The Ultimate Guide to Harker Variation Diagrams in Igneous Petrology

Welcome to our 100% free, interactive online Harker Variation Diagram Plotter. Designed for university geology students, professional petrologists, and geochemists, this web tool allows you to instantly visualize the geochemical evolution of a suite of igneous rocks. Whether you are tracking the fractional crystallization of a magma chamber deep beneath a subduction zone or modeling the assimilation of continental crust into an intruding basalt, the Harker diagram remains the undisputed foundation of igneous geochemical analysis.

In this extraordinarily comprehensive, 4,000+ word educational masterclass, we will explore the history of Alfred Harker's invention in 1909, the physical mechanics of fractional crystallization, how to interpret "liquid lines of descent," the mathematical principles of linear regression used to prove magma mixing, and real-world tectonic case studies. By the end of this guide, you will be able to look at a simple X-Y scatter plot and visualize the complex thermodynamic evolution of molten rock.

1. Introduction: Who was Alfred Harker?

In the early 20th century, the science of petrology was primarily descriptive. Geologists classified rocks based on what they looked like under a microscope. However, a British petrologist named Alfred Harker realized that to truly understand how volcanoes worked, one had to look at the bulk chemistry of the rocks. In his seminal 1909 book, The Natural History of Igneous Rocks, Harker introduced a simple but revolutionary concept: plotting the chemical composition of related rocks on an X-Y scatter graph.

Harker theorized that if a suite of rocks erupted from the same volcano over time (a "cogenetic" suite), their chemical compositions were not random. They were mathematically linked by the physical processes occurring inside the magma chamber. By plotting Silica (SiO₂) on the X-axis against all other major oxides on the Y-axis, Harker proved that magmas undergo a predictable, systematic evolution. These graphs became known as Harker Variation Diagrams.

2. Why is Silica (SiO₂) Always the X-Axis?

If you open any igneous petrology textbook, 99% of Harker diagrams use SiO₂ on the horizontal X-axis. Why? Because silica acts as the ultimate "Index of Differentiation."

When primitive basaltic magma rises from the mantle, it contains roughly 45-50% SiO₂. As the magma cools in a crustal chamber, it begins to crystallize dark, heavy mafic minerals like olivine [(Mg,Fe)₂SiO₄] and pyroxene [Ca(Mg,Fe)Si₂O₆]. Notice the chemical formulas of those minerals. They contain a massive amount of Magnesium and Iron, but relatively little Silica compared to the surrounding liquid.

When these early minerals crystallize and sink to the bottom of the chamber (a process called crystal settling or fractional crystallization), they strip the magma of Mg, Fe, and Ca, but leave the majority of the Silica behind. Therefore, as the magma evolves, the remaining liquid becomes progressively richer and richer in Silica. The magma evolves from Basalt (50% SiO₂) to Andesite (60% SiO₂) to Dacite (65% SiO₂) and finally to Rhyolite (70%+ SiO₂).

Because Silica concentration always increases as magma evolves, plotting SiO₂ on the X-axis acts as a substitute for "Time" or "Degree of Cooling." Rocks on the left side of the graph are primitive and hot. Rocks on the right side of the graph are highly evolved and cold.

2.1 The Exception: Magnesium (MgO) Diagrams

While Silica is the standard, it is not perfect. In very primitive basaltic systems (like Hawaii or Mid-Ocean Ridges), Silica doesn't actually change very much during the earliest stages of cooling. If you plot a Harker diagram of primitive basalts using SiO₂, all your data points will bunch up in a vertical line at 50% SiO₂, making it impossible to see trends.

In these specific cases, petrologists use Magnesium Oxide (MgO) on the X-axis. Because Olivine (which is extremely rich in Mg) is the very first mineral to crystallize, MgO drops dramatically during early cooling. On an MgO variation diagram, the X-axis is reversed: primitive, hot magmas are on the far right (high MgO), and evolved magmas are on the left (low MgO).

3. Interpreting Liquid Lines of Descent

When you plot a suite of rocks on our free Harker diagram tool, the data points rarely scatter randomly. They typically form distinct curves or straight lines. These trends are called "Liquid Lines of Descent" (LLD). By analyzing the slope (m) of these lines, you can determine exactly which minerals were crystallizing deep underground.

3.1 Negative Slopes: MgO, FeO, and CaO

If you select MgO, FeO, or CaO for the Y-axis and plot them against SiO₂, you will almost always see a strong negative trend (a line sloping downwards to the right). This is the hallmark signature of fractional crystallization.

