Sunday 29 November 2020

Megaloceros giganteus: behind the antlers

Megaloceros giganteus galloping about: a rare scene in palaeoart, but probably a common sight in life. Famed for its size and headgear, we mostly ignore the fact that Megaloceros was among the most cursorial of all deer. What else about this amazing animal is hidden behind those antlers?
Megaloceros giganteus, otherwise known as the Irish elk or giant deer, can genuinely be regarded as a game-changing species for the study of extinct life. First discovered in 1695, it was a founding taxon of palaeontological science - a species that sent early scientists a clear message about the existence of life before the age of humanity and the reality of extinction. The significance of the Irish Elk - which, of course, is not strictly Irish nor anything to do with any living 'elk' species - was perhaps first fully grasped by Baron Georges Cuvier, who published a series of studies on giant deer in the late 1700s and early 1800s. Since then, Megaloceros has been an unwavering presence in paleontological media of all kinds: the focus of numerous scientific studies, a case study for animal evolution, and a frequent star of palaeoartworks and popular texts. From a historic perspective, Megaloceros has demonstrated strong cultural staying power.

Among the earliest life reconstructions of Megaloceros is this small illustration from an 1854 issue of the periodical Die Gartenlaube. Its presence may have been influenced by the Crystal Palace Megaloceros covered in the same magazine earlier that year. Although a relatively unremarkable restoration in most respects, it shows the same regal pose that has stereotyped portrayals of Megaloceros for nearly two centuries. If you've drawn or sculpted Megaloceros at some point, chances are it was standing somewhat like this.

An upshot of Megaloceros always being 'around' is that it's easy to overlook some of the cooler aspects of its palaeobiology. Its size and impressive antlers have become the principle takehome about its existence, such that other facts about its anatomy and ecology are ignored or forgotten. Recently, I've had cause to dig into the palaeobiology and research history of Megaloceros and, wow, what an animal. Cervids are genuinely fascinating anyway: charismatic, complex and awesome animals which, behind their soft, doleful eyes, have a physiology and life history evolved in partnership with Michael Bay. And it's within this context of deer being awesome, unusual animals that we can view Megaloceros as a kind of "superdeer": a species that represents the known extreme of many trends in cervid evolution, and that should be known for much more than its size and antler spread.

Giant deer 101

M. giganteus is a well-known taxon represented by thousands of fossils spread across a wide but latitudinally-restricted distribution through Pleistocene and Holocene deposits of Europe and Asia (Lister 1994; Geist 1999; Vislobokova 2012; Lister et al. 2019). Like many Ice Age species, its distribution was continually shrinking and expanding in response to glacial growth and it never occupied the full extent of its known range at any one time. As with the mastodons discussed here recently, M. giganteus was an Ice Age species that avoided the cold. Although an adaptable animal that seems to have lived in a range of environments, Megaloceros appears to have thrived mostly during interglacials when the well-vegetated, herbaceous habitats it preferred were at their broadest (Vislobokova 2012). It lived in largely open, flat and well-watered settings termed ‘boreal parkland’ or ‘boreal steppe-woodland’, which were situated between cold steppe-tundra in the north and warmer, arid regions to the south. This parkland habitat provided Megaloceros with a diversity of grazing and browsing opportunities thanks to the presence of pine and spruce trees, as well as dense shrubs and herbs. The precise habitat preferences of Megaloceros changed over time and is reflected in antler shape: early subspecies had smaller, more erect antlers and seems to have preferred slightly more forested settings, for example (Lister 1994). Access to good quality food and water were likely critical to Megaloceros life histories due to the nutritional demands of growing those huge antlers each year (Geist 1999), and their need for highly productive plant communities likely prevented them from attaining the wider geographic ranges of other famous Pleistocene megafauna (Vislobokova 2012).

Distribution of M. giganteus sites recently compiled by Lister and Stuart (2019). Note the relatively limited latitudinal range.
Staying put in food-rich environments is probably one reason Megaloceros became so large and well-ornamented. Although not actually the largest extinct deer, the biggest stags are estimated to have massed an impressive 550-600 kg (or more) and stood about 2 m at the shoulder. Their antlers were the biggest of any cervid, weighing 35-45 kg and spanning up to 3.5 m. Reinforced skull bones and mandibular mineral storage were just some of their cranial adaptations to housing these immense organs (Lister 1994). Our knowledge of Megalceros does is comparably poor compared to that of stags, as our giant deer record strongly skews towards male specimens. So far as can be ascertained from a handful of skulls - the only elements we can routinely refer to either sex - females were of generally similar size to males, albeit about 80% of their mass and lacking antlers (Geist 1999). It's thought that our male-dominated fossil record reflects stags foraging separately to females and seeking aquatic vegetation rich in nutrients for antler growth. This, in turn, brought them more routinely into settings conducive to fossilisation. Although some fossil localities contain dozens or hundreds of Megalceros fossils, they are generally rare outside of these exceptional sites. This may reflect genuinely low populations of Megaloceros in most ancient environments (Lister 1994; Geist 1999), perhaps as a consequence of their resource needs.

Discussions and depictions of Megaloceros are heavily biased towards male specimens but - guess what? There were females too. Their fossils are significantly rarer than those of males but seem to have been of similar linear dimensions, though they were perhaps more slender overall. This skeleton is on display in the Natural History Museum, London.

Since the mid-1800s scholars have discussed where Megaloceros fits within deer phylogeny. Although firmly placed within Cervinae - the Old World deer clade - opinions differ over which extant deer are its closest relatives: Cervus (e.g. red deer, wapiti, sika deer - Kuehn et al. 2005) or Dama (fallow deer - Lister 1994; Geist 1999; Lister et al. 2005; Hughes 2005). The weight of morphological and DNA evidence for the Dama link is probably heavier at present, although this only resolves part of our struggles: how Megaloceros is related to other extinct deer is also uncertain. The composition, and even existence of a Megaloceros-line clade, Megacerini, is currently debated, as is whether Dama should be regarded as the last surviving megacerine (see Vislobokova 2012 for a review).

Extreme deer, extreme speed

Prehistoric animals tend to become popular because they represent some kind of biological extreme - the biggest, the heaviest and so on. It's curious, therefore, that the running capacity of Megaloceros is neither widely remarked upon nor commonly depicted in palaeoartworks. A wealth of anatomical data shows that Megaloceros was among the most cursorial of all deer - a species adapted for running far and long over wide, open ground (Geist 1999; Janis et al. 2012). Deer exhibit a range of running styles but Megaloceros seems to have been most suited to long-distance galloping with low limb ground clearance, in contrast to species which bound at height over rough terrain or high-step at speed through dense vegetation. Reindeer and wapiti are similarly adapted to long-distance running, though neither is comparable in cursorial adaptation to Megaloceros. Along with cursorial limb proportions, Megaloceros also has an evident muscle bulk consistent with running habits, highly 'hinged' shoulders permitting an enhanced forelimb stride length, and an expanded trunk volume for an enlarged heart and lungs (Geist 1999). Relatively small hooves indicate a preference running over firm ground. I'm not aware of any specific estimates of Megaloceros running speed, but Geist (1999) considered Megaloceros comparable to the fastest living deer - perhaps 80 kph (50 mph) - and assumed similarly commendable levels of stamina. The increased bulk of Megaloceros would have made for slower acceleration but, once at cruising speed, it might have been capable of running for hours and hours without stopping.

Many aspects of Megaloceros osteology - from limb proportions to chest volume, antler configuration and knee morphology - are consistent with a strongly cursorial lifestyle. It's ironic that this energetic, fast-running deer is often depicted in static poses in palaeoart when dynamic and active compositions typify the genre as a whole. Image by James St. John, from Wikimedia Commons, CC-BY 2.0.

For Geist (1999), a running lifestyle is integral to explaining the development of Megaloceros size and antler spread. It stands to reason that the largest deer will also be the fastest and fittest, and antler size is a direct, honest reflection of stag health. These factors alone could have promoted the evolution of enormous antlers in a large cervid living in uncluttered habitats, but cursoriality encourages antler size in other ways too. For mammals, a consequence of evolving fast running is that juveniles have to be strong and coordinated enough to keep up with their parents, especially if running is critical to escaping danger. Megaloceros fawns must thus have also been capable runners from a very early age, requiring their parents to not only give birth to highly developed, precocial offspring but also supply them with milk rich enough to sustain long-distance, high-velocity running. Lo and behold, studies show that antler size correlates not only with offspring health and size, but also the amount of fat and nutrients in milk (Geist 1999). Put together, these factors mean Megaloceros wasn't a fast-running deer that happened to have huge antlers: its antlers were a direct consequence of its hyper-cursorial lifestyle.

But seriously though, those antlers

As is already evident, it's difficult to write about Megaloceros without frequent mention of its headgear, so let's tackle that topic head-on. There's a lot more to talk about here than just size. Antler structure, function and history of interpretation are also fascinating. Megaloceros antlers are extremely long - up to 1.7 m each - and broaden into great palms in their distal regions, with huge tines erupting from the burr, beam and palm margins. The first tine, which emerges just above the antler base, is broadened to differing extents in different subspecies, and is especially large in early representatives of the lineage (Lister 1994). This likely served as some sort of eye protector during combat (see below).

