Friday 26 March 2021

Dinosaur fossils and Chinese dragons: ancient association or modern wishful thinking?

Dinosaurs and dragons: is there a better fit between fossils and folklore? Common wisdom is that Chinese dinosaur species, perhaps including Yutyrannus huali, became associated with local dragon lore when ancient people used their bones for medicinal purposes, but how well-evidenced is this popular idea?

Geomythology is a discipline that most of us are familiar with even if we've not heard of this term: the study of possible associations between real geological phenomena and myths and legends. The idea that certain fossils were somehow involved in the creation or development of mythical creatures is a subject we've discussed several times at this blog, including the purported fossil associations of griffins, cyclopes, giants and unicorns. Although proverbial kernels of truth underlie some of these proposals, many examples - including famous claims about Protoceratops and griffins, and elephant skulls and cyclopes, are actually nowhere near as well-supported as their popularity implies. A lot of geomythological hypotheses persist primarily because of uncritical retellings and a lack of sceptical examination.

Today, we're going to take aim at another claimed geomythological association: the proposal that fossils of Mesozoic dinosaurs were mistaken for dragon bones in ancient and historic China. This idea only has a slim footing in academic literature but is very popular, being brought up just about whenever the early history of dinosaur discoveries is raised and being bolstered by mentions in reputable outlets like Smithsonian Magazine and BBC Earth. A common issue with arguments in favour of this hypothesis, whether written in a technical paper or posted to Youtube, is vagueness: beyond pointing out the antiquity of fossil discoveries in China and the ancient origin of dragon myths, little more is said. But we don't need to be vague: we actually have extensive documentation of historic 'dragon bones' in the form of numerous historical texts, illustrations, and records of interaction between 19th and 20th-century scientists with the dragon bone industry. Collectively, these create excellent insights into what 'dragon bones' really were and what fossil animals they belonged to. So how does the dinosaur-dragon association stand up against this evidence?

Cropped dragon from the 17th-century painting Miracle of the Dragon. Chinese dragon lore shares many features with dragon myths in other Asian cultures, so they likely have a principally folkloric, not palaeontological, origin. You can see the entire and spectacular Miracle of the Dragon painting at the Cleveland Museum of Art website.

Dragon mythology has existed in China since at least the Shang Dynasty (1520 - 1030 BCE) - in other words, for as long as we've had decent records. It likely originated among the myths of other cultures rather than being a novel creation based on fossils: Hogarth (1979) suggests that the older Buddist naga dragon myth has so many cultural similarities with Chinese dragon lore that the two stories must be linked. This said, ancient texts also show that Chinese scholars knew of fossil bones thousands of years ago, and it's primarily these documents that are used to make the case for the dinosaur-dragon association (e.g. Needham 1959; Dong 1988; Sarjeant 1999; Spalding and Sarjeant 2012; Rieppel 2019). Dong (1988) and Rieppel (2019), for example, suggest that scholar Qu Chang may have been discussing dinosaur fossil beds in what is now Sichuan Province in c.350 CE. Rieppel (2019, p. 236) provides a direct translation:

"In Wucheng Countym there is a mountain called Somber Warrier Mountain, also called Three Corner Mountain, that has six bends and six rises. Dragon bones are taken from it. It is said that dragons flew up from these mountains, but when they found heaven's gates closed, they could not enter, and thus fell dead in that place, and later sank into the earth. That is why one can dig out dragon bones."

Dong (1988) further notes that Jurassic strata in this area may be the real-life source of Chang's 'dragon bones', which seems a reasonable prima facie argument.

