Wooden Liars of the Windowsill: Britain's Enduring Weather Houses
There is a certain kind of British shop window that has never quite joined the twenty-first century. Between the wind chimes and the 'Gone Fishing' plaques sits a little chalet-shaped box with two doors and two occupants: a woman in a dirndl and a man in a hat, balanced on a tiny see-saw, waiting to be right about the weather.
They are almost never right. That is rather the point. The weather house is not a scientific instrument so much as a small wooden argument you can hang on a wall, and Britain — a country that has made a national sport of discussing drizzle — has been quietly collecting them for well over a century.
A Very Small House With Two Opinions
The design is almost always the same. A pitched roof, a balconied front, two arched doorways. When the air is dry, the man strides out. When it turns damp, he retreats and the woman emerges, umbrella implied. Some models reverse the genders, which has caused family arguments in gift shops since roughly 1900. A few posh versions have a barometer in the gable and a thermometer down the side, as though the little couple needed management.
They turn up everywhere. Garden centres in Shropshire. Hardware shops that also sell rat traps. Static caravans in Rhyl. A café in Northumberland where the weather house is bigger than the specials board. They are the furniture of a damp island.
A Short History of Blaming the Continent
The weather house is usually traced to the Black Forest region of Germany, where woodcarving, cuckoo clocks and a healthy trade in whimsical forecasting all flourished from the eighteenth century onwards. Victorian British travellers brought them home as souvenirs, and by the early twentieth century they were being turned out by the thousand, some imported, some made here.
What's charming is how little the object has changed. A household in 2026 can own a weather house that is functionally identical to one bought by a holidaymaker in 1926 — same mechanism, same stubbornness, same faint smell of varnish. There are few things on a British high street you can say that about.
How the Thing Actually Works (Spoiler: String)
Here is the whole trick, and it is a good one. Inside the house, a length of catgut — or horsehair, or in cheaper modern versions a synthetic strand — is fixed at one end and twisted under tension at the other. A short arm connects it to the pivot that carries the two figures.
Catgut is hygroscopic. In dry air it shrinks and untwists, letting the man swing forward. In humid air it absorbs moisture, lengthens and twists the other way, sending him back inside and the woman out. It is not predicting rain. It is reacting to the damp already in the air, which is a subtly different and much more British arrangement: the forecast arrives after the weather has already made up its mind.
Some models use a bi-metallic strip instead, which responds to temperature rather than humidity — which is why a few weather houses cheerfully announce sunshine in a downpour, having mistaken warmth for dryness.
Why Yours Is Probably Lying
Three common failures, in order of frequency.
- The string has died. Catgut perishes. After a decade or three it stops responding to anything, including weather, and the figures settle wherever gravity left them in 1998.
- The pivot is jammed. Dust, warping, a spot of overenthusiastic repair with superglue. The figures are stuck mid-argument.
- It never worked. A significant number of weather houses sold today are purely decorative, with the figures pinned or glued in place. If your man has never moved, not once, in four years, he is not forecasting. He is a figurine with a job title.
The Kettle Test
You can settle the matter in ten minutes, and it makes an excellent excuse to stop tidying up.
- Let the house sit at room temperature for an hour and note which figure is out.
- Boil the kettle and hold the house a safe distance above the steam — close enough to feel the warmth and damp, far enough that nothing warps or scalds. A washing-up bowl of hot water and a tea towel tent works just as well and is gentler.
- Give it three to five minutes. A working string should swing the figures across within that window.
- Now dry it: a hairdryer on the cool setting, or an hour on a sunny windowsill. It should swing back.
No movement in either direction means the mechanism is gone. Movement in one direction only means the string is tired and the pivot is stiff. Movement within seconds means you own a bi-metallic version and you should adjust your expectations accordingly.
Three Windowsills Worth a Wander
Names withheld, because the pleasure of these places is stumbling on them.
- A garden centre off a fast A-road in north Cornwall where a whole wall of weather houses has been turning, out of sync, since the 1980s. Watching sixteen of them disagree is genuinely mesmerising.
- An ironmonger in a Suffolk market town with a window display that appears not to have been rearranged since about 1974 — weather house, seed potatoes, one mousetrap, a hand-written card reading 'WORKS'.
- A seafront café on the north Wales coast where the weather house hangs by the door and the owner swears by it, despite all available evidence.
In Defence of the Unreliable
A barometer tells you what the atmosphere is doing. A weather house tells you what a piece of string is doing, and then dresses it up as a man in a hat. In an age of hyper-local forecasts delivered to your pocket with 90% confidence, there is something restful about a device that simply reacts, damply, and gets it wrong with style.
Buy one. Test it with a kettle. Ignore it. That, historically, is what Britain has always done.