids have curving edentulous anterior jaw with very robust medial teeth ( Young 1964; Wellnhofer 1991).

Klobiodon rochei can be distinguished from most pterosaurs listed above with little difficulty. Its teeth lack the serrations or multiple cusps of Austriadactylus , Preondactylus , Eudimorphodon , Caviramus , or Carniadactylus . Dimorphodon and Campylognathoides lack laniaries and have more closely spaced alveoli. Wukongopterid teeth have tighter spacing, straighter jaws and do not possess laniaries. The curvature of the jaw and the size of the teeth show that Klobiodon belongs in Rhamphorhynchidae (sensu Lü et al. 2011) , the parent clade of Scaphognathinae and Rhamphorhynchinae .

In Klobiodon the longest tooth is much larger relative to the depth of the jaw than is common for scaphognathines ( Cheng et al. 2012; Bennett 2014). At least five rhamphorhynchine taxa are known with lower jaws preserved and can be compared directly with Klobiodon . Of these five taxa, the lower jaws of Bellubrunnus and Qinglongopterus are crushed dorsoventrally and difficult to compare to Klobiodon ( Hone et al. 2012; Lü et al. 2012). Angustinaripterus lacks complete teeth but the alveoli are of a relatively consistent size ( He et al. 1983). This lack of medial/anterior tooth differentiation distinguishes it from Klobiodon . Dorygnathus and Rhamphorhynchus show some size differentiation in the teeth. In Rhamphorhynchus this differentiation is subtle with a slight increase in the tooth size towards the middle of the jaw followed by a gradual decrease posteriorly. This gives it a dental profile not dissimilar to Scaphognathus but distinct from Klobiodon . In contrast, Dorygnathus shares several similarities with Klobiodon . Like Klobiodon the jaws of Dorygnathus can possess a gentle curvature throughout their length, developing into a robust and well-developed prow ( Padian 2008b; Fig. 10 View Fig ). Dorygnathus also possesses a well-developed differentiation between small medial teeth and large somewhat recurved laniaries. These similarities do suggest that Klobiodon is more closely allied to Dorygnathus than other rhamphorhynchine pterosaurs, however, Klobiodon does possess a combination of characters that distinguish it from Dorygnathus . While both pterosaurs have small medial teeth, in Dorygnathus they are smaller, more closely spaced and less robust than in Klobiodon ( Padian 2008a) . In Klobiodon the posterior most laniary is 1.4–2.4 times the size of the first medial tooth. In Dorygnathus (SMNS 55886; Padian 2008b), the posterior most laniary is between 2.6 and 3.7 times the size of the first medial tooth. In the Vienna specimen of Dorygnathus (NHMW 1911/0001/0023 and is close to the size of the Klobiodon holotype [ Padian 2008b]), the ratio is 3.3. This difference is a consequence of the medially positioned teeth being much larger and more robust in Klobiodon compared to Dorygnathus , with a tooth length/alveolar width ratio in the anterior medial teeth of at least 1.3. In the above Dorygnathus specimens, the ratio ranges between 2.6 and 6, depending on the mesiodistal width of the tooth. The laniaries of Klobiodon are robust, with broad alveoli while those in Dorygnathus are not only more strongly anteriorly inclined but comparatively thinner mesiodistally. In Klobiodon the longest tooth in the jaw is at least 1.5 times the dorsoventral depth of the jaw at its deepest point. Contrasting this, in NHMW 1911/0001/0023 the tooth/jaw ratio is around 2.6 and in Dorygnathus UUPM R 156 the ratio increases to 3. This is due to Klobiodon ’s laniaries being shorter relative to the depth of the jaw whereas Dorygnathus has elongate laniaries set in a comparatively narrower mandible ( Padian 2008b). Within Scaphognathinae the most comparable animals are the two species of Scaphognathus . Scaphognathus crassirostris Goldfuss, 1831 possess smaller less procumbent teeth and is thus a poor comparison with Klobiodon . Scaphognathus robustus on the other hand has more comparable laniaries, being larger and more procumbent. This may suggest a possible relationship, but the taxa can be distinguished on several characters. Unlike Klobiodon , S. robustus lacks the strong size differentiation between the anterior and medial teeth seen in Klobiodon , with the most posterior tooth being almost 1/ 1 in scale to the anterior fangs. In S. robustus the longest tooth/jaw depth ratio is 0.8, less than is seen in Klobiodon and is due to the greater depth of its jaw.

While the elongated laniaries, convex anterior prow and deep jaw may indicate a relationship between Klobiodon and S. robustus , the prominent differentiation between the large anterior laniaries and small medial teeth are more like Dorygnathus which also possess a convex prow and anteriorly oriented fangs. The similarities with both taxa, combined with the continued lack of broad consensus in the structure of non-monofenestratan phylogeny (e.g., Lü et al. 2012; Andres and Myers 2013; Vidovic and Martill 2014) recommends caution is assigning the new genus to either Scaphognathinae or Rhamphorhynchinae . Klobiodon rochei is therefore conservatively identified here as a member of the Rhamphorhynchidae rather than assigned to either subgroup.

