Type species: Cruxicheiros newmanorum gen. et sp. nov.

Etymology: From Latin crux, cross; and Greek cheiros, hand; intended as “cross hand”, a version of the locality name.

Diagnosis.—As for the type and only species.

Etymology: After the Newman family, owners of Cross Hands Quarry, Warwickshire, United Kingdom.

Type material: Holotype: WARMS G15770, a partial right femur. Paratypes: WARMS G15771, additional theropod material from the type locality that probably represents the same individual as the holotype: an anterior dorsal or posterior cervical vertebra; a dorsal neural arch; a partial dorsal vertebra; the anterior half of a middle-distal caudal vertebra; a partial right scapulocoracoid; a partial left ilium; the proximal end of a left pubis; numerous rib and bone fragments.

Type locality: Chipping Norton Limestone Formation (Lower Bathonian Zigzagiceras zigzag Biozone; Torrens 1980) of Cross Hands Quarry (National Grid Reference [NGR] SP 269 291) near Little Compton, Warwickshire.

Type horizon: The Chipping Norton Limestone Formation of the Little Compton area is characterised by sandy ooidal and bioclastic limestones, yielding a marine bivalve fauna (Sellwood and McKerrow 1974; Sumbler 2002). Abundant terrestrial plant material (Sumbler 2002) indicates significant terrigenous input into a shallow marine environment and a potentially appropriate mechanism for derivation of the dinosaur remains.

Diagnosis.—Tetanuran theropod with autapomorphic proximomedially inclined ridge within the trochanteric fossa of the femur. Differs from the contemporaneous Megalosaurus bucklandii in posessing low proportions of the dorsal neural spines, transversely broader dorsal neural spines, a prominent posterior flange of the femoral caput and a lower ratio of anteroposterior length to mediolateral width of the pubic peduncle (1.60 in Megalosaurus and estimated between 1.00 and 1.10 in C. newmanorum). Few informative comparisons can be made between C. newmanorum and Dubreuillosaurus or Poekilopleuron from the lower Bathonian of France because of limited overlap between known material. However, Dubreuillosaurus shows two large internal chambers separated by a midline septum in pneumatic vertebrae and an unfused scapula and coracoid (perhaps due to immaturity of the holotype) (Allain 2005) whereas C. newmanorum vertebrae have more numerous internal pneumatic chambers and the scapulocoracoid suture is fused and swollen. Poekilopleuron also differs from C. newmanorum as it lacks prominent anterior spur on the caudal neural spines.

Remarks.—Collections records for the specimens are scant and there is no formal record of their association such as a quarry map. However, the relative sizes and the absence of repeated elements are consistent with the hypothesis that they represent a single individual. Additionally, the remains were collected over a short time period of one or maybe two field seasons (BIRMG, unpublished collections records), and all preserve an identical matrix of sandy bioclastic ooidal limestone as well as residual patches of re-deposited stalagmitic calcite. They therefore probably correspond to a point locality. Finally, diagnostic remains of large-bodied theropods other than Megalosaurus are rarely preserved in British Bathonian deposits (Benson 2010), but most of the Cross Hands Quarry specimens are demonstrably different from Megalosaurus. If the material does represent multiple individuals then it is very unlikely that it represents more than one of these rare non-Megalosaurus taxa.

The following specimens from Cross Hands Quarry are listed in the accessions register at BIRMG: metacarpal; coracoid; 3 ribs; fibula; vertebra; 2 vertebrae; left ilium, left femur fragment; lower femur fragment; and pubic bone fragment (BIRMG, unpublished collection records). The listed metacarpal and fibula are unaccounted for and may be lost or have originally been misidentified. Additionally, four vertebrae are present whereas only three were listed. This leaves the possibility that one of the vertebrae described here was not recovered from Cross Hands Quarry but as there is no way to identify which this is, they are all described.

