About twice a week, Marina Elliott takes a trip back in time. On those days, the Canadian anthropologist is out of bed by 5 a.m. and on the road by 6, driving to the Rising Star cave complex about an hour northwest of Johannesburg.
It's winter in South Africa and temperatures have lately been hovering around the freezing point when Dr. Elliott and her support team gather at dawn. But as they head underground, the air is a constant 18 degrees C and humid.
So begins another workday at one of the most spectacular fossil finds in a generation, where specimen after specimen of a previously unknown human ancestor is emerging from the gloom.
As the head of field operations, Dr. Elliott has become adept at navigating the dark and challenging series of rocky fissures that can barely accommodate her slender frame. Once she's in, she may spend up to seven hours crouching on her hands and knees, unearthing clues to an anthropological mystery.
"I do find it quite meditative," Dr. Elliott said. "Particularly when it's just me working in there."
Even while her assistants rotate out of the cave on hourly shifts, Dr. Elliott is never truly alone.
All around her, buried in soft soil in long-hidden chambers, are traces of the mysterious congregation of individuals who died before our species appeared, but who may ultimately help us understand the origins of what it means to be human.
"It's kind of humbling," said Dr. Elliott, who after three years of working at the site can claim to have excavated more of our ancient predecessors than anyone else on the planet.
The remains were first spotted in 2013, when recreational cavers came upon an isolated and difficult-to-access chamber in a well-known cave system.
Over the next several months, Dr. Elliott was part of an all-female team of small-bodied researchers hired by lead anthropologist Lee Berger of the University of the Witwatersrand to squeeze their way into the chamber and carefully excavate hundreds of individual bones and teeth.
Their findings, revealed last September, are the first to document what appears to be a previously unknown species of hominin, dubbed Homo naledi (after the word "star" in Sesotho), and a cousin to our own distant ancestors.
Now, with more than 1,700 fragments of bones and teeth from at least 15 individuals ranging in maturity from infants to adults and with much more still to come, the cache is offering a rare and surprising glimpse of an early member of the Homo clan.
By any measure, it is a puzzling hybrid, with a small skull that held a tiny ape-like brain but with feet well adapted to upright walking and hands that look as though made for tools.
But even more exciting – and controversial – is the hypothesis the team has put forward to account for how the skeletons got there.
With no evidence of an alternate route into the chamber, no stone tools, no signs of daily living and no bones of other animals present, the researchers reason that the hominins may have been deliberately interring their dead.
If so, they have left evidence of a social ritual that predates, by hundreds of thousands if not millions of years, anything similar from our own direct ancestors.
The notion has been greeted with skepticism, "given our current understanding of how brain size relates to sophistication of social behaviour," said Mark Collard, an anthropologist at Simon Fraser University in Burnaby, B.C., who has studied the findings.
But despite his own misgivings, Dr. Collard agrees that the Rising Star team has made a "decent case" that could ultimately shift scientists' ideas about how far back in time what we think of as the human spark first began to glow.
Time is the operative word when it comes to the mystery of Homo naledi since one of the things Dr. Elliott and her colleagues did not provide with their initial results was a well-determined age for the fossils. Mana Dembo, a postdoctoral researcher who works with Dr. Collard, has tried to remedy that by creating a family tree for the species based on a mathematical comparison of skull measurements.
Dr. Dembo made the measurements in 2014, when she was one of 30 scientists invited to South Africa to examine the Home naledi fossils up close.
"There was an electric feeling in the room," said Dr. Dembo of the intense four-week period when so many researchers from around the world were working side by side on the new finds. "It was an exciting thing to be a part of."
Her results, published in the current edition of the Journal of Human Evolution, have added to the growing discussion about Homo naledi.
They suggest the species is indeed a member of the Homo group, but one that branched away after larger hominins had already wandered off into Asia. The new analysis also suggests Homo naledi may be just under a million years old.
This is oddly recent for a human relative with such a small skull. But that number could be overturned in a few months when the results from a geochemical analysis are expected to finally provide a direct age for the bones.
Even that may not resolve the fundamental mystery of Homo naledi. While a younger age raises questions about why such a species emerged when it did, an older one would make the idea of a deliberate burial even harder to accept.
"It could be telling us that our ideas about how much brain you need to be complex are not right," said Debi Bolter, a professor of anthropology at Modesto Junior College in California. "It starts you thinking in a whole different way about the evolution of cognition."
Meanwhile, Dr. Elliott said, a slew of new findings from another location in the cave is likely to generate more excitement in the coming months. As her unique journey through time continues, it shows no signs of letting up or of giving her and her colleagues any reason to ask "Are we there yet?"