A medieval glass piece, no matter how old it is, does not flow. This claim that has circulated for decades in textbooks and chemistry classrooms rests on a stubborn misinterpretation: cathedral stained-glass panes are indeed thicker at the bottom than at the top, but not because the glass sagged under its own weight for eight centuries. The cause lies elsewhere, and it is far more down-to-earth.
À retenir
- Why is cathedral glass thicker at the bottom than at the top?
- How did researchers measure the impossible to validate a theory?
- An eight-century-old myth has just been debunked by equations
A Correct Observation, a False Conclusion
The idea is appealing because it rests on a real and verifiable fact: yes, many panels of Romanesque or Gothic stained glass show uneven thickness, thicker at the base than at the top. But this asymmetry existed from the moment they came out of the furnace, long before the piece was set into its lead frame. The medieval glaziers did not know how to produce flat panes: these sheets were inherently thicker in thickness from their origin.
The technique involved is called crown glass blowing or sleeve blowing. A lump of molten glass was rolled, stretched, and flattened before being formed into a disk and cut into panes. These sheets were thicker at the edges, and installed so that the heavier side ended up at the bottom: artisans placed, deliberately, or by plain practical sense, the strongest part toward the ground to stabilize the whole. It’s construction-site logistics, not a failure of material physics that unravels century after century.
Another source even specifies the structural reason for this choice: it was the manufacturing method of the time that did not allow for truly flat stained glass but produced panes thicker on one side, and the thicker side naturally being placed at the bottom for stability of the window. A stained-glass window weighs a lot. It’s only sensible to have the weight borne by the sturdier area rather than the other way around.
The calculations that bury the myth
Chemistry eventually decided the matter, and not just once. As early as 1998, an article published in the American Journal of Physics directly attacked the legend. The popular notion that medieval cathedral windows thickened toward the bottom by slowly flowing like a liquid does not hold water; even taking into account the specific chemical composition of stained-glass glass, it would take longer than the age of the Universe for the glass to sag perceptibly.
The Brazilian researcher Edgar Zanotto, who led this study, shares an anecdote that reveals the international reach of the myth. He initially thought it was a purely Brazilian legend, before hearing the same story from Argentine colleagues, and later finding it in school textbooks and even in the Encyclopedia Britannica. A reminder of how a well-told false idea can travel faster than scientific truth.
Twenty years later, in 2017, a team combining Penn State and the glass manufacturer Corning sought to push the analysis even further, this time focusing on the exact chemical composition of medieval glass, namely that of the Westminster Abbey stained glass. The result is almost comic: their findings showed that the viscosity of this glass was in fact 16 orders of magnitude lower than previously thought. The ancient glass is indeed far less “frozen” than believed at the atomic scale. And yet the conclusion remains unequivocal: despite these low values, the viscosity of the glass remains far too high to observe a measurable viscous flow over a human lifetime.
The numbers are dizzying. According to the team’s calculations, medieval glass would flow at most about one nanometer per a billion years, i.e., 0.000000001 nanometers per year—a theoretically measurable value but utterly unobservable in practice. The authors themselves do not mince words in their conclusions: “this result confirms that the persistent myth of room-temperature glass flow remains nothing more than a myth.”
Why this story refuses to die
The real culprit behind this confusion is the very definition of glass at the molecular scale, a definition flexible enough to fuel all sorts of approximations. The origin of the myth remains unclear, but the confusion probably arises from a misinterpretation of the amorphous atomic structure of glass, where atoms do not adopt a fixed crystalline arrangement: the structure of the liquid and that of solid glass are very close, but thermodynamically they are not the same. Glass lacks a precise melting point, but a glass transition temperature, and once cooled below this threshold, it retains its amorphous structure while adopting the physical properties of a solid rather than a supercooled liquid. The pedagogical shortcut (“glass is a liquid that flows very slowly”) has come to dominate over scientific nuance.
This story has a merit that is often overlooked: it pushed researchers to measure something almost impossible to measure with precision. To validate their model, the Corning team had to manufacture a sample faithfully reproducing the cathedral glass composition and measure its viscosity at an extreme value—a heroic measurement that required substantial patience, in the words of the study’s principal author. An eight-century-old myth will therefore have at least served this purpose: to push science to extend the boundaries of what it can measure.
Sources : fr.quora.com | forums.futura-sciences.com
