Picture the scene the rock describes. Broad, braided rivers spilling sand across a plain, feeding into shallow tidal flats and, at times, into windblown dunes. This was the early Silurian, roughly 430 million years ago, when Western Australia sat pressed against Antarctica near the edge of a southern sea. The trace fossils here are so close a match to ones found in Antarctica that they help prove the two landmasses were neighbours. The sandstone records that drowned-and-dried world in stacked layers - the angled bedding of moving water, the ripples of an ancient shore - more than a kilometre thick in places, with its true base buried out of sight beneath even older Proterozoic rock.

The headline acts are the walking traces. The most common, named Diplichnites, are paired rows of pits that can run across a rock surface for many metres - the patient stride of a many-legged animal crossing open ground. Some appear to have been made above the waterline, which is why these trackways are read as among the earliest evidence of fully terrestrial animals: creatures not just living in water but venturing onto land. The likely walkers were arthropods - myriapods, eurypterids, the strange euthycarcinoids, and relatives of the horseshoe crab. No one watched them do it. We have only their tracks, and the tracks are enough to rewrite the timeline of when life first came ashore.

For all its abundant trackways, the Tumblagooda has yielded essentially one creature's actual remains: Kalbarria, a small early arthropod named for the town, carrying eleven pairs of legs. Bodies almost never fossilised here, and the reason is its own small story. The sand was coarse and oxygen-rich, so anything that died and was not promptly eaten simply rotted away before it could turn to stone. The burrowers, in other words, got there first. What we are left with is a paradox: a place crowded with the signs of animals and nearly empty of the animals themselves.

Pinning an age on the formation has been a century-long argument. Early guesses ran from the Cretaceous to the Cambrian; later work using spores, microscopic acritarchs, tooth-like conodonts and the magnetism frozen into the rock pulled the estimate toward the Silurian and Ordovician. The honest answer remains a range - somewhere between four and five hundred million years - because the techniques disagree and the rock guards its secrets. Researchers keep trying new approaches, hoping that crystals grown inside the sandstone might one day yield a sharper date, though so far the rock has refused to cooperate. Part of the problem was simply reaching it: until the 1970s, the roughly 600 miles of track from Perth were mostly dirt, and the formation went largely unstudied until 1948. Some of the planet's most important early footprints sat unexamined, in plain sight, for a very long time. They are still here, in the gorge walls above Kalbarri - a record of the moment life first tested the land, written in stone and waiting to be read.