Old Windows Are Thicker at the Bottom. Not Why You Think

Stand in front of an old cathedral window and look closely. In a lot of really old panes you will notice something strange: the glass looks thicker near the bottom and thinner near the top. For more than a century, tour guides, teachers and even some textbooks have offered the same explanation. Glass is secretly a liquid, they say. Not a solid at all, but an incredibly slow-moving fluid, oozing downward under its own weight, one impossibly slow drip at a time. Give it a few hundred years and gravity quietly pulls the glass toward the bottom of the frame. It sounds almost poetic. There is one small problem with that lovely story: it is not true. Glass does not flow. It never oozed down those old window frames, and it is not plotting to slide off your kitchen table while you sleep. Start with the question everyone eventually asks: is glass a solid or a liquid? The honest answer is neither, in the simple sense most of us learned in school. In a typical solid, like ice or a grain of salt, the atoms line up in a neat repeating pattern, like tiles on a floor. In a typical liquid, the molecules are disordered and slide freely past one another. Glass is strange because its atomic arrangement looks disordered, almost like a liquid frozen mid-splash. Scientists call this an amorphous solid — amorphous meaning without a regular repeating shape. So glass has the messy, liquid-like internal structure, but it behaves mechanically like a solid. It is rigid. It holds its shape. It shatters instead of oozing when you drop it. That combination is exactly why the myth took root: if scientists themselves describe the inner arrangement as liquid-like, it is an easy leap to imagine actual visible flowing. But liquid-like structure at the atomic level does not mean liquid-like movement at the human level. Glass reaches that frozen state through the glass transition. When molten glass cools, its atoms do not have time to settle into the neat crystal pattern something like quartz would form. Instead the material grows thicker and more resistant to movement as the temperature drops, until it locks into a rigid structure without ever crystallising. Think of a crowded dance floor where the music slows so gradually that everyone freezes in whatever position they happen to be standing in, rather than filing out neatly through the exits. So does glass flow at all? Here scientists want to be precise rather than smug, because the truth has a small asterisk. Materials scientists have shown that amorphous solids can in theory experience an extremely tiny amount of atomic rearrangement over immense timescales, and there is genuine ongoing debate about the deepest theoretical description of the glassy state. That nuance is real. But do not let it fool you into thinking the cathedral window story has any truth in it. Researchers at Corning tested this directly using genuine medieval glass, including samples connected to Westminster Abbey dating back to the 1200s, calculating how much a pane could possibly have flowed under gravity across its entire lifetime. The result was almost comically small: roughly one nanometre of movement per billion years. A nanometre is about a hundred thousand times thinner than a human hair. Even across the entire seven-hundred-year history of that abbey, the theoretical flow would be far too small to detect with the naked eye, or honestly with most laboratory instruments either. There is also a wording problem at the root of all this. An older phrase used in many textbooks called glass a supercooled liquid, and that label was meant as a technical description of how glass forms — cooling past its melting point without crystallising — not a claim that the material keeps behaving like liquid water forever afterward. Somewhere along the way, readers understandably mixed up the origin story of glass with its ongoing behaviour on a windowsill. So if glass is not the culprit, why do so many old windows genuinely look thicker at the bottom? The real explanation is wonderfully human, and honestly more interesting than the myth. For most of history, before machines could roll out perfectly uniform sheets, panes were made by hand using methods like crown glass and cylinder glass, techniques refined in medieval Europe with crown glass famously perfected by French craftsmen around Rouen in the 1320s. In the crown method a glassblower gathered molten glass on a blowing tube, formed a bubble, attached a solid rod called a pontil to the opposite side, cracked off the blowpipe, then spun the whole piece rapidly, reheating it as needed to keep it workable. Centrifugal force flattened that bubble into a wide circular disk, sometimes five or six feet across. Here is the key detail: the spinning disk was never even in thickness.