90 million years ago, a stealthy predator slipped in and out of burrows that it dug in the shadows of the dinosaurs, in what would later become Argentina.

Today this creature is known as Najash -- named after a monstrous biblical snake.

And, for the most part, it had all of the classic traits that you expect a snake to have, like a long, sinuous body and ribs for days.

But this ancient snake also had ... legs.

Which sounds weird, right?

Like a snake with limbs?

But what's even stranger about this creature is: We're not at all sure where it came from.

OK, first, let me answer the question that's probably on your mind already: Yes!

Najash is considered a true snake, even though it had legs -- two of them, toward the back of its body.

It's one of the earliest known snakes in the fossil record found with limbs intact.

But it was by no means the only one.

Snakes-with-legs have been described in such disparate places as Great Britain, Morocco, Romania, and Wyoming.

Some of these reptiles, like Najash, had two legs.

Others might have had four.

But most of them are known from only a few bones -- a piece of a jaw here or a vertebra there.

And because of this, the evolutionary history of snakes is among the most mysterious of all the vertebrates.

Biologists know that snakes diverged from lizards, probably as early as the Jurassic Period, about 200 million years ago, and eventually radiated into the 3,000 species that we have today.

But we don't really know which group of lizards gave rise to the snakes -- or when or why they lost their legs in the first place.

We can't answer these questions, because we don't have enough data.

Most of the very oldest snake fossils that have been found are just fragments.

But, we do have a couple of good, if competing, theories about where snakes came from.

Some scientists think that the first snakes descended from burrowing lizards.

Others think they might have come from from lizards that swam in the open ocean.

That's because the way that snakes move seems to work best for either burrowing through soil, like a worm does, or propelling through water, as eels do.

So, let's start with that.

The consensus is that snakes came from lizards.

But ... where does one animal stop and the other one start?

Like what is the difference between a snake with legs and a ... lizard?

Well, most of us would identify a snake by its long, slinky body.

But herpetologists actually define snakes not by their slinkiness but by their mouthy-ness.

Specifically, snakes are identified by adaptations in their skulls that allow them to unhinge their jaws.

For example, several parts of a snake's skull are smaller than they are in other reptiles.

And its two lower jaw bones aren't fused together, like they are in other vertebrates.

Instead, snakes just have a little cartilage there.

These adaptations work together to provide maximum flexibility, so snakes can swallow things bigger than their own heads.

Which we do not recommend that you try at home.

Because you can't do that.

Snakes are also unique in that they don't have ... ears.

No ear holes, no ear drums, and no inner ear bone that other reptiles have.

Instead, what was once their ear bone became fused to their jaw, where it allows snakes to detect vibrations in the ground.

So, you'd think that knowing all this could help us figure out where snakes came from: We just have to find out when these adaptations first appeared.

But it turns out, that's not easy.

Because there's evidence that seems to support each hypothesis -- that snakes evolved from terrestrial lizards, and that they came from aquatic lizards.

Advocates of the aquatic lizard hypothesis, for example, think that snakes may actually be distant cousins of the fiercest predators of the Cretaceous seas -- the mosasaurs.

Mosasaurs were themselves descended from land-dwelling lizards.

And like snakes, they developed long bodies and jaw adaptations that allowed them to open their mouths wide for a crushing bite.

And for decades, researchers looked for an evolutionary link between snakes and mosasaurs in one of the most-studied snakes of the Cretaceous - - a two-legged snake known as Pachyrhachis Pachyrhachis was found in the West Bank, near Ramallah -- a region that was underwater when the snake lived there 95 million years ago.

And studies of its anatomy have found that it probably moved the way an eel does, by whipping its tail from side-to-side, to get short bursts of speed.

Some paleontologists have argued that snakes like Pachyrhachis descended from mosasaurs, because they share a lot of similarities in the shape of their skulls, especially around the jaw.

Plus, Pachyrhachis and other aquatic snakes that have been found in Ramallah were, for a long time, thought to be the oldest snake fossils ever found.

And if the oldest snakes were marine reptiles, then snakes must have come from the sea, right?

Well, in the last 10 years or so, a bunch of new discoveries have shaken up this idea - - as well as the rest of the snake family tree.

Paleontologists have recently described at least half a dozen previously unknown species of ancient snakes -- many of which are even older than Pachyrhachis, and most of them lived on dry land.

Among these new species is Najash, which lived around the same time as Pachyrhachis, but many are even older.

Some go back as much as 167 million years, like Eophis, which was described from Great Britain in 2015.

It's now considered the oldest snake ever found, and it was terrestrial.

So, to some experts, the notion of snakes originating on land makes a lot more sense.

For one thing, many of the classic traits that snakes have, like jawbones that pick up vibrations, seem better suited for life on the ground.

And not having legs can offer a lot of advantages if you're a reptile that burrows through the dirt.

After all, the skinnier you are, the skinnier your burrow can be, which makes it easier to evade predators.

But what about genetics?

Can molecular evidence settle the Great Snake Debate?

Well, in 2013, scientists used gene data from more than four thousand living snake and lizard species to construct a new family tree for these groups of reptiles.

And the results suggested that snakes' closest relatives are probably the Varanid lizards, a group that includes monitor lizards and Komodo dragons.

That might sound like score one for the terrestrial hypothesis.

But the problem is, Komodo dragons and their kin are also thought to be the closest living relatives of the mosasaurs.

So, depending on how you read them, these new genetic data could support EITHER the terrestrial hypothesis OR the aquatic hypothesis.

Which is kind of frustrating!

But also fascinating!

Ultimately, all we know is that snakes came from lizards, but we don't know the details of their remarkable transformation.

And we know that there were snakes before the likes of Najash -- the leggy reptile from Argentina -- but we don't understand which lineage of lizard they came from.

So in the end, snakes pose one of the greatest evolutionary mysteries in the history of animal life.

Not even modern genomics has been able to tell us how and when they became a thing.

At least, not yet.

Thank you for watching this particularly mystifying episode of Eons.