Glass: Liquid or Solid -- Science vs. an Urban Legend

©1996Florin Neumann


I compiled this short article to address the widespread urban legend according to which glass is a liquid. It consists chiefly of a selection of quotes from recent works by specialists in materials science which state unambiguously that glasses are amorphous solids. I also speculate that at the origin of the legend may have been a misreading/mistranslation of an influential paper by Gustav Tamman. Additionally, I appended to the article an extract from an ASTM method of discriminating between a liquid and a solid.



"What is glass... is it a liquid or a solid?"

This question is a perennial staple of science-oriented online discussions among non-scientists. It is caused by a widespread urban legend [1], which has wormed its way even in some high school textbooks, and has managed to fool even apparently intelligent and well-informed laymen.

An Urban Legend

The legend usually appears in any of the following forms:

The prolonged survival of this legend, chiefly among English speakers (and particularly among North Americans) is puzzling -- especially when one considers that glass and glassy materials are readily available, and one can easily verify if one can pour a gallon of glass, or drain a pint of obsidian.

The Antique Windowpanes Story

The question of antique windowpanes has been addressed by Plumb, 1989[2]. He noted the following:

[...W]hy are the panes of antique window glass thicker on the bottom than the top? There really are observable variations in thickness, although there seem to have been no statistical studies that document the frequency and magnitudes of such variations. This author believes that the correct explanation lies in the process by which window panes were manufactured at that time: the Crown glass process.

In other words, while some antique windowpanes are thicker at the bottom, there are no statistical studies to show that all or most antique windowpanes are thicker at the bottom than at the top. The variations in thickness of antique windowpanes has nothing to do with whether glass is a solid or a liquid; its cause lies in the glass manufacturing process employed at the time, which made the production of glass panes of constant thickness quite difficult.

What the Scientists Say

But is glass solid or liquid? For those who can't trust their own judgment or the evidence of their own senses, this section contains a selection of quotes from distinguished scientists who specialize in the study of amorphous solids in general and glassy materials in particular.

I selected them with three things in mind:

  1. All of them were extracted from papers published in the past 5 years (1991 or later). They represent essentially the current scientific consensus in this field.

  2. None of them require in order to be understood more knowledge of physics than I would expect from a high-school graduate. Materials science and thermodynamics are highly formalized disciplines and have complex technical jargons. It would be all too easy to intimidate any layman or even scientists trained in other fields by showing off something like the Debye-Scherrer equation or alluding to the "nonergodicity of glass".

  3. All quotes (except one) are from monographs dealing with the subject of glass or amorphous solids, which, for those who wish to do a bit of reading on their own, should be more easy to find than specialized journals.

And now, to our work alive.

Aristotle's teachings on the elements [...] have survived into our times in the form of the generally accepted classification of substances by their state of aggregation. The solid, liquid, and gaseous states have been supplemented by the addition of a fourth state of matter, plasma. [...] However, a new classification may seem sensible against the background of results obtained from [modern] structural research, [which would] differentiate between predominantly ordered crystalline solids and substances of disordered structure.

[Feltz, 1993, p. 1]

Note that some physicists do not accept plasma as a distinct fourth state of aggregation, while others propose more than four states. It has been suggested that glasses form a distinct state of aggregation, the vitreous state. This proposition has been generally ignored (see Gutzow and Schmelzer, 1995).

Glass is an amorphous solid. A material is amorphous when it has no long-range order, that is, when there is no regularity in the arrangement of its molecular constituents on a scale larger than a few times the size of these groups. [...] A solid is a rigid material; it does not flow when it is subjected to moderate forces [...].

[Doremus, 1994, p. 1]

Glass includes all materials which are structurally similar to a liquid. However, under ambient temperature they react to the impact of force with elastic deformation and therefore have to be considered as solids.

[Pfaender, 1996, p. 17]

Amorphous substances, like crystalline solids, are usually characterized by certain areas of short-range order [...] A long-range order, as in crystals, does not exist in amorphous substances. The designations 'amorphous' and 'noncrystalline' describe the same fact [...].

Glasses are noncrystalline or amorphous substances. Nevertheless, the term vitreous state is restricted to (i) solids obtained from melts, or (ii) solids produced by other methods and obtained in a compact form or as thin coherent films [...].

