It’s just a little dumpling. I don’t know what else to call it. The baby, this most tragic and precious little thing, is curled up tight as if still inside a leathery shell that decayed away more than 251 million years ago. The protomammal perished before it was even born but I still want to hold it in my palm and offer it everything.

I’ve never seen a fossil quite like this one. It was exciting enough, back in 2018, when paleontologist Eva Hoffman and colleagues described dozens of Kayentatherium babies. The small bones represented the first protomammal perinates anyone had ever uncovered. Their size, apparent helplessness, and number all suggested that their parent stayed nearby and raised them on milk. But these Lystrosaurus embryos, they tell a very different story for one of the most resilient species that’s ever existed.

Lystrosaurus doesn’t look like an especially close relative of ours. The animal had a turtle-like beak flanked with tusks, its form often likened to a reptile-like pig. Still, if we follow the family tree, Lystrosaurus is more closely related to us than to any reptile, one of the many forms of protomammals that thrived in the Permian world of more than 252 million years ago. When an especially terrible mass extinction ground down the majority of known vertebrate species alive at the end of the Permian, Lystrosaurus was one of the few survivors. The protomammal even seemed to thrive in the shaken, struggling Triassic world. Why this animal? What made Lystrosaurus so special? The dumplings embody an important clue.

Described by paleontologists Julien Benoit, Vincent Fernandez, and Jennifer Botha, the three embryonic Lystrosaurus were found in the Early Triassic rock of South Africa’s Karoo Basin. The little Lystrosaurus had lived after the mass extinction, but still many millions of years before dinosaurs and other reptiles would undergo their great evolutionary burst. Among the three, the specimen NMQR 3636 is the most stunning of all. The embryo’s skull is clearly visible in the oval fossil, but the CT scans reveals that much of the Lystrosaurus is still encapsulated in the rock. The baby is curled up much like a little bird in an egg, yet there’s no calcified eggshell. If the bones fossilized in position like this, then a hard eggshell should have, too. The conspicuous absence, the authors suggest, indicates Lystrosaurus had a leathery eggshell. Protomammals, specifically relatives of our direct ancestors who were around during this same early Triassic time, laid soft, leathery eggs.

The oval of rock containing the embryo is a little short of three inches long. About the size of a chicken egg. Even though this might seem small to us, it’s still a fairly large egg - and inconsistent with what experts would expect if Lystrosaurus was lactating for many small offspring.

The embryos from the Karoo indicate that, at least compared with Kayentatherium and egg-laying monotremes like the platypus, Lystrosaurus laid large eggs that enclosed the development of large, well-formed babies. The bigger the egg, the more yolk, the longer a baby starting as a single cell can grow and develop, favoring more precocial offspring able to run around on their own soon after hatching. Such babies also tend to grow quickly and reach sexual maturity earlier than altricial babies that need prolonged care. Hatching ready to take on the world, and reproducing early, may have allowed Lystrosaurus to weather some of the environmental upheavals, the heat and reduced oxygen and stricken botanical remainders, that closed out the Permian. Especially when conditions were so inescapably hard that juvenile mortality would have been high, even moreso than it usually is, reproducing early may have helped the protomammal hang on.

If bearing precocial young was such a boon, though, what were our ancestors doing at the time? Our only other point of comparison is Kayentatherium, a beast closer to the origin of all mammals but lived tens of millions of years after the little Lystrosaurus were folded into the fossil record. Were our protomammal ancestors - among the weasel-like cynodonts - reproducing like Lystrosaurus? Or had they already begun to have large, milk-fed litters? If the latter is the case, then we have more wrinkles to uncover. It may not have been just the reproductive strategy itself that allowed both cynodonts and dicynodonts like Lystrosaurus to hang on, but how reproductive particulars were complemented by body size, diet, and growth rate. Either way, finding an egg-laying protomammal that reproduced in a different way from even our modern, egg-laying mammals underscores that the fossil record is as endlessly full of variety as our modern world. Of course different protomammal lineages lived and reproduced in different ways, the dicynodonts and cynodonts as different from each other as pigs are from weasels today. One confirmed difference indicates so many others we haven’t found yet, details that are critical for considerations of how life continued to spin off even more variety.

For the moment, though, I’ll set aside the high-minded consideration of what the little embryos mean for big evolutionary questions. They’re resting too soundly to bother them with such concerns. It is stunning that we get to know these little creatures at all. The first time the sun touched them, they were already fossils.