Remarks. —NHMUK PV OR 47991 was first mentioned by Waterhouse (1878) in a vertebrate acquisitions list for the NHMUK, who confirmed the specimen was donated by Robert Marsham. In the list Waterhouse (1878: 34) can be quoted as listing NHMUK PV OR 47991 as “mandible, right ramus, of Pterodactylus raptor , of Stonesfield Slate, Stonesfield”. The name Pterodactylus raptor was taken from an unpublished manuscript by Richard Owen. No formal description or figure reference has been associated with this name but Ingles and Sawyer’s (1979) compilation of NHMUK illustrations referenced NHMUK PV OR 47991 as being illustrated in Folio 201A of Richard Owen’s unpublished Collection of Drawings under the name Pterodactylus raptor (Michael Hanson, personal communication 2017). Lydekker (1880) placed NHMUK PV OR 47991 in Rhamphocephalus and identified it as a large example of Rhamphocephalus depressirostris . He argued that it possessed two characteristics of the species: five teeth in the dentary and a concave ventral jaw line. It was considered the best example of Rhamphocephalus depressirostris in the NHMUK collections and was noteworthy for its relatively large size. Lydekker’s (1880) description of NHMUK PV OR 47991 was written immediately after his new diagnosis of Rhamphocephalus depressirostris and this may explain some confusion on the part of Benton and Spencer (1995) who mistakenly identified NHMUK PV OR 47991 as the type specimen for Rhamphocephalus depressirostris .

Stratigraphic and geographic range.—NHMUK PV OR is most likely from the Stonesfield Slate Member of the Taynton Limestone Member, collected from the Stonesfield region of Oxfordshire.

Genus Rhamphocephalus Seeley, 1880 (nomen dubium)

Remarks.—The genus Rhamphocephalus is identified in this study as a non-diagnostic thalattosuchian crocodylomorph (see Non-pterosaurian section below). This assignment raises questions as to the taxonomic identification of the holotypes of “ Pterodactylus ” bucklandi and “ Rhamphorhynchus ” depressirostris , two isolated jaws described below.

“ Pterodactylus ” bucklandi Meyer, 1832 (nomen dubium)

(= Rhamphorhynchinae indet.)

Fig. 2A View Fig .

1832 Pterodactylus bucklandi ; Meyer 1832: 27–30. 1859 Rhamphorhynchus bucklandi ( Meyer, 1832) ; Huxley 1859: 658–

670. 1859 Pterodactylus bucklandi Meyer, 1832 ; Owen 1859b: 169. 1871 Rhamphorhynchus bucklandi ( Meyer, 1832) ; Phillips 1871: 224. 1888 Rhamphocephalus bucklandi ( Meyer, 1832) ; Lydekker 1888:

34–636. 1907 Rhamphocephalus bucklandi ( Meyer, 1832) ; Seitz 1907:289–6291. 1978 Rhamphocephalus bucklandi ( Meyer, 1832) , Wellnhofer 1978: 41. 1995 Rhamphocephalus depressirostris ( Huxley, 1859) ; Benton and

Spencer 1995: 128. 1996 Rhamphocephalus bucklandi ( Meyer, 1832) , Unwin 1996: 293. 2000 Rhamphocephalus bucklandi ( Meyer, 1832) ; Riqlés et al. 2000:

351. 2003 Rhamphocephalus bucklandi ( Meyer, 1832) ; Sayão 2003: 335. 2003 Rhamphocephalus bucklandi ( Meyer, 1832) ; Unwin 2003: 177. 2008 Rhamphocephalus bucklandi ( Meyer, 1832) ; Barrett et al. 2008: 68. 2012 Rhamphocephalus bucklandi ( Meyer, 1832) ; Buffetaut and Jeffery

2012: 1. 2012 Rhamphocephalus bucklandi ( Meyer, 1832) ; Steel 2012: 1341. 2013 Rhamphocephalus bucklandi ( Meyer, 1832) ; Witton 2013: 125. Holotype: “ Huxley 1859: fig. 2”, a lower jaw in left lateral view. The whereabouts of the specimen is unknown. Type locality: Smith’s Quarry, Sarsden, Oxfordshire, UK. Type horizon: Great Oolite Group, Bathonian, Middle Jurassic.

Material.—NHMUK PV R 28610 , 32752 , 37765 , 38014 , 38015 , 38016 , 38017 , 38019 , 38020 , 38025 , 40126 , 47994 , 47999 a, 1028, 1029, 1030, 1824, 2637, 6749, 6750 ( Steel 2012) ; OUM J.28275, J.28537, J.283043 ( Huxley 1859), isolated pterosaur limb elements from the Stonesfield region of Oxfordshire with no cross over with the holotype of “ Pterodactylus ” bucklandi .

