Every museum built in the last thirty years has an atrium, and most of them have the same one: full height, roof-lit, with a stair doing something expressive in the middle of it. It is the space that appears in the competition render and on the cover of the annual report. It is almost never the best room in the building.
In this case the best room is a top-lit hall of about 180 square metres, reached from a corridor the public is not encouraged to use, and it is where the institution actually works.
What the hall does
Nominally it is a handling and staging space: crates arrive, works are unpacked, conditions are checked, and exhibitions are laid out flat before they go up. It has a north-facing roof slot running its full length, a floor that takes a pallet truck, and no services in the ceiling because everything is fed from a perimeter trunking run at waist height.
That combination — even daylight, a clear ceiling, and a floor you can abuse — turns out to be the most useful set of properties a room in a museum can have. Curators use it to rehearse hangs. Conservators use it because the light is consistent enough to judge surface. Education uses it in the mornings when nothing else is happening. The building’s own staff describe it, without irony, as the nicest space they have.
None of that was the brief. The brief said “unpacking”.
Why the galleries cannot do this
The galleries are good galleries. They are also, by design, controlled to within a narrow band: 50 lux on paper, tight humidity, no daylight without a filter and a blind and a sensor. Those constraints exist for excellent reasons and they make the rooms unsuitable for anything except looking at objects under supervision.
The atrium has the opposite problem. It has plenty of daylight and no control at all, which makes it usable for events and useless for work. It is also acoustically live — a full-height volume with hard surfaces on five sides — so any activity in it is audible in the two galleries that open off the balcony.
The staging hall sits between the two: daylight without direct sun, a volume tall enough to feel generous and low enough to stay quiet, and a specification that tolerates use. It is a workshop that happens to be beautiful, which is the oldest good idea in museum architecture and the one most reliably cut.
The generosity was structural, not intentional
It is worth being precise about how the hall came to be as good as it is, because the honest answer is that nobody designed it that way.
- The roof slot exists because the plant room above needed to be set back from the parapet, leaving a void that was cheaper to glaze than to fill.
- The clear ceiling exists because the services engineer moved the ductwork to a perimeter route to avoid a structural downstand.
- The floor specification came from a loading requirement for the crate lift, not from any thought about durability.
Three unrelated decisions, none of them about quality of space, produced the room everyone likes. This happens more often than the profession admits, and it is the strongest available argument for tolerance in a design: leave the generous option in place even when the reason for it disappears.
What a second version should keep
If the institution builds an extension, and it is planning to, the temptation will be to reproduce the hall as a public space. That would probably ruin it. Its usefulness depends on being unbookable, uncatalogued, and slightly out of the way.
What is worth reproducing is the specification: north light from a slot rather than a lantern, a ceiling with nothing in it, a floor with no finish to protect, and services at handling height. Those four moves are cheap. They are also almost impossible to justify in a value engineering meeting, because each one reads as an absence rather than a feature.
The room does not appear on the visitor plan. It should probably stay that way. But it belongs in the drawings that get published, because it is the part of the building that answers the question the atrium only poses.
