Water and its physical properties — The World, with the Evidence

A large white block of ice floats on the sea. — A recent iceberg in “Iceberg Alley”, near Mawson in Antarctica, photographed in November 1981 by Robert Ricker (NOAA). Dr. Robert Ricker, NOAA/NOS/ORR., Public domain, source

Ice floats. Almost nothing else does that.

Almost every substance contracts as it cools and becomes denser, and its solid form sinks in its liquid form. There are a few exceptions — silicon, bismuth, gallium — but they are rare.

Water is one of them, and it is an exception with enormous consequences.

The anomaly

As it cools, water contracts like everything else — up to a point.

Then, a little above zero, at around four degrees (3.98 °C to be exact), it starts to expand again. This is where fresh water is at its densest.¹

If you keep cooling it, it freezes, and the ice takes up more room than the water that formed it — about a tenth more. That is why a full bottle bursts in the freezer.²

The result: ice is less dense than liquid water. It floats.

We know this from a measurement that is easy to repeat. Take a quantity of water, measure its volume precisely at different temperatures, and plot the curve: the volume decreases as the water cools, reaches a minimum at around four degrees, then rises again. A thermometer and a graduated container are enough. The Genevan Jean-André Deluc had noticed it; the Scot Thomas Charles Hope established it in 1805.³

A word about a phrase you often come across: water is not magic. The word closes the question instead of opening it, when in fact there is a mechanism — and this page sets it out. These properties are not there in order to make life possible: it is the other way round — life developed in a world where water behaves like this, and it took advantage of it.

Why

This behaviour has an explanation, and it requires no magic formula.

Water molecules attract one another, and they do so in a particular way: each one has specific places where it can hook on to its neighbours, rather like a jigsaw piece.

In liquid water, the molecules are constantly moving, hooking on, letting go, hooking on again: they are packed close together, in disorder.

In ice, they become still, and each one hooks on to its neighbours in every direction it can. This forms a regular lattice — and this lattice leaves gaps, like scaffolding. Neatly arranged, the whole takes up more room than in disorder.⁴

Nothing more. No magic: a molecular shape, and a way of hooking on.

What difference it makes

A lake freezes from the top.

In autumn, the surface water cools, becomes denser, and sinks. This continues until the whole lake is at around four degrees. Then the surface water, as it cools further, becomes less dense: it stays at the top, and that is where it freezes. The layer of ice insulates what lies beneath, and the water at the bottom stays liquid, at around four degrees. The creatures that live there survive the winter.⁵

If ice sank, it would build up on the bottom, where almost nothing would melt it. Lakes in cold regions would probably freeze from the bottom up, and never thaw completely.

One clarification, because the scope matters: this reasoning holds for fresh water. Seawater, because of the salt, does not have this density maximum at four degrees — it keeps getting denser until it freezes, at around minus two. Sea ice still floats, because ice is always less dense than water. But the sea does not behave exactly like the lake.⁶

The other properties of water

The way water molecules hook on to one another explains more than why ice floats. It explains almost everything water does, and it is a rare case where a single cause accounts for a long list of effects.

Water dissolves a great many things. Each water molecule has a slightly positive side and a slightly negative side — it is said to be polar — which allows it to surround and separate the particles of a large number of substances. For this reason it is called the universal solvent, which is an overstatement: it dissolves neither oils nor fats, which are not polar. That is why a vinaigrette always separates.
Water heats up and cools down slowly. It takes a lot of energy to raise its temperature, because part of it goes first into undoing the bonds between molecules. As a consequence, the oceans take a long time to warm up and a long time to cool down, which moderates the climate of coasts and delays the seasons by the sea.
Water carries heat away as it evaporates. The molecules that leave are the fastest; those that remain are therefore colder. This is the principle behind sweating, and the reason you feel cold getting out of a swimming pool wet, even in full sunshine.
Its surface behaves like a stretched skin. The molecules at the top are attracted by half as many neighbours as the others, and the surface pulls itself in as tightly as possible. This surface tension allows certain insects to walk on water and makes drops round.
Water rises on its own in a thin tube. It clings to the walls — this is adhesion — and each molecule pulls the one behind it — this is cohesion. Together, the two produce capillarity.⁷

This is what allows sap to reach the top of a tree without any pump. Capillarity alone would not be enough; what really does the pulling is evaporation through the leaves. Each molecule that escapes from a leaf pulls the next one, and the column of water rises unbroken from the roots, sometimes over a hundred metres.

One and the same molecular peculiarity, six seemingly unrelated consequences. This is the kind of convergence that gives confidence in an explanation: it does not account for a single fact, but for many at once.

A case where researchers disagree

One claim is widely repeated: hot water sometimes freezes faster than cold water. It is called the Mpemba effect, after a Tanzanian school pupil who put it to a visiting university lecturer, with whom he published it in 1969 — but the question is much older: it can already be found in Aristotle.⁸

Researchers have tested this many times, and they disagree with one another.