Another ancient account, cited in some popular articles, recalls the recovery of a 'dragon bone' during the construction of a canal in 120 BCE. This discovery was so significant that the waterway was named "Dragons Head Canal" after its discovery (McCormick and Parascandola 1981). Several other records of such antiquity are known and make a fine case for ancient Chinese cultures interacting with fossils, but they each have the same problem: none are detailed enough to demonstrate that they pertain specifically to dinosaur fossils (Delair and Sargeant 1975; Buffetaut 1987; Sarjeant 1999). Dinosaurs don't just occur anywhere and everywhere fossils are found, so without descriptions or illustrations of the bones, or specific geographic information we can use to track down these ancient sites, we can only guess what these first mentioned 'dragon bones' were. Qu Chang's discussion of the Sichuan Province is interesting because - as noted by Dong - this is a rich ground of dinosaur fossils, but the same province also bears plenty of non-dinosaur fossil sites, including fossil-rich Pleistocene gravels (Delair and Sarjeant 2002) and sinkhole deposits (Buffetaut 1987). Without more data, we simply cannot say what fossils these ancient documents pertain to, and the involvement of dinosaurs is equivocal.

An honest-to-goodness collection of 'dragon bone' teeth obtained from drugstores during the late 19th century, as figured by McCormick and Parascandola (1981). No doubt some readers are already identifying some of these elements: many are quite obvious to anyone generally familiar with vertebrate anatomy.

And that, really, is that. If you're hoping that our tale now turns to more concrete evidence of the Chinese dragon-dinosaur link, you're out of luck: these loose, arm-wavy arguments are the best we have to offer. The crux of our problem is that 'dragon bone' in ancient China was a catch-all term for a variety of fossil types that were initially only categorised as being bone, tooth or horn. By the 5th century, colour and texture were also factored to determine dragon gender and bone quality (McCormick and Parascandola 1981), but something resembling modern taxonomy was never applied. Economic value and pharmaceutical use were the main interest in 'dragon bones' so it was along these lines that they were categorised: details of anatomy and morphology were basically unimportant. We know this thanks to medical documents dating back as far as the third century BCE. Dragon remedies were considered powerful agents that could treat dysentery-like symptoms, heart, liver and kidney conditions, forgetfulness, anxiety, epilepsy and even disembowelment (McCormick and Parascandola 1981; Buffetaut 1987). You may have eaten dragon bone raw, boiled it, fried it, cooked it in rice wine, or else crushed it into powder to combine it with other ingredients. Ground bone sprinkled into tea eventually became the most common way of taking your dragon medicine (Hargarth 1979).

Alongside the medicinal use of dragon remains, the nature of the dragons themselves were also routinely discussed by Chinese physicians: principally, whether the dragons these bones were harvested from were alive or dead (some argued that dragons shed their bones like a snake sheds its skin). Little interest was shown in the geographic and geological origins of the fossils themselves, with some exceptions, such as a medical text from 456-536 CE that outlined dragon bones being extracted from caves along rivers (McCormick and Parascandola 1981). Although still vague, this helps us in our investigation to some extent by implying that some dragon bones were not especially old. Fossils collected from caves tend to be thousands of years old, not millions, so Mesozoic dinosaurs were unlikely to be among the fossils referenced in this text.

Dragon bones were just one of several fossil commodities available in historic China. As early as the 12th-century Chinese scholars had correctly identified that animal remains were capable of becoming petrified (Wen-chung 1956) and specimens of fossil shellfish and fish were commercially available for decorative and pharmaceutical purposes (Needham 1959; Edwards 1967). Fossil fish were so sought after that a market in forged specimens was established (Buffetaut 1987) and dragon bones were also in extremely high demand. This might be one reason why their source localities were not widely recorded - suppliers may have been cagey about their suppliers to protect their commercial interests. Contrary, perhaps, to Western expectations, dragon bones were big business for the Chinese pharmaceutical industry and were given premium prices (Hargarth 1978; Duffin 2017). They were sold in huge quantities from drugstores around the country and even exported en masse to neighbouring regions. Buffetaut (1987) mentions an 1885 manifest detailing some 20 tonnes of dragon bones passing through a Chinese port in a single year, and even in the 1980s several tonnes of 'dragon bones' were being exported internationally. At one time Chinese dragon bones were available in Chinese communities across Japan, Tibet, the East Indies, the Philippines, and the USA (Koenigswald 1952; McCormick and Parascandola 1981). And yes, you can still walk into some Chinese pharmacies and buy dragon bone today.

