Fossils provide evidence of among the oldest signs of human collecting behaviour. Amongst the first evidences of an interest in fossils, by pre-sapiens humans, are flakes made of coral-bearing chert found at Swanscombe, in Kent (424–374 ka before present), England, inferred to have been carried approximately 193 km from an outcrop of Jurassic (Portlandian) rock at Tisbury in Wiltshire.Footnote 1 From the standpoint of the establishment of interchange networks in the Upper Palaeolithic, in continental Europe there is clear evidence of widespread collection and the occasional long-distance transport of fossils.Footnote 2 During Classical antiquity, the engagement with fossils became societally deeply embedded and fostered a number of mythological narratives.Footnote 3 Greek and Roman literature contains references to the remains of large skeletons, marine organisms and plants, many of which are fossils. For example, fossil palatal teeth of fishes appear to have been first mentioned by Pliny the Elder in the first century a.d. He gave the name ‘Batrachites’, or frogstone, to these fossils.Footnote 4 This article documents the first ichthyosaur remains discovered in a securely dated Roman deposit, which also represent the earliest known deliberately collected ichthyosaur fossil.
The specimen (Find no. 4022; Fig. 1) is an isolated anterior caudal centrum. It was found in 2024 by Colchester Archaeological Trust at the site of the former Essex County Hospital (ECH), Lexden Road, Colchester (NGR TL98925 24885) during archaeological work preceding the construction of new housing. Ichthyosaurs (Ichthyopterygia) are a well-studied group of secondarily aquatic marine reptiles of the Mesozoic. The earliest scientific account of an ichthyosaur was a paper published in 1814 by Sir Everard Home, a British surgeon, in which he described and illustrated a large ichthyosaur skull and some incomplete postcranial remains unearthed by Joseph and Mary Anning in the Lias Group at Lyme Regis, Dorset, in 1811 and 1812.Footnote 5 In two papers in 1819, Home discussed three additional specimens, including the first almost complete skeleton.Footnote 6 Prior to this first dissemination of scientific findings, ichthyosaurian fossils had been illustrated as early as 1699, and other material was collected in the eighteenth century in southern Germany and England, but its affinities had remained unheeded.Footnote 7 Home’s pioneering work prompted a major phase of the collecting and reporting of many new ichthyosaur finds, initially from the Lower Jurassic of England, and thereafter from most of the Mesozoic from various localities worldwide.Footnote 8
Ichthyosaur centrum obtained from a late second-century Roman pit in Colchester (Find no. 4022): (A) anterior articular, (B) posterior, (C) dorsal, (D) ventral and (E) left lateral views.
The dolphin-like ichthyosaur (‘fish lizard’) was the first large prehistoric reptile identified in the world. It played a prominent role in the contentious scientific and religious debates of the nineteenth century regarding the Earth’s age, the origin of species, the causes of extinction and the understanding of ancient gigantic animals that were not accounted for in the Bible.Footnote 9 In the Roman period, although the true nature of the isolated vertebra was not recognised, it could have been valued due to the prevalent Greco-Roman superstitions and folklore. Alternatively, it could have been kept as a curio.
The fossil from Colchester was recovered from the fill of a sub-rectangular, steep-sided pit, F. 2323, that measured in plan 1.28 m × 1.05 m. Half of the pit was excavated to a depth of 0.6 m, being the formation level of the development site, and the undisturbed base of the pit is now under a new access road. F. 2323 was observed to be cut by two other pits of different morphologies and by a ditch. The size, shape and depth of F. 2323 suggest it may have been a quarry pit.
In addition to the ichthyosaur centrum, pit F. 2323 contained 68 Roman pottery sherds, 18 fragments of Roman ceramic building material (brick, tegulae and imbrex), a moderate-sized archaeological mammal bone assemblage, a fragment of Purbeck Marble veneer, an iron nail and a complete copper-alloy toilet spoon (ligula), having a small, round, flat scoop.Footnote 10 The archaeological sherds suggest the deposition date is in the third quarter of the second century a.d. This age designation is based on the presence of various pottery types, including a sherd of Colchester samian (fabric BACO), a Cologne colour-coated (fabric KOL CC) beaker, a local black-burnished (fabric GB) Cam 37A/38A bowl dating from c. a.d. 120 onwards, and a small quantity of plain central Gaulish Lezoux samian (fabric BACG) dating to the second century a.d., including a Drag. 27 cup and a Drag. 18/31 or 31 dish. Notably, pottery types that emerged during the early to middle third century a.d. at Colchester, such as Nene valley colour-coated wares (fabric EA) and Hadham oxidised wares (fabric CH), were absent.