An additional specimen, a fragmentary right tibia (OUMNH J.29831), was collected from Cross Hands Quarry by Mr. P. Stewart and donated to the OUMNH in 1980. This specimen is very poorly preserved and its affinities within Theropoda cannot be determined. It is not referred to Cruxicheiros newmanorum herein as the details of its collection, from where and which horizon in the quarry it was collected, cannot be established at present. The specimen should be considered as Theropoda indet.

Stratigraphic and geographic range.—Only known from a single locality in the Lower Bathonian of Warwickshire, United Kingdom.

Description

Vertebrae.—A fragmentary posterior cervical or anterior dorsal vertebra is preserved as a transverse slice from close to the posterior end of the bone (Fig. 1A). This reveals the neural canal and several large internal chambers (camerae) in the centrum. These camerae are probably pneumatic in origin, arising from invasion of the bone by pneumatic diverticulae via large pneumatic foramina (“pleurocoels”) in the lateral surfaces of the centrum, as is common in theropods (Britt 1993). They are similar to the confirmed pneumatic camerae of Sinraptor (Britt 1993: fig. 21.1–3, 22.11) but are smaller and more numerous than those of megalosauroids (= spinosauroids) such as Marshosaurus (Britt 1993: fig 19.3–4) and Torvosaurus (Britt 1991, 1993), in which two large, main chambers are usually divided by a midline septum. Vertebrae with thin external walls and very large numbers of small internal chambers are termed “camellate”. These are known in neoceratosaurs and carcharodontosaurian allosauroids among basal theropods (Britt 1993), which therefore differ markedly from C. newmanorum.

A dorsal centrum and transverse process (Fig. 1B) is mostly concealed by attached matrix. The centrum lacks internal pneumatic cavities, indicating that pneumatic foramina (= pleurocoels) were absent. A prominent anterior centrodiapophyseal lamina is visible. A dorsal neural spine and left transverse process (Fig. 1D) also have a large block of attached matrix in the space between them. The transverse process is oriented dorsolaterally, and the combined anterior and posterior centrodiapophyseal laminae form a stout longitudinal ridge along its ventral surface. Some matrix is still attached to the bone surface adjacent to this ridge anteriorly and posteriorly, giving the impression (Fig. 1D₁) that large foramina are present. However, they are absent. The neural spine is abraded posteriorly, but a thick, rugose, ligament ossification is present on the anterior surface, as in various basal tetanurans, including Marshosaurus (DMNH 3718) and the allosauroids Aerosteon (Sereno et al. 2008), Allosaurus (Madsen 1976), Neovenator (Brusatte et al. 2008) and Tyrannosaurus (Brochu 2002). Less prominent ligament ossifications are present in other large theropods such as Torvosaurus (Britt 1991). The distal end of the dorsal neural spine of WARMS G15771 is abraded but expands transversely (Fig. 1D₄), indicating that the distal end is almost complete. In the British Jurassic theropods Megalosaurus (Benson 2010) and Metriacanthosaurus (OUMNH J. 12144), and in Ceratosaurus (Madsen and Welles 2000) and Sinraptor (Currie and Zhao 1994), the neural spine is over 1.5 times the height of the centrum and substantially longer than the transverse processes. However, in Cruxicheiros the neural spine is approximately the same length as the transverse process, indicating relatively low proportions as in various other theropods including Allosaurus (Madsen 1976), Marshosaurus (DMNH 3718), Torvosaurus (Britt 1991) and Piatnitzkysaurus (Bonaparte 1986).

The anterior half of a middle-distal caudal vertebra (WARMS G15771) has a shallowly concave anterior articular surface and short prezygapophyses. A lump of limestone matrix adheres to the dorsal surface of the neural arch anteriorly. On the posterior surface of this lump, a section through the low anterior portion of the neural spine is preserved which indicates that an anterior spur of the neural spine was present. This feature is a tetanuran synapomorphy (Rauhut 2003).

