Tempering is the management of six crystal forms of one fat

One fat, several ways of being solid

Cocoa butter is unusual among edible fats in that it is polymorphic: the same molecules can pack into six distinguishable crystal arrangements, conventionally numbered I to VI, and each one melts at a different temperature and behaves differently in the mouth and on the shelf. This is the entire reason tempering exists. Melted chocolate left to cool on its own will crystallise, and it will crystallise into whichever forms happen to nucleate first, which is not the one anybody wants.

The desirable arrangement is form V. It melts at roughly 34°C, which is just below body temperature, and this is why properly tempered chocolate is hard at room temperature and then goes suddenly and completely liquid in the mouth rather than turning waxy. It also contracts as it sets, which releases a moulded piece cleanly, and it produces a dense reflective surface and the sharp report when a bar is broken. Form VI is more stable still, but it forms slowly over weeks and is associated with bloom rather than gloss.

The lower forms are faster and worse

Forms I through IV melt between roughly 17 and 28°C. They nucleate more readily than form V, which is the awkward fact at the heart of the process: left to itself, cocoa butter takes the quick route. Chocolate that sets in these forms is soft at room temperature, dull to look at, greasy on the tongue, and it will not release from a mould because it does not contract properly. It is entirely edible and it is not what anyone means by tempered chocolate.

So tempering is a selection problem rather than a heating problem. The task is to end up with a melted mass that contains a modest population of stable form V seed crystals and none of the unstable ones, so that when the whole thing cools, those seeds template the rest of the fat into the same arrangement. Everything in the standard procedure follows from that single objective.

What the temperature curve is doing at each stage

The first step takes the chocolate above about 45 to 50°C. This is not simply melting; it is erasure. Every crystal form, including the stable ones, has to be destroyed so that the starting point is genuinely uniform, since a surviving cluster of form IV will happily seed the whole batch into form IV.

The chocolate is then cooled — for dark chocolate to somewhere near 27 to 28°C — which is low enough that crystallisation begins in earnest. Both stable and unstable forms appear during this stage, which is expected. The final step brings the temperature back up, for dark chocolate to roughly 31 to 32°C, and this is where the selection happens. That temperature is above the melting point of the unstable forms and below the melting point of form V, so the undesirable crystals melt away while the good ones survive. Milk and white chocolate contain milk fat, which softens the mixture, and are worked one to two degrees cooler throughout.

Seeding is the same logic by a shorter road

Tabling — pouring two thirds of the melted chocolate onto a stone slab and working it with a scraper until it thickens — was the traditional method, and it does the cooling stage very efficiently because a thin film on cold stone loses heat fast while the constant movement distributes crystals evenly. It is also messy, and it needs a cool room.

Seeding achieves the same end differently. Solid tempered chocolate is stirred into the melted batch, and because that solid chocolate already contains form V crystals, it supplies the seed population directly instead of requiring you to grow it. The melting of the added pieces also pulls the temperature down. The method is more forgiving in a domestic kitchen, though it depends entirely on the seed chocolate being properly tempered to begin with, which is easy to assume and worth checking.

Reading the result

The traditional test is to dip a knife or a strip of paper into the chocolate and leave it somewhere cool. Correctly tempered chocolate sets within a few minutes, evenly, with a matt-to-glossy surface and no streaks; untempered chocolate stays tacky for far longer and dries with a dull, mottled look. There is no way to judge it from the bowl, and the temperature reading alone is not sufficient evidence, because a batch can sit at 31°C with no seed crystals in it at all.

Ambient conditions matter more than people expect. A warm kitchen keeps the working chocolate above its setting range and stretches everything out; a cold one thickens it in the bowl as unstable crystals accumulate, and it will have to be warmed slightly and brought back. Humidity is a separate hazard entirely, since any condensation reaching the chocolate causes it to seize. Tempering is not difficult so much as unforgiving of inattention, which is a different problem.