Pit F. 2323 produced 73 isolated mammal bones, principally postcranial elements, and two isolated teeth, weighing a total of 1.130 kg. The bone was generally in good condition, although it was heavily fragmented from post-depositional damage and butchering. The only complete bone was a bovid phalanx. Cattle (Bos taurus), domestic pig (Sus domesticus) and ovicaprids were the most common taxa present. A horse (Equus caballus), represented by a worn third molar, and a possible cat (Felix catus), represented by an ischium fragment with a complete acetabulum, were also identified. Evidence of butchery, including knife cuts, chop marks and scoop marks, was present on some cattle bones and a sheep/goat dentary and scapula. Some fragments of cattle humerus, radius and metacarpals show deliberate bone breakage. On a few of the bones canid gnawing was noted. In terms of the distribution of archaeological bone and other material, this was not uniform and, based on the ichthyosaur centrum’s relative position in the pit’s fill, there was nothing to distinguish the archaeological bone from it. The pit F. 2323 assemblage appeared typical of much of the animal bone recovered from pits and non-grave features at ECH, which also contained cattle, sheep/goat and pig remains. The fragmentary state of the bone in F. 2323 is suggested to indicate a degree of residuality of the finds from the feature. This is also suggested by scavenging by dogs, indicating that at least some of the bones were accessible prior to being buried in the pit.
The Colchester centrum is one of two finds of the remains of a large fossil vertebrate uncovered to date in Roman Britain. The second is an isolated plesiosaur dorsal vertebra, found in the King’s Hedges area of Cambridge.Footnote 11 This bone was found among a number of Roman items, including pottery and animal bones. It is of interest that both fossils were discovered in the East Anglian region of England in similar archaeological contexts.
Probable provenance location and preservation
In this section, we examine the most plausible location of origin of the fossil centrum. This was possible to assess based on the characteristics of a thin patch of adherent green micaceous very fine sandy siltstone in the central region of its posterior articular surface. A description and taxonomic interpretation of the centrum are provided in Appendix A.
Microscopic analysis revealed the presence of glauconite-quartz grains within the siltstone, evident from their green colour and an association with pinpoints and a peripheral zone of ferric oxide. Limonite, a mixture of hydrous ferric oxide minerals, is a common product of glauconite decay driven by aerobic alteration processes.Footnote 12
In the south-eastern area of England, this lithology is typical in many members of the Aptian Lower Greensand Group, which is composed of four formations: the Atherfield Clay Formation, the Hythe Formation, the Sandgate Formation and the Folkestone Formation. The general lithology of the lowermost Cenomanian Cambridge Greensand Member, which is a remainié deposit comprised of glauconite-rich, fine ‘chalk mud’, is similar.Footnote 13 Both the Lower Greensand Group and Cambridge Greensand Member have yielded ichthyosaur fossils.Footnote 14 Locally to Colchester, Greensand chert from the Medway catchment is an important clast type in the Blackwater terrace gravels, the Mersea Island Member and the High- and Low-level East Essex Gravels south of the Blackwater.Footnote 15 However, these gravels have not yielded Greensand fossils. Elsewhere in East Anglia, the Lowestoft Formation diamicton at Pakefield has produced ichthyosaur material, but not Greensand derived.Footnote 16
The vertebra specimen displays significant wear and erosion, and most of its surface is comprised of trabecular bone. This demonstrates it was excavated by weathering and then underwent attritional wear, possibly by sand. The extent of wear indicates that it was subjected to abrasion processes for an extended period. Such conditions have been demonstrated to degrade isolated bones and teeth progressively.Footnote 17
During the Roman period, large quantities of Kentish ragstone, a grey-green, sandy and glauconitic limestone from the Hythe Beds, were obtained from outcrops primarily in the region of Maidstone, near Sevenoaks, and on the coast, where the Hythe Beds outcrop near Folkestone and Hythe.Footnote 18 In coastal settings, the greensand was likely mainly sourced from readily accessible wave-worn platforms.Footnote 19 This origin of the stone is confirmed by rounded marine-eroded blocks used as a facing stone in some sections of the later third-century walls of Roman Canterbury. Prominent examples of the use of greensand include at the Kentish Saxon Shore forts at Reculver and Richborough,Footnote 20 the early Saxon Shore fort at Bradwell,Footnote 21 the late first-century bath house at Beauport ParkFootnote 22 and the second-century Circus in Colchester.Footnote 23 In the light of the above, given that the fossil centrum had undergone extensive natural abrasion, the foreshore/intertidal zone at a site on the east Kent coast is deemed its most likely place of origin.