Pectoral girdle.—A right scapulocoracoid fragment (Fig. 2A) preserves portions of the fused scapula and coracoid. The area of fusion is swollen and heavily ossified, as in Megalosaurus (Benson 2010), and the glenoid faces posterolaterally. The ventral margin of the coracoid foramen is preserved, forming a suboval opening on both the medial and lateral surfaces of the bone, as it does in most other theropods. The specimen is highly abraded and it is not possible to determine the morphology of the coracoid tubercle or posteroventral process, although this process is clearly present. Fragments of the scapula are also preserved (Fig. 2B). These are not adequate to assess the proportions of the blade. However, they do indicate that it was mediolaterally narrow with a cross-section that tapers dorsally and ventrally (Fig. 2C₂) as in other theropods.

Pelvic girdle.—The left ilium is 477 mm long as preserved, missing the anterior blade, posterior and medial blades, supracetabular crest, and ischial peduncle (Fig. 3A). The pubic peduncle and periphery of the iliac blade were damaged during preparation, so the bone surface is stripped away around the margins of the bone. A low, swollen median ridge is present on the lateral surface of the blade. This ridge is a tetanuran synapomorphy that is widely-distributed among basal members of the clade (Bonaparte 1986; Benson 2009; Zhao et al. 2010). Bone surface preservation is not sufficient to determine whether associated nutrient foramina were present in C. newmanorum. The supracetabular crest is damaged but seems to project ventrolaterally as a shelf, as in tetanurans other than Chuandongocoelurus and Monolophosaurus (Zhao et al. 2010). As preserved, the pubic peduncle is slightly longer anteroposteriorly (95 mm) than its maximum mediolateral width (90 mm). However, medial and posterior portions are broken off and it may have been as wide as long, or only slightly longer than wide. In Megalosaurus and most other tetanurans the pubic peduncle is at least 1.3 times as long anteroposteriorly as it is wide mediolaterally, but in the megalosaurid Eustreptospondylus (OUMNH J. 13558, J.29774; Sadleir et al. 2008) and the allosauroid Fukuiraptor (Azuma and Currie 2000) it is approximately as broad as long, resembling the condition in non-tetanurans. This may also have been the case in C. newmanorum.

The proximal end of the left pubis is abraded. The smooth anterior margin of the obturator “foramen” is preserved, but preservation is not good enough to determine if it was open as a notch or closed posteriorly to form a foramen.

Hindlimb.—A right femur is preserved in three pieces. One piece comprises the head and proximal part of the shaft (Fig. 3C), another comprises a portion of the shaft including the fourth trochanter (Fig. 3D), and the third comprises the abraded distal end (Fig. 3E). The proximal and posterior surfaces of the head and posterior surface of the caput are well-preserved. The head measures 185 mm mediolaterally and 95 mm anteroposteriorly at the caput. The greater trochanter is narrower than the caput, so the outline of the head in proximal view narrows from medial to lateral as in non-coelurosaurian theropods (Hutchinson 2001). The proximal surface of the head bears a shallow, anteromedially oriented groove, the proximal articular groove, which is present in non-neotetanuran theropods and many non-theropods (Hutchinson 2001). The posterior flange of the caput is prominent and extends further posteriorly than the posterior surface of the head (Fig. 3C₄), as in most basal theropods. However, this is unlike the low posterior flange of most megalosaurids, including Afrovenator (UC OBA 1) and Megalosaurus (Benson 2010). The trochanteric fossa, the groove bounding the posterior flange laterally on the posterior surface of the femur, is anteroposteriorly broad. Within this fossa, nearly parallel to the posterior flange, is a prominent ridge. This ridge is not present in any other basal theropod femur (e.g., Madsen 1976; Bonaparte 1986; Holtz et al. 2004; Sadleir et al. 2008) including the Bathonian taxa Dubreuillosaurus (MNHN 1998–13) and Megalosaurus (Benson 2010), and is considered here as an autapomorphy of C. newmanorum. The base of the lesser trochanter is preserved along with the lateral part of the anterior surface of the greater trochanter region. However, as it is very incomplete it is not possible to determine the morphology of the lesser trochanter.