The size and elaboration of M. giganteus antlers placed it at the heart of a historic discussion about the fitness, evolution and extinction of fossil animals. M. giganteus stags were considered examples par excellence of orthogenic evolution in the late 19th and early 20th century when, as reviewed by Gould (1974), it was viewed as a species that had evolved itself into a corner: the result of a runaway, one-way evolutionary process that encouraged the creation of vastly oversized and biologically untenable antlers. Eventually, it was suggested, the antlers became so large and heavy that their owners were forever being caught in vegetation, mired in mud and bogs, or even suffered catastrophic brain haemorrhages caused by redirected blood flow from antler velvet. Extinction was inevitable for such sorry creatures.

Of course, even an elementary grasp of modern principles of natural selection shows these ideas as naive, quaint and totally wrong - there is no conceivable mechanism through which species can evolve structures that are so hazardous to their health. But it wasn't until the 1930s that a sensible alternative explanation for Megaloceros antler size was proposed (positive allometry), and we waited until the 1970s for someone to actually test whether Megaloceros antlers were actually oversized, cumbersome organs or consistent in proportion to what we'd expect from living deer. This study came in the form of Stephen J. Gould's famous 1974 paper "The origin and function of 'bizarre' structures: antler size and skull size in the 'Irish Elk,' Megaloceros giganteus", which demonstrated that Megaloceros antler dimensions were not only consistent with those predicted from living deer, but that their evolution was explainable through regular old sexual selection - no crazy, uncontrolled evolutionary mechanism required. Further work has shown that, relative to body size, Megaloceros antlers were no heavier than those of fallow deer and, indeed, were actually lighter than those of reindeer (Geist 1999). Gould's results were an important grounding of Megaloceros in the wider context of deer biology. Rather than seeing it as a weird, dead-end outlier, Gould showed that Megaloceros represented the known limit of antlers operating as 'honest' signals of stag health and virility. Their size is a predictable outcome of cervine evolution should species develop into large-bodied, open-country specialists.

Were Megaloceros antlers exceptionally large? In an absolute sense, yes, but in a relative sense, no: they're appropriate for a deer of its body size. This graph from Gould's landmark 1974 paper is a little crude - neither shoulder height nor antler length are the best measure of body proportions (masses would be preferable) but this study marked the beginning of more informed and quantified investigations into the evolution, function and significance of giant deer antlers.

Furthermore, ideas that Megaloceros antlers were somehow evolving out of control ignores nuances of their structure and evolution. There's no doubt that - as with all stags - giant deer males were physiologically stressed by antler production. Even assuming development occurred over several months, centimetres of growth were required each day and it's doubted that enough calcium and phosphate could be processed directly from their food fast enough to meet demand (Moen et al. 1999). Megaloceros responded with the standard cervid adaptation of borrowing bony material from its skeleton, and physiological models suggest up to 80% of their available resorbable calcium was required to achieve a full antler spread - about twice that of moose (Moen et al. 1999). Assuming access to suitable plants, these resorbed minerals could be replaced in time to rebuild a strong, durable skeleton for the rut, but osteoporosis remained a risk for stags unable to recapture that material. It's thus unsurprising that Megaloceros antlers are, on average, not quite as solid as those of other deer. All antlers are a blend of compact bone (very dense and heavy, equating to a high mineral cost) and a spongier component (lighter and less dense, so lower mineral cost), and Megaloceros evolved a higher ratio of lower-density bone to better manage the physiological demands of antler growth. Even this was not enough to prevent their antlers from becoming a burden at times, however: average antler size is seen to reduce on occasion throughout their evolutionary history. These responses are the exact opposite of what we'd expect in an orthogenic 'runaway evolution' model, and are evidence of how ordinary selection pressures were keeping Megaloceros anatomy in step with environmental conditions.

Megaloceros antlers were exceptional communication devices, appearing enormous from multiple angles, and especially so from the front. Unlike other deer, Megaloceros would not have to pose to show off its antler palms: they were unmissable however you saw it. Screengrab of a 3D scan by National Museums Scotland that is loads of fun to play with (and forms a useful reference for artists!).

The relatively low mineral value and size of Megaloceros antlers has led to discussions about their function: were they used for visual communication only? Palmate antler regions, which occur today in species like fallow deer and moose, are thought to serve primarily in communication, and it stands to reason that this applied to Megaloceros too. But while living palmate stags use head and neck motion to draw attention to their antlers (Gould 1974; Geist 1999), Megaloceros headgear was rotated in such a way that stags could exhibit their huge palms by simply standing still. This precluded the need to wave 40 kg of bone around for days on end, which was probably terrific news for the neck and head tissues of Megaloceros stags (Gould 1974). This said, Megaloceros skulls, necks and shoulders were heavily augmented to support their antlers (the skull roof was 30 mm thick! - Lister 1994) and it's probable that they were capable of moving them around with speed and precision, perhaps even violently. Although some authors (Gould 1974) have regarded Megaloceros stags as incapable of wrestling with one another, studies of their antler histology and stress distribution suggest such matches were possible (Lister 1994; Klinkhamer et al. 2019). The effects of clashing antlers have yet to be modelled - this being fraught with variables that are difficult to estimate at present - but Megaloceros antlers perform well under pushing and twisting regimes provided the stresses are primarily experienced in their proximal regions. They appear to have been especially resistant to rotational forces, and we might imagine stags locking their antlers together at close range, faces low to the ground, shoving and twisting each other to the floor (Lister 1994). Their antlers are so long that forces incurred at their tips might have been amplified to potentially bone-snapping levels however, so it's possible Megaloceros may have avoided especially violent, unpredictable fighting styles. Again, this is consistent with trends seen across deer: as stags become larger, and species become more sociable, their fights generally become more ritualised and lower risk. Megaloceros may be the ultimate example of this correlation (Geist 1999).

Life appearance

Charles Knight's 1906 take on Megaloceros in a traditional, red deer/wapiti form. Cervus-like reconstructions have typified this genus since at least the 1850s, when Waterhouse Hawkins reconstructed Megaloceros for the Crystal Palace Park. Image from Wikimedia Commons, in public domain.
Although no soft-tissue Megaloceros fossils are known, a reasonable amount has been inferred about its life appearance through predictions of ecology and interpretation of Palaeolithic art (e.g. Lister 1994; Geist 1999; Guthrie 2006; also see Darren Naish's summary at Tetrapod Zoology). Although these differ in some details, a consensus has formed around some aspects, including the presence of a large shoulder hump (a diagnostic feature of Megaloceros cave art) and a fallow deer-like prominent laryngeal region. These features seem common to both males and females, implying a use beyond male competition and sexual signalling. It's tempting to link large larynxes with deep, guttural calling, akin to that produced by Dama, and with both genders sporting them, perhaps Megaloceros was an especially vociferous deer? Palaeolithic artists depicted the shoulder hump as large and bulging in both genders, so it probably does not merely represent the augmented, antler-carrying shoulder and neck muscles of stags. Geist (1999) speculated that it represents a fatty hump: an energy reserve for periods of seasonal stress that would not, by being concentrated in one place, cause the same overheating risk as distributing fatty tissues across the body, as occurs in some reindeer. Zebu cattle, white rhinos, bison and buffalo are similarly adapted running species (Geist 1999). There is no artistic evidence of a mane or bell, as seen in some other large cervids, although some Palaeolithic art hints at longer hairs on the hump.

A take on Megaloceros body patterning based on a very literalistic reading of certain Palaeolithic artworks - note the stripe running from the shoulder to hindlimb on the stag. The shoulder hump on the doe should probably be larger.
Predictions about colouration have also been presented, although no-one should be under the illusion that we have a robust idea of colours and patterns in this species. Our most informative insights into this are cave art showing dark colours at the shoulder hump which taper into dark stripes extending across the body and neck (diagonally across the flank from the shoulder to the knee region, as well as vertically across the neck-torso junction). A further dark stripe or band is recorded at the top of the neck, possibly in relation to the enlarged laryngeal region. Another vertical stripe may have framed a rump patch - a feature consistent with Dama. How we interpret these drawings isn't 100% clear, with the full-body restorations of Geist (1999) and Guthrie (2006) differing in some details. Geist regarded Megaloceros as pale all over with a conspicuously dark transverse body stripe and dorsal midline, while Guthrie portrayed more regionalised colouration with darker hindquarters grading into paler hues around the shoulder, framed by prominent stripes. Lister (1994) was more conservative, merely noting the likelihood of the dark shoulder hump.

One of the most informative Palaeolithic artworks of Megaloceros is featured in the Megaloceros Gallery in France's famous Chauvet Cave. Unlike other examples, it seems to show shaded areas that could reflect darker body regions, similar to the banded Coelodonta depicted on the opposite wall. A second Megaloceros is featured a little further in, also showing a dark neck band. Alas, other Megaloceros panel drawings lack similar details, precluding determination of their reliability. Cropped from Google Arts and Culture - make sure to check out the Megaloceros Gallery in 3D!

The challenges of interpreting Megaloceros colouration from cave art alone are many. Opinions differ on how many panel drawings we have - perhaps up to 40, but of which maybe only 15 or 16 are confidently identified as M. giganteus. Most are body outlines without details of colours or patterns and, to my knowledge, only 4-5 pieces feature those dark humps and body stripes. This is enough to assume that they are recording something about Megaloceros life appearance, but what, exactly, might that be? Are the lines radiating from the shoulder humps stripes, or something else, like the boundaries between colour regions? Are those simpler 'body outline' artworks actually depicting Megaloceros with uniform colouration? Do these differences reflect seasonal or regional colour variation, or else varying artistic approaches? The question of how literally we should take Palaeolithic depictions is not new, and it doesn't have a straightforward answer (Guthrie 2006).