It is noteworthy that East Wear Bay, Folkestone, was the location of a significant Greensand stone-quarrying industry that was active until at least the fourth century a.d.Footnote 24 It is thus a potential location from which the centrum originated. East Wear Bay has yielded an important fauna of ichthyosaurs, plesiosaurs, pliosaurs, turtles and pterosaurs, which occur throughout the section at this site.Footnote 25 The ichthyosaur material from this locality is represented by teeth, vertebral and limb elements; it is assignable to the ophthalmosaurid Pervushovisaurus campylodon,Footnote 26 or regarded as Platypterygiinae indet.Footnote 27
Discussion
Nearly all fossils from Roman sites in Britain are marine invertebrates of late Mesozoic age. Most were probably acquired through beachcombing, as in the case of the jet industry in Roman Yorkshire,Footnote 28 with some fossils collected inland. The fossils may have been considered as interesting or exotic objects. However, contextual archaeology indicates that some may have had a ritual or superstitious purpose. For example, the deliberate deposition of fossil echinoids (Echinocorys sp.) was documented at a settlement at Studland, Dorset, where 11 fossils were intentionally buried beneath rectangular huts and a roundhouse.Footnote 29
Elsewhere in England there is an association between find spots of fossils, temple sites and graves.Footnote 30 Three echinoids and a sponge were found at a temple site at Wanborough, three echinoids at a temple site at Titsey and four echinoids at a temple at Farley Heath.Footnote 31 Six fossils are from graves, four from cremation burials, one from an inhumation and one from a probable inhumation.Footnote 32 They are a belemnite, an ammonite, a crinoid stem, two unidentified fossils and an echinoid. The echinoid, from a child cremation burial in Colchester, is a small ‘bead-like’ form (Gauthieria spatulifera), preserved as an internal flint cast. It formed part of a burial group that additionally included an amber bead, a glass cabochon and two dice.Footnote 33
Two sites in continental north-west Europe, the Romano-Celtic temples at Essarts, La Mare-du-Puits and Saint-Aubin-sur-Gaillon,Footnote 34 provide further evidence of the active selection and re-use of fossils. Excavations at both sites uncovered curated hoards of polished stone axe-heads and fossil sea urchins, sometimes deposited with Roman Venus figurines. Fossils found at Roman temple sites could have proffered a specific image or symbolic meaning, as might prehistoric axe-heads at these sites, of which eight examples are known from temple sites in Britain.Footnote 35 In the case of the axe-heads, it is unclear whether they were votive, ritualised, or apotropaic deposits.
The main categories of artifacts and animal bone from F. 2323 were essentially the same as found in most ditches and pits distributed in all parts of the ECH site. However, the presence in F. 2323 of the ichthyosaur centrum, co-occurrence of a ligula, and of a possibly disparate associated animal bone group might feasibly indicate the presence of some form of structured (ritual) deposition of this somewhat atypical grouping. In Britain, probable structured deposition has been documented at numerous sites, but notably in the contents of pits, wells and shafts at the major urban Roman settlements of Silchester (Calleva Atrebatum), Dorchester (Durnovaria) and London (Londinium).Footnote 36 The phenomenon is also observed in ditches and quarries, for example at Reigate Road, Epsom,Footnote 37 and in pits, ditches and ‘working hollows’ in a Romano-British enclosure complex south of Wetherby in West Yorkshire.Footnote 38
Structured deposition provides evidence for persistent, widespread, ritual practices in Roman Britain that originated from Iron Age traditions, and the incorporation of Roman influences in the religious landscape, often to the degree that the native–Roman divide is blurred.Footnote 39 Some of the contents of structured deposits at Reigate Road, Epsom,Footnote 40 occurring at this locality in quarries and large pits, are comparable to the pit F. 2323 finds. At this site, a circular to irregular in plan 3 m long, 2.6 m wide and 1.86 m deep pit, possessing very steep sides, interpreted as a quarry, yielded cattle and ovicaprid bones, a badger bone, a partial cat skeleton and the remains of several canids. It also contained a bone hairpin or needle fragment and a shaped red deer antler tine, presumably a pick, with a transverse mounting hole for attaching it to a haft.Footnote 41 Quarry 2 at Epsom contained the remains of at least five partial canids, horse, deer and corvid remains. In Quarry 2, a copper-alloy dupondius of Vespasian (a.d. 77–78), a copper-alloy as of Antoninus Pius (a.d. 154–155), a contemporary copy of a copper-alloy nummus of Constantius (a.d. 353–360), a first-century iron stylus of Manning Type A N9,Footnote 42 the spring of a mid-first- to early second-century Colchester derivative broochFootnote 43 and a copper-alloy ligula were recovered in addition. At Epsom evidence of butchery was rare, limited to occasional chop and knife marks, thus contrasting to the treatment of many bones from pit F. 2323, with the carcasses subjected to little dismemberment.Footnote 44 However, at Wattle Skye many deposited cobbles were found in conjunction with fragmented pottery, animal bone, quern stones and Roman coins, that were considered to be potential feasting remnants.Footnote 45 It was likely, however, that much of this activity represented dumps of refuse and midden material.