Description. —The holotype (“ Huxley 1859: fig. 2”) of Rhamphocephalus bucklandi ( Fig. 2A View Fig ) is an isolated jaw with 5 alveoli and a short anterior prow. The holotype was (and remains) lost but as the other specimens figured by Huxley (1859) are accessioned in the OUM collections, it is likely it may have been held there at one point. Based on Huxley’s (1859) figure 2, it is a rhamphorhynchine jaw due to its anterior prow, its being relatively shallow and its low number of teeth. As we cannot locate the specimen, our only reference point being a drawing (which the holotype of Rhamphocephalus prestwichi highlights can be a poor representation of the actual specimen) and Huxley’s (1859) only identifier being it is morphologically distinct from R. depressirostris , we cannot recommend a further detailed analysis. As there is no crossover between Huxley’s (1859) figure and any of the appendicular material assigned to the genus by Lydekker (1888), we also identify the name as inapplicable to anything but “ Huxley 1859: fig. 2” and transfer all remaining material to either Pterosauria indet. or Rhamphorhynchinae indet. (see SOM 1). This unfortunately puts “ Pterodactylus ” bucklandi in a nebulous position. We can confidently state that the number of teeth, slim jaw and distinct prow would place the taxon within Rhamphorhynchinae . We can also say that the teeth seem less broad and the jaw less dorsoventrally thick than Klobiodon but unless the holotype is one day uncovered, we regrettable cannot test the taxon any further.

Remarks.—Placed within Rhamphorhynchinae indet. due to its prowed mandibular symphysis and alveoli indicating broad, well-spaced teeth. The holotype has been lost.

“ Rhamphorhynchus ” depressirostris Huxley, 1859 nomen dubium) =?Scaphognathinae indet.)

Fig. 11 View Fig .

1859 Rhamphorhynchus depressirostris ; Huxley 1859: 658–670.

1888 Rhamphocephalus depressirostris ( Huxley, 1859) ; Lydekker 1888: 34–36.

1978 Rhamphocephalus depressirostris ( Huxley, 1859) ; Wellnhofer 1978: 41.

1994 Rhamphocephalus depressirostris ( Huxley, 1859) ; Evans et al. 1994: 307.

1995 Rhamphocephalus depressirostris ( Huxley, 1859) ; Benton and Spencer 1995: 128.

1996 Rhamphocephalus depressirostris ( Huxley, 1859) ; Unwin 1996: 293.

2003 Rhamphocephalus depressirostris ( Huxley, 1859) ; Unwin 2003: 177.

2008 Rhamphocephalus depressirostris ( Huxley, 1859) ; Barrett et al. 2008: 69.

2012 Rhamphocephalus depressirostris ( Huxley, 1859) ; Buffetaut and Jeffery 2012: 1.

2012 Rhamphocephalus depressirostris ( Huxley, 1859) ; Steel 2012: 1341.

2013 Rhamphocephalus depressirostris ( Huxley, 1859) ; Witton 2013: 125.

Holotype: GSM 113723 View Materials , a three-dimensional lower jaw symphysis preserved in left lateral view.

Type locality: Smith’s Quarry (51°54’8.62”N 1°34’44.74”W), Sarsden,

Oxfordshire, UK.

Type horizon: Fuller’s Earth Formation, Bathonian, Middle Jurassic.

Material. — Holotype and NHMUK PV R 40126 , isolated limb material; Taynton Limestone Formation (Bathonian, Middle Jurassic), Stonesfield , Oxfordshire ( Steel 2012).

Description. — GSM 113723 is a three dimensional and nearly complete lower jaw symphysis with partial rami ( Fig. 11 View Fig ). It is 87 mm long and contained within a block of oolitic limestone typical of the Stonesfield Slate Member. The medial left and lateral right rami are both obscured by matrix, as is the dorsal aspect of the symphysis. It has been damaged with the right ramus broken off and reattached. The anterior symphysis is missing, exposing its oval cross-section and thin bone walls. The ventral posterior symphysis is bowed indicating the presence of a prow. In ventral view the symphysis has a relatively deep sulcus at its posterior boundary. There are five alveoli preserved on the rami, with the fifth alveolus of both rami bearing an in-situ tooth. There are no alveoli preserved posterior of these teeth. The alveoli are 5–6 mm mesiodistally and spaced 3–8 mm apart, with the spacing increasing posteriorly. The anterior alveoli are slightly splayed, giving the dorsal symphysis an undulose margin and hinting that the teeth projected anterolaterally. The preserved teeth are 15 mm long, thin and peg-like, with teeth directed anteriorly at ~70° relative to the jaw line.

The elongate jaw, thin bone walls, the dental arrangement and smooth bone texture confirm GSM 113723 as a pterosaur. The low tooth count and simple elongate teeth distinguish GSM 113723 from the more complex tooth pattern of basal non-rhamphorhynchid pterosaurs. The height and number of the medially placed teeth also serves to separate it from Jurassic monofenestratans, which tend to have numerous relatively low medial teeth ( Wellnhofer 1978, 1991; Wang et al. 2010). GSM 113723 is confidently placed in Rhamphorhynchidae due to the narrow mandibular symphysis, reduced tooth count and wide spacing of the teeth. The apparent lack of teeth posterior to the fifth alveolus may ally it with Scaphognathinae as these pterosaurs have no more than 5–6 teeth in the lower jaw ( Unwin 2003). Tall, peg-like teeth, wide medial spacing and relatively short prows are also found in scaphognathines ( Cheng et al 2012; Bennet 2014; Zhou 2014), being similar to the jaw of Scaphognathus (= Jianchangnathus ) robustus ( Bennett 2014) , a scaphognathine from the Tiaojishan Formation of China.