Glasses have numerous properties in common with crystalline solids, such as hardness and elasticity of shape [...]. The term 'amorphous solid state' has a more comprehensive meaning broader than that of the 'vitreous state'. All glasses are amorphous, but not all amorphous substances are glasses.

[Feltz, 1993, pp. 7-8. Author's emphasis.]

An amorphous (or synonymously, non-crystalline) material can be defined as one which is topologically disordered and which does not exhibit either the long-range translational order (periodicity) characteristic of single crystals, or the long-range orientational order characteristic of quasicrystals. Within this definition, such materials could be either solid or liquid, and this distinction is essentially simply one of timescale. A material is a solid when there is no observable long-range translational diffusive motion during the duration of the experiment; in other words, dynamic disorder is absent.

A glassy (or synonymously, vitreous) material is an amorphous solid that exhibits a glass transition. (Thus, by definition, all glasses are amorphous, but not all amorphous solids are necessarily glassy.) The glass transition is marked (as a function of temperature) either by a change in slope of extensive thermodynamic quantities (e.g., volume or entropy) or, equivalently, as a discontinuity in derivative quantities (e.g., specific heat or thermal expansivity).

[Elliott, 1994, pp. 75-76]

Most liquids crystallize rapidly at a well-defined temperature Tf (melting point or liquidus temperature) with a marked change in volume--usually a decrease. If the melt is completely free from crystal nuclei or foreign particles, it can be supercooled to some extent [...]. Glass-forming melts can be supercooled to an unusually high degree, even when nuclei are present. [...] The viscosity of the supercooled melt continues to increase as the temperature is reduced until a range of temperatures [around a point called Tg] is reached, below which the material is for most practical purposes a solid. This [range] is called the transformation range [...]. Only below this range is correct to refer to the material as glass. [...] Below Tg a glass, like a supercooled liquid, has a higher free energy than a crystalline phase or a mixture of crystalline phases. However, because structural rearrangements can occur only extremely slowly at temperatures well below Tg, the glass is stable for all practical purposes.

[Rawson, 1991, pp. 1-2]

[...] glasses are frozen-in non-equilibrium systems. Non-equilibrium systems cannot be described in the framework of classical thermodynamics [...].

[Gutzow and Schmelzer, 1995, p. 67]

As kinetically frozen forms of liquid, glasses are characterized by a complete lack of long-range crystalline order and are the most structurally disordered types of solid known.

[Jeanloz & Williams, 1991, p. 659]

The structure of vitreous and amorphous solids is based on short-range order. This has now become established beyond doubt. The bonding properties of atoms lead to the formation of defined subassemblies with a certain symmetry. These are often identical to the short-range order units already known from the structural analysis of crystalline forms of the corresponding compounds.

[Feltz, 1993, pp. 89-90]

The structural-chemical conditions for glass formation can be regarded as a combination of energetic interactions and steric factors [...]. It has been found that a degree of covalent bonding is an essential feature of all glass-forming substances.

[Feltz, 1993, pp. 95-96]

[...] Amorphous solids are characterized by a topological disorder, so there is no long-range order (or periodicity) in their structure. However, this does not mean that amorphous solids are structurally completely random (i.e., gas-like) at all length scales. In fact, covalent materials, in particular, exhibit a rather high degree of structure organization at length scales corresponding to several atomic separations [...]. On the other hand, materials characterized by non-directional centro-symmetric interatomic interactions, e.g. metals or completely ionic materials, are intrinsically much more disordered even at short length scales.

[Elliot, 1994, p. 79]

Origin of an Urban Legend?

How did the "glass is a supercooled liquid" urban legend originate? It is possible it began with an erroneous reading of an influential book by Gustav Tammann (1861-1938), a German physicist who was among the first to study glass as a thermodynamic system (Tammann, 1933). I was unable to locate a copy of Tammann's book to verify this, so the following is speculation. One or two papers I consulted attributed to Tammann the statement "Glass is a supercooled [or undercooled] liquid." But, from other papers, it appears that what Tammann actually wrote was "Glass is a frozen supercooled liquid" [my emphasis]. My speculation is that an author misquoted Tammann, and this misquotation was repeated by later authors who, since copies of Tammann's book are rather rare, did not refer directly to Tammann.