These qualities also have bearing on our investigation of dragon bone origins. Their widespread availability and mass-exportation tell us that they were not difficult fossils to find, extract or transport. Rather, they must have been superabundant, robust fossils that could be easily dug and cleaned using the basic technology of historic rural China (Wang et al. 2020). If so, this conclusion also doesn't help the dragon-dinosaur hypothesis. Mesozoic dinosaur bones can be abundant but, on account of their age and the lithification of the sediments they occur in, they are often challenging to excavate and can be very fragile. Most of us will be familiar with the careful excavation, jacketing and preparation processes that accompany modern dinosaur discoveries - it's unlikely this sort of time-consuming care was practised thousands or hundreds of years ago in the Chinese countryside.

1505 illustration of 'dragon bones' from the Bencao pinhui jingyao, more recently featured by Duffin (2013). This illustration is significant for showing us what constituted a 'dragon bone' some 500 years ago, several centuries before scientists saw and began documenting these remains.

It's through China's traditional medicine industry that 19th-century European scholars first documented China's vertebrate fossils. It was this process that transferred dragon bone mythology into scientific reality. Europeans visiting China bought and sent 'dragon bone' specimens to their home institutions at a grand scale, cumulating in thousands of bones and teeth being sent to Western museums (Buffetaut 1987). This material was exciting to 19th-century scientists and attracted the attention of premiere scholars, including Richard Owen, who provided the first scientific description of Chinese vertebrate fossils in 1858 (see Owen 1870) and returned to this topic several times in his career. What Owen and others discovered was that Chinese 'dragon bones' was a catch-all term for fossil debris: shells, small bones, bone fragments, teeth and horns. Some bones had evidently been more complete when found but were broken to be sold, especially if it meant relatively valuable teeth could be freed from the less desirable jawbones (Koenigswald 1952). Some bones retained traces of their source sediment but exact locality data were unrecorded.

Despite the lack of provenance and quality of the fossils, European and (later) American palaeontologists were able to identify numerous species among the dragon bone collections. The Devonian brachiopod Spirifer verneuili was among the most common elements sent to European collections - a fossil also thought to have curative properties in traditional Chinese medicine - but vertebrate material was also present in abundance: specifically, Neogene and Pleistocene mammals. By far the most common vertebrate elements were the teeth of the Miocene-Pleistocene equine Hipparion, of which thousands of examples were sent to Europe alone (Koenigswald 1952; Buffetaut et al. 1987). Many other species were present too: some 60 mammal species were identified in one large sample analysed in Munich (Koenigswald 1952). Over time, 'dragon bones' from across China were found to include fossil horses, giraffes, elephants, rhinos, tapirs, hyenas, deer (their antlers being 'dragon horns'), bears and hominids (Koenigswald 1952; McCormick and Parascandola 1981; Buffetaut 1987; Duffin 2008, 2013; Wang et al. 2020). Among the most exciting finds to come from the drugstore dragon bones were hominid remains, which helped catalyse interest in Asia as a source of human ancestry. The discovery of Peking Man can be directly related to 'dragon bones' and the first known fossils of everyone's favourite enigmatic giant fossil ape, Gigantopithecus, were purchased over a drugstore counter (Koenigswald 1952; Wen-chung 1956; McCormick and Parascandola 1981). That mammal bones were being sold as dragon remains for hundreds of years before European scientists became interested is documented by early 16th-century illustrations of 'dragon bones' which are unmistakably mammal teeth and deer antlers (above, Duffin 2013).

The anatomical reality of Hipparion, a mid-sized, three-toed Miocene-Pleistocene horse, might seem millions of miles away from that of a Chinese dragon, but the teeth of this genus constituted the overwhelming majority of 'dragon teeth' reviewed by 19th and 20th-century scientists. Strange as it seems, the fossils of this horse have far greater relevance to Chinese dragon mythology than any Mesozoic dinosaur.
Excited by the flow of fossils from China but frustrated by their quality and lack of geological context, Western explorers and palaeontologists eventually gained access to dragon bone sites through the same networks that supplied Chinese drugstores (Koenigswald 1952; Buffetaut 1987; Wang et al. 2020). Among these were riverside caves in Yunnan Province and sinkholes in Sichuan Province, both of which were richly stocked with Pleistocene mammal fossils. Given that both riverside caves and Sichuan Province were mentioned by early accounts of fossil discovery in China, accessing these sites goes some way to closing the loop in our dragon bone story, although whether they are the exact same sites referenced in these ancient writings will forever be unknown.