Select Palaeolithic art of Megaloceros as interpreted by Guthrie (2006) - arrows indicate features of interest such as shoulder humps, expanded laryngeal regions etc. Note the absence of obvious colour and pattern data in most depictions - what significance, if any, might that have? This compilation of Guthrie's work is swiped from Tetrapod Zoology.
But while Palaeolithic art does not give us a clear answer on Megaloceros life appearance, ideas of it being a highly-specialised 'extreme' deer species help us explore likely colour schemes a little further. We can look to deer and other ungulates that match Megaloceros in behaviour and ecology to make some broad-brush predictions of what an especially large, well-ornamented, communicative, open-habitat and cursorial deer species may have looked like. Geist (1999) calculated that an energetic 600 kg deer living in an open, shadeless environment could be sensitive to solar insolation, especially during sustained ruts and runs, and predicted light-colours across the body to reflect sunlight and avoid overheating. Thoughts that Megaloceros may have been subject to heat stress are not baseless: some readers may recall that moose are highly prone to overheating, even in winter, despite living in shadier settings and having a higher surface area:volume ratio than Megaloceros (Dussault et al. 2004; van Beest and Milner 2013). Geist also predicted a rump patch - a feature seen in Dama as well as possibly one piece of ancient art - as this feature has convergently developed in a number of social, cursorial ungulates. The possibility of dark body regions or stripes has some functional explanation, too, with Geist (1999), Guthrie (2006) and Lister (1994) all noting the utility of dark patterning in ornamented deer, and Geist suggesting this would be especially effective in an open environment against a light sky. It's difficult to know how correct these predictions are of course, but I admire the fact that they consider the unique lifestyle of Megaloceros rather than, as has been palaeoart tradition until recently, merely dressing Megaloceros as a red deer or wapiti.

Demise and extinction

Charles Knight's depiction of a Megaloceros stag being hunted by early humans. Did we contribute to the extinction of giant deer? Possibly, although evidence for our interaction with this species is uncommon. From Knight (1949).
It seems remiss to have written all this about Megaloceros without mentioning modern ideas of why it is no longer with us. This remains a topic of ongoing research, and a synthesis provided by Lister and Stewart (2019) gives a good overview of the current state of the debate. Our vast sample size and detailed understanding of Quaternary sediments means that the history of Megaloceros can be read in detail, providing a substantial (though still developing) insight into the major factors that contributed to its demise. Once thought to have been an exclusively Pleistocene species that held out longest in northwest Europe, recent discoveries and dating of Megaloceros sites in Siberia have not only confirmed its survival well into the Holocene, but shown that Asia held Megaloceros populations longer than anywhere else - as recently as 7,600 years ago.

With concepts that Megaloceros antlers were some sort of doomed anatomy pulling giant deer to extinction firmly disregarded, more sensible discussions about Megaloceros demise consider environmental change, their high productivity demands, and human hunting likely extinction causes. Generally, it seems that local Megaloceros extinctions coincide with habitats changing from parkland habitat to wooded or tundra-like settings, both of which heralded reduction in optimal foraging conditions as well as losses of plants rich in calcium. Megaloceros populations in those areas may have thus dwindled or been forced to find more suitable places to live. There is evidence that Megaloceros anatomy responded somewhat to nutrient-stressed conditions, but ideas that dwarfed or otherwise stunted populations developed in islands or other refugia have not been bourne out: late-surviving Megaloceros were only something like 2% smaller than earlier specimens. The lack of more dramatic body size reduction may be related to predation pressures to stay large and fast, especially as new data shows that humans co-existed with Megaloceros for longer than was once thought. Evidence for our hunting giant deer is relatively rare so we may not have been a major factor in their extinction, but as suitable habitats reduced, even low hunting frequencies may have been enough to push Megaloceros over the edge.

As is always the case with a well-studied species, there's much more we could say about Megaloceros. We are still learning about its palaeobiogeography, behaviour, evolution and extinction from new discoveries and detailed examination of archived specimens. Digging into the literature on these animals has proven genuinely fascinating. As someone used to studying Mesozoic reptiles, the amount of information and insight we have on this species is almost overwhelming, and it seems that understanding its palaeobiology is challenged more by having to consider so much data rather than, as I'm more accustomed to, having too little information to draw meaningful conclusions. And yet, despite an excellent fossil record and centuries of study, some major aspects of its biology remain poorly understood or contested. Food for thought, indeed, about our capacity to interpret fossil tetrapods from more typical palaeontological datasets.

Enjoy these 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 like a trivial 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, paintings 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!


  • Dussault, C., Ouellet, J. P., Courtois, R., Huot, J., Breton, L., & Larochelle, J. (2004). Behavioural responses of moose to thermal conditions in the boreal forest. Ecoscience, 11(3), 321-328.
  • Geist, V. (1999). Deer of the World. Swan Hill Press, Shrewsbury.
  • Gould, S. J. (1974). The origin and function of 'bizarre' structures: antler size and skull size in the 'Irish Elk,' Megaloceros giganteus. Evolution, 191-220.
  • Guthrie, R. D. (2005). The Nature of Paleolithic Art. The University of Chicago Press.
  • Hughes, S., Hayden, T. J., Douady, C. J., Tougard, C., Germonpré, M., Stuart, A., ... & Say, L. (2006). Molecular phylogeny of the extinct giant deer, Megaloceros giganteus. Molecular phylogenetics and evolution, 40(1), 285-291.
  • Janis, C. M., Shoshitaishvili, B., Kambic, R., & Figueirido, B. (2012). On their knees: distal femur asymmetry in ungulates and its relationship to body size and locomotion. Journal of Vertebrate Paleontology, 32(2), 433-445.
  • Klinkhamer, A. J., Woodley, N., Neenan, J. M., Parr, W. C., Clausen, P., Sánchez-Villagra, M. R., ... & Wroe, S. (2019). Head to head: the case for fighting behaviour in Megaloceros giganteus using finite-element analysis. Proceedings of the Royal Society B, 286(1912), 20191873.
  • Knight, C. R. (1949). Prehistoric man: The great adventurer. Appleton-Century-Crofts.
  • Kuehn, R., Ludt, C. J., Schroeder, W., & Rottmann, O. (2005). Molecular phylogeny of Megaloceros giganteus—the giant deer or just a giant red deer?. Zoological Science, 22(9), 1031-1044.
  • Lister, A. M. (1994). The evolution of the giant deer, Megaloceros giganteus (Blumenbach). Zoological Journal of the Linnean Society, 112(1-2), 65-100.
  • Lister, A. M., Edwards, C. J., Nock, D. A. W., Bunce, M., Van Pijlen, I. A., Bradley, D. G., ... & Barnes, I. (2005). The phylogenetic position of the ‘giant deer’ Megaloceros giganteus. Nature, 438(7069), 850-853.
  • Lister, A. M., & Stuart, A. J. (2019). The extinction of the giant deer Megaloceros giganteus (Blumenbach): New radiocarbon evidence. Quaternary International, 500, 185-203.
  • Moen, R. A., Pastor, J., & Cohen, Y. (1999). Antler growth and extinction of Irish elk. Evolutionary Ecology Research, 1(2), 235-249.
  • van Beest, F. M., & Milner, J. M. (2013). Behavioural responses to thermal conditions affect seasonal mass change in a heat-sensitive northern ungulate. PloS one, 8(6), e65972.
  • Vislobokova, I. A. (2012). Giant deer: origin, evolution, role in the biosphere. Paleontological Journal, 46(7), 643-775.

Wednesday 26 August 2020

The "palaeontological folklore" of mastodon hair

The American mastodon Mammut americanum is one of the most iconic members of the North American megafauna. A frequent subject of museum displays, books and technical research for over two centuries, we can all immediately conjure mental images of this long-bodied, relatively short-legged elephant-like animal. Nearly all of us will imagine mastodons the same way: that is, covered with a thick layer of brownish hair in a fashion reminiscent of its even more iconic cousin, the woolly mammoth. This is simply how we've come to understand mastodon life appearance through centuries of artistic and literary reinforcement (examples in familiar books include Špinar and Burian 1972; Benton 2015; Prothero 2017). Many texts don't even bother citing academic sources evidencing the claim: the hairy mastodon concept has been repeated often enough and long enough to be established fact. Zebras have stripes, lions have manes, and mastodons had a thick, brown covering of hair.

Zdenek Burian's 1964(?) take on the American mastodon: an entirely typical restoration of this species that would pass as a credible restoration at any point during the last two centuries.

It might come as a shock, therefore, to learn that the foundation of evidence behind our shaggy mastodons was actually entirely baseless for almost two centuries, and that this widely accepted concept has only gained a small amount of support within our lifetimes. This isn't a new revelation, either. American anthropologist and author Loren C. Eisley, a key figure in unravelling the strange history of mastodon soft-tissues, held no punches when describing mastodon hair as "items of palaeontological folklore" in 1945 (p. 108). Eisley's interest in mastodons was driven by ideas of their survival into recent centuries, where so-called mastodon soft-tissues discovered in the 19th century were interpreted as evidence of mastodons dying out just hundreds of years ago. Eisley published rebuttals to this concept several times during the 1940s and, in his 1946 Science article Men, Mastodons, and Myth, he specifically delved into the peculiar history of mastodon soft-tissue discoveries. Through historical detective work, Eisley uncovered a series of erroneous interpretations, failed corrections, and even deliberate subterfuge from the early days of American fossil exploration. A brief summary of his findings are provided below, but be sure to check out Eisley's account yourself for the full picture.