Until about 20 years ago supercooling a glass melt was the only way to obtain glass, and the behaviour of melts as they passed through the glass transition (i.e., solidified) was very different from crystallization. But solid-state physics was almost entirely based on the study of crystalline solids, which made the behaviour of glass melts appear paradoxal. To emphasize this a professor would state "Glass is a liquid which has lost the ability to flow", and some undergraduate, with his mind more on the Friday night date than on the physics of glass, would remember only "glass is a liquid"... Perhaps now we can finally put this legend to its well-deserved rest.


Glasses are amorphous solids. There is a fundamental structural divide between amorphous solids (including glasses) and crystalline solids. Structurally, glasses are similar to liquids, but that doesn't mean they are liquid. It is possible that the "glass is a liquid" urban legend originated with a misreading of a German treatise on glass thermodynamics.


  1. See the alt.folklore.urban FAQ ('Frequently Asked Questions' document) on the World Wide Web or by ftp.

  2. Also available online, courtesy of Bryan Derksen.


Doremus, R. H. (1994) Glass Science, 2nd Edition. John Wiley & Sons, New York, 339 pp. ISBN 0471891746.

Elliott, S. R. (1994) Amorphous Solids: An Introduction. In: Catlow, C. R. A. (eds.), "Defects and Disorder in Crystalline and Amorphous Solids", NATO Advanced Studies Institutes Series; Series C, Mathematical and Physical Sciences, 418, Kluwer Academic Publishers, Dordrecht: 73-86. ISBN 0792326105.

Feltz, A. (1993) Amorphous Inorganic Materials and Glasses. VCH Verlagsgesellschaft mbH, Weinheim/VCH Publishers, New York, 446 pp. ISBN 3527284214/1560812125.

Gutzow, I. and Schmelzer, J. (1995) The Vitreous State: Thermodynamics, Structure, Rheology, and Crystallization. Springer Verlag, Berlin, 468 pp. ISBN 0387590870.

Jeanloz, R. and Williams, Q. (1991) Solid-State Physics: Glasses Come to Order. Nature, 350: 659-60.

Pfaender, H. G. (1996) Schott Guide to Glass, 2nd Edition. Chapman & Hall, London, 207 pp. ISBN 0412719606.

Plumb, R. C. (1989) Antique windowpanes and the flow of supercooled liquids. Journal of Chemical Education, 66(12): 994-996.

Rawson, H. (1991) Glasses and Their Applications. Royal Institute of Metals Book, Royal Institute of Metals, London, 499: 166 pp. ISBN 0901462896.

Tammann, G. (1933) Der Glaszustand. Voss, Leipzig, 123 pp.


The following is an extract from ASTM Standard D4359-90: "Standard Test Method for Determining Whether a Material Is a Liquid or a Solid". It prescribes a relatively simple test to determine whether a material whose flow characteristics are in doubt is liquid or solid. The test requires fairly simple implements and can be carried out at home or in a school lab. (For more information on ASTM visit their home page).

ASTM (1996) "D4359-90: Standard Test Method for Determining Whether a Material Is a Liquid or a Solid."


In: ASTM "Annual Book of ASTM Standards", Section 6, Vol 6.01, pp. 500-501. American Society for Testing and Materials, West Conshohocken, PA. ISBN 0803122705 [Set] 0803123094 [Section] 0803123108 [Volume].


3. Summary of Test Method.

The material under test is held at 100deg.F (38deg.C) in a tightly closed can. The lid is removed and the can inverted. The flow of the material from the can is observed to determine whether it is a solid or a liquid.


5. Apparatus.

5.1. Can, quart (1L), friction top with lid, diameter approximately 4'1/4 in (108 mm) height 4'3/4 in (120 mm)

5.2. Oven, maintained at temperature of 100deg.F (38deg.C)

5.3. Tripod or ring stand

5.4. Watchglass, tared

5.5. Stopwatch

5.6. Ruler.


9. Interpretation of results.

9.1. A material that flows a total of 2 in (50 mm) or less within 3 min is considered a solid. Otherwise it is considered a liquid.