The drop stones in the tropical rocks were among the first clues that something strange once happened to planet Earth.

Drop stones are rocks that land on the seabed, sometimes with so much force that the sediment deforms. But there shouldn’t have been any drop stones in these rocks. Glaciers are the most usual source; ice sheet bellies collect rocks like ticks, then shed them when they put to sea. But the drop stone–bearing rocks were formed under what were intermittently hot tropical waters, evident from the bands of limestone interspersed with them. Surely there couldn’t have been glaciers in the tropics, right? Right?

In 1989, California Institute of Technology geologist Joe Kirschvink took a look at this and other evidence that had been kicking around for a few decades and minted a new hypothesis: snowball Earth. The idea is that our planet was once wholly encased in ice up to a kilometer thick around 650–700 million years ago. Surface temperatures everywhere were well below zero, and life, in whatever simple form it then took, had to cope.

And the evidence suggests this catastrophe happened around that time not once, but twice. The first entombment seems to have lasted some 58 million years—which, I feel obligated to point out, is over 24 times as long as T. rexes existed (a mere 2.4 million years). The second snowball, 10 million years later, lasted another 5–15 million years. Although partial glaciations would creep into the temperate zone on a regular schedule hundreds of million of years later in close proximity to our own time, as far as we know, ice would never again consume the Earth.

A study of new Chinese fossils published this January adds an interesting detail: cave fungi may have helped drag the planet out of the second snowball. If true, it would also be noteworthy because the earliest agreed-on terrestrial fungal fossils date from over 200 million years later.

If you accept that the planet was frozen solid, it follows that once all that heavy ice melted, the unburdened land rebounded and bathed in fresh air. Rainwater falling on new naked rock weathered the surface but also seeped into cracks, creating caves.

It is inside what they claim are the remains of these cavities in the Ediacaran Doushantuo rock layers of China that scientists reported in Nature Communications in January they have found both cave formations and pyrite-fossilized filaments that look to them—and I agree—a heck of a lot like fungi.

There are branching and fusing curved filaments—including A- and H-shaped structures—and little branch buds that sometimes appear to be seeking each other out; hollow spheres solo or in chains (spores?) both integrated into filaments and at their termini; and two different gauges of fiber, implying at least two species. The fibers also lack internal walls called septa that often divide such tubes into cells.

Several kinds of fungi today possess this exact suite of characters, while no other group of organisms does, the authors say. Furthermore, the curved and bent filaments seem to rule out any abiotic look-alikes. Studies of physical fungal fossil imposters shows that they are uniformly wide, while real fungi tend to be narrower and may come in multiple sizes.

Outside evidence suggests a fungal interpretation is plausible. Molecular clocks, which use calculated rates of DNA mutation to estimate when various groups of organisms evolved, suggest fungi could be 0.9–1.5 billion years old.

The authors hypothesize that soon after the cavities formed, structures like stalagmites, stalactites, and grapelike botryoids coated their walls, colonized and catalyzed by fungi and other microorganisms; large spheres in the fossils penetrated by the filaments could be some kind of symbionts—or food. Notably, modern cave formations bear similar microbes, including fungi that resemble the fossils.

Modern fungi are known for the ability of some to mine rock and extract nutrients. If moldy rock pockets in fresh-outta-snowball Earth were geographically widespread, such life would have accelerated weathering of continental rocks already underway at the surface and the delivery of phosphorous, a fertilizer, into the ocean. The resulting algal blooms would have pumped oxygen into the air. Notably, atmospheric oxygen levels seem to have been much lower than today’s prior to snowball Earth.

It wouldn’t be too long before life forms capable of capitalizing on all that new free redox power seized the opportunity, an event called the Cambrian explosion. Before snowball Earth, the most complex life—and the sum total of over 3.5 billion years of evolution—seems to have been a sponge. After snowball Earth, titanosaurs and dawn redwoods and humongous funguses appear. Was there a connection, and did it involve all that extra oxygen? This is a topic of active research.

More broadly speaking, the fact that we’ve only recently realized our planet probably went through a 58-million-year Mr. Freeze phase is unsettling as much for what it implies about what we don’t know as what we do. I once heard a lecture by a classicist who claimed that, given the pitiful nature of the remaining evidence, what we know about the ancient human world is akin to what could be discovered about the Palace of Versailles by peering through a keyhole.

The same could be said of paleontology, particularly the paleontology of life before bones and shells, or the geology of Earth under conditions distant and alien. Geologist Paul Hoffman told that after Joe Kirschvink first developed the snowball hypothesis and shared it with him, it took them a while to do anything about it because such a scenario was so different from known Earth history that they had no idea if any particular piece of evidence in the rock record was for or against it.

And if discerning major events in Earth history has proven difficult where we do have a rock record, what else do we not know about the history and owners’ manual of our planet because the relevant rocks don’t survive or aren’t currently at the surface? For instance, much of Earth seems to lack a distressingly large one-billion-year chunk of the geologic record, a glaring omission that geologists refer to somewhat ominously as the Great Uncomformity. They debate why it’s missing, but the simple fact that it is missing worries me more. It’s like Earth went on a bender for 25 percent of its lifespan, and “totally can’t remember” what happened or where its keys are.

So, we have to be grateful for the memories our planet does have. The drop stones and the rock pockets paint a vivid picture if the interpretations given here are correct: a Europa-like planet metamorphoses into a freshly scraped set of weathering continents filled with moldy caves, quietly fertilizing the oceans and oxygenating the atmosphere just prior to the greatest explosion of life the planet would ever see.

This is an opinion and analysis article; the views expressed by the author or authors are not necessarily those of Scientific American.