Lies, damned lies, and mastodon hair

The story of mastodon soft-tissue begins in 1800 Newburgh, New York, where a mastodon tooth and associated sample of coarse, 'dun brown' hair was extracted from a bog on the farm of a Mr. A. Colden. This specimen, said to have been so rotten that it decayed to dust within days, was a significant find. It was not only the first alleged mastodon hair, but also the first indication that some Pleistocene giants might have been covered in fur. It's among the more credible accounts of mastodon hair from the 19th century and, for many, was the best evidence of mastodon skin even as new finds were made later in the same century.

Charles Knight's awesome 1897 Mammut americanum painting. So far as I can tell, Knight's work is among the earliest restorations of this species and establishes the shaggy coat we'd come to associate with it. This is one of my favourite Knight paintings: just look at that landscape.

Shortly after the Colden Farm discovery, reports of mastodon soft-tissue then came thick and fast. In 1805 a mastodon was reported by Shawnee native Americans as having a long nose and mouth - interpreted, naturally, as a fossilised trunk. A report of a fossil mastodon stomach with gut content was reported at around the same time, as were more specimens with small amounts of hairy skin. Especially large sheets of skin were reported in 1839. These were reportedly so well preserved that they included arteries and sinew, but were too fragile to collect in one piece - only small fragments could be extracted. Collectively, such remains formed a significant dataset regarding mastodon soft-tissues, from internal organs to external features. They are the seed from which the concept of hairy, brown mastodons grew, and by the end of the century hairy mastodons were described in textbooks (e.g. Hutchinson 1893 - note that Hutchinson's book contains a rare Joseph Smit illustration of a nearly hairless mastodon) and featured in influential Charles Knight artworks (above),

But if this evidence is so great and extensive, why are none of these specimens better known? Are they not on display in museums, or at least illustrated in a book or paper? It turns out that there's a good reason you've never seen them: they're all completely bogus. Not a single example of mastodon soft-tissue reported in the 19th century was accessioned to a museum, virtually none were examined by individuals with paleontological expertise, and no detailed reports were ever made. Some clearly never existed and those that did were almost certainly erroneous takes of otherwise unremarkable objects.

Eisley's investigation assumes that some of these misinterpretations were honest mistakes by inexperienced parties, or perhaps over-interpretations of field reports (e.g. the 'long nose' specimen account makes no actual mention of soft-tissues, and could pertain to osteological observations alone). He attributes these errors to the early 1800s being a time of great excitement about the then newly-discovered Siberian frozen mammoths, and anticipation that fossil American proboscideans would return remains of similar quality. This might explain why some cases, such as with the alleged mastodon stomach specimen, are just outright strange. Gut tissues are among the first organs to decay when animals die and it would be very odd for a stomach to survive, alone, after the rest of the animal had rotted away. This discovery sounds a lot like the sort of over-zealous interpretation that might be made by someone naive about taphonomy and, indeed, this specimen was quickly subjected to rebuttals and corrections from more experienced scholars. 

But other accounts were likely dishonest from the start - tall-tales to excite interest in fossil specimens set for display in private museums and touring shows. The early 19th century was a time before public museums existed in the USA, so fossil remains were exhibited to the public through private enterprise: hype and publicity-seeking were important to making such efforts financially viable. It's in this context that we have to view the 1839 Missouri mastodon reported to have extensive amounts of preserved skin, arteries and sinew, which was reported by non-other than Albert Koch: the same museum proprietor responsible for reconstructing Basilosaurus as a sea serpent and a Missouri mastodon as the 'Missourium' - a monstrous assemblage of mastodon and wooden blocks toured for public display in the mid-1800s. Koch's exaggerating, fraudulent approach to the conveyance of paleontological data robs all credibility from his accounts, and no-one can take his unverified claim of giant skin samples seriously. The fact that he never followed up on these seemingly remarkable soft-tissue remains is further evidence that they never existed.

Albert Koch's "Missourium" - a composite and distorted mastodon skeleton augmented with bits of wood to make the vertebral column longer. The Missourium specimen was collected around 1840, and I wonder if Koch's accounts of giant skin samples were associated with its discovery.

But what of the Colden Farm find which - while still entirely anecdotal - at least mentioned the unstable nature of its mastodon hair, and thus accounted for its absence in current collections? This 1800 discovery gains additional credibility in pre-dating our knowledge of frozen mammoths in Russia, and must therefore have been an unbiased, honest interpretation of alleged mastodon material. But, again, particulars of this find are peculiar: isn't it strange that an entire mastodon disappeared to leave only a solitary tooth and a patch of hair behind? This is taphonomically very odd, but was not an isolated incident: other fragmentary bones with patches of hair were also recovered from swamps in the same area. The American geologist James Hall provided an explanation of such sites which is far more consistent with our understanding of animal decay patterns. As part of a wider survey of the geology of New York state, Hall found the swamps yielding these fragmentary mastodon remains were filled with a hair-like algae known as conferva. It was a good match, size and morphology wise, for the alleged mastodon hair, and when desiccated, it turned - you guessed it - 'dun brown'. When describing a mastodon site in 1843, he wrote:

In a small muck swamp in Stafford, Genesee County, a small molar tooth was found several years since. Its situation was beneath the muck, and upon a deposit of clay and sand. A large quantity of hair-like confervae, of a dun brown color, occurs in this locality; and so much does it resemble hair, that a close examination is required to satisfy one's self of its true nature.

Hall 1843 (from Eisley 1946, p. 522)

Hall was one of the few truly experienced and qualified individuals to write about mastodon soft-tissues in the 19th century, so his assessment is of real interest to this story. This is not to discount the insight of the pastors, farmers and businessmen behind other accounts, but Hall's explanation certainly sits better with our modern understanding of taphonomy, as well as the fact that those New York swamps have - even today - yet to yield a single scrap of incontrovertible mastodon hair. For Eisley, if there's any truth to these early takes on mastodon hair, Hall nailed it: the New York mastodon hair samples were simply misidentified dry algae.

'Conferva' is not a widely used term nowadays, but once pertained to a great number of filamentous green algae species. It's easy to see how examples like the above might be mistaken for mastodon hair by naive parties. Image by Anne Dixon, from 1843-45, borrowed from Getty Museum.

Without any specimens to examine, we cannot be certain today whether Hall and Eisley were correct, but their work clearly shows that 19th century claims of mastodon hair are suspicious. This is the line taken by at least some modern authors writing about mastodon hair (e.g. Hallin 1989; Haynes 1991; Larramendi 2015) but, as we know from history, most people ignored both Hall in the 19th century and Eisley in the 20th to perpetuate the discredited concept of hairy mastodons. How did such questionable data become the established, unquestioned truth about mastodon life appearance? Eisley (1946) attributed this to the illusory truth effect, where the repetition of a claim by perceived authorities makes it seem factual and truthful, regardless of the underlying evidence. In this case, enough scientists, museums, books and other media have towed the hairy mastodon line to transform folklore into 'fact', seemingly without anyone wondering where the real evidence of mastodon hair was. As Eisley put it:

In the midst of this constant repetition of what, through the sheer prestige of age, has come to be accepted as undeniable fact, it has never been pointed out that American institutions of science do not possess the tangible evidence which alone could justify such wholehearted faith in the exact appearance of this long-vanished beast.

Eisley 1946, p. 517.

I struggle to think of a case where interpretations based on a comparably feeble palaeontological dataset have been rehashed so uncritically for so long, so regularly and so publically. There is ordinarily some pushback against wholly baseless ideas of extinct animal life appearance, even if only among specialists, but I can only find a handful of articles promoting non-hairy takes on mastodon life appearance from the last century. In light of dedicated efforts by the likes of Eisley to set the record straight - published in Science, no less - it's really quite baffling that we've unwaveringly promoted hairy, brown mastodon for so long.

Finally: a real specimen

The situation around mastodon soft-tissues has changed somewhat today. In the 1980s, Pleistocene mammal expert Kurt Hallin published two abstracts and one popular article about the first genuine chunks of mastodon skin, both of which were covered in hair (Hallin and Gabriel 1981; Hallin 1983, 1989). But if you're hoping that this finally gives us real insight into mastodon life appearance, you're out of luck. To my knowledge, this specimen has never been described or illustrated beyond these short works, and a scanning electron microscope image of a single hair is all that's been figured of it (below). This short paragraph by Haynes (1991) provides one of the more detailed overviews I could find:

Specimens of what appears to be carbonized skin holding together bundled and fine hairs interspersed with hollow, coarser hairs may be the only preserved Mammut soft tissue currently known. These specimens were recovered by Krut Hallin in association with cranial fragments found near Milwaukee, Wisconsin (K. Hallin 1989 personal communication). The preserved guard hairs are hollow, a common enough trait in mammals, including woolly mammoths and African elephants. The underfur appears similar to that of semi-aquatic mammals such as the otter and beaver (Hallin 1983, 1989; Hallin and Gabriel 1981), in that it is very fine and wavy, and grows in dense bundles.
Haynes 1991, p. 34.

