The Dragon Man Skull from Harbin: China's Mystery Human

The fossil is a cranium, the skull without its lower jaw, and it is unusually complete for a human relative this old. Most of what we know about ancient human groups in Asia comes from teeth, jaw pieces and skull fragments, so a braincase and face in one piece is rare. Scientists call it the Harbin cranium, after the city where it was reportedly found. The headlines call it Dragon Man, after the species name.

It also has a museum number. The skull is catalogued as HBSM2018-000018(A) and belongs to the Geoscience Museum of Hebei GEO University in Shijiazhuang, China, according to the 2021 species paper by Ni and colleagues. The "2018" in that number is the year the skull reached the university, which is its own story.

The way the skull came to light is not the usual archaeology story of trowels and trenches. As later accounts tell it, the skull came out of the ground in 1933 near Harbin during the building of a bridge, while the region was under Japanese occupation. The man who came across it hid it in an abandoned well. It stayed there for decades. In 2018 the skull was recovered and donated to Hebei GEO University, a sequence described in Archaeology Magazine and in a commentary in Nature Ecology & Evolution.

The 2021 species paper records the donation to the university. The 1933 find and the well come from later recollections, so they are best read as the reported history of the fossil. Because nobody excavated it and recorded the layer it lay in, its exact original location is uncertain. That missing record matters for the next question everyone asks.

The skull is at least 146,000 years old. That figure comes from direct uranium-series dating of the fossil itself, a method that measures the slow decay of uranium taken up by bone, published by Shao and colleagues in The Innovation. It is a minimum age. The true age could be older.

The same team tried to work out which deposits the skull had come from. They compared the chemistry of the fossil with sediments around Harbin and linked it to lake deposits, which geologists call lacustrine sediments, dated to roughly 138,000 to 309,000 years ago. That puts the skull in the late Middle Pleistocene, a stretch of the Ice Age when several kinds of humans lived across Europe, Africa and Asia.

Why not a single number? The dating paper explains that the uncertain find spot makes a narrower age difficult. With an excavated fossil, scientists can date the layers above and below it. With a fossil that spent decades in a well, they date the bone and match its chemistry as closely as they can. If you want a feel for how these methods work on any old bone, our piece on how old bones are dated walks through the basics.

In 2021, Qiang Ji, Xijun Ni and colleagues published a study proposing the Harbin cranium as the holotype, the reference specimen that defines a species, of *Homo longi*. The name translates as "Dragon Man," and it stuck in the press at once.

Their argument rested on anatomy. The team found that the skull showed an unusual combination of features, and their analysis placed it as a distinct Middle Pleistocene lineage. In their family tree, that lineage sat close to the line leading to modern humans and to the Neanderthal line, while belonging to neither. In plain terms, they saw a third branch of large-brained humans living in East Asia at the same time as early Neanderthals in Europe.

That was the state of things for four years. A new name in human evolution always draws discussion, and this one did, partly because the skull had no excavation record. The next step came from inside the skull itself.

On June 18, 2025, two papers led by the geneticist Qiaomei Fu and colleagues tied the Harbin individual to the Denisovans.

The first, in *Science*, looked at ancient proteins. The team recovered 95 endogenous proteins, meaning proteins that came from the person's own body and from no later contamination. Among them were three amino-acid variants known from Denisovans. When the researchers built a protein-based family tree, Harbin grouped with Denisova 3, the Denisovan whose high-coverage genome came from Denisova Cave.

The second paper, in *Cell*, went after DNA. The team could recover no DNA from the tooth or from the petrous bone, the dense part of the skull near the ear that often keeps DNA best. They turned instead to dental calculus, the hardened plaque on the teeth, and recovered mitochondrial DNA from it. That DNA fell within known Denisovan mitochondrial variation and was related to an early Denisovan branch previously identified in southern Siberia, according to the PubMed record of the Cell study.

One limit belongs right beside those results. The team did not obtain a complete nuclear genome, the main body of DNA that carries most of a person's ancestry. Mitochondrial DNA passes down the maternal line only. Taken together, the two studies strongly support a Denisovan affiliation, which is how Nature's news coverage described the result while noting that some researchers were still debating the interpretation.

Denisovans were an archaic human population first identified through DNA, which is unusual in this field. In 2010, a *Nature* study led by Svante Pääbo, with David Reich and colleagues, showed that a finger bone and molars from Denisova Cave in the Altai Mountains of southern Siberia, Russia, belonged to a previously unknown human group, distinct from both modern humans and Neanderthals. For years, that was most of what anyone had: small bones, a few teeth and a great deal of genetic data.

The cave itself had many tenants over time. The Smithsonian Human Origins Program summarizes sediment DNA evidence placing Denisovans there as early as about 250,000 years ago, with Neanderthals appearing later and modern humans arriving around 45,000 years ago. The Smithsonian's page on ancient DNA and Neanderthals adds that Denisovans and Neanderthals interbred, and that Denisovan ancestry survives in some living people. If the Neanderthal side of this family interests you, our story on the Neanderthal skull picks it up from there.

Before Harbin, scientists could read Denisovan DNA in detail but had almost no idea what a Denisovan head looked like. A nearly complete cranium changes that. It gives the clearest direct evidence yet for Denisovan skull anatomy, which means other puzzling fossils from Asia can now be compared with a known shape.

It also changes the map. Denisova Cave sits in southern Siberia, and Harbin lies far to the east in northeastern China. A Denisovan-related person living near Harbin more than 146,000 years ago shows that these people ranged across a large area of Asia. The mitochondrial link to an early Siberian branch, reported in the *Cell* paper, points the same way.

For anyone curious how much a single bone can say about a life, this is a good case. Our piece on what bones tell archaeologists covers the everyday version of the same questions: age, health and diet, read from what survives.

So, which is it? Both names are in use, for different reasons.

The 2021 name *Homo longi* remains part of the scientific literature, proposed by Ji, Ni and colleagues on the strength of the skull's anatomy. Some researchers argue that Harbin should instead be treated as a Denisovan, without a separate species of its own. The 2025 studies themselves describe the individual as Denisovan or Denisovan-related, and the Nature Ecology & Evolution commentary "Face to face with a Denisovan" summarized the protein result as grouping Harbin with Denisova 3.

What everyone agrees on is the core: a skull more than 146,000 years old, from northeastern China, belonging to a person of an archaic human population that genetics had known for fifteen years before it had a face. What remains open is the name on the label.

The skull is held by the Geoscience Museum of Hebei GEO University in Shijiazhuang, China, under the number HBSM2018-000018(A). The university received it in 2018. If you hope to see it, contact the museum directly about display and access, since a research specimen of this importance is not always on public view.

Harbin is one of several famous human fossils from China. The Zhoukoudian cave near Beijing gave the world Peking Man, and Spain's Sima de los Huesos holds the world's largest collection of Middle Pleistocene human fossils. For a different branch altogether, the small-bodied people of Flores in Indonesia are told in our story on Homo floresiensis.