GSM 113723 is identified here as?Scaphognathinae but the question remains: can it be assigned to a genus or species? Neither Huxley’s (1859) or Lydekker’s (1888) definition of Rhamphocephalus depressirostris (robust jaw, shallow anterior rostrum, reduced curvature in the mandible, 5 teeth) are diagnostic by modern taxonomic standards. Unfortunately, the taphonomy of the jaw limits morphological taxonomy. The preserved tooth is clearly distinct from Klobiodon but is still similar to what we see in several Jurassic pterosaur taxa (see above) and lacks the distinct ratios inform the definition of Klobiodon . There is a deep mandibular sulcus which can be seen at the posteroventral prow but this character could be ontological. Overall, morphological definition is limited. We also cannot suggest the erection of a chronotaxon, not only because it runs counter to the taxonomic practices of this research but also because named scaphognathines occur in contemporaneous formations (e.g., Cheng et al. 2012). Therefore, GSM 113723 is not currently considered a strong candidate for a generic or specific name despite representing a distinct mandibular morphotype within the assemblage. Regrettably this puts “ Rhamphocephalus ” depressirostris in a similar nebulous position as “ Pterodactylus ” bucklandi , if for different reasons. However, if GSM 113723 can at some point undergo further preparation or perhaps CT-scanning, we may reveal currently obscured diagnostic characters.

Remarks. —Alleged attribution to Scaphognathinae Hooley, 1913 is based on the morphology of the teeth and consistent increase in the alveolar spacing towards the posterior end of the jaw.

Other rhamphorhynchid remains

Other identifiable rhamphorhynchid fossils consists of axial and dental remains collected from the Taynton Limestone Formation of Stonesfield and the Chipping Norton Formation of Chipping Norton, both in Oxfordshire (Bathonian, Middle Jurassic). Full locality information is available in SOM 1.

Scapulocoracoids.— OUM J.28294 ( Fig. 12B View Fig ) is a partial left scapulocoracoid preserved in posterolateral view, heavily worn around the glenoidal region. The scapula is 37 mm long and the coracoid is 23 mm long. The coracoid possesses a low biceps tubercle that extends 2 mm ventrally and 6 mm proximodistally. The glenoid is posterolaterally positioned and limited to the scapula. The angle formed between the scapula and coracoid is at 75°. A second complete scapulocoracoid, OUM J.28295 ( Fig. 12A View Fig ) is near complete and preserved in anteroposterior view. It is missing the proximal ends of both the scapula and the coracoid as well as the surface bone of the glenoid. In the case of the glenoid, this exposes the internal trabeculae. The scapula is 28 mm long and the coracoid being 40 mm long. The angle between them is approximately 75°. The restriction of the glenoid to the scapula identifies them as non-pterodactyloid. The proximal coracoid is slenderer than Campylognathoides or Dimorphodon ( Buckland 1829; Padian 2008a). The low biceps tubercle is like Dorygnathus , Darwinopterus , and Campylognathoides ( Padian 2008a; Lü et al. 2010) but lower than other basal pterosaurs. The angle of 75° falls into the range shared by several basal pterosaurs ( Wild 1984; Padian 2008b; Witton 2013). The glenoid limited to the scapula and approximately 10 mm proximodistally. The lateral scapulocoracoid is less inclined than in Rhamphorhynchus but in a similar position to Dorygnathus and Scaphognathus . The coracoid distinguishes it from Darwinopterus where it is wider dorsoventrally and does not taper proximally ( Lü et al. 2010). Based on the angle formed by the scapula and coracoid, along with the thin coracoid morphology, the scapulocoracoids are identified as? Rhamphorhynchidae indet.

A complete right scapulocoracoid ( OUM J.28297) is preserved in medial view ( Fig. 13 View Fig ) and is the most complete pterosaur shoulder element in the assemblage, with a 62 mm long scapula ramus and a 60 mm coracoid ramus. The two elements form an angle of ~70°. The glenoid is restricted to the scapula and is ~ 16 mm long. The acrocoracoid process is rounded and extends 4.5 mm distal of the glenoid. The coracoid bows slightly behind its scapular articulation. The sternocoracoidal joint is 3 mm wide dorsoventrally and is more rounded on its ventral surface. The biceps tubercle is very low, extending 1 mm ventrally and 7–9 mm proximodistally. OUM J.28297 is identified as a non-pterodactyloid by the glenoid restriction and the coracoid not exceeding the scapula in length. However, the scapula is of sub-equal length to the coracoid whereas in most basal pterosaurs, the scapula is between 124–60% the length of the coracoid ( Buckland 1829; Wild 1984; Padian 2008a, b). In Sericipterus ( Andres et al. 2010) and Rhamphorhynchus the scapula and coracoid are of sub-equal lengths. Although this suggests a rhamphorhynchine affinity, the same situation occurs in Darwinopterus ( Lü et al. 2011) . However, Darwinopterus has a larger biceps tubercle than OUM J.28297 while in rhamphorhynchines the size of the biceps tubercles is variable ( Padian 2008b; Andres et al. 2010; O’Sullivan and Martill 2015). Based on the ratio of the scapula to the coracoid and the profile of the biceps tubercle, OUM J.28297 is tentatively identified as? Rhamphorhynchinae indet.