30 years on, these brief reports remain our only direct evidence of mastodon hair (Haynes 1991; Larramendi 2015), leaving the life appearance of the American mastodon not significantly advanced from Eisley's dismantling of 19th century discoveries from 75 years ago. While the Wisconsin hair specimen represents a potential step forward, it remains in a scientific grey area for having never being described or illustrated in detail, and Hallin's interpretation has never been subjected to peer review. Its significance to mastodon palaeobiology and life appearance thus remains an open question, as does the nature of mastodon skin in general. After all, the Wisconsin specimen only represents skin from the cranium, and we really need skin from different regions of extinct animal bodies to be confident about their full appearance. With such little data to hand, the whole body integument of mastodons remains mysterious (Haynes 1991; Larramnedi 2015).

Hallin's (1989) SEM shot of mastodon hair from Wisconsin - that's it on the left. To my knowledge, this is the only published image of mastodon hair.

Could mastodon still be hairy, though?

But this is not to say that we should just give up on restoring mastodon life appearance, however. These are relatively well-understood fossil animals, and we can surely say something about their skin from details of their anatomy, ecology and palaeobiogeography. If we are to play this predictive game, perhaps the most important first step is to realise that woolly mammoths - which are clearly the historic inspiration for mastodon art - are not great analogues for American mastodon. The familiarity of woolly mammoths makes it easy to forget they were genuinely strange, specialist species adapted to extreme cryoarid Pleistocene habitats (Boeskorova et al. 2016). They relied on thick, three-tiered hairy coverings, generous adipose layers, shoulder fat humps, small ears, shortened tails, and trunk tip 'hoods' to protect themselves from extremely cold winters and periodic food shortages (Boeskorova et al. 2016). American mastodon, by contrast, were not Ice Age specialists. They actually evolved during the warmer Pliocene conditions pre-dating the Pleistocene glaciations and, even during the Ice Age, avoided the extreme climates endured by woolly mammoths. The distinct habitat preferences of mastodon and woolly mammoths are demonstrated by their remains being very rarely found in the same fossil horizons (Graham 2001; Hodgson et al. 2008). Woolly mammoths were denizens of steppe, tundra and forest habitats in northern regions (specifically, Alaska, the upper midwestern and northeastern United States, and the northern Atlantic coastal plain), while mastodon preferred wetter woodlands in more southerly locations (the eastern US (extending as far south as Florida), southeastern Canada, and parts of Mexico) (Haynes 1991; Graham 2001; Newsom and Mihlbachler 2006). Mastodon thus enjoyed a range of habitats and climates: northernmost populations inhabited boreal forests, living alongside moose and beavers, while those of Florida and Mexico inhabited relatively balmy swamps and woodlands, sharing their environment with reptiles and amphibians (Hine et al. 2017). Without soft-tissue mastodon fossils we can't truly assess their cold adaptations, but Larramendi (2015) noted that Mammut tails are long for proboscidians. This contrasts with the especially short tails of woolly mammoths and might have implications for mastodon thermal energetics - in other words, they weren't feeling the cold enough to shorten their tails (Larramendi 2015). These details of ecology, biogeography and anatomy demonstrate how different mastodon and woolly mammoths were, and caution against our porting of mammoth skin to their stockier cousins.

Woolly mammoths are very familiar to us, but we should not forget that they were not 'normal' giant animals. They were specialists adapted to life in extremely cold, dry habitats and, like most specialists, their lifestyle required a glut of anatomical weirdness. We should not readily assume that they are suitable model species for other giants. This 2018 image shows M. primigenius meeting a family of Neanderthals.

Having taken mammoths out of the picture, we can focus on a key question about mastodon: would they have benefitted from a full-body covering of hair? To answer this, we have to consider their giant size, and what that means for their thermal energetics. I'm increasingly of the opinion that that, when considering the life appearance of gigantic, multi-tonne extinct animals, we should be justifying the presence of thick layers of insulation, not the removal of it, and I feel that approach has merit here. American mastodon averaged body masses of 8 tonnes, making them larger than not only their distant woolly cousins (Larramendi 2015) but also considerably heavier than an average living elephant of any species. As discussed at length in previous posts, even at these smaller masses, elephants are simply so big that they really don't feel cold that much, to the extent that some African populations endure months of sub-freezing nights despite their lack of hair (Haynes 1991). Indeed, we know that elephants likely rely on these cooler periods to survive, as their size and low surface area-volume ratio present them with numerous challenges related to heat loss during the day. Among other issues, they cannot shed heat as quickly as it is generated during exercise, so their bodies regularly warm to dangerous hyperthermic temperatures. With little direct control over their body temperatures - they cannot even sweat or pant - they rely on features of their environment - bodies of water for drinking and trunk spraying, wallowing, cool nighttime temperatures etc. - to remain cool (Wright and Luck 1984; Weissenböck et al. 2012; Rowe et al. 2013).

Given that mastodon represent an even stockier, heavier and shorter-legged variant on the elephant body plan, they surely faced similar, or even more pronounced, challenges on this front. Cooler Pleistocene temperatures would have alleviated these concerns somewhat, but were unlikely to solve them outright. Elephants generate their own climate envelope around their bodies when exercising so that, even away from their very hot natural ranges, they overheat when exercising for long periods (Rowe et al. 2013). The insulative effects of fibrous integuments compound considerably with body size - a layer of fuzz over a one-tonne animal has a significantly greater insulative effect than the same layer on a 1 kg animal (Fariña 2002; Porter and Kearney 2009). Thus, even if smaller mammals living alongside mastodon required fur coats, mastodon themselves may not have needed them. Based on the thermal tolerances of living animals, we can predict that a blanket of fur all over a mastodon would have a significant impact on their thermal physiology, likely pushing their thermal neutrality (crudely explained, their ambient temperature 'comfort zone') many degrees below zero. Consider that Fariña (2002) calculated thermal neutrality for a naked 4 tonne Megatherium as -17°C*: I'm not sure what the thermal neutrality of a mastodon is, but their additional 4 tonnes of mass would surely push their thermal neutral figure even lower, maybe much lower. These values are only indicative because they assume dry, windless conditions, and both wind and rain make animals more vulnerable to cold, but they give a sense of how cold-resistant multi-tonne animals are. I'm guessing that Floridan mastodons felt pretty hot a lot of the time.

*Before you ask, this calculation also assumed a placental-typical metabolic rate, so the comparison to mastodon is apt.

Another question we might ask is how prone animals inhabiting generally cold habitats, such as boreal forests, are to heat stress. Does the cold in these settings negate any risk of overheating? Weighing up to 750 kg, moose and elk - members of Alces - are some of the largest modern animals to inhabit boreal forest, and their thermal energetics are well studied. They thus provide useful insight into how large animals cope with cold climates. Winters in boreal forests can get seriously cold - an average day might be -20°C - but moose are actually still prone to heat-stress during these months (Dussault et al. 2004; van Beest and Milner 2013). These aren't just occasional occurrences, either: moose have to manage their heat-stress continually via habitat choice and behavioural modification. If they don't, the associated energy drain to cool their bodies dramatically impacts their mass and health over the winter months (van Beest and Milner 2013). If these cervids find their whole-body fur coats stifling in boreal forest winters, how would mastodon - creatures ten times their size or more - cope with similarly hairy skin? I appreciate that this comparison is very crude - deer are not proboscidians, and modern boreal forests might be climatically different to those of the Pleistocene - but it demonstrates that we should not automatically equate cold climates with cold-stressed animals, and that we should think about the physiological implications of insulating fossil giants, even in frosty settings.

An American mastodon in the cypress swamps of Pleistocene Florida, sporting a hairy face and shoulders, but a largely hairless body. This looks odd compared to our standard fully-hairy restoration, but is consistent with our cranial mastodon hair sample and the thermal physiology of very large proboscideans. Happily, there's enough hair left for an obvious Simpsons gag. Mmm... mastodon fresh.

Putting all this together, there are clear grounds to question whether mastodon needed full hair coats. Consider the evidence: we know that elephant-like mammals already struggle more than other species with heat loss; that an 8 tonne animal is going to have very low thermal neutrality even without a fur coat; that mastodons were not universally denizens of especially cold settings; and that at least some parts of their anatomy - their tails - show opposing conditions to hairy, cold-adapted relatives. Based on this, I'm wondering if parts of the mastodon body - maybe even large portions of it - had reduced or absent hair to negate heat stress (above). This may have been especially important for mastodon living in warmer southern regions or areas with hot summers. Perhaps seasonal or geographic variation in hair growth and distribution was also employed, so that mastodon living in the warm south were almost as hairless as living elephants, and those exposed to cold winters were a little hairier? Of course, we cannot ignore the data from the Wisconsin hair specimen, which shows at least some mastodon faces were hairy, but we should be cautious about extrapolating this to a full-body covering. Proboscidian faces are actually one of the few parts of their bodies that are vulnerable to the cold, as their ears and trunks can develop frostbite (Haynes 1991). The face is thus one of the regions we might expect hair to be present in mastodon enduring cold winters, and it may not reflect the thermal demands of the rest of the body. Clearly, what's needed here is a dedicated study factoring mastodon mass, body surface area, metabolism, and heat loss against a nuanced consideration of Pleistocene climates: there are lots of equivalent studies on other fossil animals, and it would be great to get some real data and investigation on this. Until then, I find Asier Larramendi's (2015) comment on this matter an excellent summary: "...the relatively long tail... and the massive body of M. americanum suggest that the prevalent ideas that these animals were covered with a thick coat of fur are probably exaggerated".

