Gallery and print store

Friday, 17 May 2019

The science of the Crystal Palace Dinosaurs, part 2: Teleosaurus, pterosaurs and Mosasaurus

It's time to continue our palaeoartistic discussion of the famous 1850s Crystal Palace prehistoric animal sculptures. As you'll know if you've read part 1 of this series, I've been supplying artwork and palaeoart notes to the Friends of Crystal Palace Dinosaurs charity for their new 'about the statues' web pages. What started as a small project has ballooned into several weeks of illustration, research and writing based around these charming and fascinating early palaeoartworks, culminating in this series of blog posts. As before, we'll be reviewing the models in roughly sequential order. Previously, we looked at the Dicynodon, "Labyrinthodon" and marine reptiles, and today we'll be covering the Jurassic Teleosaurus, the Jurassic and Cretaceous pterosaurs, and the Late Cretaceous Mosasaurus. Those familiar with the Crystal Palace prehistoric menagerie will recognise a dinosaur-shaped hole in that line-up but, have no fear: we'll be covering them next time, and then the mammals after that.

As before, I want to point out that the following notes are expanded and referenced versions of content I've provided for the Friends of Crystal Palace website, and readers are encouraged to check out those pages to supplement the dedicated palaeoartistic assessment provided here. Also, while I don't want to labour the need for increased interest and investment into the Crystal Palace sculptures - there's lots of that in the introduction to part 1 - I want to remind readers that these models, now approaching 170 years old, need a lot of care and maintenance. Efforts to restore the models are underway, and you can help by donating money or volunteering your time to keep the site maintained.


The Crystal Palace Teleosaurus sculptures as seen from the Secondary Island. Both Teleosaurus have been restored recently and look very handsome, the aim is to get all the models looking this good in the next few years. Note that these models are meant to be surrounded by water, but last year the entire site was covered in lush vegetation. Photo from 2018.
Two Teleosaurus bask on the banks of the Secondary Island as part of the Jurassic Oolite scene that also includes Megalosaurus and (before they went missing) small pterosaurs. As superficially crocodylian-like animals, Teleosaurus was perhaps the most straightforward fossil reptile for Hawkins to restore. The two Teleosaurus models are, in terms of detailing, some of the finest reptile restorations in the Crystal Palace arrangement. The general proportions of Teleosaurus are captured correctly with its short limbs, long body and narrow, gracile jaws, and their depiction as basking on a shoreline is consistent with predictions of teleosaurid habits. Teleosaurids were part of a marine radiation of crocodylian-line reptiles but they still bore limbs instead of flippers, and they lacked well-developed tail fins. In not being as specialised for an open water existence as subsequent marine crocodylomorphs, it’s entirely possible, maybe even likely, that teleosaurids returned to shore for rest and procreation.

There are several anatomical peculiarities about Hawkins’ teleosaurids that not only conflict with modern understandings of these animals, but also those of Victorian Era palaeontologists. Owen (1854) stressed the great similarity between teleosaurs and modern crocodylians - specifically the gharial - and it appears that modern crocodylians ultimately informed Hawkins’ take on these marine crocodylomorphs more than their actual fossils. Teleosaurid scalation was quite different from all modern crocodylians in that their dorsa bore two medial rows of large scutes from their necks to their tails, and their bellies were covered with a shield of interlocking scutes. This was well known to Victorian anatomists (e.g. Owen 1842), but Hawkins seems to have used Crocodylus scute arrangements instead, depicting several rows of smaller dorsal scutes and omitting any trace of belly armour. Over at Tetrapod Zoology, Darren Naish has suggested that saltwater crocodiles may have been specifically referenced for this arrangement. Gharial cranial proportions seem to have been used to model the posterior head, this region being short and narrow (like a gharial) instead of long and expanding outwards towards the back of the skull (as in teleosaurs).

