Why old window glass has uneven thickness

Do old windows become thicker at the bottom because the glass slowly runs downward? That is the familiar explanation the author wants to challenge. We should separate the observation from the proposed cause. Some old panes have uneven thickness; it does not follow that the material flowed after the panes were installed.

The first paragraph explains how the misconception may have arisen. Glass does not have the fixed crystal structure people might expect in a solid. Its atoms remain arranged in an amorphous way even after it cools. That makes liquid and solid glass similar in one structural respect, but similarity in atomic arrangement does not make their physical behavior identical.

The transition also differs from the familiar idea of water freezing at one precise temperature. Glass changes over a range, typically spanning a few hundred degrees Celsius. Below the lower end, it takes on the physical properties of a solid while retaining its amorphous structure. So the important distinction is not simply fixed crystals versus flowing liquid. Glass can lack that fixed arrangement and still behave as a solid.

The new study then addresses the window explanation directly. Certain solids, including glass, can flow slightly under gravity. That admission might initially seem to help the popular belief, but the next question is how much movement could occur in the available time. Zanotto calculates that medieval cathedral glass would need longer than the age of the universe to change thickness noticeably. A cathedral that is only several hundred years old has not supplied anything close to that time.

This is the useful distinction: possible in principle does not mean sufficient to explain the observed result. We do not need to claim that solid glass never moves. We need to ask whether its movement could account for these visibly uneven windows.

Composition affects the rate. Even the relatively fluid germanium oxide variety would need a period measured in trillions of years before visible sagging occurred (paragraph 3). Impurities in medieval glass could speed the process somewhat, but not enough to rescue the explanation. A faster rate can still be far too slow for the event we are trying to explain.

The research also supports what many scientists had already reasoned. Glass needs heating to at least 350 degrees Celsius to have more than a negligible ability to flow. The ordinary window is not undergoing the sort of heating that would make substantial flow plausible. The study adds calculations to an existing scientific reason for rejecting the myth.

Then why are some panes thicker at the bottom? The final paragraph supplies the alternative that the earlier discussion left us wanting: they probably began uneven. Older glassworkers blew molten glass into a globe, flattened it into a disk, and whirled it. Picture the resulting pane being cut from material with ripples and thicker edges. Those differences arise during manufacture, rather than through centuries of later sagging.

Installing the thicker portion at the bottom would make structural sense. The author presents that as a reasonable explanation, not a report that every medieval installer followed an identical documented rule. Later, drawing glass into sheets made the thickness more uniform; today's floating process on molten tin produces an extremely flat surface.

The manufacturing history therefore has a job in the argument. It provides a probable cause for the observation after the flow explanation has been ruled out on the relevant timescale. The main point combines those two steps: old glass did not noticeably sag into its present shape; older production methods probably created that shape.