From folklore to parable?

Of course, the main point of this article isn't to explore mastodon appearance in detail, but to stress that the nature of their skin is nowhere near as well-evidenced or understood as we've historically implied. I feel that those of us involved in education, research and palaeoart have a job on our hands to revise our misplaced confidence about mastodon life appearance, and to become more open-minded and explorative about what these iconic animals may have looked like. We should also take a moment to think about how this matter differs from our conventional issues with depictions of fossil animals. While it's not uncommon to bemoan how prehistoric animals are discussed and communicated to the public, we mostly do so because of pop culture influences - a film, TV show or book that perpetuates old tropes or fabricates something ridiculous about a prehistoric species. But there are no sensationalised media behind the perpetuation of hairy mastodons: these misleading depictions have come directly from decades of academic texts, officious museum displays and professionally produced palaeoartworks repeating the same unverified facts over and over, all the while ignoring direct rebuttals of the sketchy data behind this depiction. This one's on us, in other words. Perhaps it's time to turn this the "palaeontological folklore" of mastodon hair into a "palaeontological parable": a story to remind us to check and verify even the most basic and well-known information about our subjects from time to time, to ensure we're actually communicating science, and not rehashing mastodon-sized tall tales.

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 like a trivial 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, paintings 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!


  • Benton, M. J. (2014). Vertebrate palaeontology. John Wiley & Sons.
  • Dussault, C., Ouellet, J. P., Courtois, R., Huot, J., Breton, L., & Larochelle, J. (2004). Behavioural responses of moose to thermal conditions in the boreal forest. Ecoscience, 11(3), 321-328.
  • Eiseley, L. C. (1946). Men, mastodons, and myth. The Scientific Monthly, 62(6), 517-524.
  • Eiseley, L. C. (1945). The mastodon and early man in America. Science, 102(2640), 108-110.
  • Fariña, R. A. (2002). Megatherium, the hairless: appearance of the great Quaternary sloths (Mammalia; Xenarthra). Ameghiniana, 39(2), 241-244.
  • Graham, R. W. (2001, October). Late Quaternary biogeography and extinction of proboscideans in North America. In The World of Elephants: Proceedings of the 1st International Congress, Rome, Italy (pp. 16-20).
  • Hallin, K. F. (1983). Hair of the American mastodon indicates an adaptation to a semiaquatic habitat. In American Zoologist (Vol. 23, No. 4, pp. 949-949).
  • Hallin, K.F. (1989). Wisconsin's Ice Age tuskers: Ice Age elephants and mastodonts. Wisconsin Academy Review, 35, 6-10.
  • Hallin, K.F. & Gabriel, D. (1981). The first specimen of mastodon hair. Geological Society of America 34th Annual Meeting of the Rocky Mountain Section, Abstracts with Program. 13(4), 199.
  • Haynes, G. (1991). Mammoths, mastodonts, and elephants: biology, behavior and the fossil record. Cambridge University Press.
  • Hine, A. C., Martin, E. E., Jaeger, J. M., & Brenner, M. (2017). Paleoclimate of Florida. Florida's Climate: Changes, Variations, & Impacts.
  • Hutchinson, H. N., & Woodward, H. (1893). Extinct monsters. Chapman & Hall.
  • Hodgson, J. A., Allmon, W. D., Nester, P. L., Sherpa, J. M., & Chiment, J. J. (2008). Comparative osteology of late Pleistocene mammoth and mastodon remains from the Watkins Glen site, Chemung County, New York. Mastodon Paleobiology, Taphonomy, and Paleoenvironment in the Late Pleistocene of New York State: Studies on the Hyde Park, Chemung, and North Java Sites. Palaeontographica Americana, 61, 301-367.
  • Larramendi, A. (2015). Shoulder height, body mass, and shape of proboscideans. Acta Palaeontologica Polonica, 61(3), 537-574.
  • Porter, W. P., & Kearney, M. (2009). Size, shape, and the thermal niche of endotherms. Proceedings of the National Academy of Sciences, 106(Supplement 2), 19666-19672.
  • Prothero, D. R. (2016). The Princeton field guide to prehistoric mammals (Vol. 112). Princeton University Press.
  • Rowe, M. F., Bakken, G. S., Ratliff, J. J., & Langman, V. A. (2013). Heat storage in Asian elephants during submaximal exercise: behavioral regulation of thermoregulatory constraints on activity in endothermic gigantotherms. Journal of Experimental Biology, 216(10), 1774-1785.
  • Špinar, Z. V., & Burian, Z. (1972). Life before man. McGraw-Hill Companies.
  • van Beest, F. M., & Milner, J. M. (2013). Behavioural responses to thermal conditions affect seasonal mass change in a heat-sensitive northern ungulate. PloS one, 8(6), e65972.
  • Weissenböck, N. M., Arnold, W., & Ruf, T. (2012). Taking the heat: thermoregulation in Asian elephants under different climatic conditions. Journal of Comparative Physiology B, 182(2), 311-319.
  • Wright, P. G., & Luck, C. P. (1984). Do elephants need to sweat?. South African Journal of Zoology, 19(4), 270-274.

Friday 31 July 2020

Benjamin Waterhouse Hawkins: the first grand master of palaeoart

Few of us need an introduction to the Crystal Palace dinosaur sculptures (here shown during construction in 1854), or their artist, Benjamin Waterhouse Hawkins, but not many of us know much about Hawkins or his work beyond this one installation, thus overlooking a major contributor and pioneer in the early history of palaeoart. Image from the Friends of Crystal Palace Dinosaurs blog.

Most of us would agree that palaeoart history is marked by a few key figures that, for one reason or another, define their generation. Knight, Burian, and Paul are perhaps the most considered of these, each producing an iconic portfolio of work that set the bar for other artists and recast how we might view and consume palaeoartworks. But the professional grandfather to all these famous figures is Benjamin Waterhouse Hawkins, a familiar name to anyone interested in palaeoart or vertebrate palaeontology, and yet one that doesn't always receive the accolade and praise it perhaps deserves. Hawkins - whose life spanned 1807-1894 - became the first great palaeoartist at a time when both palaeoart and palaeontology was still finding their feet. We all know of Hawkins' work at Crystal Palace, where he built 33 life-sized restorations of prehistoric mammals, reptiles and amphibians as part of the grand educational and commercial extravaganza, but many of us do not know much about him, his other works, or even how groundbreaking and unique his Crystal Palace sculptures were, demonstrating the sort of good practise and insight that we regard as essential in modern palaeoart.

[page image]
[page image]
[page image]
Hawkins was a skilled anatomist who wrote several books on animal anatomy and form. These images, from his 1860 book A comparative view of the human and animal frame, show his mastery of osteology and anticipate the now-common convention of a dark soft-tissue silhouette behind our skeletal reconstructions.

An appreciation of Hawkins is helped by learning something of his life. Although featuring frequently in accounts of Victorian-age natural history, the details of Hawkins' life had not been pieced together in real detail until his great great, great-granddaughter Valerie Bramwell and historian Robert M. Peck synthesised his personal and professional records into a concise biography (Bramwell and Peck 2008). Hawkins' existence was an eventful one, including both great professional success and tragedy, as well as a complex and somewhat mysterious personal life. The latter is perhaps the most cryptic and unexpected part of the Hawkins story, as he somehow sustained nearly four decades of bigamy (juggling two wives and 10 children, seven of whom survived infancy) without his families discovering the deception. His strange domestic life was found out in the mid-1870s, not long before a string of personal tragedies and a stroke brought Hawkins' life to a sad, inauspicious end. But despite his bizarre private dealings, Hawkins seems to have been a well-liked, generous man known as hardworking, personable and charming, equally confident of his abilities as an artist and anatomist but also reverential and respectful to his peers, especially the academics and intellectuals he frequently worked with.

Hawkins never secured permanent or long-lasting employment but instead moved from commission to commission and project to project. Though he dabbled in a few other careers, his chief profession was creating art of animals - initially living, and eventually extinct. Known best today for his sculpture, Hawkins only began working in 3D during the 1840s, well after he had established himself as an expert painter and illustrator. He was regarded as a high-quality, experienced natural history artist early in his career such that, even before he reached his career peak - the Crystal Palace geological restorations - he was sought-after as an illustrator of zoological specimens. He produced art for many of the biggest names in contemporary palaeontology and biology, including William Buckland, Gideon Mantell, Richard Owen, Thomas Henry Huxley, Charles Darwin and Joseph Leidy (interestingly, Hawkins could have worked with another big name, Edward Cope, but he disliked and refused to work with him - see Desmond 1976). Hawkins' association with Darwin and Huxley is of interest because he became an outspoken anti-evolutionist in later life, a position he expressed directly and clearly in many of the popular public lectures he delivered in his post-Crystal Palace career.

