In 1890, Dutch palaeoanthropologist Eugene Dubois, exploring the dense rainforest that covers the Padang Highlands of Sumatra, entered a particular cave and found two human-looking fossilised teeth.
The cave was named Lida Ajer. Dubois made some sketches in a notebook, then took the teeth, along with some other fossils, including gibbon and orangutan teeth, back home.
More National News Videos
The discovery that rewrites Australia's history
Artefacts found in Kakadu national park show that Aboriginal people have lived in Australia for a minimum of 65,000 years, 18,000 years longer than the previous estimate.
And there they remained, largely unregarded. Dubois – who today has a research institute named after him – wasn't the most meticulous note-taker, and did not have today's hyper-accurate dating tools at his disposal. Thus, the teeth – an incisor and a molar – never featured in any serious investigation of settlement in South East Asia.
In 1948, they became, briefly, the focus of new interest, when Dutch palaeontologist Dirk Albert Hooijer took a close look at them and pronounced them definitely human.
And there things stood – until Australian scientist Dr Kira Westaway got curious.
Dr Westaway, senior lecturer in the Department of Environmental Science at Macquarie University in Sydney, specialises in dating fossils and artefacts linked to Homo sapiens and related hominids. And Hooijer's finding that the teeth were human piqued her interest.
"This project started back in 2008," Dr Westaway says. "You could say it's been a bit of a labour of love."
You could also say that's a considerable understatement, glossing over a nine-year slog during which Westaway recruited 23 co-authors based at universities in Australia, Indonesia, America and the Netherlands – two of whom are now deceased – and suffered two cruel rejections by science journal Nature before a successful third submission.
This month, the peer-reviewed journal published the paper and beneath its dry, technical prose lies a ripping yarn involving modern Australian science, treks through Indonesian rainforests, a dogged refusal to take no for an answer, and a fax machine.
Focus on new mystery
In fact, what the scientists have been able to find is 20,000 missing years.
When research was published in July definitively dating Aboriginal occupation of a site in Kakadu to 65,000 years ago – 18,000 years earlier than previously thought – the discovery threw a new mystery into sharp focus. How did they get there?
The dominant theory – known as the "recent single origin hypothesis" – holds that Homo sapiens migrated from Africa around 100,000 years ago and then spread through Asia, eventually reaching every continent except Antarctica.
The Kakadu findings, however, threw a significant chronological spanner into the works. People were living there, at the Madjedbebe rock shelter, 65,000 years back, yet no one had ever found incontestable evidence that humans reached the islands of south-east Asia, the only route to Australia, before 60,000 years ago.
It was Dr Westaway and her co-authors who have found the missing piece of the puzzle.
Several sites in Indonesia have produced fossils of Homo erectus (the first, appropriately enough, unearthed by Dubois in 1891) and, of course, in 2004, sensationally, researchers found Homo floresiensis, the "hobbit".
Could Hooijer's conclusions about the teeth found in 1890 be wrong? Could the Lida Ajer teeth belong to either of these species, rather than humans? Dr Westaway got hold of detailed scans and images from Holland and went over them with, um, a fine-toothed comb. She confirmed his conclusions.
Her next proposition was logical enough: dating the rock layer in which Dubois found the molar and incisor, as well as the orangutan teeth and (hopefully) others newly discovered at the same spot, would provide, for the first time, an exact date for when humans had lived in the cave.
There was just one problem: no one knew where it was. Undaunted, Dr Westaway and colleagues headed off.
"It was really full on," she laughed. "We went over there, and we thought it would be really easy. We thought, it was only 100 years ago, the locals would still have word-of-mouth and know which cave was Lida Ajer.
"But it was really difficult. We went to a number of sites and they weren't right. Then we had some weird stories about the cave being in the lowlands and used to house buffalo. That wasn't right, either.
"So we had a few days of just running around in circles and not getting anywhere."
Exasperated, Dr Westaway had an idea. She contacted a colleague, Dr Gerrit van den Bergh from Wollongong University, who in turn got in touch with the Eugene Dubois Foundation and asked for the original notebook. Pages from it were duly faxed to Sumatra. "It was a really small cross on a really big map," said Dr Westaway, "but we started looking more determinedly."
A couple of days later, they were successful. "It was my Eureka moment. As soon as I saw the entrance and the big calcite column inside I knew it was exactly the place Dubois had drawn in his notebook."
After that, it was all as simple as fieldwork up a mountain in a jungle can be. Fossils and rock samples were sourced, and later dated using several methods involving luminescence and isotope decay.
And thus, eventually, after several years, rejections and rewrites: the bombshell. The teeth were between 73,000 and 63,000 years old – more ancient by many millennia than all other human remains found in south-east Asia.
"We found the missing 20,000 years," said Dr Westaway.
And that wasn't all. Palaeoanthropologists have always assumed that ancient humans migrating into new territory stuck to the coast, where open land and food are plentiful. Whoever lost Dubois' teeth had been living in enclosed rainforest – where food is scarce and mostly at the top of tall trees. It is by far the earliest evidence for humans living in this type of environment.
And – the cherry on the cake – the upper range of Dr Westaway's dating leaves plenty of time for continued migrations to reach the Northern Territory.
Kakadu study leader Chris Clarkson is delighted. "I think the new dates for the modern human teeth lend wonderful support to the new dates from Madjedbebe for the colonisation of Australia around 65,000 years ago," he said.
"It is quite astounding that these two new independent findings have come out within a month of each other and are so mutually supporting."