The Colchester ichthyosaur and the Cambridge plesiosaur vertebrae are the first fossil bones procured from a Roman deposit in the north-western Provinces of the Roman Empire. Given that the manner of deposition of the Colchester find in pit F. 2323 is not fully established, the reason why it was valued is difficult to ascertain. One possibility could be linked to the manner in which understandings of larger fossil vertebrates were amalgamated with mythical ideologies.Footnote 46 During Classical antiquity, within the area of tradition and popular culture, myth-making about heroes and monsters offered a strong interpretive framework for, in particular, the understanding of large teeth and bones.Footnote 47 For example, Pausanias notesFootnote 48 that a large scapula, said to be of Pelops, was displayed at Olympia in the Pelopion shrine. This scapula may have originated from nearby fossil-bearing deposits.Footnote 49 Pausanias also refers to an enormous sea-monster’s skull that was kept at the sanctuary of Asklepios at Sikyon.Footnote 50
Classical mythology includes numerous tales of transformation, several of which are centred upon the transmutation of objects, including bones, into stone. In consequence, the physical characteristics of fossil bones as a field of interest received the attention of scholars. For example, in his treatise On Stones, according to Pliny the Elder, ‘Theophrastus states that fossil ivory mottled black and white is found, that bones are produced from the earth, and that stones resembling bones come to light.’Footnote 51 Hence, it is not improbable that, alternatively, the Colchester fossil centrum was a curio, being a lithified bone. If so, the find, as mundane as this interpretation seems, evinces a good background knowledge of a palaeontological site, and it strongly attests to the timeless appeal of collecting fossils.
Finally, mention should be made regarding the possibility that the Colchester ichthyosaur centrum was kept as connected with Greco-Roman mythology in relation to nineteenth-century responses to what appeared an almost mythical creature that required explanation. Fossilised ichthyosaur remains were at first viewed as almost inexplicable, at odds with familiar categories of thought.Footnote 52 This, perhaps, is understandable. The unfamiliar and idiosyncratic appearance of Mary Anning’s initial ichthyosaur find had suggested a creature of myth and legend. However, over time the ichthyosaur became a catalyst for a number of significant scientific debates in fields including comparative anatomy, palaeontology, and other branches of natural history, particularly as connected with the study of geology.Footnote 53 Yet, in the nineteenth century, ichthyosaurs, along with other prehistoric reptiles, were often referred to as ‘monsters’ in both scientific and non-scientific writings. They were frequently depicted as a miscellany of various animal parts, comparable to the Greek mythological chimera. For example, Figuier, citing Georges Cuvier, assigned to the ichthyosaur ‘the snout of a dolphin, the head of a lizard, the jaws and teeth of a crocodile, the vertebrae of a fish, the head and sternum of a lizard, the paddles like those of a whale, and the trunk and tail of a quadruped’. Footnote 54
Henry De la Beche initiated a tradition of visually depicting the imagined ferocity of the ichthyosaur. Thus, for example, the first English edition of Jules Verne’s Voyage au centre de la Terre (1864), published in 1871 as A Journey to the Centre of the Earth,Footnote 55 portrays an epic battle between two immense marine reptiles; one is a plesiosaur, while the other is described as ‘the world-renowned ichthyosaurus’. It is unknowable whether the Colchester centrum was curated because of its perceived connection to classical folklore or mythology, or if it was merely considered a curiosity. However, its appeal likely stemmed from its unique and puzzling features, the appreciation of which people in the nineteenth century were still cognisant.
