In 1992, a farmer named Wu Anai in Zhejiang Province, China, decided to pump the water out of a pond near his village. Local legend held that the ponds dotting the landscape around Longyou were bottomless — cursed, some said. Wu Anai didn’t believe in curses. He believed there might be fish worth catching. What he found instead, after seventeen days of pumping, was a cathedral.

Beneath the muddy water lay an enormous hand-carved cavern: soaring pillars, smooth walls, angled ceilings — all cut from solid siltstone with a precision that, even today, nobody can fully explain. Over the following years, twenty-three more caves were discovered in the surrounding area. Together, they represent one of the largest underground excavations of the ancient world, and one of the most poorly understood. No historical text mentions their construction. No legend names their builders. They simply exist, silent and vast, waiting under the rice paddies of eastern China.

The Scale of the Problem

The Longyou Caves are not small. The largest single grotto — Cave No. 1 — covers roughly 2,000 square meters with a ceiling height that reaches over 30 meters. Across all 24 discovered caves, the total excavated volume has been estimated at around 36,000 cubic meters of stone removed. Some researchers have pushed that number higher. To put it in crude terms: you’re looking at roughly a million tons of rock, carved out by hand, transported somewhere, and disposed of — all without leaving a trace in any surviving record.

The caves are located near the village of Shiyan Beicun, about 260 kilometers southwest of Shanghai. They were carved from relatively soft Cretaceous-period siltstone, which is workable but far from trivial to excavate at this scale. Each cave follows a similar design pattern: a roughly rectangular footprint, load-bearing pillars left in place to support the ceiling, and walls covered floor-to-ceiling in parallel chisel marks arranged in a consistent decorative pattern. The marks run at approximately 60-degree angles, creating a series of uniform bands that look almost machine-made. They aren’t. Every groove was cut by hand.

Yang Hongxun, a researcher at the Chinese Academy of Social Sciences and one of the first archaeologists to seriously study the caves, estimated that the excavation would have required approximately a thousand workers laboring day and night for six years — assuming they had adequate lighting, ventilation, and a system for removing debris. That’s a conservative estimate. It also raises the obvious question: where did the stone go? A million tons of siltstone doesn’t vanish. Yet no corresponding spoil heaps have been identified in the surrounding area.

What We Know (and What We Don’t)

The honest answer is: not much, and quite a lot, respectively.

What we know: radiocarbon dating and geological analysis place the caves’ construction somewhere around 200 BCE, during the late Warring States period or early Qin Dynasty. This makes them roughly contemporaneous with the construction of the first Great Wall and the Terracotta Army. China was not short on ambition during this era. The siltstone itself has been dated, and the tool marks are consistent with iron chisels available in that period.

The caves contain no inscriptions. No pottery. No tools. No human remains. No religious iconography. No graffiti. Nothing that would indicate who built them, why, or when they stopped. This is the part that keeps researchers up at night. A construction project of this magnitude — likely state-sponsored, given the labor requirements — should have left some bureaucratic footprint. Ancient China was obsessive about record-keeping. The Qin Dynasty in particular maintained detailed registries of labor, taxation, and public works. And yet: silence.

Several hypotheses have been floated over the years, none of them entirely satisfying.

Military storage or shelter. The caves could have served as wartime depots or refuges during the chaotic Warring States period (475–221 BCE). The region saw significant conflict, and underground storage would have been strategically valuable. But the caves show no evidence of habitation or storage — no soot from fires, no wear patterns on floors, no niches or shelves.

Mining operations. Perhaps the caves were quarries, and the stone was the product, not the byproduct. This is plausible in theory — siltstone has construction applications — but the interior finishing argues against it. You don’t decorate a quarry. You don’t carve parallel chisel patterns at precise angles across every surface of a hole you’re just pulling rock out of.

Imperial tomb preparation. Some researchers, including Jia Gang of Zhejiang University, have suggested the caves might have been intended as elaborate burial chambers, possibly for a regional king or the Yue state nobility. The Warring States period produced extravagant tombs. But again: no burial evidence, no funerary objects, no inscriptions marking them as sacred or royal space.

Water management. A more recent hypothesis ties the caves to ancient hydraulic engineering — underground reservoirs or flood-control systems for the nearby Qu River. The caves do accumulate water naturally, which is why they were flooded for so long. But their design features (the pillars, the decorative finishing, the sheer depth) seem wildly over-engineered for water storage.

