Liaoning fossil swimmer was forming a third set of teeth 120 million years ago

Everything here rests on one skeleton. It came out of the Jiufotang Formation in 2016 ^[5], and the description appeared in Nature Communications in 2026 ^[8], about ten years later ^[18]. One well-preserved specimen is enough to show that a trait existed. It is a smaller base for saying when a trait evolved across mammals, and Phys.org's write-up notes that the whole account concerns a single fossil ^[20].

The tooth evidence is a snapshot of replacement in progress. CT scans showed teeth that had erupted and were being worn down by chewing, with their replacements still growing inside the jawbone ^[9]. At one position, a tiny enamel cap sat beneath a replacement tooth. The team takes it as a third generation beginning to form ^[10]. The same account says evidence of three generations appeared in all four tooth types: incisors, canines, premolars and the front molar-like teeth ^[11]. The press summary does not say how many positions show a third crown. One cap and a full row of them would justify very different confidence in that sentence.

Humans, dogs, cats and horses are diphyodonts, with baby teeth followed by adult teeth ^[3]. Paleontologists long assumed that pattern arose early in mammal history ^[4]. The paper's title calls Dongoconodon a polyphyodont eutriconodontan from the Early Cretaceous ^[8]. If the third generation holds up, two-set replacement was not universal among mammals 120 million years ago ^[1]. I think the finding is secure as a description of one animal and provisional as a revision of the rule. The thing this doesn't tell you is the direction of change. One jaw cannot separate an old pattern this lineage kept from one it evolved again, and settling that needs comparable scans of its relatives.

The second finding concerns hearing. The scans show a thin bar of bone, ossified Meckel's cartilage, joining the lower jaw to the middle ear bones ^[12]. In modern mammals those bones sit free of the jaw and work only for hearing ^[13]. The fossil suggests the order in which the connection came apart: the cartilage began to detach from the jaw before it separated from the ear bones ^[14].

The swimming reconstruction is built by comparison with living animals. The team measured the long leg bones and set their proportions against a database of 124 living mammal species ^[7]. Broad hands and feet with flattened finger bones resemble the webbed feet of platypuses, and the boxy tail bones closely match those of nutria ^[15]. From that, the authors conclude the animal paddled with its forefeet and steered with its tail ^[16]. This is an inference from shape. Living swimmers have these proportions; the fossil cannot show the animal swimming. "The new fossil shows adaptations for semiaquatic life, with hands and feet like the platypus that were similarly webbed," the research team wrote ^[17].

The authors draw a wider conclusion from finding two oddities in one animal. "This diversity indicates that key mammalian traits did not necessarily evolve in a tightly linked fashion," the researchers wrote, according to Phys.org ^[19].