Lemmysuchus obtusidens, a teleosaurid of similar size as the Crystal Palace Teleosaurus, but the skull and dentition are more robust. Note the osteoderm arrangment along the back and proportions of the skull - teleosaurids had a much longer and broader temporal region than Hawkins captured, perhaps because of his referencing living crocodylians. As an aside, note that this artwork has a pretty obvious homage to another classic piece of Victorian palaeoart, John Martin's 1840 The sea dragons as they lived.
The result is a blend of crocodylian and teleosaurid anatomy, and it is not obvious why Hawkins did not reflect the anatomy of teleosaurids more faithfully when this effort was made for other models. Subsequent artworks by Hawkins (e.g. his 1862 posters produced for the Department of Science and Art - see Rudwick 1992) show the same erroneous interpretations, so he may have simply been misinformed or unaware of what teleosaurids really looked like. To be fair to Hawkins, he is far from the only palaeoartist to apply scalation patterns from modern crocodylians to their ancient, distant relatives. Perhaps he just assumed - as many still do today - that all crocodylian-like animals are and were pretty much alike. It’s more interesting that Owen knew the precise nature of teleosaurid scale arrangement - even down to the differences between species - and yet Hawkins’ models turned out as they did. Did Owen not tell him? Did Hawkins ignore him? Was there a practical reason for why the design couldn't be altered? Even more intriguingly, despite Owen being unafraid of pointing out aspects of the models that he disagrees in his 1854 sculpture guide, he made no mention of this error. Is this evidence of Owen not having much regard for these models, or maybe not wanting to draw attention to a mistake he could have fixed? I'm not sure we know what happened there, but this obvious difference between Hawkins' models and Owen's knowledge certainly fits wider evidence that Owen gave very little input to the Crystal Palace project (Secord 2004; Dawson 2016).


Two sets of pterosaurs were created for the Crystal Palace display: two larger individuals representing pterodactyloids from British Cretaceous Chalk deposits, and two smaller animals from the Jurassic Great Oolite Group. As originally positioned, these models flanked the dinosaur sculptures in the centre of the Geological Court. Alas, the original models of the smaller pterosaurs went missing in the 1930s (McCarthy and Gilbert 1994) and fibreglass replacements installed in 2002 were stolen and destroyed by vandals in 2005. The larger models are still standing but are in disrepair, owing to both their relative delicacy and continued vandalism.

The large Chalk Pterodactylus models of Crystal Palace, as photographed in 2018. The models are replete with fine anatomy that cannot be appreciated from afar, such as variable scale sizes and individually crafted digits. Close inspection reveals many obviously avian-inspired features, and I suspect they were not closely based on the excellent pterosaur fossils known to early 19th century palaeontologists.
Images of the pterosaur models in their prime show that all four shared a similar bauplan: long necks terminating in smallish, toothed heads; scaly skin; large torsos; crouching hindlimbs and - of course - enormous, membranous wings supported by long forelimb bones and hypertrophied wing fingers. In this respect Hawkins’ pterosaurs were fairly typical of pterosaur restorations from the 1800s, and they are somewhat different to how we imagine them today. Contrary to Owen’s confident and entirely unfounded guidebook assertion that pterosaurs were scaly (Owen 1854), the German zoologist Georg August Goldfuss had recorded solid evidence of pterosaurs having a fuzzy, hair-like covering as early as 1831. The absence of fibrous integuments on Hawkins’ pterosaurs is not an example of Owenian arrogance, however. Just about everyone ignored Goldfuss’s claims about pterosaur skin and it was not until the 1970s that pterosaur fuzz would be an accepted fact of their palaeobiology. Goldfuss has recently been vindicated by modern fossil imaging techniques proving beyond all doubt that his soft-tissue interpretations were accurate (Jäger et al. 2018). This is one of two places were Goldfuss’ research could have enhanced the portrayal of prehistory at Crystal Palace had some of his papers not been so widely ignored by other researchers (see discussion of Mosasaurus, below).

