Glass is a slow-flowing liquid
Glass is an amorphous solid, not a liquid. The frequently cited observation that old medieval windows are thicker at the bottom reflects manufacturing techniques of the era, not glass slowly flowing under gravity. Physical calculations confirm that room-temperature glass cannot flow on any meaningful timescale, including over centuries.
What we know
The claim that glass is technically a very slow-flowing liquid is a persistent piece of popular science trivia, often supported by pointing to old European church and cathedral windows that are noticeably thicker at the bottom edge than at the top, with the explanation offered that gravity has caused the glass to gradually flow downward over the centuries since installation. This explanation is intuitively appealing and widely repeated, but it does not survive scrutiny of either glass physics or the actual historical manufacturing record.
Materials scientists classify glass as an amorphous solid, a state of matter that lacks the long-range, repeating crystalline atomic structure found in materials like ice or metal, but that nonetheless behaves mechanically as a rigid solid under normal conditions, maintaining its shape indefinitely and resisting deformation the way any solid does, rather than flowing or deforming under its own weight the way an actual liquid would. The often-cited comparison of glass to a "frozen liquid" refers to the disordered molecular arrangement inherited from the glass's molten manufacturing state, a description of atomic structure, not a claim about its ongoing physical behavior at room temperature.
Physicists have directly calculated the theoretical viscosity of glass at room temperature and found it to be so extraordinarily high, many orders of magnitude greater than even extremely viscous liquids like pitch or tar, that any actual gravitational flow would require timescales vastly longer than the age of the universe to produce even a barely detectable amount of movement. This calculation directly rules out the possibility that a window installed several centuries ago could show any measurable, gravity-driven thickness change within a human historical timeframe, regardless of how visually persuasive the old-window observation might seem.
The real explanation for uneven thickness in historic windows lies in the manufacturing methods used at the time. Medieval and early modern glassmakers commonly used techniques such as the crown glass method, in which molten glass was spun into a disc, naturally producing a sheet that was thicker near the center (where the initial molten blob was attached to the spinning rod) and thinner toward the edges. When artisans cut this unevenly thick disc into individual window panes, the resulting rectangular pieces of glass were themselves unevenly thick from one edge to the other, purely as an artifact of the original disc's geometry. When glaziers installed these panes into window frames, evidence suggests that they often, though not universally, installed the thicker edge at the bottom, whether for practical structural stability, aesthetic preference, or simply convenience, meaning the thickness variation observed today was already present in the glass at the time of installation rather than developing gradually afterward.
Materials science research examining the properties of ancient glass artifacts directly, including detailed structural analysis of glass objects that are thousands of years old, such as Roman-era glass, has found no evidence of gravity-driven flow deformation even over these far longer timescales, providing direct empirical confirmation to accompany the theoretical viscosity calculations. The myth persists partly because the phrase "glass is a slow-moving liquid" is a memorable, almost poetic-sounding piece of trivia, and partly because the visual evidence of old windows really is uneven in a way that seems, at first glance, to support the story, even though the true explanation lies entirely in how the glass was made and installed centuries ago rather than in any physical process continuing since then.
Common claims
- Glass is technically a very slow-flowing liquid.Not supported
- Old church windows are thicker at the bottom due to gravity flow.Not supported
- Glass is an amorphous solid.Accurate
- Medieval glassmaking methods explain uneven pane thickness.Accurate
Evidence hierarchy
All sources
- Is glass a liquid or a solid?University of California, Berkeley physics department · 2010
- Crown glass manufacturing historyCorning Museum of Glass · 2019
- Viscosity and structural relaxation in glassJournal of Chemical Physics / AIP Publishing · 2011
- Ancient glass structural stability studiesJournal of Non-Crystalline Solids / ScienceDirect · 2016