Humeri.— OUM J.23043 ( Fig. 14A View Fig ) is a complete left humerus exposed in dorsal view, 90 mm long proximodistally with a diaphysis 10 mm wide anteroposteriorly at its medial point with a length/width ratio of 9/1. The medial process and distal articulation are preserved primarily as an external mould on the rock surface, but with a clearly distinct outline. The diaphysis develops a gentle anterior bowing distally, angled at approximately 165° and distally has a round anterior articulation which extends 8 mm anterior to the shaft. The medial process has a triangular outline and is 10 mm long anteroposteriorly. The DPC is 13 mm anteroposteriorly and 19 mm proximodistally. The anterior margin is rounded, giving the DPC a short tonguelike shape ( Fig. 14A View Fig ). OUM J.23043 is one of the largest Jurassic pterosaur humeri known, 7% larger than the largest Dorygnathus humerus and 13% that of the largest Rhamphorhynchus ( Wellnhofer 1975; Padian 2008b). It can be distinguished from via the morphology of the DPC and medial process. Eudimorphodon ( Wild 1984) , Caviramus ( Stecher 2008) and Campylognathoides ( Padian 2008a) have enlarged quadrangular DPC. Wukongopterids have similar morphologies to OUM J.23043 but the DPC is less robust ( Lü et al. 2011). The medial process of aurorazhdarchids and basal ctenochasmatoids are distally deflected away from the proximal articulation ( Wellnhofer 1978; Vidovic and Martill 2014). OUM J.23043 lacks the medial pinching of the DPC seen in Nesodactylus Colbert, 1969 and some examples of Rhamphorhynchus . The DPC of Scaphognathus is a like OUM J.23043 with a comparable placement of the medial process. However, the proximal margin of the DPC is straighter and the DPC is less strongly deflected ( Bennett 2014). Sericipterus has a higher length/width ratio than OUM J.23043 ( O’Sullivan et al. 2013) and a more elongate DPC ( Andres et al. 2010). The humerus of Dorygnathus is like OUM J.23043 ( Padian 2008b) but the DPC is generally more elongate. Ultimately OUM J.23043 does not perfectly correlate with any described pterosaur humerus but the DPC, medial process and the gentle curvature of the diaphysis is most like Scaphognathus ( Cheng et al. 2012; Bennett 2014), and thus it is assigned to?Scaphognathinae indet. Scaphognathine humeri make up at least 6% of the total wingspan ( Bennett 2014), suggesting a total wingspan of 1.5 m.

NHMUK PV R 40126b ( Fig. 14B View Fig ) is a small 15 mm long humerus, and 1.5 mm wide medially. It is almost complete, missing only the proximal humeral head, medial process and some of the distal articulation. The DPC is 2.4 mm anteriorly, semi-tongue shaped with a broad sub-rectangular anterior margin and a slight proximodistal pinching towards the extremity of the DPC. NHMUK PV R 40126b is identified as? Rhamphorhynchinae indet based on the slight elongation and minor pinching of the DPC. At 15 mm long NHMUK PV R 40126b is one of the smallest pterosaur humeri known and based on similarly sized examples of Rhamphorhynchus from the Solnhofen Limestone ( Wellnhofer 1975), is most likely a juvenile with a 290–340 mm wingspan. NHMUK PV R 40126c ( Fig. 14C View Fig ) is a right humerus preserved in dorsal view. It is 8 mm long proximodistally with a 6 mm wide diaphysis, giving it a length/width ratio of 13:1. It is missing a large section of its humeral head and distal diaphysis. The distal articulation is not preserved. The diaphysis curves anteriorly at 170–175°. NHMUK PV R 40126c does preserve the medial process. The DPC has a rounded anterior margin. The high length/width ratio in combination with the anteriorly bowed shaft of NHMUK PV R 40126c is indicative of it being a possible rhamphorhynchine identification. NHMUK PV R 40126c is identified here as Rhamphorhynchidae , most likely? Rhamphorhynchinae . The possibility that NHMUK PV R 40126c represents an adult example of NHMUK PV R 40126b and that the straighter shaft in the latter specimen is an ontogenetic feature cannot be ruled out. However, without any intermediaries the two morphologies are considered sufficiently different to rule out an ontogenetic transition. Thus, NHMUK PV R 40126a–c confirms the presence of a second large humeral morphotype that can be assigned to Rhamphorhynchinae rather than Scaphognathinae.

NHMUK PV R 28160a is a left pterosaur humerus preserved in dorsoposterior view ( Fig. 14D View Fig ). It is 14 mm proximodistally with a 1–2 mm medial diaphysis. It is anteriorly bowed through an arc of ~160°. The medial process and DPC both project into and are partially overlain by the matrix. The diaphysis has several “dimples” across the posterior margin corresponding with an immature bone texture as illustrated by Tumarkin-Deratzian et al. (2007). This, combined with its small size, indicates that NHMUK PV R 28160a is a juvenile. The DPC proximally deflected ventral margin. It is similar in overall morphology to OUM J.23043 and is thus also identified as?Scaphognathinae indet. It should be noted that Unwin (2015) considered NHMUK PV R 28160a as a possible monofenestratan humerus but did not explain his reasoning in detail.