One of the most detailed accounts of a dragon bone locality was provided by Walter Granger in his visit to a series of Pleistocene sinkholes in Sichuan Province during the 1920s. These contained abundant, well-preserved Pleistocene mammals that were excavated by local farmers lowering themselves into pits and digging fossils from soft mud. Once hauled to the surface, the fossils were cleaned with water before being stacked in rough piles to dry (Buffetaut 1987). Granger specifically noted that the fossils were treated roughly, there being no need to worry about their condition given their pharmaceutical destiny (Koenigswald 1952). Granger's description of recent-ish fossils being extracted from loose sediments is exactly the sort of low-tech set-up we'd expect to be supplying the dragon bone industry, and similar sites were reported by other explorers and geologists. Surprisingly, there are still dragon bone sites waiting to be found: some fossils brought back from China in the 19th and 20th centuries have yet to be matched to their source rocks. This is an active area of research as many of those exported bones are now type specimens that should, ideally, have established provenance (Wang et al. 2020) - a cautionary tale par excellence for the importance of labelling your specimens!

But wait wait wait: this is an essay about dinosaurs and dragons, and I'm waxing on about mammals. It's here, facing centuries of excellent documentation about the reality of 'dragon bones', that the lack of a substantiated link between dinosaurs and Chinese dragons is especially obvious. To my knowledge, no dinosaur bones were identified among the dragon bone samples sent to Europe and North America during the 19th and early 20th centuries, no illustrations of obviously dinosaurian fossils are known among historic works, and no dinosaur localities were discovered through drugstore dragon bone networks. Instead, China's dinosaur sites were made known to science through traditional geological prospecting and amateur discoveries in the early 20th century (Buffetaut 1987) - no dragons required. In contrast, we know for a fact that Chinese dragons were associated with extinct mammals, many of which are relatively familiar types - horses, hyenas, bears and so on. It's strange to juxtapose these animals with the fantastic dragons of Chinese mythology, but that's what our evidence shows.

Part of the surprising pervasiveness of the Chinese dinosaur-dragon link is that the real history of dragon bones is already associated with exciting, popular fossil species, including Homo erectus and Gigantopithecus blacki. In this light, it's hard not to see the perpetuation of the speculated, unsupported dinosaur-dragon history as overriding the genuinely interesting and significant contributions that Chinese 'dragon bones' have made to palaeontological research.

Of course, if we're being true sceptics on this, we also have to point out some important caveats. Firstly, our historic record of dragon bones is good, but it does not cover the entirety of China's geographic, historic and paleontological resources: we cannot rule out any and all use of dinosaur bone as dragon remains in historic times. It would actually be surprising, given the appetite for dragon bone and the abundance of dinosaur material in China, if no-one found and interpreted a dinosaur or another type of fossil reptile in this way. So who knows - evidence of this may turn up in future.

Secondly, although the historic use of dinosaur bones in traditional medicine is not well-evidenced, a surprise discovery in 2007 revealed that a central Chinese village has been using dinosaur bones in medicine for at least a few decades. It differs from older historic accounts in that these villagers have been collecting their fossils locally and not through drugstores, and makes me wonder if other small communities exploited dinosaur fossils in the same way. Maybe our records and analyses are biased towards the history of the national, commercialised end of dragon bone use? It's an interesting idea, but is also currently an unknown.