The Chisel Marks

If the caves have a signature, it’s the marks on their walls. Every interior surface — walls, ceilings, pillars — is covered in parallel grooves, each about 10 centimeters wide, carved at consistent angles. The pattern shifts direction between surfaces, creating a visual effect somewhere between corduroy and a fingerprint. The marks are so uniform that early visitors assumed they were made by machines.

They weren’t. Analysis by Zhu Danian of Tongji University and others has confirmed the marks are consistent with hand-held chisels, likely operated in a rhythmic, standardized motion. This implies not just organized labor but trained labor — workers who had been taught a specific technique and applied it consistently across tens of thousands of square meters. The uniformity suggests either a single sustained campaign or a rigidly maintained tradition passed between work crews.

But here’s the thing that nobody has adequately explained: the marks appear to be decorative, not functional. They don’t follow the most efficient excavation pattern. They required extra labor. Someone made a deliberate decision that the interior surfaces of these caves should be beautiful — or at least patterned — even though (as far as we can tell) almost no one ever saw them.

Why finish the inside of a cave that would immediately be filled with groundwater? Were the builders aware the caves would flood? Did they care? Or were the caves intended to be kept dry, maintained, used — and something interrupted that plan?

The Silence in the Archives

This is, for my money, the strangest part of the whole mystery. Not the engineering. Not the scale. The silence.

China has one of the deepest continuous literary traditions on Earth. The Shiji (Records of the Grand Historian), written by Sima Qian around 94 BCE, covers centuries of Chinese history in extraordinary detail. Regional gazetteers, compiled and maintained for millennia, record everything from floods to famines to local construction projects. The bureaucratic apparatus of the Qin and Han dynasties tracked labor conscription with near-modern precision.

And none of it mentions the Longyou Caves.

There are a few possible explanations. The caves may predate the local record-keeping systems. If they were built during the Warring States period by the Yue or Chu states, the relevant records could have been destroyed during the Qin unification — the infamous burning of books and burying of scholars in 213 BCE targeted precisely this kind of regional documentation. Alternatively, the caves may have been a classified military project, deliberately kept off the books. Or the records may simply be lost, rotted away in some provincial archive that didn’t survive the centuries.

But absence of evidence, as the saying goes, is not evidence of absence. Nor is it an explanation. The gap in the historical record doesn’t tell us what the caves were — it just tells us that whatever they were, someone either didn’t think it was worth recording or made sure it wouldn’t be.

Preservation and Tourism

Since their discovery, the Longyou Caves have become a tourist attraction and a source of local pride. Five of the caves have been opened to the public, with lighting installations and walkways. The site was designated a national-level protected monument in 2001 and has been on China’s tentative list for UNESCO World Heritage status.

But preservation presents its own challenges. The siltstone is soft and vulnerable to erosion. Exposure to air after two millennia underwater has accelerated weathering in some chambers. Algae growth, facilitated by the tourist lighting, has begun to discolor wall surfaces. Researchers from Zhejiang University and the Chinese Academy of Sciences have been studying stabilization methods, but the tension between access and preservation is a familiar one.

Meanwhile, serious archaeological investigation remains limited. The caves have attracted more popular attention than academic funding. Much of the published research comes from Chinese-language journals with limited international circulation. Western archaeological literature has been slow to engage with the site, partly because of access issues and partly because the caves don’t fit neatly into existing frameworks for understanding ancient Chinese infrastructure. They’re not tombs. They’re not temples. They’re not mines. They resist categorization, and academia doesn’t always know what to do with things that resist categorization.

What Remains

Twenty-four caves. A million tons of missing stone. No records. No names. Walls covered in patterns that served no structural purpose, carved by thousands of hands that history forgot.

The Longyou Caves are not the only unexplained underground structure in the world — the Barabar Caves in India, the underground cities of Cappadocia, the Hypogeum of Malta all carry their own unanswered questions. But what sets Longyou apart is the combination of scale, precision, and total documentary absence. These aren’t rough-hewn shelters or natural caverns adapted for human use. They are engineered spaces, built with intention, finished with care, and then apparently abandoned to the rising water table.

Were they completed? Were they ever used for their intended purpose? Or did the political upheaval of the Qin unification — the same upheaval that may have destroyed their records — also halt their construction, leaving them as monuments to a project that never reached its conclusion?

The water that Wu Anai pumped out in 1992 had been sitting in those caves for centuries, maybe longer. It kept the stone intact. It preserved the chisel marks. It also kept the secret. And now that the water is gone, the caves are talking — but in a language we haven’t yet learned to read.