The Paleo Diet: Regret the fossil error. It wasn’t the first

At its best, paleontology opens windows into trillions of other lifetimes spent swimming, scuttling, stomping and soaring across this planet. Scientists, the press and the public alike tend to tell and retell these success stories, lionising intrepid researchers. The most impressive specimens are enshrined in museums. But possibly just as important is when scientists get something wrong, badly, and somebody sets the record straight.

In the last pre-Covid lockdown days of 2020, for example, Gregory Retallack, a paleontologist from the University of Oregon, and a few colleagues toured a famous set of Indian cave paintings. Afterward, they announced they had discovered something that previous visitors had overlooked: a 550-million-year-old fossil called Dickinsonia from the dawn of animal life. The dramatic find drew outside scrutiny. Last December, a team led by Joseph Meert, a paleontologist at the University of Florida, studied the same site. “When we found the fossil, some alarm bells went off in my head,” Dr. Meert said. First, the specimen looked different than it had in pictures from 2020: Part of it had rubbed off. Second, the team kept noticing giant honey bee nests on the surrounding rocks.

Then it clicked: This wasn’t a Dickinsonia at all. Neither was it a fossil. The pattern on the cave wall was just a bit of waxy material left behind by a bee nest, the team reported in December, in the same peer-reviewed journal that had vetted the original finding. Another study, recently accepted to the Journal of the Geological Society of India, arrived at the same result. Dr. Retallack is now working on a formal correction. “It is rare but essential for scientists to confess mistakes when new evidence is discovered,” he wrote to the Florida team, once its researchers contacted him with their new analysis.

This discovery-that-wasn’t joins a long, ignominious history of paleontological misfires. These range from outright misclassifications to pseudofossils (where a non-biological process made a pattern that only looks biological) and dubiofossils (weird, ambiguous rocks that are probably not as important as they’re cracked up to be).

Like Tolstoy’s unhappy families, each misidentified fossil comes with its own unhappy story. Many rocks that look lifelike but aren’t — like mineral nodules that resemble fossil poop and supposed “dinosaur eggs” and “dinosaur footprints” — are screened out the very first time a real paleontologist looks at them. Others are just old mistakes, relics of a more primitive scientific past. Still other errors or misreadings persist in fringe sources. Occasionally, though, they penetrate modern scientific enterprise, even through peer review from other experts, especially when key evidence is ambiguous.

Each of the examples below is ambiguous in another way, too: as both a scientific failure and a demonstration of how science advances by publicly correcting mistakes.

In the 1670s, the English chemist Robert Plot made perhaps the first ever scientific illustration of a dinosaur fossil. He suspected that the specimen was part of a femur bone. But it was big — perhaps, Plot reasoned, belonging to a Roman war elephant, or a giant human described in the Bible.

Almost a century later, the illustration was reprinted in a natural history volume compiled by a physician, alongside a new, fairly self-explanatory caption that compared it to the dangly bits of an ancient human. But these were no reproductive organs: While the specimen itself has been lost, it was in fact part of a femur of a carnivorous dinosaur, maybe Megalosaurus. In 1981, two different ancient species named by the early 20th-century German paleontologist Baron Friedrich von Huene — mercifully, already deceased at the time — were both shown to be cases of mistaken identity. One supposed mammal tooth was actually a bit of the mineral chalcedony. The other, a dinosaur jaw, turned out to be a chunk of petrified wood that mollusks had burrowed into.

In 1864, Canadian geologists announced the discovery of Eozoon canadense, the “dawn animal of Canada,” a wavy, striated set of rock patterns they claimed came from the fossilized shells of giant cellular organisms. The find filled a gap in the theory of evolution: Until Eozoon canadense, there had been no prior fossil evidence for life on Earth before 540 million years ago.

In the following decades, though, evidence mounted that the patterns were just layered, bent rock forged by high temperatures and pressures. Eozoon’s proponents never quit arguing that it was a real fossil, but they eventually died. In the meantime, other very old fossils (like real examples of Dickinsonia) emerged to fill the gap in the fossil record.

In 2019, one team announced the discovery of a new Triassic horseshoe crab-like species. But the researchers were corrected the following year: What had looked like a separate animal was actually the severed head from a known fossil cicada.

Differentiating impostor fossils from the real deal can come with much higher stakes. In 1996, scientists proposed that they had found a microfossil in a Martian meteorite. President Clinton even held a news conference discussing the implications of the discovery, footage of which was edited into the 1997 movie “Contact.”

Since then, scientists have documented many chemical and geological processes that can “grow” intricate, tiny structures without any life involved. Some of the oldest claimed fossils on Earth might fall into this category — and similar patterns could show up in the first rocks returned from Mars.