A reconstruction drawing of Pohlsepia mazonensis (Credit: Dr Thomas Clements, University of Reading)
A Hidden Set of Teeth Just Rewrote 150 Million Years of Octopus History
In A Nutshell
- A fossil long called the world’s oldest octopus has been reclassified as a decomposed nautilus relative, not an octopus at all.
- Synchrotron X-ray imaging revealed a hidden radula (tooth ribbon) with 11 visible tooth structures, a pattern matching nautiloids, not octopuses.
- The reclassification removes the fossil calibration point that had pushed octopus origins back into the Paleozoic era, over 250 million years ago.
- Evolutionary evidence now points to modern octopus lineages arising later, during the age of dinosaurs.
For 25 years, a smudgy, palm-sized fossil pulled from a strip mine in Illinois held a title few questioned: the oldest octopus ever found. It became a textbook example, a key calibration point for genetic dating studies of cephalopod evolution. Now an international team has used high-powered X-ray technology to peer beneath the fossil’s surface and find something its describers never could see: a hidden set of teeth showing the creature wasn’t an octopus. It was a decomposed relative of the nautilus, that ancient, spiral-shelled animal still drifting through tropical oceans.
This reclassification, published in Proceedings of the Royal Society B, strips the fossil of its identity and removes the calibration point behind claims that octopuses appeared before 252 million years ago. Without this fossil anchoring the timeline, the evolutionary story of octopuses snaps back into line with other evidence pointing to a later origin, during the age of dinosaurs.
Decades of Doubt Surrounded the ‘Octopus’ Identity
Named Pohlsepia mazonensis, the fossil was first described in 2000 from the Mazon Creek deposit in Illinois, a site known for preserving soft-bodied creatures inside lumps of iron-rich rock. Dated to roughly 311 to 306 million years ago, it appeared to show a sac-like body, paired eye spots, fins, and arms, features that led its describers to classify it as a deep-sea octopus and push the history of octopus-like animals back over 150 million years.
Doubts crept in almost immediately. Researchers noted the fossil lacked features expected of that type of octopus, including an internal shell remnant and a single row of suckers. A 2019 study found no pigment structures in the supposed eye spots, structures that readily preserve elsewhere in Mazon Creek fossils. Some scientists questioned whether the creature was even a mollusk. Excluding the fossil from genetic dating analyses pushed the origin of modern octopuses to between 200 and 145 million years ago, matching the oldest undisputed octopus fossils. The debate stayed stuck without new physical evidence.
Hidden Teeth Exposed a Nautilus Identity, Not an Octopus
Lead author Thomas Clements, of the University of Leicester and University of Reading, and colleagues brought the fossil to the SOLEIL synchrotron facility in France, a ring-shaped particle accelerator generating intense X-ray beams. A focused beam at the surface excited chemical elements in the fossil and rock, causing them to emit signals that revealed structures invisible to the naked eye.
Mapping iron across the mouth region initially hit a wall, since the fossil and rock shared the same iron content. But the sharpness of the scan produced a high-resolution surface map anyway. Hidden beneath the surface sat a radula, the ribbon-like strip of tiny teeth mollusks use to scrape and process food, with beak-like jaw structures nearby. That alone confirmed something uncertain for years: the creature was definitely a mollusk.
Detail in the teeth sealed the case. Researchers counted at least 11 tooth structures per row, including a central tooth and pairs of side and curved outer teeth. Modern octopuses have 7 or 9 elements per row, with markedly different teeth. Nautiluses have 13, and the fossil’s teeth closely matched Paleocadmus pohli, a nautilus relative already known from the same deposit. Researchers think a small pair of plates was probably lost to preservation, a gap common in other Mazon Creek nautilus fossils, which would bring the total to the full 13-part nautiloid row.
Decay Explains the Octopus-Like Appearance
Why did this nautilus relative look so much like an octopus? Decay is the answer. Not every carcass at Mazon Creek was buried quickly; some sat decomposing on the seafloor first. Many supposed octopus features, the arm-like structures, fin-like lobes, and eye spots, were either too degraded to interpret or explainable as decayed nautilus anatomy. Researchers propose the ‘fins’ could instead be decayed retractor muscles or digestive gland lobes already documented in living nautiluses.