The rarely seen posterior end of the remaining complete Chalk Pterodactylus. Note the deep chest and pelvic region, bird-like limb posture and blunt, deep tail. This is the body of a bird, not a pterosaur. The triangular structure on the left of the image is the wing of the neighbouring pterosaur. Photo from 2018.
Less explicable mistakes in the pterosaurs are their overall proportions. The very first pterosaur fossils known to scientists were complete skeletons showing that their heads were longer than their bodies, and yet Hawkins - like virtually all artists of the 1800s - restored them with small heads and massive torsos. My assumption is that this reflects Hawkins modelling his pterosaurs with an avian physique rather than with exacting attention to fossil material. Close inspection of his models reveals many bird-like details that are inconsistent with pterosaur remains, such as an emphasis on ventral flight muscles, deep bellies, horizontally-held thighs, an extensive pelvic region with a short, poorly defined tail, and a proportionally small head. It is also noteworthy that Hawkins anchored the wing membranes to the body alone, leaving the hindlimbs free. This creates the impression of a bird-like wing arrangement, even though most of his peers were illustrating more accurate bat-like membranes that stretched all the way to the hindlimb (e.g. see illustrations by Soemmerring (1817); Newman (1946) and even Owen (1859)). Though Hawkins observed some aspects of pterosaur form correctly - such as posing one animal quadrupedally, and placing the full length of the foot on the ground - the overall impression is of a goose or swan-like creature with pterosaur features, not a precise reflection of the fossils he had available to him. Hawkins was not alone in making such mistakes, and very few pterosaur restorations from the early decades of palaeontology have accurate body proportions. Indeed, it took a full century before pterosaurs were restored in ways that reflected their anatomy more precisely (Seeley 1901).

A modern take on Cimoliopterus cuvieri, the large toothed Chalk pterosaurs imagined by Hawkins for Crystal Palace. It took us years to get there (and some of us are still in transit), but the message that pterosaurs have enormous skulls and tiny bodies is finally being reflected in art.
The exact identification of the Chalk pterosaurs is left somewhat vague by Owen (1854), but he indicates that large bones (indicating 5-6 m wingspan individuals) of animals such as “Pterodactyluscuvieri were influential on the models. The Oolitic species, in contrast, are precisely identified as “Pterodactylus bucklandi”. None of these taxa were well known in the 1800s (and, indeed, they remain poorly known today) so it’s likely that Hawkins based all four models on complete Pterodactylus antiquus skeletons known from Germany. Aspects such as the long necks and low, homodont teeth are consistent with this hypothesis. Today, we recognise that the species informing the Crystal Palace models are distantly related to Pterodactylus and had very different proportions. The large Chalk species were long-winged soaring specialists with sophisticated dentition adapted for snagging fish, as in Cimoliopterus cuvieri, above. “P. bucklandi” was probably a long-tailed, short-necked animal with procumbent teeth, although the fossils bearing this name are too poor to consider it a valid species (O'Sullivan and Martill 2018). Whatever “P. bucklandi” was, it may have looked somewhat like the contemporary Oolite species Klobiodon rochei, below.

Hawkins' Oolite pterosaurs were based on scrappy material once referred to "Pterodactylus bucklandi", which later became subsumed into the genus "Rhamphocephalus". We now know that "Rhamphocephalus" has little taxonomic utility, being based on teleosaurid fossils and containing no valid pterosaur species. The Oolite species Klobiodon rochei is a rhamphorhynchid, the sort of animal that the "Pterodactylus bucklandi" material likely represents.

