Scelidosaurus, Owen, 1859

Scelidosaurus : discovery and naming

The first indication of terrestrial fossil reptiles, rather than the ubiquitous specialized marine ichthyosaurs and plesiosaurs ( Conybeare & De La Beche, 1821; Conybeare, 1822, 1824; Norman, 2000a), in the Liassic strata of West Dorset, was a collection of bones obtained by James Harrison (1819–64) who had retired prematurely to Charmouth due to illhealth ( Lang, 1947). Charmouth was then a hamlet near the mouth of the River Char and situated less than a mile along the coast east of Lyme Regis ( Fig. 1 View Figure 1 ). The fossil specimens collected from the foreshore and coastal cliffs (The Spittles-Black Ven – Fig. 1 View Figure 1 ) at Charmouth were loaned to Richard Owen in 1858. Although Owen made brief reference to this material as ‘ Scelidosaurus ’ in print ( Owen, 1859, 1860, 1862), it was three years after receipt of this material that Owen (1861) formally described and illustrated these fossils. The specimens were collectively given the binomial Scelidosaurus harrisonii Owen, 1861 (= ‘Harrison’s limbed reptile’). Owen’s choice of this name reflects a wish to differentiate this saurian (with limbs adapted for terrestrial locomotion) from the flippered or paddle-limbed aquatic reptiles that were more often discovered in the area.

From a strictly nomenclatural perspective the generic name first appeared in print in 1859 ( Owen 1859) and again the following year ( Owen 1860). However, neither report mentioned the full binomial nor designated (or illustrated) any of the material associated with the new name. The generic name Scelidosaurus was, therefore, according to modern guidelines for nomenclatural stability, a nomen nudum from 1859 until the date of the more detailed publication in 1861 ( Newman, 1968).

Some of the original specimens named Scelidosaurus may have been found in water-rolled boulders or as fragments on the foreshore beneath Black Ven, while others were most probably exposed during the quarrying of nodules (or slabs) of the carbonate-rich marlstone and shale that crops out as distinctive layers (hence ‘Lias’) along the cliffs of this coastline. Quarrying along the Charmouth–Lyme Regis coast was at this time an important commercial enterprise. The purer limestone could be burnt to yield quicklime (calcium oxide), an essential ingredient in the production of mortar and stucco, and also served as a ‘disinfectant’ (for burials) and as a soil conditioner for farming. The cliffs below Charmouth also became of particular importance industrially because a new form of mortar: ‘hydraulic cement’ (that would harden underwater) could be manufactured from the Black Ven Marl. This marl combined the ideal proportions of limestone-to-clay for hydraulic cement production. The present-day Charmouth Heritage Coast Centre on the seafront has been built on the foundations of the original mid-19 th century hydraulic cement works.

The renowned fossil discoveries made by Mary Anning (1799–1847) promoted interest in the collection of fossils from the steadily eroding coastal cliffs along this coastline ( Torrens, 1995) and undoubtedly fostered the interest in fossils shown by Harrison. Harrison’s reptile fossils comprised a mixed assortment of bones indicating the presence of animals with a range of body sizes and, as was only much later recognized, belonging to different dinosaur taxa. The specimens comprised a large partial femur (NHMUK OR109560 – Fig. 2A View Figure 2 ), an incomplete, but articulated, knee joint of equally large size, comprising the distal end of a femur and its associated proximal tibia and fibula (NHMUK OR39496 – cemented in life position by marlstone – Fig. 2B View Figure 2 ) and a poorly preserved ungual (claw) phalanx (NHMUK OR109561 – Fig. 2B View Figure 2 ). None of these items had any provable association.

There was also an apparently associated group of much smaller limb bones and vertebrae from a fourth individual ( LYMPH 1998 6.1-6.7 – Fig. 3 View Figure 3 ) and a larger, apparently near-complete skull ( NHMUK R1111 – Fig. 4 View Figure 4 ). Owen’s account of this material included fine lithographs of all these specimens ( Owen, 1861: tabs I– VI; see Figs 2–4 View Figure 2 View Figure 3 View Figure 4 ), but he failed to designate a holotype from among this extremely disparate collection of specimens. Therefore, Owen had inadvertently created a hypodigm for the name S. harrisonii based upon a set of syntypes belonging to as many as five separate animals .

