In 1887, a scientist in Germany announced the discovery of a fossil animal that was totally new to science.

The thing was clearly a turtle.

I mean, it had a shell, with a flat plastron on the bottom that was fused to a carapace on top - so it had all the requirements for what makes an official turtle.

But this was not only a whole new genus and species of turtle, it was also the oldest turtle that had ever been found at the time.

And this discovery ended up spawning a debate over turtles that would last for more than a century.

Like, where did turtles come from?

What lineage gave rise to these weird reptiles with beaks for mouths and retractable necks?

And ... how did the turtle get its shell?

Well, I don't want to spoil it for ya, but the answers would eventually cause scientists to rethink the entire history of reptile evolution.

The German turtle that started all this was called Proganochelys, which means "early shell," and it lived about 210 million years ago, in the late Triassic Period.

At about a meter long, it was pretty big - about half the length of the largest species alive today.

But unlike modern turtles, Proganochelys hadn't yet developed the ability to retract its head under its shell.

The fact that its shell was found intact was really helpful, because that made it easy to identify this animal as a true turtle.

But at the same time, it left a lot of open questions about how that shell evolved, and why it resembled modern turtles in some ways, but not in others.

So the discovery of this creature kicked off a debate about which major taxonomic group, or clade, of reptiles that turtles belonged to.

Now, most reptiles fall under the clade known as eureptilia, or "true reptiles."

This includes stuff like lizards, snakes, dinosaurs, and birds.

But there's also, parareptilia, or "side reptiles."

These are some of the earliest reptiles, all of which are now extinct - like the spiky-cheeked procolophonids and the mesosaurs, which were probably the first aquatic reptiles.

Now, by and large, which clade you put turtles in, depends on where you think its shell came from.

Starting in the late 19th century, most paleontologists thought that turtles belonged in the clade we now call parareptilia, because they shared the same basic skull structure with other parareptiles.

And by the late 1940s, this theory became even more specific, suggesting that turtles were related to a particular group of extinct parareptiles called pareiasaurs.

These are sometimes called the "ugliest reptiles," but we don't judge here Bradysaurus, for example, lived around 260 million years ago, in the Permian Period, and it was covered in a layer of hardened scutes.

Likewise, a later pareiasaur called Anthodon had more developed bony plates that formed a layer of armor a lot like a turtle's shell.

So it's not too hard to see why a lot of paleontologists thought that pareiasaurs were closely related to turtles - and were maybe even their direct ancestors.

The idea was that over time, the scutes found in pareiasaurs could have fused into a solid protective layer, eventually combining with the ribs to form a shell.

But, you know how scientists are!

No one likes a good, rigorous, centuries-long argument more than they do!

Enter the developmental biologists -- they study how living things grow throughout the course of their lives.

Starting in the late 1920s, some of these biologists studied the embryonic growth of modern turtles.

They found that, as turtles develop, bones that basically function as a set of lower ribs widen and fuse, forming the plastron.

Then another set of bones up top, which are the "normal" ribs, do the same, widening and fusing to form the carapace.

So these scientists proposed that turtle evolution took a similar path, with widened ribs forming first, and the shell later.

By the 1990s, the idea had really taken off.

Keep in mind, this was basically the opposite of what the paleontologists were proposing - that the shell on top evolved from scutes that fused together before combining with the ribs below.

And if the developmental biologists were right, turtles wouldn't be parareptiles; they'd be much more similar to eureptiles.

So, both sides had some evidence to support their cases.

But it was hard to resolve the debate without more, and older, fossils.

Then, in 2008, researchers found one in China.

It was shaped a lot like a turtle, but it only had the bottom part of the shell - the plastron.

And it didn't have a carapace It was a shell-less turtle!

They called it Odontochelys, and it dated to about 220 million years ago - around 10 million years before Proganochelys from Germany.

The fact that it had a plastron without a carapace was pretty strong evidence for the developmental biologists' hypothesis, that the plastron evolved first.

But there was more.

Unlike all other known turtles, Odontochelys had ... teeth.

Its name actually means "toothed turtle."

And those teeth looked nothing like pareiasaur teeth.

Pareiasaurs had teeth with lots of little cusps on them, like human molars.

But this turtle's teeth were more like pegs.

So it looked like this ancient turtle didn't belong in the parareptile group.

It was a eureptile.

Eventually, this would be supported by several other genetic studies done in the last few decades, that compared the turtle genome with that of other reptiles, and placed turtles in a subgroup of eureptilia.

But while all that was going on, more species got added to the turtle family tree.

In 2010, the discovery of Odontochelys led researchers to reexamine another ancient species, called Eunotosaurus.

It lived about 260 million years ago in Africa - - long before Proganochelys, and the previously found Odontochelys The first fossil of this species was originally found back in 1892, but most experts at the time didn't think it was a turtle ancestor, because it didn't have a shell.

But it did have wide, flat ribs.

And modern paleontologists noticed that it bore more than a passing resemblance to Odontochelys.

Finally, in 2015, researchers discovered yet another early turtle in Germany, which they called Pappochelys, or "grandfather turtle."

It lived about 240 million years ago, and it had wide ribs, with a set of bones below them that were partially fused - but not to the point that they formed a plastron.

So old grandpa turtle seemed to mark a kind of transitional stage between Eunotosaurus, with its wider ribs, and Odontochelys, with its full plastron.

Together, these discoveries helped fill out the timeline of turtle evolution, and it became clear that the first step in the evolution of turtle shells was the formation of wider ribs.

But there was still the question of why.

Why did turtles acquire these weirdly wide ribs in the first place?

What purpose did they serve?

And why did they eventually develop into shells?

Well, in 2016, paleontologists again took a closer look at Eunotosaurus -- which is now considered the oldest of the turtle ancestors - - and noticed something funny about it.

Most of us think of turtles as being adapted for life in the water, with webbed feet or flippers.

But Eunotosaurus seemed to have a lot of adaptations for burrowing through dirt.

Its head was shaped basically like a shovel.

Its front legs were stronger than its rear legs, and it had giant claws that would have been great for digging.

So, the evidence pointed to life as a burrower.

And this could help explain why it had those wide ribs ... and where turtle shells came from.

Researchers proposed that wider ribs would've been useful as an anchor when Eunotosaurus Eunotosaurus was digging with its front legs Wider ribs provide a more stabilized trunk, which would have made it easier for the turtle to keep its body in one place while it was digging.

Other burrowing animals, like anteaters, have similar adaptations.

Problem is, having wide ribs with such short legs makes it awfully hard to walk.

Or at least, walk quickly.

In fact, the reason that turtles are so famously slow is that their giant ribs make it hard for them to swing their little legs forward.

So, ancestral turtles needed extra-wide ribs for digging, but that also slowed them down.

Now they needed more protection.

And that's when the ribs started to fuse into a plastron, which eventually became a full, protective shell.

Over time, the evolutionary purpose of the turtle's shell changed, from digging to protection, and turtles as we know them became a thing.

Of course, none of this is fully resolved.

In theory, it makes sense for wider ribs to have evolved for digging, but right now that's just based on what we've seen in one species.

We still need a lot more evidence.

Turtles' exact place among the eureptiles isn't settled either.

A lot of researchers think they're more closely related to a clade that includes animals like crocodiles and birds.

But others argue that they're closer to a different group that includes lizards and snakes.

Hopefully it won't take another 130 years, for an answer to that debate.

But either way, we now know that, on the tree of life, the turtles' branch isn't turtles all the way down.

It's stacked with a diverse array of reptilian characters, some of which have no shells, others of which have partial shells, and others still who sport the full, beautiful shells and famously slow gaits that we know today.