About the ground under your feet: we've never actually seen it. Everything we know about Earth's interior comes from listening. An earthquake somewhere punches a wave through the planet, and when that wave arrives at a monitoring station on the far side, its timing and its shape tell a story about what it passed through. It's a CT scan run on a scale nobody designed. And lately the scan keeps finding things.

The latest version of this story, recently reported, involves six distinct structures sitting deep underground, detected through exactly this kind of seismic eavesdropping. I wrote about this line of research a few weeks back when two giant blobs near the core were making the rounds, and I mentioned then that the specifics of these deep structures were still thin. Six of them, mapped as separate objects, is a step past that. Not a step toward an answer, exactly. A step toward better questions.

The technique, if you want the clinical name, is seismic tomography. Same math hospitals use, different patient. When a wave crosses material that's denser, hotter, or chemically different from its surroundings, it speeds up or slows down, and researchers can reconstruct the boundaries of whatever it hit. Over decades, thousands of quakes and millions of arrival times pile up into a fuzzy image of the mantle. Most of that image makes sense: slabs of old ocean floor sinking, plumes of hot rock rising, the ordinary circulation of a planet still cooling off after 4.5 billion years.

Then there are the parts that don't.

The six structures in question don't slide neatly into the standard picture. Reports put them at depths of nearly 1,800 miles, which is roughly 2,900 kilometers, deep in the neighborhood where the mantle meets the outer core, and well past the reach of any drill. Their sizes, their depths, their density signatures resist the easy explanations, and the researchers behind the work have been careful not to oversell what they have. That caution is worth respecting. Deep-Earth findings are pure inference, stacked inference, inference all the way down, and I said this before and I'll keep saying it: when a scientist in this field describes themselves as stunned, the reaction is earned, because they know better than anyone how indirect the evidence actually is. You can't go look. You listen or you know nothing.

What could they be? The honest answer is a shrug with footnotes. Candidates include ancient sunken tectonic slabs parked in the mantle, pockets of primordial rock that never mixed with the rest of the planet, chemically distinct reservoirs older than the continents. The primordial option is the one that keeps me up, in a good way. The idea that chunks of the original Earth, the planet as it was assembled, might still exist down there as unmixed lumps, like flour that never got stirred into the batter, is the kind of thought that makes 3 AM feel like a reasonable time to be reading about mantle viscosity.

And here I'll admit I'm speculating freely, because this next part is mine, not the researchers'. If some of these structures really have been down there for billions of years, I can't help wondering what else they've been doing. Long-lived lumps of rock that deep could plausibly have left fingerprints on volcanic hot spots, on mountain building, maybe even on the slow churn of carbon between the surface and the interior. That's a late-night chain of what-ifs, not a finding. But the possibility that the history of everything walking around up here could be partly written by lumps of rock nobody has ever seen and nobody can fully identify, that part I'll stand behind as a thought worth losing sleep over.

I'll be watching for the follow-up papers, the ones with the composition models and the depth estimates rather than the headlines. Because I've said it before and it's still true: I'd take an honest 'nobody knows yet' over a confident wrong answer any night of the week. Six mysteries underfoot. Sleep well.