Before Hawkins recreated extinct dinosaurs, he was a prolific and sought-after illustrator of the modern kind, along with many other types of animals. This illustration of a White-crested Kalij was produced for John Edward Gray's Illustrations of Indian Zoology (1830-1835). (From Bonhams; incidentally, original Hawkins' lithographs like this sell for hundreds or even thousands of pounds to private collectors, a sad contrast with the under-financed conservation of his Crystal Palace works.)

Hawkins was well-known for his expertise in animal anatomy and form, which he learned primarily from zoological museum specimens as well as drawing captive animals (Bramwell and Peck (2008) recount an urgent trip to Paris in 1849 to sketch a newly born giraffe calf). As would happen later with Charles Knight, this expertise with living animals was instrumental in his later work at reconstructing extinct ones. Hawkins authored several books on animal anatomy and from the 1850s onward spent a large amount of time - and commanded a high fee - for grand public lectures. At his lecturing acme, Hawkins produced large drawings of prehistoric animals on stage (of such size that a ladder was required to reach the top of the board canvas) to bring his restorations direct to his audience. A gleaming reputation among British academics meant he enjoyed a warm reception in the US where he was offered many auspicious platforms and opportunities by well-regarded officials and academic institutions. These included the production of (sadly never realised) Crystal Palace-like prehistoric animals installations in Central Park and the Smithsonian, and production of the first-ever mounted dinosaur skeleton (Hadrosaurus), of which several casts were made (of which only one skull survives).

Hawkins' Central Park Workshop, c. 1869, with an assembled Hadrosaurus mount and the beginnings of a Laelaps skeletal reconstruction alongside - note the mount outline to the right of the Hadrosaurus ;skeleton. This terrific image (reproduced from Bramwell and Peck 2008) is a treasure trove of detail: look out for various mouldings of reptile skin, including a carnivorous species on the left, real animal skeletons in the background and unmounted bone replicas on the floor (are these bits of Laelaps?). There's also a lack of vermin - Hawkins' Crystal Palace workshop was shown with rats and birds scampering over the floor, and described as lying among a muddy swamp. The content of this workshop was destroyed by order of corrupt New York politicians, which devasted Hawkins.

There's a lot more we could say about Hawkins career, but you get the point: Hawkins wasn't a flash-in-the-pan artist who happened to land the Crystal Palace gig, nor was he someone who just knew how to throw clay and concrete around to someone else's design. He was a distinguished and respected intellect with expertise in natural history and anatomy, and brought considerable experience to his palaeoart commissions. Indeed, historians are uncovering increasing evidence that Hawkins was the principle intellect behind many of the Crystal Palace sculptures. We touched on this in my previous discussions of the Crystal Palace palaeoartworks but it's worth repeating: Richard Owen, lauded by many (e.g. Phillips 1854; Owen 1894; Desmond 1979; Beaver 1986; McCarthy and Gilbert 1994) as the brains and overseer of the Crystal Palace models, was actually barely involved and may have even held disdain for this grand project (Secord 2004; Dawson 2016). Numerous pieces of evidence, the best of which are the private correspondence from Hawkins, Crystal Palace officials and others, show that Owen only visited the models once during their construction for the famous 1853 New Years's Eve banquet held in the clay Iguanodon. Most tellingly, Hawkins is on record as stating that Owen "afforded no assistance" while the models were being built (Dawson 2016). Owen later attempted - as was his custom - to take greater credit for his role in the project, but his actual contributions seem restricted to input on some early clay maquettes constructed by Hawkins; a short, incomplete guide book about the sculptures (Owen 1854); some promotional duties; and allowing his name to be used to give the restorations scientific authenticity. We can thus draw two conclusions: 1) having been largely abandoned by his consultant, Hawkins' expertise and anatomical confidence were probably instrumental in seeing the models realised and completed on time; and 2) the sculptures are not - as so often stated - Owenian theory brought to life, but Hawkins' personal take on the works of Owen, Mantell and Cuvier. This probably explains why, as we discussed previously, many details of the models are at odds with Owen's ideas, as well as the nitpicking tone his 1854 sculpture guidebook.

File:Benjamin Waterhouse Hawkins Moas of Prehistoric New Zealand ...
Hawkins' 1870s take on the moa, produced for College of New Jersey (now Princeton University). The moa is not a creature we intimately associate with Hawkins, but its research history - where its form was predicted from a small amount of fossil material - underpinned Hawkins career as a palaeoartist. From Wikipedia.

It's in this largely unguided context that we have to view many of the details in Hawkins' Crystal Palace work as the product of a genuine palaeoart master, and especially so given their incorporation of an important but ultimately flawed 19th century way of interpreting extinct animals. Hawkins was a student of the then-in vogue Cuvierian philosophy of anatomical correlation: that is, the idea that whole animals could be reconstructed with some degree of precision from very few anatomical remains. This concept, explored in depth by Gowan Dawson in his 2016 book Show Me the Bone, gained traction after successful demonstrations in the early 1800s by Cuvier and then, most famously, by Owen's seemingly miraculous reconstruction of the moa from a femoral fragment. Anatomical correlation became the principle philosophy guiding the earliest visualisations of all poorly known extinct creatures and became celebrated as a scientific marvel in popular and academic spheres alike. Although some scholars - including Mantell - developed misgivings about anatomical correlation by the time the Crystal Palace project began in August 1852, Hawkins relied on this technique to restore his dinosaurs as well as Dicynodon, Mosasaurus and Labyrinthodon, none of which were known from more than scraps of bone at the time. In lieu of a reliable consultant, such predictive restorations might have been disastrous, had they ever been achieved at all, but Hawkins' anatomical expertise instead saw the creation of genuinely lifelike, plausible-looking restorations. Moreover, as discussed in my previous posts, his restorations of Crystal Palace are often more precedent than they first appear, anticipating not only many genuine aspects of prehistoric animal anatomy but also demonstrating sound, logical palaeoart approaches that we can approve of today.

The buffalo-like shoulder hump of Hawkins' Megalosaurus has a slightly complex backstory. Sometimes logically assumed to reference the tall-spined vertebrae of Altispinax dunkeri (a species once part of the Megalosaurus taxonomic complex), it was actually a Hawkinsian speculation that large-skulled dinosaurs needed vast neck and shoulders musculature to support their heads. Initial interpretations of Altispinax seemingly proved Hawkins right, but, of course, we've since realised that the Altispinax vertebrae belong much further down the body, possibly representing a Concavenator-like sail.

Using the dinosaur sculptures as an example of this, it's evident that Hawkins was anticipating the sort of anatomy that giant, somewhat mammal-like giant reptiles might have had to support their vast frames, and was moving well beyond the brief lifestyle and functional commentary provided by Owen and Mantell. He predicted, for instance, that the large heads of dinosaurs would need additional support from their shoulder regions, leading to the portrayal of both Megalosaurus and Iguanodon with shoulder humps of varying size (this is especially obvious in the Megalosaurus, but a shoulder prominence is also discernable in the reclined Iguanodon, while the shoulders of the standing animal bulge into a voluminous, rhino-like neck). This was an entirely sensible prediction that mirrors how we approach reconstructing fossil animals today: when we see robust and large osteological features, we assume they had correspondingly developed soft tissues for support and motion. In this respect Hawkins was ahead of some later artists who essentially ignored aspects of functional morphology and biomechanics in their work, leading to emaciated, peculiarly proportioned animals, such as the emaciated Stout and Kish dinosaur reconstructions from the 1970s and 1980s. Moreover, Hawkins was proved correct in assuming dinosaurs with large heads needed additional neck support, once (as it would turn out, erroneously) during his lifetime when Owen referred the tall-spined dorsal vertebrae of Altispinax dunkeri to the shoulder of Megalosaurus, specifically commenting on their use to support the head (Owen 1856) and, later, by the genuine shoulder and skull anatomy of ceratopsids and certain theropods, which bear augmented anatomy related to skull support.

Hawkins also predicted that dinosaur torsos may differ markedly depending on diet and habits, moreso than they do in conventional reptiles. The herbivorous Iguanodon has a vast belly that spreads in the reclined sculpture, simulating the weight of a large animal resting on a voluminous gut. In contrast, the Megalosaurus has a taut, narrow torso without a significant gut region, more in line with that of a carnivore. At this point in history virtually nothing concrete was known about dinosaur toro dimensions, so these reconstruction choices - also borne out by later discoveries - were sensible predictions of the functional properties of dinosaur guts. A third example concerns his dinosaur musculature, which was modelled after a mammalian, rather than lizard-like fashion. Though mammals are not the best reference for dinosaur myology, Hawkins was right in assuming that large, powerful animals needed large, powerful limbs, and realised that mammals were a more appropriate model than modern reptiles. In this regard, Hawkins was more insightful than some of his successors, who would restore dinosaurs with skinny, lizard-like limbs ill-suited to their (by then well-known and obvious) limb girdles (see artwork by Cope, Knight, Smit etc.). He also went so far as to add features in his art that we associate with especially thorough extinct animal reconstructions today, such as skin webs linking limbs and body, tissue deformation, pose-based muscle bulges and so on.

It's been said before, but it's worth saying again: the reclining Crystal Palace Iguanodon is an amazing work of animal reconstruction. This 30-tonne concrete and brick model captures minutiae of superficial musculature and other nuanced features of anatomy. The world had not seen anything like this in 1854, and even today it knocks the stuffing out of the detailing seen on many dinosaur sculptures.