Conclusion
It is suggested that the east Kent coastal outcrop of Lower Greensand, i.e. between Folkestone and Hythe, is the probable provenance location of the centrum. Given the centrum’s abraded state, it was probably collected from the foreshore or a beach. The fossil is important in debates about where Greensand building material (Kentish ragstone) in Colchester, commonly found in Roman contexts, was obtained. It helps to verify a likely east Kent coastal source, in addition to Greensand originating from the upper Medway valley. The individual who collected the centrum may have been involved in the quarrying industry on the east Kent coast that utilised Folkestone Greensand.Footnote 56 At Folkestone, Roman greensand quarrying activity spanned from the second to the fourth century a.d.
The ichthyosaur centrum from Colchester, together with the King’s Hedges plesiosaur dorsal vertebra, are significant in regard to understanding how large vertebrate fossils were interpreted during the Roman period in north-western Europe. During the Classical, Hellenistic and Roman periods, large vertebrate fossils were widely perceived as the remains of mythological creatures, giants or heroes.Footnote 57 Fossil finds were often curated and displayed as curiosities or sacred relics, serving as physical evidence that validated existing myths. The ichthyosaur centrum may have been collected for this reason or for its interesting and curious form.
Acknowledgements
We express our special gratitude to Alec Wade for excavating pit F. 2323 and recovering the centrum. We also thank Chelsea Hodges for identifying it during finds processing. Dr Matthew Loughton conducted the analysis of the Roman pottery assemblage in the F. 2323 pit, and Alec Wade summarised the animal bone assemblage. Emma Holloway prepared the figure. Many thanks are due to Professor Nelson Fernández, who carefully read the text and critically and constructively improved upon it. We would like to thank two anonymous reviewers for helpful comments on the submitted article.
Appendix A: Description and taxonomic interpretation of the ichthyosaur centrum
The centrum (Fig. 1) is strongly amphicoelous, substantially shorter (anteroposteriorly) relative to its height, and the articular surfaces are distinctly subtriangular in outline. It has a single apophysis in its lower half and is thus interpreted as an anterior caudal. In ichthyosaurs, this position is in the tail stock close to the pelvic girdle elements. Most of the bone surface has been eroded off, apparently due to the effect of natural weathering. Remaining areas of cortical bone occur inside the anterior and posterior articular surfaces, in the floor of the neural canal, and in the right dorsolateral side of the element. The essentially flat ventral surface of the centrum is broad and smooth with no keel. Viewed ventrally, the central portion of the posteroventral edge of the centrum is gently concave, whereas the anteroventral edge is slightly convex across most of its width. The dorsal and ventral edges of the anterior articular surface are positioned at about the same level as their counterparts on the posterior articular surface. The apophysis is set upon the upper half of an extensive gently convex posteroventral surface.
The left apophysis is barely laterally expanded to form a broad vertically aligned oval in ventrolateral view and lies close to the posterior edge of the centrum. Due to a broken-off portion, the right apophysis is largely missing.
On the left and right dorsolateral surfaces of the centrum, the neural arch facets, that border the floor of the neural canal, are raised above the dorsal surface of its main body on low sub-vertical ridges that extend along its entire length. The right ridge is much lower than the left ridge, due to differential growth or possibly preservational distortion. Since in ichthyosaurs the neurocentral suture did not close, the neural arch had become detached and lost.
The isolated centrum from Colchester is generally similar to those of a wide range of ichthyosaurs, and it exhibits two character states known in all ichthyosaurs, namely it is notochordal and amphicoelous.Footnote 58 It is not possible to identify the subclade to which the centrum can be assigned with confidence, as there are no ichthyosaurian subclade-defining caudal synapomorphies thus far proposed. However, although weakly developed, the Colchester centrum displays the distinct trait of a concave posteroventral edge of the centrum. This could be construed as an ophthalmosaurid apomorphic condition, and therefore we provisionally refer the centrum to Ophthalmosauridae incertae sedis. The Ophthalmosauridae is a family of thunnosaurian ichthyosaurs that range from the Middle Jurassic to the Late Cretaceous (Bajocian-Cenomanian).
