Bowen's Reaction Series: Mineral Crystallization

In the 1920s, Norman Bowen heated natural rock powders to melting, then cooled them slowly in the laboratory. He found that minerals did not all crystallise at once; they appeared in a consistent order, with high-temperature minerals forming first and low-temperature minerals forming last. He summarised this order as Bowen's reaction series.

The series has two branches that run in parallel from high temperature to low temperature. Together they predict which minerals you will find in a cooling magma — and, just as importantly, which minerals you will not find together.

The discontinuous branch tracks iron-magnesium silicates that crystallise one after another, each reacting with the melt to produce the next as temperature falls:

1. Olivine (~1200 °C) — the highest-temperature common silicate. Rich in iron and magnesium, it crystallises first from mafic and ultramafic magmas.
2. Pyroxene (~1100 °C) — as the melt cools and reacts with olivine, pyroxene replaces it. Olivine is consumed if equilibrium is maintained.
3. Amphibole (~1000 °C) — with more cooling and water in the melt, amphibole forms from pyroxene.
4. Biotite mica (~800 °C) — the last Fe–Mg silicate, stable only in cooler, wetter, more evolved melts.

Each step is a reaction: the earlier mineral is unstable in the cooler melt and transforms into the next one. That is why the branch is called discontinuous — the crystal structures change abruptly.

The continuous branch is the plagioclase feldspar series. Unlike the Fe–Mg minerals, plagioclase keeps the same crystal structure all the way down, but its composition shifts gradually:

• At high temperature: calcium-rich plagioclase (anorthite, CaAl₂Si₂O₈). It has more calcium and aluminium.
• As temperature falls: the plagioclase progressively incorporates more sodium, becoming sodium-rich plagioclase (albite, NaAlSi₃O₈). The change is smooth and continuous.

A single plagioclase crystal may have a calcium-rich core (formed early) and a sodium-rich rim (formed later) — a texture called zoning that records the cooling history directly.

In nature, crystals often sink to the bottom of a magma chamber or are otherwise separated from the melt. When that happens, the removed crystals no longer react with the remaining liquid. The melt is left depleted in the elements that went into those crystals — a process called fractional crystallization.

Removing olivine (rich in Fe and Mg) leaves the melt richer in silica, aluminium, and sodium. As cooling continues, pyroxene crystallises and is removed, then amphibole, then biotite. Each step pushes the residual liquid toward higher silica content. A magma that started as mafic basalt can evolve into an intermediate andesite or even a felsic rhyolite if enough crystals are removed.

This is the primary mechanism by which a single magma body produces a range of compositions — and it is the reason intermediate and felsic rocks exist at all.

When solid rock is heated, melting does not happen all at once. The lowest-temperature minerals melt first, extracting a liquid that is richer in silica and dissolved volatiles (especially water) than the original rock; sodium enrichment is a secondary effect. The leftover solid is depleted in these elements and becomes more mafic.

This is exactly the reverse of fractional crystallisation. In crystallisation, the first minerals to form are high-temperature (mafic). In partial melting, the first minerals to melt are low-temperature (felsic). A basaltic magma produced by partial melting of peridotite is the complement of the olivine-rich solid left behind.