One team repeated the experiment under carefully controlled conditions, examined the data published before them, and concluded — “somewhat sadly”, the authors write — that no convincing observation supports this effect. They note in passing that the question, put this way, is vague and imprecise: hot water can always be made to cool more quickly, by supplying it with more energy.⁹

Another researcher replied that the team had asked the wrong question: the Mpemba effect is not about the rate of cooling, but about when the water freezes. These are not the same measurements.¹⁰

One detail of the first team’s experiment explains why this disagreement has lasted so long. They found that the result depended on where they placed the thermometer: a difference of one centimetre in height could be enough to create a false impression of the Mpemba effect.¹¹

The reason is simple and profound. In a cup that is cooling quickly, the water is not at the same temperature throughout. Talking about “the” temperature of the cup then has only an approximate meaning — and that is precisely what one is trying to measure.¹²

When a quantity has no single value, two careful teams can obtain two different results without either of them being wrong.

The question is a familiar one: what exactly does this figure measure?

So the right answer to give here is neither yes nor no. It is: researchers disagree, and here is what they disagree about. Such examples are rare, and they are valuable.

The authors’ phrase deserves notice: “somewhat sadly”. They would have preferred it to be true. That did not stop them publishing the opposite.

What this site has already established about water

The three facts that follow are established, with their sources, in Entries 1.7, 7.10 and 9.9.

On Venus, under a pressure ninety times greater than ours and at more than four hundred degrees, there is no liquid water: only traces of vapour (1.7). It is not the substance alone that decides: it is the conditions.

Water circulates: our page on the water cycle (7.10) explains that the total amount of water on Earth does not decrease. But it also explains that in a given place — an aquifer, a river — there may be much less of it left than before. “Running out of water” is about the place, not the planet.

Irrigating without draining damages the land: our page on managing water (9.9) tells the story of salinisation, the ancient techniques that avoided it, and an attribution that has never been proven.

Note 1

The density of pure water occurring at 3.98 °C, which is 1.0000 g cm-3, or 1000 kg m-3. Water when cooled down contracts normally until the temperature is 3.98 °C, after which it expands.

Observatoire de Paris, “Maximum density of water”, An Etymological Dictionary of Astronomy and Astrophysics (English-French-Persian)

This means that below 3.98 °C cooling results in the water molecules spacing out and both the liquid and the solid water below this temperature expand.

The Open University, “3.2 The density of fresh water and seawater”, The oceans, OpenLearn

See the full note ↓

Note 2

Upon freezing, the density of ice decreases by about 9 percent.

U.S. Geological Survey, “Water Density”, Water Science School

See the full note ↓

Note 3

Deluc discovered that water attains its maximum density at 39° F (4° C)

Encyclopaedia Britannica, “Jean-André Deluc”

the first to determine, in 1805, that water has its greatest density at about 39.5 degrees F

Linda Hall Library, “Thomas Charles Hope”, Scientist of the Day

See the full note ↓

Note 4

Ice has a very different structure than liquid water in that the molecules align themselves in a regular lattice rather than more randomly as in the liquid form. The lattice arrangement allows water molecules to be more spread out than in a liquid, and, thus, ice is less dense than water.

U.S. Geological Survey, “Water Density”, Water Science School

See the full note ↓

Note 5

In wintertime, the surface of lakes found in many northern regions is frozen. However, the water under the ice is slightly warmer, and the water at the bottom of the lake is warmer yet at 4 °C to 5 °C (39.2 °F to 41 °F). […] As air temperatures drop in the fall, the temperature of the lake water cools to 4 °C; therefore, this causes fall turnover as the heavy cold water sinks and displaces the water at the bottom.

Clark, Mary Ann, Douglas, Matthew and Choi, Jung, Biology 2e, OpenStax, section 44.2 “Biogeography”

In lakes and ponds, ice will form on the water's surface creating an insulating barrier that protects the animals and plant life in the pond from freezing.

Clark, Mary Ann, Douglas, Matthew and Choi, Jung, Biology 2e, OpenStax, section 2.2 “Water”

See the full note ↓

Note 6

Sea ice forms more slowly than freshwater ice for two main reasons. First, the freezing temperature of salt water is lower than freshwater; ocean temperatures must reach -1.8°C (28.8°F) to freeze. Secondly, in contrast to fresh water, the salt in ocean water causes the density of the water to increase as it nears the freezing point.

National Snow and Ice Data Center, “Science of Sea Ice”, Sea Ice section, NSIDC

See the full note ↓

Note 7

While there is no net charge to a water molecule, water's polarity creates a slightly positive charge on hydrogen and a slightly negative charge on oxygen, contributing to water’s properties of attraction.