Wing phalanges.—There are several WPIs within the Bathonian pterosaur assemblage ( Fig. 2 View Fig ; see SOM 1), although their preservation is variable. The WPIs are of a seemingly uniform morphology, being relatively large and robust elements between 101–142 mm long. Each WPI diaphysis is bowed medially such that the proximal and distal termini are angled at 170–175° relative to the midpoint. The presence or absence of a posterior longitudinal groove is indeterminate in all specimens given their preservation, however, NHMUK PV R 40126 C ( Fig. 2B View Fig ) is not only free of surrounding sediment but broken medially, revealing a triangular cross-section with a relatively thick bone wall and a more rounded anterior surface. Taxonomic identification of these phalanges is problematic given their isolated nature. The consistent robustness and slightly bowed diaphysis suggest they are most likely from a single taxon with distinctive anterior bowing. This is unusual as most pterosaurs have straight proximal phalanges ( Wellnhofer 1991; Witton 2013). Unwin (2003) suggests that a bowed phalanx may be a dsungaripteroid apomorphy but notes that a dsungaripteroid presence in the Taynton Limestone is problematic as presently there is no substantive evidence for Dsugaripteroidea in strata older than Kimmeridgian ( Fastnacht 2008). Anterior bowing is, however, found in several Dorygnathus specimens ( Padian 2008b) and it is therefore not a dsungaripteroid apomorphy. The phalanges of dsungaripteroids are more elongate ( Young 1964) than those of the British Bathonian pterosaurs while those of Dorygnathus (e.g., SMNS 56255; Padian 2008b) are shorter and somewhat broader, as in NHMUK PV R 40216. He we identify the phalanges as Rhamphorhynchidae indet. Based on their similarity to the phalanges of Dorygnathus the wingspans of these pterosaurs are estimated to be between 1.4 m and 2 m.

Mandibles.—NHMUK PV R 1824 ( Fig. 15 View Fig ), is broken transversely and divided into part and counterpart. The jaw is missing most of the anterior symphysis. It is 100 mm long with rami approximately 13 mm deep dorsoventrally. Both rami preserve five alveoli but with no teeth. The alveolar sections of the rami comprise approximately 49% of the total ramus length. The posterior alveoli are dorsally placed with the more anterior alveoli becoming more laterally placed. The alveoli range between 4–5 mm mesiodistally and are set 6–8 mm apart. The taxonomic identity of NHMUK PV R 1824 is difficult to determine. The number of medial teeth posterior to the symphysis, their relatively uniform size and spacing suggest it is a rhamphorhynchid but it is unclear if it is distinct from or synonymous with Klobiodon rochei . While the holotype of Klobiodon rochei has similar anterior alveolar spacing, the rami of NHMUK PV R 1824 appears more gracile than Klobiodon rochei , and the rami lack the gentle curve seen in Klobiodon . The rami remain relatively straight but curve ventrally proximal to the posterior articulation. It is tentatively identified as Rhamphorhynchidae indet.

