Drimolen Homo erectus skull adds an adult at 2 million years

Key Takeaways

The fragment	Jesse M. Martin and colleagues tentatively assign skull piece DNH 127 to Homo erectus. It is part of the right side and back of an adult skull from Drimolen Main Quarry in South Africa.
The date	The paper places that evidence at about 1.98 million years ago, inside deposits dated to about 2.04 million to 1.95 million years ago. A La Trobe University account calls the fragment roughly 200,000 years older than adult fossils from Dmanisi, Georgia.
The neighbors	Of the 18 fossils in the study, 14 are assigned to Paranthropus robustus and three cannot be placed in a genus. Martin said Paranthropus outnumbered Homo erectus at the cave by more than 10 to one. The paper does not give that ratio.
Still open	The authors say no well-preserved, fairly complete adult Homo erectus cranium is known from Drimolen or the nearby caves they compared. DNH 127 came from a loose block, so they call its age likely rather than directly measured.

On Sept. 25, 2026, researchers published an adult Drimolen Homo erectus fossil from a cave northeast of Johannesburg. In a paper in Annals of Human Biology, Jesse M. Martin of La Trobe University and colleagues tentatively assign skull fragment DNH 127 to Homo erectus. They call that evidence about 1.98 million years old.

What the Drimolen Homo erectus fragment shows

The paper says DNH 127 is not a full skull. It is a partial right temporal bone, a partial right parietal bone, and a small piece of occipital bone, meeting behind the ear. A hominin is a member of the group that includes modern humans and their extinct relatives. Two traits are incompatible with Paranthropus, the heavy-jawed hominin common in this cave, the authors write.

The mastoid, the bump behind the ear, is moderate in size, the paper says. The seam between the side bones is straight, which the paper calls characteristic of Homo erectus. The shape most closely resembles skull D2700 from Dmanisi, Georgia, they write. "Based on this mosaic of traits, we tentatively assign DNH 127 to Homo erectus sensu lato," the paper says. Sensu lato means in a broad sense.

A catalogue entry dates the piece to 2008. It came from a loose block beside the DNH 7 pinnacle, and the authors think that block broke off the pinnacle. They treat it as likely the same age as other hominins there, not as a fossil dated where it lay.

Co-authors are A. B. Leece of La Trobe, Stephanie Baker and Andy I. R. Herries of the University of Johannesburg, and David S. Strait of Washington University in St. Louis. Herries is also at La Trobe and Witwatersrand, and Strait is also at Johannesburg. The paper sets out to describe 18 skull and jaw pieces and to test whether Homo erectus was at the quarry about 2 million years ago, and whether Drimolen's Paranthropus differs from later fossils nearby.

A 2020 date, and a later comparison

The new paper says the earliest unequivocal Homo erectus at Drimolen is still a child's skull. In Science on April 3, 2020, Herries and colleagues dated the main layer to about 2.04 million to 1.95 million years ago. They said cranium DNH 134 was deposited at about 2.04 million years ago and extended the species' known range by about 200,000 to 150,000 years.

The new paper treats that child skull as the earliest unequivocal appearance of the species. It does not claim a still earlier date. Martin said, in an account ScienceDaily credited to La Trobe, that DNH 134 "belonged to a child aged between two and four, and there was uncertainty about what the adult version may have looked like."

That account said DNH 127 is roughly 200,000 years older than adult fossils from Dmanisi, which it placed at about 1.8 million years ago. Earth.com reported the same comparison on Oct. 7, 2026. Ferring and colleagues dated the earliest Dmanisi occupations to 1.85 million to 1.78 million years ago. The difference between those published ranges is about 100,000 to 260,000 years. The 2026 paper compares DNH 127 with D2700 and does not calculate that gap.

Paranthropus robustus in the same quarry

Of the 18 specimens, excavated between 1992 and 2019, the authors assign 14 to Paranthropus robustus, one to Homo erectus, and three they cannot place in a genus. They compared the bones with fossils from Swartkrans and Kromdraai.

Martin said, in the La Trobe account, that "2 million years ago P. robustus outnumbered Homo erectus at Drimolen by more than 10 to one." The paper does not state that figure. It says the high share of Paranthropus fits the ratio at Swartkrans and earlier counts of Drimolen teeth.

The team also uses its earlier name for the Drimolen population, the subspecies Paranthropus robustus ukusa. It places the comparison samples at younger than 1.95 million years ago, citing a 2022 study by Herries. The La Trobe account called Swartkrans about 1.8 million years old, and Martin said Paranthropus teeth and chewing muscles became larger over that span.

The paper is narrower. It says many differences sit in the chewing apparatus, and that later populations have been argued to bite harder. Jaw DNH 5 has a wide notch and a joint above the point where a chewing muscle attaches, but no formal bite-force test has been done. Jaw DNH 6 came from miners' rubble, so its age is unknown. If it is older than 1.95 million years, the authors say, its size could weigh against a simple story of jaws growing through time.

What DNH 127 cannot settle

No well-preserved, relatively complete adult Homo erectus cranium is known from Drimolen, Swartkrans, Kromdraai, and Sterkfontein Member 5 together, the authors write. They also note disagreement over other early Homo fossils, including SK 847 and StW 53, and they see no clear second Homo species in this quarry's sample.

Fossils found in place under a magnetic reversal dated to about 1.95 million years ago can be tied to that clock. DNH 127 cannot. The quarry is in the Cradle of Humankind, a UNESCO World Heritage site about 40 kilometers northeast of central Johannesburg. Two other pieces in the study had been stored as animal bone until Martin and Leece recognized them in 2015. Scans of the Drimolen fossils can be requested from Witwatersrand and the Ditsong National Museum of Natural History.