The sugar stages measure water content, not heat

A thermometer in a syrup is measuring concentration

The named stages of sugar cookery look like a list of temperatures and are really a list of concentrations. Pure water boils at 100°C at sea level. Dissolve something in it and the boiling point rises, because the dissolved molecules interfere with the escape of water molecules into vapour, and the more concentrated the solution the higher that boiling point climbs. A sugar syrup therefore reports its own composition through its temperature.

As the syrup boils, water leaves as steam and the sugar stays behind, so the solution becomes steadily more concentrated and the boiling point rises with it. When a recipe tells you to take a syrup to 118°C, it is not asking for that heat for its own sake. It is asking for the water content that produces that reading, because that is what determines how the syrup behaves once it cools.

The traditional names describe cooled behaviour

Before thermometers were common in domestic kitchens, cooks tested syrups by dropping a little into cold water and handling the result, and the names survive because they describe exactly what the cooled sugar does. Thread stage, around 110°C, pulls into a fine strand. Soft ball, roughly 112 to 116°C, collects into a lump that flattens between the fingers — the concentration behind fudge and Italian meringue. Firm and hard ball, up to around 130°C, give progressively more resistant lumps, which is caramel-chew and nougat territory.

Soft crack, near 135 to 143°C, produces threads that bend before they break. Hard crack, roughly 149 to 154°C, snaps cleanly and is where brittles, hard sweets and spun sugar live. Beyond that the sugar has almost no water left, and at around 160°C and above it stops merely concentrating and begins to decompose, which is caramelisation proper. The cold-water test is still a perfectly reliable method, and it has one advantage over a thermometer: it tests the syrup rather than the instrument.

Why the last few degrees move so fast

Anyone who has watched a syrup climb has noticed that the early stages take an age and the late ones arrive in seconds. That is a direct consequence of what is being measured. At the beginning there is a great deal of water and a large amount of energy goes into evaporating it while the temperature barely moves. As the water runs low, each further degree requires much less evaporation, and the reading accelerates.

This is why the difference between a syrup for a chewy caramel and a syrup for a brittle can be under a minute of inattention, and why sugar work rewards having everything prepared before the pan goes on the heat. It also means altitude matters. Water boils lower where the air pressure is lower, so every stage temperature drops by roughly a degree for each three hundred metres or so of elevation, and a recipe written at sea level will overcook a syrup in the mountains.

Crystallisation is the failure mode that stalks all of this

A boiling syrup is a concentrated solution looking for an excuse to return to crystals, and a single seed will do it. A grain of undissolved sugar clinging to the side of the pan, a speck of dust, an aggressive stir at the wrong moment — any of these can start a chain of crystallisation that turns a clear syrup into an opaque grainy mass in seconds.

The standard defences all attack the same problem. Washing the pan sides down with a wet brush removes stray crystals. Not stirring after the sugar has dissolved avoids agitating the solution. Adding an interfering sugar — glucose syrup, or an acid such as lemon juice or cream of tartar that inverts some sucrose into glucose and fructose — puts molecules of a different shape into the mixture, and they physically obstruct the neat lattice that sucrose crystals need. That is the mechanism behind the corn syrup in so many sweet recipes, and it has nothing to do with flavour.

Where crystals are wanted instead

Not every sugar preparation is trying to avoid crystallisation. Fondant and fudge depend on it, and the skill there is producing crystals so small that the tongue reads them as smooth rather than gritty. The method is to cook the syrup to the target concentration, then cool it substantially before agitating it, because a cool supersaturated syrup nucleates in many places at once and yields a large number of tiny crystals rather than a few large ones.

Beat a fudge while it is still hot and you get a coarse, sandy texture, which is one of the most common failures in home sweet-making and one of the most avoidable. Same syrup, same concentration, entirely different product — decided by when the spoon went in. That is the general shape of sugar work: the composition sets the possibilities and the handling decides which one you get.