Mammals these days have hair, birds have feathers and reptiles have neither.
But the relative nakedness of today's reptiles may not have been the case 247 million years ago, when at least one reptile species appeared to be quite experimental with its embellishments.
That is according to fossils of a creature that lived at the beginning of the Triassic period and sported a curious crest running down its spine.
The crest, while feather-like, pre-dates feathers by millions of years.
A new study published in Nature has described the crested reptile, which is dubbed Mirasaura grauvogeli.
Stephan Spiekman, study lead author and a researcher at the State Museum of Natural History in Stuttgart, Germany, said he was fascinated by the creature's complex appendages — particularly since it did not belong to the group of animals that eventually grew feathers.
"Mirasaura was a type of reptile that was very different from dinosaurs and their close relatives," Dr Spiekman said.
Stephen Poropat, a palaeontologist at Curtin University who was not involved with the research, called the fossils an "extraordinary" find.
"If we saw it with modern eyes, as a living animal, it would be something totally unique," he said.
The fossil treasure trove lying in wait
The Mirasaura fossils were discovered by French palaeontologist Louis Grauvogel in the 1930s in north-eastern France.
He found rocks containing small reptilian skeletons, alongside fan-like depressions made from plumes up to 15 centimetres long.
Professor Grauvogel thought the mysterious fans were fish fins or insect wings, and did not connect them to the reptilian skeletons found in the same fossils.
In 2019, decades after his death, Professor Grauvogel's collection was acquired by the museum in nearby Stuttgart, and researchers realised that 80 of the fossils belonged to the same animal.
The researchers named the creature Mirasaura grauvogeli, which means "Grauvogel's wondrous reptile".
But they could not immediately tell what the mysterious crests on the creatures' backs were made of.
"It was clear pretty quickly that these appendages were not feathers or hairs in the traditional sense," Dr Spiekman said.
The team used a number of techniques, including powerful X-rays and electron microscopes, to reconstruct what the creature looked like and figure out the composition of the crests.
They discovered remnants of tiny cells called "melanosomes" preserved in the fossil crests, which make the colours in hair, skin and feathers.
The melanosomes were shaped similarly to those found in feathers, Dr Spiekman said.
"This could indicate that the tissue that they are composed of is quite similar to that of feathers."
But the appendages were clearly different to feathers: they did not have any of the key branching features feathers develop as they grow.
"This shows that the structures grew in a largely different way than feathers," Dr Spiekman said.
Anatomical analyses of the reptile's skeleton also showed that it belonged to a different group of reptiles to the one that eventually became the dinosaurs.
Crest prompts evolutionary questions — and answers
John Long, a palaeontologist at Flinders University who was not involved in the research, said that the existence of Mirasaura had profound implications for understanding how feathers and skin evolved.
"Prior to this discovery, it was thought that only birds and dinosaurs evolved feather-like structures, and mammals modified those feather-like structures to develop hair," Professor Long said.
"This is a third group of animals yet discovered that actually have feather-like structures as part of their skin."
Dr Poropat said the discovery helped explain another highly controversial fossil find from a similar time period, called Longisquama insignis.
The extinct Longisquama reptile, which was discovered more than 50 years ago, also had feather-like appendages, but its fossils were not preserved well enough to identify them properly.
This led to fierce debates about whether or not the Longisquama reptile had real feathers.
"The finding of Mirasaura demonstrates Longisquama's structures are not feathers, because the Mirasaura reptile's are very similar and they're not feathers either," Dr Poropat said.
"I'm quite relieved to see that Longisquama finally has a home."
What was the crest used for?
The researchers believe the reptiles used the crest to communicate with each other.
"It seems most likely that it was used for signalling to other members of the same species, either to attract a mate, or to warn off a competitor, or both," Dr Spiekman said.
Professor Long said that we could only speculate about the crest's function, but he agreed that communication was a likely reason.
"Certainly today, visual signalling is incredibly important for birds and for many reptiles," he said.
Dr Poropat pointed out the Mirasaura reptile lived a few million years after the catastrophic Permian mass extinction.
This may have favoured the evolution of more bizarre behaviours and body structures because there was less competition.
"The animals that did survive that extinction event were basically free to experiment with all sorts of different modes of life."
While the Mirasaura reptile was not a precursor to the dinosaurs, its existence suggests that other Triassic creatures could also have had strange features that did not survive fossilisation.
"Because soft tissue is so seldom preserved, we are getting an incomplete picture of the past almost inevitably," Dr Poropat said.
"It is quite possible that we can't just apply scaly skin to most Triassic reptiles or most Triassic mammal forerunners."
