You've heard that glass is a super-slow liquid that oozes down over centuries, which is why old windows are thicker at the bottom. It's a great story and completely wrong. Glass is a solid — its molecules are locked in place, just jumbled instead of neatly stacked. Old panes are thick at the bottom because glassmakers couldn't make them perfectly even, and they installed the heavy edge down. Glass doesn't flow.
Most people think old cathedral glass is thicker at the bottom because glass slowly flows downward. In fact glass is an amorphous solid, jumbled like a liquid but frozen rigid, and room-temperature creep would take longer than the age of the universe.
What's actually happening
It is one of the most repeated pieces of science trivia: visit an old cathedral, and a guide will point to a window that is visibly thicker at the bottom and explain that glass is really a liquid, frozen in ultra-slow motion, that has oozed downward over the centuries. It sounds plausible, it has the satisfying shape of a hidden truth, and it is wrong in every part. Glass does not flow at room temperature, not slowly, not at all on any timescale that matters.
The confusion comes from what glass actually is. Most solids are crystals — their atoms sit in a neat, repeating grid. Glass is different: its molecules are jammed together in the disordered jumble of a liquid, but frozen solid, locked rigidly in place. This in-between nature (an "amorphous solid") is why people reach for the "liquid" label. But "arranged like a liquid" is not "behaves like a liquid". Mechanically, glass is as solid as a rock. Careful estimates of how fast it could possibly creep at room temperature give numbers vastly longer than the 14-billion-year age of the universe — a medieval window has had nowhere near enough time to sag a measurable hair.
So why are the old panes uneven? Because of how they were made. Before modern float glass, window panes were spun out by hand (crown glass) or blown and flattened, producing sheets that were genuinely thicker on one side. Glaziers, sensibly, installed the heavier edge at the bottom for stability — so most old windows are thick-at-the-bottom by choice, and a fair number are thick at the top or sides, which the "flow" theory can't explain at all. The clinching evidence: Roman and Egyptian glass, far older still, shows no sagging whatsoever. The window didn't flow. The glassmaker just couldn't make it flat.
Glass doesn't flow: old panes are uneven because of how they were made and mounted, and Roman glass ten times older shows no sag at all.
- 1Next time you hear the cathedral-window story, ask: if glass flows down, why are some old panes thicker at the top or the side?
- 2And: why does 2,000-year-old Roman glass show no sag, when it's had ten times longer than any cathedral window?
- 3Neither question has an answer under the "flow" theory — because the unevenness was there the day the glass was made.
Common questions
Not on any timescale that matters. Careful estimates of room-temperature creep give numbers vastly longer than the 14-billion-year age of the universe, so a medieval window has had nowhere near enough time to sag a measurable hair.
Before modern float glass, panes were spun or blown by hand and came out genuinely thicker on one side. Glaziers installed the heavier edge at the bottom for stability — but a fair number are thick at the top or sides, which the flow theory cannot explain.
Yes. Roman and Egyptian glass, far older than any cathedral window, shows no sagging whatsoever. The window did not flow; the glassmaker just could not make it flat.