These points demonstrate how excellent Hawkins was at solving anatomical 'problems' thrown up by fossil remains with an appropriate and sensible corresponding morphology. They belie an experience and instinct with anatomy and biological functionality that remains essential to high-quality palaeoart today, and this was undoubtedly a major factor in how these largely speculative creatures look so realistic. Hawkins' predictions were not always correct of course, but his decisions were logical given the material available to him - the best we can aim for in any palaeoartwork, regardless of its vintage. Needless to say, plenty of people were amazed by his work, and both its popularity and seeming scientific authenticity saw it referenced and replicated by artists for decades to come.

Because Hawkins is mainly known for the Crystal Palace models few people realise that he enjoyed a significant post-Crystal Palace career revolving around reconstructing extinct life. His skills were sought out by various bodies for use in museums, posters and other artworks. During lengthy visits to the United States, he almost re-realised the grandeur of the Crystal Palace models (first in Central Park, later the Smithsonian), but circumstances were never on his side and these projects never materialised or - in the case of the Central Park Paleozoic museum - were sabotaged. Hawkins continued to produce palaeoart until at least the 1870s where he completed what could be considered as his second most substantial set of palaeoartworks: 17 paintings of different geological periods for the College of New Jersey (now Princeton University). These artworks capture the evolution of Hawkins' personal ideas on prehistoric life and are also worth tracking down as exemplary takes on prehistory from the late 19th century. Among the most interesting is the 1877 painting Jurassic Life of Europe where Hawkins revisited his Crystal Palace subjects for (so far as I know) the final time. Hawkins got a lot of professional mileage from the Crystal Palace project and many of his subsequent artworks referenced his sculptures in detail, down to their posing, but by 1877 he must have realised that many of his reconstruction choices were no longer tenable. Both the now-hornless Iguanodon and Megalosaurus bear visibly short forelimbs, long, bird-like hindlimbs, and relatively slender necks with somewhat smaller, less bulky heads. Although still quadrupedal, these restorations indicate the influence of new dinosaur discoveries on Hawkins' classic dinosaur interpretations. This 1877 work might be criticised for his animals not having attained the status of true bipeds, as Hawkins realised was appropriate for both Hadrosaurus and "Laelaps", but consider that Megalosaurus was still considered to have powerful, bulky shoulders at this time (a holdover from Owen's interpretations) and that, while Iguanodon was considered by Mantell (1848) to have relatively gracile limbs capable of non-supportive functions (e.g. grasping vegetation), an explicit case for Iguanodon bipedality had not yet been made - the famous Bernissart Iguanodon skeletons would not be found until a year after Hawkins completed his painting, in 1878. Jurassic Life of Europe is surely among the last palaeoartworks where overtly Mantellian and Owenian interpretations of dinosaurs could be said to have some validity.

Hawkins' 1877 painting Jurassic Life of Europe, produced for the College of New Jersey (from Princeton University Art Museum).

Jurassic Life of Europe and its sister paintings demonstrate Hawkins' adaptability and humility as a palaeoartist. Much of Hawkins fame came from the Crystal Palace Company promoting their dinosaurs as authentic, wholly accurate creations, so their adaption to fit more modern ideals was an admission that his most revered work was not without some significant errors, and that the predictive principle at the heart of the geological court was not infallible. Concerns that the Crystal Palace sculptures were flawed were not new at this point (Dawson 2016) but, by the late 19th century, they were subject to increasing vicious comments from detractors. Othniel Marsh, speaking in 1895, remarked
The dinosaurs seem... to have suffered much from both their enemies and their friends. Many of them were destroyed and dismembered long ago by their natural enemies, but, more recently, their friends have done them a further injustice by putting together their scattered remains, and restoring them to supposed lifelike forms... So far as I can judge, there is nothing like unto them in the heavens, or on the earth, or in the waters under the earth. We now know from good evidence that both Megalosaurus and Iguanodon were bipedal, and to represent them as creeping, expect in their extreme youth, would be almost as incongruous as to do this by the genus Homo.
Othniel C. Marsh, 1895 (quoted in Desmond 1976)

Marsh and other late 19th century critics were, of course, not being entirely fair to Hawkins by comparing his work to their contemporary knowledge of prehistory, and not judging it by the standards of decades prior. Perhaps, for those living through the rapid paleontological advancements of the late 19th century, Hawkins' attempts to reconstruct fossil animals from fragmentary remains, and the genuine belief that they were credible takes on their true form, seemed premature, arrogant and foolish. There may be some truth to this: the 19th century acceptance of anatomical correlation is surely a major case of Dunning-Kruger effect, where scholars had yet to be humbled by the bewildering anatomical diversity present in Deep Time and believed their mastery of modern natural history equipped them to make bold, confident predictions about the past. But it's surely also the case that, while the science of his time was flawed, Hawkins' work was as true to that science as could realistically be expected. Like the kids say, don't hate the player.

Mark Witton on Twitter: "Yes, they seem to linger on in mainland ...
The final extinction of Hawkinsian dinosaurs: the frontispiece to Camille Flammarion's Le monde avant la création de l'homme& (1886). In this (possibly) last use of Hawkins' restorations in a non-historic, non-ironic context, a Crystal Palace Iguanodon (right) meets an early interpretation of Stegosaurus (left) - a representation of New World dinosaurs coming into focus. By this time, new European fossils had already made Hawkins' reconstructions long in the tooth, and American specimens were showing how wide of the mark his predictions were.

Hawkins' death in 1894 occurred without note, despite his significant contributions to science, public outreach, and education. Thankfully, his legacy has not gone neglected and increasing research into his life and work has seen his stock as an artist and intellect grow considerably, at least among academics. But it remains the case that showcasing Hawkins' work to even palaeoart fans sees it quickly dismissed, as by Marsh, as inaccurate and thus now worthless takes on prehistory. It's surely the case, however, that drawing attention to the inaccuracies of Hawkins' palaeoart is the most superficial and least interesting observation one can make of it, and defies the obvious fact that palaeoart of any kind must be evaluated in an appropriate historic context. Of course Hawkins' work is inaccurate, and pointing it out does not make one look smart or insightful. Rather, look closely and we see the work of someone who, with no prior experience in prehistoric animal restoration, took early palaeaort into a different league of quality and popularity; pioneered principles of soft-tissue restoration and attention to detail that we consider essential today; and demonstrated that even the most unusual extinct animals, with sufficient knowledge of living species, could be restored to plausible, life-like forms. The outdated science underlying Hawkins' work may make not it the most accessible of vintage palaeoart, but it's quality, the knowledge that informed it, and historic significance leaves little doubt that Hawkins was a true master of the genre, and worthy of deeper study and appreciation by palaeoart enthusiasts.


In June 2020 the face of the Crystal Palace Megalosaurus, surely one of the most iconic works of Victorian palaeoart on the planet, was significantly damaged. The survival of these works, and Hawkins' legacy, is a continued fight against the elements and the thoughtless people who routinely clamber over them. Image from the Friends of Crystal Palace Dinosaurs website.

Having just written about Hawkins at length, it would be remiss not to mention the terrible damage that recently occurred to the iconic Crystal Palace Megalosaurus, where the front of the face was broken off through suspected vandalism. The Friends of Crystal Palace Dinosaurs charity is now working to repair the damage with Historic England, Bromley Council and specialist conservators, as well as to address issues with the security of the island that houses most of Hawkins' sculptures. Note that this incident has nothing to do with the new bridge to the island, which has not yet been installed and, in any case, will be a rotating structure inaccessible to the public most of the time. Instead, the chief security issue is the low water level around the island, making it easy to access. As you can imagine, fixing this damage and ensuring the security of the Crystal Palace models is an expensive undertaking, so please support the charity if you can. If you really think Hawkins' work is worth preserving long-term, you can set up a recurring monthly donation - a small amount each month can quickly build into a substantial contribution, and gives the FCPD more reliable income over time.


  • Beaver, P. (1986). The crystal palace: A portrait of Victorian enterprise. Phillimore & Company.
  • Bramwell, V., & Peck, R. M. (2008). All in the bones: a biography of Benjamin Waterhouse Hawkins. Academy of Natural Sciences.
  • Dawson, G. (2016). Show me the bone: Reconstructing prehistoric monsters in nineteenth-century britain and America. University of Chicago Press.
  • Desmond, A. J. (1976). The hot-blooded dinosaurs: a revolution in palaeontology. Dial Press.
  • Mantell, G. A. (1848). XIII. On the structure of the jaws and teeth of the Iguanodon. Philosophical Transactions of the Royal Society of London, (138), 183-202.
  • McCarthy, S., & Gilbert, M. (1994). The Crystal Palace dinosaurs: The story of the world's first prehistoric sculptures. Crystal Palace Foundation.
  • Owen, R. (1854). Geology and inhabitants of the ancient world (Vol. 8). Crystal palace library.
  • Owen, R. (1856). The fossil Reptilia of the Wealden Formations. Part III, Megalosaurus bucklandi. Palaeontographical Society. Monographs, 9, 1-26.
  • Owen, R. (1894). The Life of Richard Owen by His Grandson the Rev. Richard Owen,... with the Scientific Portions: An Essay on Owen's Position in Anatomical Science by the Right. John Murray.
  • Phillips, S. (1854). Guide to the Crystal Palace and park. Crystal Palace Library.
  • 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.