Mosasaurus hoffmanni

Lurking behind the dinosaurs and pterosaurs, sometimes almost out of sight, is a partial restoration of the giant mosasaur Mosasaurus hoffmanni. The choice to depict a mosasaur at Crystal Palace was unusual and very progressive. Although their skulls have been known since the 1700s mosasaurids were rarely depicted in palaeoart in the early 19th century, probably owing to being overshadowed by the better-known ichthyosaurs and plesiosaurs. Mosasaurs would become more routinely featured in artwork after palaeontologists excavated their skeletons from the American midwest in the 1860-70s, but it’s rare to see a restoration of a mosasaur of 1850s vintage. Indeed, I’m fairly certain that the Crystal Palace Mosasaurus is among the first attempts at restoring the life appearance of one of these animals. Hawkins’ take was influential enough that other artists heavily referenced his sculpture (e.g. see 1860s work by Louis Figuier), but the discovery of superior mosasaur fossils just two decades after Crystal Palace Park opened dated the model relatively quickly.

A view of the Crystal Palace Mosasaurus that Luis Rey would approve of. Note the pterygoidal teeth, left forelimb and extensive body - there is a lot more to this model than just a head, as is sometimes reported. The low water level exposes the uniquely incomplete nature of the restoration. This could reflect a lack of complete mosasaur skeletons in the 1850s, but Hawkins also made complete Dicynodon despite just having skulls, so a lack of material may not have phased him. Photo from 2018.
The holotype of Mosasuarus hoffmanni is an enormous, disarticulated skull, and it was this material that largely informed the Crystal Palace reconstruction. The Mosasaurus is the only deliberately incomplete restoration in the entire Geological Court, comprising just a head and neck, a nondescript back, and a single flipper. This is sometimes attributed to Hawkins not having sufficient material to complete his restoration, and yet his Dicynodon, which were also only known from skulls at the time, were given entirely speculative limbs, bodies and tails. An absence of fossil material was no deterrent to Hawkins' vision, and the fact that he made one forelimb suggests he had ideas about what the front end of the animal should have looked like at least. The idea that Hawkins was attempting to hide unknown anatomy underwater is inconsistent with his otherwise bold reconstruction approach, so might other factors - perhaps time and money - have influenced the execution of this model? We know that the models were not ready for the opening of Crystal Palace Park in 1854 (McCarthy and Gilbert 1994), so it’s not inconceivable that Hawkins might have been looking to cut corners somewhere.

The head of Hawkins' Mosasaurus is wide and boxy, consistent with Owen’s (1854) interpretation of the holotype skull dimensions being 2.5 x 5 feet. As with “Labyrinthodon”, Hawkins accurately captured the palatal teeth of Mosasaurus. He seems to have modelled the soft-tissues on monitor lizards, these being realised as close relatives of mosasaurs as early as 1800. Extensive lips, voluminous tissues around the eye socket, and laterally-facing, posteriorly-positioned nostrils are obvious monitor features, and the skin texture of the body shows large ossicles surrounded by smaller basement scales. This recalls the osteoderm-studded skin of certain large and robust monitor species, including komodo dragons and white throats.

Mosasaurus hoffmanni, as restored in 2019. Our understanding of mosasaur life appearance has changed in recent years: rather than crocodile-like swimming lizards, they were highly convergent with whales, sharks and other strongly pelagic fish. The dorsal fins in this restoration are conjectural, but based on their presence in numerous large marine vertebrates.
Hawkins’ take on M.hoffmanni is pretty insightful given the material he and Owen had to work with. Though wide of the mark with many anatomical details, he was on the right track by essentially interpreting them as giant, waterborne monitor lizards. Today we understand mosasaurs as being fully committed to a marine existence - the lizard equivalent of toothed whales. Among their many marine adaptations were shark-like caudal fins, streamlined paddles and smooth skin formed of tiny scales (e.g. Lindgren 2013). But hindsight shows there were some aspects of Hawkins’ sculpture that could have been more precise. In 1845 Georg August Goldfuss described a near complete and undistorted skull of “Mosasaurus maximiliani” (now classified as M. missouriensis) from the midwestern United States. This skull demonstrated a narrower cranial profile than assumed by Hawkins and Owen, and also that the nasal openings were on the dorsal surface of the snout. Williston (1914) reports that scholars generally ignored Goldfus’s Mosasaurus work for decades, with his observations effectively being repeated by Leidy, Cope and Marsh as they uncovered American fossil vertebrates in the 1860s and 1870s. This might absolve Owen and Hawkins for overlooking this important find, though it was nevertheless a second occasion where Goldfuss' work could have futureproofed some of the Crystal Palace models.