Cranial material.— UMZC T.718 ( Fig. 16 View Fig ) is a semi-3D posterior skull, 52 mm anteroposteriorly and 37 mm dorsoventrally. It is catalogued under the genus “ Rhamphinion jenkensi ” Padian, 1984b , an Early Jurassic American genus from the Kayenta Formation of north-eastern Arizona. While the holotype of Rhamphinion is from the same region of the skull, there is very little proportional similarity between the two fossils (see Padian 1984: fig. 1 for comparison) and as the identification has no accompanying descriptive text, this identification is considered here to be unsupported. It preserves the quadrate, quadratojugal, jugal, postorbital and squamosopostorbital bar. The quadrate is 32 mm dorsoventrally and 1.6 mm medially in lateral view. It is elongated, strap-like in posterior view and angled posteriorly at 118°. Ventrally it has a well-developed and rounded condyle for articulating with the mandibular glenoid. The condyle is ~ 3.8 mm wide. Only the postorbital and quadrate processes of the squamosal are preserved but their full extent is obscured by fractures and some fusion of the skull. The postorbital is tri-radiate, 8 mm anteroposteriorly and 11 mm dorsoventrally. The squamosoparietal bar is elongate and sub-rectangular, 11 mm long anteroposteriorly and 4 mm wide dorsoventrally. The quadratojugal is sub-triangular, with slightly elongated jugal and quadrate processes. The quadratojugal is well developed, 2.3 mm dorsoventrally and 6.7 mm anteroposteriorly, with the jugal being the largest element in the specimen. Including its processes, it is 34 mm anteroposteriorly and ~ 6 mm dorsoventrally. It is a sub-quadrangular bone, although here the maxillary process is more of a flange, being large and sub-rounded where it borders the antorbital fenestra. The postorbital process is the longest of the four but the point of contact with the postorbital is one of the most damaged regions of bone, obscuring its true length. It is estimated to be 8–16 mm long. The postorbital and lacrimal processes form an angle of ~70° around the ventral margin of the orbit. UMZC T.718 preserves three fenestrae in varying degrees of completeness. The superior temporal fenestra is missing its dorsal half. It is bounded by the squamosal and postorbital and is 14 mm anteroposteriorly. The ventral margin is smooth and sub-oval. The anterior border is more vertical than the posterior, and the anteroventral margin is shallower than the posterior. The inferior temporal fenestra is the only complete fenestra in the specimen. It is 28 mm dorsoventrally and 14 mm anteroposteriorly. The ventral margin is smooth and sub-oval. The anterior border is more vertical than the posterior, and the anteroventral margin is shallower than the posterior. Its boundary includes the quadrate, quadratojugal, jugal, postorbital and squamosal. It possesses an irregular piriform morphology outline, with the dorsal half being wider with a relatively straight dorsal boundary in comparison to the ventral half’s thinner, more angular appearance. The orbit is missing its dorsal half but is clearly the largest of the three fenestrae. It is bounded by the jugal ventrally and the postorbital posteriorly. Like the inferior temporal fenestra, the orbit is piriform with the ventral boundary approximately 10 mm wide in comparison to the 24 mm dorsally positioned widest point. UMZC T.718 is identified as a non-monofenestratan based on its low-lying antorbital fenestra, the relatively shallow angle of the quadrate and its tetraradiate jugal. Generic identification is problematic given the limited material: the angle of the quadrate at 118° distinguishes it from Dimorphodon ( Buckland 1829) , Campylognathoides ( Padian 2008a) and Anurognathus ( Bennett 2007) but is comparable to several other basal pterosaurs, including Eudimorphodon , Austriadactylus , and Dorygnathus ( Wild 1984; Padian 2008a, b). The angle formed by the dorsal processes of the jugal is typical of basal pterosaurs more derived than Campylognathoides or Eudimorphodon ( Wellnhofer 1975; He et al 1983; Wild 1984; Padian 2008b; Bennett 2014). The broad sub-rounded inferior temporal fenestra is distinct from most basal pterosaurs ( Wild 1984; Padian 2008a) but a similar condition is seen in several rhamphorhynchids ( He et al. 1983; Padian 2008b; Cheng et al. 2012; Bennett 2014). Based on the angle of the quadrate combined with the ventral orbital and inferior temporal fenestra morphology, UMZC T.718 is identified here as? Rhamphorhynchidae .

OUM J.28409 ( Fig. 17 View Fig ) is an 83 mm long isolated and near-complete left maxilla and a fragment of the posterior premaxilla. Much of the premaxilla and the posterior maxilla are absent. The surface texture appears irregular, but this is a taphonomic artefact as most of the external bone wall has been eroded away anteriorly, revealing the internal trabeculae. The centrally positioned nasal process is angled posteriorly at 70–90°. While the posterior margin of the antorbital fenestra is not present in the maxilla, the preserved section of the premaxilla defines the anterior border of the nares. The nares is approximately 23 mm anterioposteriorly. The antorbital fenestra is at least 27 mm along the same plane. It has five maxillary alveoli. The smallest is 4 mm mesiodistally, with the subsequent alveoli approximately equidimensional at around 7 mm. OUM J.28409 is identified as pterosaur based on a combination of the nares and antorbital fenestra being proximal to each other and the thin bone walls. It is further be identified as a non-monofenestratan as it possesses differentiated nares and antorbital fenestra. The possession of five relatively widely spaced maxillary tooth pairs is a rhamphorhynchid characteristic and OUM J.28409 is identified as Rhamphorhynchidae indet.

Teeth.—Numerus isolated teeth ( Fig. 18 View Fig ) have been collected from the Taynton Limestone Formation and accessioned within pterosaur collections. However, due to the difficulty in identifying ex-situ teeth with a simple slender conical, gently curved morphology, many of these have been misidentified. Gently recurved teeth with a slightly sigmoidal shape accessioned as pterosaur might more correctly be assigned to Teleosauridae ( Massare 1987) . Pterosaur teeth can be identified by the restriction of the enamel to the higher parts of the tooth crown while the base of the crown is exposed dentine. The exposed dentine sometimes extends laterally to a median point of the crown ( Witton 2013). Six teeth thought to be pterosaurian, lacking carinae and seemingly possessing restricted enamel are illustrated here ( Fig. 18 View Fig ). They share a similar morphology of being relatively elongate and distally recurved, making them likely to be rhamphorhynchine.

Monofenestrata remains

Monofenestrata is a recently erected group ( Lü et al. 2010) including the transitional Wukongopteridae and the derived Pterodactyloidea . Wukongopteridae includes several taxa showing mosaic evolution with animals possessing derived skull morphology but plesiomorphic bodies. The recognition of this group has not only revolutionised our understanding of pterosaur evolution but complicates Middle Jurassic taxonomy. With few uniquely diagnostic features ( Witton et al. 2015) found within the group, isolated remains are easily assigned to either basal or pterodactyloid pterosaurs, and consequently, this has been a major contributor to the conservative approach adopted in this study. Pterodactyloidea is a highly diverse group including at least 11 families. Distinguished from basal pterosaurs by features such as metacarpal length, rostral index, dentition, cervical structure, pelvic morphology and numerous other features, the group includes the most well-known pterosaurs ( Wellnhofer 1991; Unwin 2005). The Middle Jurassic has gained increasing relevance in their evolution as several discoveries push the group further and further back ( Andres et al. 2014). Establishing if any monofenstratans are present in the Bathonian of Britain has considerable significance to Jurassic pterosaur diversity.

