Old window glass is not sagging, it was made that way

The theory of flowing window glass is well covered elsewhere. This is about the version you meet in practice.

The observation that started it

Walk around an old building and the glass in the oldest windows is often visibly thicker along the bottom edge than along the top. The explanation almost everyone reaches for is that glass is really a very slow liquid, creeping downwards over the centuries under its own weight.

It is a genuinely appealing idea, because it makes a solid object secretly dynamic and turns a window into a clock. The observation is real and worth explaining, which is why the myth deserves better than a flat denial. The problem is that the explanation predicts things that old buildings do not show, and it contradicts what glass is.

How the panes were actually made

Before industrial float glass, one common method was to blow a bubble, open it, and spin it on a rod until centrifugal force flattened it into a large disc. A disc formed that way is thin near the centre and thicker towards the outer rim, because the material is being flung outwards as it flattens.

Panes were then cut from the disc, so almost every pane came out with one edge measurably heavier than the other. A glazier fitting a heavier edge into a frame will normally put the thick side at the bottom, because it seats more stably and sheds water better. That single working habit accounts for the pattern people notice, without asking anything unusual of the material itself.

The panes that point the wrong way

If gravity were the cause, the thick edge would always be at the bottom, since gravity does not consult the glazier about orientation. In practice surveys of old windows turn up panes seated thick edge up, thick edge sideways, and every orientation in between. A single window with the heavy edge at the top is enough to sink the flow explanation, and such windows are not rare.

Where the story actually begins, cylinder glass, made by blowing a long tube and cutting it open, produces a different thickness pattern again and shows the same random orientations. Whenever a proposed mechanism should produce a strict rule and the evidence shows a mixture, the mechanism is usually not the one at work.

What glass actually is

Glass is an amorphous solid, meaning its atoms are frozen into a disordered arrangement rather than a repeating crystal lattice. Disorder at the atomic level is why glass has no sharp melting point and softens over a range, which is what makes it workable.

Calling it a liquid because of that disorder confuses structure with behaviour, since the atoms in room-temperature glass are locked in place and are not sliding past one another. Estimates of how long a window pane would take to visibly deform at ordinary temperatures come out at figures vastly longer than the age of the universe.

Any process that slow is not a process at all in a building that has stood for six hundred years.

The evidence from much older glass

Roman glass vessels, some of them nearly two thousand years old and far thinner than a window, have been recovered without the sagging the theory would demand. Ancient glass in museum cases sits on shelves for decades and telescope mirrors hold their figure to fractions of a wavelength for a lifetime. Optical instruments would be impossible if glass crept, because a lens that flowed would slowly ruin the image it was ground to produce.

Tested properly, the absence of drooping in objects far older and far more delicate than a window is the quiet counter-evidence that settles the question. Old glass does change, through surface weathering and chemical attack, but that is corrosion rather than flow and it does not move mass downwards.

Why the story is so persistent

It survives because it explains something people genuinely see, and a wrong explanation for a real observation is far stickier than a wrong observation. It also carries a pleasant piece of hidden-world drama, in which ordinary objects turn out to be secretly in motion.

The primary source says otherwise: the correct answer is less mystical but arguably better, because it makes an old window a record of a craftsman spinning molten glass on a rod. There is a legitimate technical sense in which glass is described as a supercooled liquid, and that phrase has done a lot of the damage. The distinction worth keeping is that the phrase describes how glass was formed and structured, not what it is doing in the wall this afternoon.

The takeaway

The window is not moving. It has simply kept the shape a spinning rod gave it three centuries ago.

Believing it was ordinary. Continuing to is the avoidable part.