That rounds up this entry in the series, but come back soon for part 3: a post dedicated to the least famous and popular Crystal Palace models, the dinosaurs. 

Enjoy monthly insights into palaeoart, fossil animal biology and occasional reviews of palaeo media? Support this blog for $1 a month and get free stuff!

This blog is sponsored through Patreon, the site where you can help online content creators make a living. If you enjoy my content, please consider donating $1 a month to help fund my work. $1 might seem a meaningless amount, but if every reader pitched that amount I could work on these articles and their artwork full time. In return, you'll get access to my exclusive Patreon content: regular updates on upcoming books, papers, painting and exhibitions. Plus, you get free stuff - prints, high-quality images for printing, books, competitions - as my way of thanking you for your support. As always, huge thanks to everyone who already sponsors my work!


  • Dawson, G. (2016). Show me the bone: Reconstructing prehistoric monsters in nineteenth-century Britain and America. University of Chicago Press.
  • Goldfuss, A. (1831). Beiträge zur Kenntnis verschiedener Reptilien der Vorwelt. Nova Acta Physico-Medica Academiae Caesareae Leopoldino-Carolinae Naturae Curiosorum, 15:61-128.
  • Goldfuss, A. (1845). Der Schädelbau des Mosasaurus, durch Beschreibung einer neuen Art dieser Gattung erläutert. Nova Acta Academa Ceasar Leopoldino-Carolinae Germanicae Natura Curiosorum 21:1-28, pl. VI-IX.
  • Jäger, K. R., Tischlinger, H., Oleschinski, G., & Sander, P. M. (2018). Goldfuß was right: Soft part preservation in the Late Jurassic pterosaur Scaphognathus crassirostris revealed by reflectance transformation imaging (RTI) and UV light and the auspicious beginnings of paleo-art. Palaeontologia Electronica, 21(3), 1-20.
  • Lindgren, J., Kaddumi, H. F., & Polcyn, M. J. (2013). Soft tissue preservation in a fossil marine lizard with a bilobed tail fin. Nature Communications, 4, 2423.
  • McCarthy, S., & Gilbert, M. (1994). The Crystal Palace dinosaurs: The story of the world's first prehistoric sculptures. Crystal Palace Foundation.
  • O'Sullivan, M., & Martill, D. (2018). Pterosauria of the Great Oolite Group (Middle Jurassic, Bathonian) of Oxfordshire and Gloucestershire, England. Acta Palaeontologica Polonica, 63(4), 617-644.
  • Owen, R. (1842). Report on British fossil reptiles, part II. Report for the British Association for the Advancement of Science, Plymouth, 1841, 60-204.
  • Owen, R. (1854). Geology and inhabitants of the ancient world (Vol. 8). Crystal Palace library.
  • Owen, R. (1859). On a new genus (Dimorphodon) of pterodactyle, with remarks on the geological distribution of flying reptiles. Report for the British Association for the Advancement of Science, 28, 97-103.
  • Rudwick, M. J. (1992). Scenes from deep time: early pictorial representations of the prehistoric world. University of Chicago Press.
  • Secord, J. A. (2004). Monsters at the crystal palace. In: de Chadarevian, S, & Hopwood, N. (eds). Models: the third dimension of science, Stanford University Press. 138-69.
  • Seeley, H. G. (1901). Dragons of the air: an account of extinct flying reptiles. Methuen & Company.
  • Williston, S. W. (1914). Water reptiles of the past and present. University of Chicago Press.

1 comment:

  1. Really interesting article. I was wondering, though; why were Goldfuss's discoveries seemingly so disregarded by his contemporaries?