Subsequently, via Harrison’s wife (James Harrison was by this time bed-ridden), Owen obtained the greater part of an articulated postcranial skeleton that was associated with the well-preserved, large skull. The skull and skeleton were embedded in 12 more-or-less contiguous broken nodules of marlstone ( Fig. 5 View Figure 5 ). This material was cleaned and partly prepared by Owen’s assistants (using chisels) so that it could be better described and illustrated ( Owen, 1863: tabs I–XI). The skull with its associated torso were subsequently purchased from the Harrisons by the British Museum (Lang, 1847) and the entire specimen was later registered as NHMUK R1111.

It is unfortunate, from a formal nomenclatural perspective, that the material described by Owen comprised not only different-sized individuals, but also different taxa ( Newman, 1968; Charig & Newman, 1992; Norman, 2000b, 2001). Owen’s first monograph provided the full binomial: Scelidosaurus harrisonii Owen, 1861 (by monotypy). Considerably later, Richard Lydekker (1888) somewhat idiosyncratically ( Charig & Newman, 1992: 281) selected as the ‘type’ (more correctly, the lectotype) the articulated knee joint (NHMUK OR39496) illustrated by Owen (1861: tab. II, 1–3; see Fig. 2B View Figure 2 ); consequently, the remaining specimens became a set of paratypes.

Acid preparation of Lydekker’s lectotype ( Fig. 2B View Figure 2 ) by Arthur Rixon enabled Newman (1968) to demonstrate that the knee joint was attributable to an indeterminate theropod dinosaur. Additional paratypes: the femoral shaft ( NHMUK OR109560 : Fig. 2A View Figure 2 ) and ungual phalanx ( NHMUK OR109561 : Fig. 2B View Figure 2 ), were also shown to be referable to other similarly indeterminable theropod individuals. In marked contrast, the associated bones of the small skeleton ( LYMPH 1998 6.1-6.7: Fig. 3 View Figure 3 ) and the well-preserved larger skull ( NHMUK R1111 : Fig. 4 View Figure 4 ) both pertained to an ornithischian dinosaur .

A long record of publications persistently associated the name Scelidosaurus with the distinctive skull and its associated articulated skeleton of an armoured ornithischian dinosaur, rather than the other poorly understood portions of the hypodigm. Given the longestablished precedent, an appeal was made to the ICZN ( Charig & Newman, 1992) for the replacement of the inappropriate lectotype designation (by Lydekker) with that of the entire skull and skeleton identified as NHMUK R1111. This recommendation was formally approved ( BZN, 1994).

ARMOURED DINOSAURS IN THE FOSSIL RECORD: PROBLEMATIC SYSTEMATICS AND PHYLOGENETICS

The description of Scelidosaurus came at an opportune moment in relation to research on armoured dinosaurs (not differentiating between ankylosaurs and stegosaurs, because this split was not recognized until the early decades of the 20 th century). The first discovery and report of dinosaur remains suggesting the presence of bony spines and plates (osteoderms) covering the body surface was published by Gideon Mantell (1833). He described and illustrated the articulated neck and anterior torso of an animal collected from Wealden rocks (Early Cretaceous) in Sussex, that he named Hylaeosaurus armatus . The next discovery (not appreciated for what it was at the time) was of an isolated tooth collected from Late Cretaceous rocks in Nebraska, and named Palaeoscincus costatus Leidy, 1856 . The formal description of Scelidosaurus by Owen (1861, 1863) was followed by the discovery in 1865 of another partial skeleton, from Wealden rocks on the Isle of Wight, of another armoured dinosaur named Polacanthus foxii Hulke, 1881 ( Nopcsa, 1905). In 1874, a much larger plated dinosaur was recovered from the Kimmeridgian (Late Jurassic) of Wiltshire: Omosaurus (= Dacentrurus ) armatus ( Owen, 1875) . Attention moved in the direction of North America following the discovery of the Kimmeridgian-aged remains of Stegosaurus and the establishment of a new ‘Order’ ( Stegosauria ) to recognize the existence of a group of closely related plated/armoured dinosaurs ( Marsh, 1877). By 1889, Marsh was able to expand on his ‘ Order Stegosauria’ and incorporated all these genera: Hylaeosaurus , Palaeoscincus , Scelidosaurus , Polacanthus , Omosaurus and Stegosaurus , along with the later-described Priconodon , Dystrophaeus , Euskelosaurus , Anthodon , Acanthopholis and Struthiosaurus ( Marsh, 1889a) . His Stegosauria thus included both stegosaurs and ankylosaurs, which today are grouped together in Thyreophora. The ankylosaurs were neither understood nor described as a distinct grouping at this time. Marsh further confused matters for a considerable time by incorporating his Ceratopsidae within the Stegosauria ( Marsh, 1889b), because of the mistaken association of some flattened osteodermal plates and spikes with one of his ceratopsians ( Ceratops montanus ) and some nodular osteoderms that he named Nodosaurus (also presumed to pertain to a ceratopsian).