A missing shell fits the same picture. Decay studies show a nautilus’s soft body can separate from its shell before burial, though researchers cannot say when that happened here; they found no shell material on the fossil. Additional Mazon Creek fossils resembling the specimen also preserved tooth ribbons tying them to nautilus relatives. Imaging found no confirmed siphon, no evidence of eight distinct arms, no arm hooks, and no pigment in the eye spots.
Researchers reclassified Pohlsepia mazonensis as a junior synonym of Paleocadmus pohli, meaning the two names refer to the same species and the older name takes priority.
Octopus Evolution Timeline Shifts to the Age of Dinosaurs
Because this specimen anchored many early genetic studies estimating when major cephalopod groups split apart, its reclassification removes the calibration point behind those ancient-origin estimates. Molecular studies now point instead to the octopus lineage arising in the Mesozoic, matching fossil evidence of gradual shell reduction in ancient octopus ancestors during the Jurassic. It also weakens the case for reading certain controversial Cambrian fossils, from over 500 million years ago, as early relatives of squid and octopuses.
Still, the fossil gains a different distinction: the only confirmed soft tissue from a nautilus relative of this age, older than any other by roughly 220 million years. A smudge in a lump of ironstone quietly rewrote its own history, and this time, the teeth did the talking.
Paper Notes
Limitations
The study acknowledges several limitations. The tooth ribbon of Pohlsepia mazonensis (PE51727) is incomplete, and the resolution of the synchrotron scan was not sufficient to identify the smallest pair of marginal plates (MP1) that would bring the total element count to 13, matching known nautilus-relative tooth ribbons. The researchers argue this gap is consistent with preservation patterns in other Mazon Creek nautilus fossils, but it remains an inference rather than a direct observation. Additionally, the fossil is compressed sideways, making it difficult to resolve the exact shape and size of each tooth element. Many of the soft tissue features, including the arm crown, fins, eye spots, and ink sac, remain ambiguous even after analysis with multiple techniques. The researchers note that detailed geographic and stratigraphic data are lacking for where the original specimen was found, and it is unclear whether it was collected in place or from waste piles. There are also limited experimental studies on the decay and breakdown of modern Nautilus soft tissues for direct comparison. The study did not section the rock, meaning the identity of certain dense mineral structures and voids within it could not be conclusively determined. Finally, the absence of shell material in or on the specimen is consistent with the nautilus reinterpretation but remains unexplained by direct evidence of how or when the shell separated.
Funding and Disclosures
The paper states that no funding was received for the article. The authors declare no competing interests. The authors also state they did not use AI-assisted technologies in creating the article. The paper is dedicated to the memory of Dr. Joanne Kluessendorf, the original describer of Pohlsepia, who passed away in 2018.
Publication Details
Title: Synchrotron data reveal nautiloid characters in Pohlsepia mazonensis, refuting a Palaeozoic origin for octobrachians | Authors: Thomas Clements, Imran Alexander Rahman, Alan R. T. Spencer, Christian Klug, Dirk Fuchs, Isabelle Rouget, Isabelle Kruta, Sebastian Schöder, Jack Wittry, Orla G. Bath Enright, and Pierre Gueriau | Journal: Proceedings of the Royal Society B, Volume 293, Issue 2068 | DOI: https://doi.org/10.1098/rspb.2025.2369 | Received: 12 September 2025 | Accepted: 11 February 2026 | Institutional affiliations include: University of Leicester; University of Reading; Natural History Museum, London; Oxford University Museum of Natural History; Imperial College London; Universität Zürich; Bayerische Staatssammlung für Paläontologie und Geologie, Munich; Muséum National d’Histoire Naturelle, Paris; Synchrotron SOLEIL, France; Field Museum of Natural History, Chicago; Staatliches Museum für Naturkunde Stuttgart; University of Portsmouth; Université de Lausanne; and Ministère de la Culture, Université Paris-Saclay.