But these points don't change the details outlined above. I wonder if some dinosaur fans will be disappointed by the lack of evidence for a dinosaur-dragon link, but I think it's a fascinating story with a terrific twist: the fantastic and iconic Chinese dragon is most closely associated, palaeontologically speaking, with familiar mammalian megafauna - including our own lineage. But beyond providing a fun historic narrative, this outcome also serves as an example of how geomythological ideas can be misguided by our own biases and wishful thinking. I suspect we inject dinosaurs into this story because we see their obvious similarity with ancient dragon depictions, but this assumption falls foul of historian's fallacy: the projection of contemporary knowledge onto people of the past. Opinions vary on how knowledgeable ancient people were of dinosaur anatomy (Lyons 2009) but we have no evidence that the appearance of Mesozoic dinosaurs was known to anyone before complete dinosaur skeletons were discovered, excavated, prepared and reassembled in late 19th century Europe. My suspicion is that ancient knowledge of dinosaurs was no greater than what could be gleaned from partially-exposed, weathered fossils or loose bones and teeth: the alternative, that people living thousands of years ago in China had - for example - concepts of what whole stegosaurs looked like is very far-fetched. Moreover, our better-evidenced examples of geomythology, which includes Chinese dragons, shows that there need not be any real resemblance between fossil species and the legendary animals they become associated with: an ammonite can be a snake, a mammoth can be a unicorn, and a fossil horse can be a dragon. These might seem ridiculous to us, but only because centuries of scientific insight makes interpreting fossils and ordering the natural world look so easy. Our modern learned interpretations of fossils were not necessarily shared with peoples living thousands or hundreds of years ago.

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!


  • Buffetaut, E. (1987). A short history of vertebrate paleontology. Croom Helm.
  • Delair, J. B., & Sarjeant, W. A. (1975). The earliest discoveries of dinosaurs. Isis, 66(1), 5-25.
  • Delair, J. B., & Sarjeant, W. A. (2002). The earliest discoveries of dinosaurs: the records re-examined. Proceedings of the Geologists' Association, 113(3), 185-197.
  • Dong, Z. (1988). Dinosaurs from China. British Museum of Natural History and China Ocean Press.
  • Duffin, C. J. (2008). Fossils and folklore. Ethical Record, 113(3), 17-21.
  • Duffin, C. J. (2013). Lithotherapeutical research sources from antiquity to the mid-eighteenth century. Geological Society, London, Special Publications, 375(1), 7-43.
  • Edwards, W. N. (1967). The early history of palaeontology (No. 658). Rudolph William Sabbott Natural History Books.
  • Hogarth, P. 1978. Dragons. Allen Lane.
  • von Koenigswald, G. H. R. (1952). Gigantopithecus blacki von Koenigswald, a giant fossil hominoid from the Pleistocene of southern China. Anthropological papers of the AMNH; v. 43, pt. 4.
  • McCormick, J. P., & Parascandola, J. (1981). Dragon bones and drugstores: the interaction of pharmacy and paleontology in the search for early man in China. Pharmacy in history, 23(2), 55-70.
  • Needham, J. 1959. Science and civilisation in China. Volume 3. Mathematics and the sciences of the heavens and the Earth. Cambridge University Press.
  • Owen, R. (1870). On fossil remains of mammals found in China. Quarterly Journal of the Geological Society, 26(1-2), 417-436.
  • Rieppel, L. (2019). Assembling the dinosaur: Fossil hunters, tycoons, and the making of a spectacle. Harvard University Press.
  • Sarjeant, W. A. 1999. The earliest discoveries. In: Farlow, J. O., & Brett-Surman, M. K. (Eds.). (1999). The complete dinosaur. Indiana University Press.
  • Spalding, D. A., Sarjeant, W. A., & Brett-Surman, M. K. (2012). Dinosaurs: the earliest discoveries. In: Brett-Surman, MK, Holtz jr., T. R., Farlow, James O. (Eds.): The Complete Dinosaur. Indiana University Press.
  • Wang, X., Jukar, A. M., Tseng, Z. J., & Li, Q. (2020). Dragon bones from the heavens: European explorations and early palaeontology in Zanda Basin of Tibet, retracing type locality of Qurliqnoria hundesiensis and Hipparion (Plesiohipparion) zandaense. Historical Biology, 1-12.
  • Wen-chung, P. (1956). New material on man's origins. East and West, 7(3), 268-270.