With the passing of Marsh, other workers started trying to resolve the affinities and relationships between these armour-plated dinosaurs. Nopcsa (1902) created a new family Acanthopholidae for Hylaeosaurus , Polacanthus and Acanthopholis , and in doing so removed them from the family Scelidosauridae (the latter having been created by Huxley in 1870). This proved to be an important step in the gradual process of recognition of ankylosaurs and stegosaurs as distinct groups of dinosaurs, and the placing of Scelidosaurus as a little apart from these other two groups. Better quality material began to be recovered from the North American Late Cretaceous after the turn of the century. Brown (1908) described Ankylosaurus and erected the family Ankylosauridae to recognize the unusual and unique characters associated with this familial grouping. However, he still retained his new family within Marsh’s Order Stegosauria . Gilmore (1914), in what became a classic monograph on Stegosaurus , reviewed the classification of plated and armoured dinosaurs and clustered what he referred to as the ‘ridge-scuted’ dinosaurs into a grouping that resembles the ankylosaurs as presently conceived. Hennig (1915) summarized the then current understanding in his Fossilium catalogus. Within the ‘suborder’ Stegosauria Hennig recognized three families of armoured dinosaurs in a format that seems much more familiar to modern eyes: Scelidosauridae for Scelidosaurus and the ‘Purbeckian’ Echinodon [the latter taxon had similar shaped teeth to those seen in Scelidosaurus and was reported to have osteoderms ‘granicones’ associated with its remains, which was later disproved by Norman & Barrett (2002)]: Stegosauridae for animals that are mostly recognized as stegosaurs today; and Nodosauridae for animals that are mostly recognized as ankylosaurs.

The constituents (and names) of the higher-level general groupings of armoured and plated dinosaurs continued to change throughout much of the 20 th century, reflecting in large measure the fragmentary nature of their remains. One voice of clarity across this period of time proved to be that of Alfred Sherwood Romer (1927). He elevated the ankylosaurs to subordinal rank and articulated the case for distinguishing ankylosaurs from stegosaurs (and all other ornithischians) on the basis of their distinctive pelvic structure and the form of their body armour. This suggestion was not followed by other workers, but was reaffirmed by Romer (1956, 1966). Indeed, it was not until the work of Coombs (1971, 1978) that a greater measure of clarity was established for the ankylosaurs, as well as Scelidosaurus .

With the advent of a cladistic approach to systematics, the classification and implied phylogenetic relationships of the armoured and plated dinosaurs has become more stable. Nopcsa’s (1915) suprafamilial category Thyreophora (introduced to replace the older term Orthopoda) was re-instated by Norman (1984) for a clade of armoured quadrupedal ornithischians that included ankylosaurids and nodosaurids, and he positioned Scelidosaurus as the basal sister taxon to Ankylosauridae . This topology was opposed by Sereno (1984), who proposed that the same clade-name Thyreophora should be used to include Ankylosauridae , Nodosauridae , Stegosauridae , Pachycephalosauridae and Ceratopsia – much closer to Nopcsa’s original conception of the Thyreophora. In response to the clear and obvious disagreement between these two analyses, Sereno (1986) tacitly accepted the topology generated by Norman (1984) and formalized a classification and phylogeny of Ornithischia , albeit without a character matrix or numerical phylogenetic analysis. This schema posited Scelidosaurus as a sister taxon to the clade named Eurypoda ( Ankylosauria + Stegosauria ; Sereno, 1986: fig.3). This topology has remained largely unchallenged since the early 1990s.