Go down the steps behind a shop on a hill street and the brickwork stops.
Past the last course the wall turns to bare rock, and the room carries on into it, cut square, the ridges of a hand pick still legible across the ceiling.
Nobody built this room. Somebody removed it.
There are more of these under the pavement than anyone kept track of, and the stone they sit in is soft enough to work with a hand tool. Which is the whole reason a modern site can put a machine straight into one without warning.
Why soft rock lets a machine through
The rock under the city is friable.
It is an Early Triassic sandstone, pinkish red to buff grey, medium to coarse grained, pebbly and cross bedded, and it runs somewhere between 230 and 560 feet thick beneath the county.
Friable is the operative word.
The grains are weakly cemented, so a pick cuts it the way a spade cuts packed earth, and it still holds a shape once cut. A room in this stone needs no lining, no arch and no timber, which is why generations dug rooms instead of building them.
That cuts both ways on a modern site.
A pile rig or a breaker meets the same weak cement, goes through the last few inches of roof and finds air. No rubble layer, no change in resistance, and the first sign is usually that the tool drops away from him.
What one of these rooms was actually for
Almost none of them were dug to be caves.
They were dug to be something, and the something is usually obvious once you are standing in it. A beer cellar sits under a pub with the barrel run sockets still cut into the walls.
A tannery has pits.
A malting has a kiln flue. A sand mine has wide flat chambers and a haulage way to the street, because the point of that one was the sand itself, sold for scouring floors and casting metal.
Then the rooms got reused.
Families lived in some of them during the worst of the industrial crowding. Others were fitted with benches and blast walls and became air raid shelters. The same hole in the rock can carry four hundred years of different jobs, and it rarely carries a sign saying which.
How 425 became 1,000
The register stood at 425 when systematic recording started in the late 2000s.
It reached 800 by 2017 and 924 by May of last year, and the thousandth entry was announced in May.
Most of that growth is archive work.
More than 4,000 historical documents have been read against the map, with a few hundred still outstanding, and each promising reference then waits for a chance to look. Deeds, sale notices and old plans put a room on the list, but a site visit is what confirms it.
A smaller effort put numbers on the shapes.
Between 2010 and 2012 a university team scanned 64 caves with a 3D laser scanner, building point clouds of billions of measurements, funded by a heritage body and a regional partnership. The gear moved between sites in trailers towed by bicycles.
The archaeologist behind the register
The work is done by the city council’s own archaeologist, Scott Lomax, who has been at it for more than fifteen years.
His finding is not that a thousand caves exist. It is that a large number of caves written off as destroyed are still down there, intact, under buildings and roads put up over them.
Recent additions read like a street directory.
A beer cellar of 1830 under Brook Street. A cave beneath Stoney Street in the Lace Market. An early nineteenth century room on Goose Gate. An eighteenth century sand mine under Mansfield Road. The council announcement lists them.
The city also writes the caves into its planning framework.
Developers must investigate and protect what is underneath, which is why so many entries arrive with a site address attached. Rock cut voids turn up under other cities too, from a tunnel near Jerusalem to the cavern under Sydney.
What the register still does not show
A thousand is a floor, not a total.
Every unread reference is a possible entry, the count has climbed for fifteen straight years with no sign of leveling off, and nobody involved describes the list as finished.
It is also not a map of the ground.
An entry can mean a confirmed and measured room, or a strong documentary case for a room that nobody has stood in since the deed was written. Those are very different things to a structural engineer, and the register does not always separate them for you.
The naming is unsettled too.
The rock these sandstone caves are cut into was long filed as the Nottingham Castle Sandstone Formation, and the geological survey has since folded that name into the Chester Formation, so older reports and newer ones can describe the same stone under two different headings.