Vertebrae.—Whereas several examples of Bathonian pterosaur vertebrae are known (see SOM 1), the most diagnostic specimens are those accessioned under NHMUK PV R 40126a. These are two isolated cervical vertebrae referred to here as CVA and CVB ( Fig. 19 View Fig ). CVA is an isolated cervical vertebra within a slab of oolitic limestone while CVB is a compressed, but three dimensional isolated cervical free of matrix. Both were described by Owen (1859b) as coming from the “Stonesfield oolite” but no locality data was provided. CVA is identified as a cervical based on its robustness, the lack of strongly developed horizontal transverse processes and its relatively elongate centrum ( Howse 1986). It is 16 mm anteroposteriorly and 12 wide mm transversely. Based on its quadrangular appearance it is identified as a cervical 3–7 ( Wellnhofer 1991). In comparison to CVB, CVA is the more poorly preserved. As well as being dorsoventrally compressed, it is heavily fractured along its lateral cervical rib articulation margins. Both the pre- and postzygapophyses are preserved but are incomplete. The neural spine has been broken, most likely during compaction. CVB is morphologically similar CVA and considered a cervical 3–7. It is 25 mm anteroposteriorly (excluding the posterior condyle) and 15 wide mm medially. It has undergone dorsoventral compression, the loss of the neural spine and the right prezygapophysis. CVB has undergone some repair or preparation as an adhesive fills several fractures in the surface. Despite this damage, CVB is well preserved and provides a clear view of the centrum in ventral view. Ventrolateral to the posterior condyle are two enlarged projections that in dorsal view are slightly dished. Andres et al. (2010) provided a brief description of NHMUK PV R 40126e, identifying these structures as postexapophyses. In the anteroventral centrum a prominent hypapophysis extends posteriorly towards the midpoint of the cervical. On either side of the neural arch are oval structures which are most likely lateral pneumatic foramina ( Wellnhofer 1991; Andres and Norell 2005). On either side of the vertebra are transverse processes, broken proximal to the centrum, corresponding to the dorsal articulation loci for the cervical ribs. NHMUK PV R 40126e have been discussed previously by Owen (1859b) and Andres et al. (2010). Owen figured CVB and identified it as “ Pterodactylus ” bucklandi but did not provide a description. Andres et al. (2010) discussed the vertebrae in comparison with the cervicals of the rhamphorhynchine Sericipterus from the Shishugou Formation in Xinjiang, China. They noted that NHMUK PV R 40126e cervicals were of a similar size to those of Sericipterus Andres et al. (2010) , arguing that they were non-pterodactyloid with the unusual feature of possessing postexapophyses. They suggested that development of postexapophyses may correlate with increasing size of cervicals.

Whereas CVA is more limited in its taxonomic information due to taphonomy, the three-dimensional preservation of CVB allows for a more detailed comparison. Howse (1986) considered the possession of cervical ribs as a characteristic of basal pterosaurs, but reduced cervical ribs are known from several pterodactyloid clades ( Zhou and Schoch 2011), including the highly derived Azhdarchidae ( Witton and Naish 2008) . Cervical ribs are not preserved in the holotype of the non-pterodactyloid monofenestratan Darwinopterus modularis Lü, Unwin, Jin, Liu, and Ji, 2010 , but there appear to be anterolateral projections on several vertebrae which may correspond to cervical rib facets. Therefore, the presence of these facets is an unreliable character for distinguishing between pterodactyloid and non-pterodactyloid vertebrae. The presence of hypapophyses in combination with the postexapophyses corresponds with Howse’s (1986) descriptions of pterodactyloid cervical vertebrae however, as Andres et al. (2010) note, in most respects these vertebrae appear typically non-pterodactyloid. Most likely these vertebrae are from a non-pterodactyloid monofenestratan, identified here as?Monofenestrata indet., which have garnered considerable attention for possessing a mosaic of features of both pterodactyloids and non-pterodactyloids ( Lü et al. 2010), as seen in NHMUK PV R 40126a.

Metacarpals.—NHMUK PV R 28160b ( Fig. 8B View Fig ) is a 41 mm MCIV with a preserved length/width ratio of 8/1. It has a prominent distal condyle above which is a piece interpreted as either a broken sliver of the shaft of MCIV or the remnants of one of the associated metacarpals I–III. Basal pterosaurs have MCIVs with a maximum length/width ratio of approximately 5:1 whereas pterodactyloids have a length/ width ratio of 9:1–20:1 ( Wild 1984; Wellnhofer 1991; Martill et al. 2013). Although a ratio of 8: 1 in NHMUK PV R 28160b is closer to the pterodactyloid condition, the proximal and distal epiphyses are relatively broader and more like those of basal pterosaurs. The non-pterodactyloid monofenestratan Darwinopterus has a MCIV with a ratio of ~10:1, with a similar constricted medial shaft along with wider proximal and distal epiphyses ( Lü et al. 2010). NHMUK PV R 28160b is therefore identified here as?Monofenestrata indet.