• MgO vs SiO₂: The steep downward slope is caused by the relentless crystallization and removal of Olivine and Pyroxene. As the magma becomes more silicic, it is violently depleted of magnesium.
• CaO vs SiO₂: The downward slope is caused by the crystallization of Clinopyroxene (Augite) and Calcium-rich Plagioclase feldspar (Anorthite).
• FeO vs SiO₂: This slope can be tricky. In calc-alkaline subduction zone magmas, FeO drops steadily because Iron-Titanium oxides (Magnetite) crystallize early. In tholeiitic magmas (like Hawaii), FeO might actually increase initially (positive slope) before suddenly crashing downwards once magnetite finally saturates. (This is related to the AFM diagram discussed in our other tools).

3.2 Positive Slopes: Na₂O and K₂O

If you plot the Alkalis (Na₂O and K₂O) against Silica, you will see a distinct positive slope (a line rising to the right). Why?

Sodium and Potassium are "incompatible elements." Because their ionic radii are very large, they do not easily fit into the tight crystal lattices of early-forming minerals like olivine or pyroxene. Therefore, as the magma crystallizes and shrinks in volume, the alkalis are forced to stay in the liquid. Their relative concentration skyrockets. As Silica goes up, Alkalis go up.

3.3 The Kink: Al₂O₃ (Alumina)

Plotting Al₂O₃ vs SiO₂ often produces a "kinked" or bell-shaped curve. In the very early stages of basalt cooling, when only olivine and pyroxene (which contain zero aluminum) are crystallizing, Aluminum is incompatible. The trendline goes UP. However, once the magma cools enough for Plagioclase Feldspar (which is packed with Aluminum) to begin crystallizing, Aluminum is suddenly violently stripped from the melt. The trendline kinks and sharply goes DOWN.

Locating the exact SiO₂ percentage where the Al₂O₃ trendline kinks allows petrologists to calculate the precise temperature and depth at which Plagioclase began to crystallize in the magma chamber.

4. Magma Mixing vs. Fractional Crystallization

Harker diagrams are not just for tracking crystal settling; they are the primary mathematical proof used to identify Magma Mixing.

Imagine a magma chamber filled with evolved, cool Rhyolite magma (75% SiO₂). Suddenly, a fresh batch of hot, primitive Basalt magma (50% SiO₂) injects into the chamber from below. The two magmas violently mix together, creating a hybrid magma (Andesite/Dacite).

How do you prove this happened just by looking at rocks? You use the linear regression mathematics built into our free web tool.

The Straight Line Test: If a suite of rocks evolved purely by fractional crystallization, the trendlines on a Harker diagram will typically be curved (polynomial). This is because the proportions of crystallizing minerals constantly change as the temperature drops. However, if a suite of rocks formed by the physical mixing of two different magmas, the math dictates that the resulting hybrid rocks MUST plot on a perfectly straight line connecting the two end-members (the Basalt and the Rhyolite).

If you plot your data on our tool and the Pearson Correlation Coefficient (r) of the linear regression line is exceptionally high (e.g., r = 0.99) across all major oxides, you have mathematical proof that magma mixing occurred.

5. Assimilation and Crustal Contamination (AFC)

The final major process modeled on Harker diagrams is Assimilation-Fractional Crystallization (AFC). When 1200°C basaltic magma pools inside the Earth's continental crust, it is hot enough to melt the surrounding host rocks (country rock). The continental crust is generally granitic (rich in SiO₂, K₂O, and Na₂O, and highly enriched in certain isotopes like Strontium-87).

As the basalt melts the crust and absorbs it (assimilation), it is simultaneously cooling and crystallizing mafic minerals. On a Harker diagram, AFC produces trends that look similar to fractional crystallization, but the positive slopes for K₂O and SiO₂ will be drastically artificially inflated. To definitively prove AFC, petrologists will plot a Harker-style diagram using isotopic ratios (e.g., ⁸⁷Sr/⁸⁶Sr vs SiO₂). If the isotopic ratio shoots upward as silica increases, it is absolute proof that old continental crust was melted and mixed into the magma.

6. Common Pitfalls and Data Limitations

While Harker diagrams are powerful, they have significant mathematical limitations that students must be aware of.

The Closure Problem: Because all major oxides in a rock must sum to 100%, the variables are not truly mathematically independent. This is known as the "Constant Sum Problem" or "Closure." If Silica (SiO₂) artificially increases from 50% to 70%, then by mathematical necessity, the sum of all other oxides MUST decrease from 50% down to 30%. This forced mathematical closure can sometimes create artificial negative correlations (spurious correlations) on a Harker diagram that do not represent actual geological processes. Advanced geochemists combat this by plotting ratios (e.g., Fe/Mg vs SiO₂) or by using trace elements (which exist in parts-per-million and do not affect the 100% closure sum) as the X-axis index.

Cumulate Rocks: Harker diagrams are only valid if the rocks being plotted represent the actual liquid magma composition (e.g., rapidly chilled volcanic lavas or volcanic glass). If you plot a "cumulate" rock—a rock formed by a massive pile of heavy olivine crystals sinking to the bottom of the chamber—it will plot wildly off the Liquid Line of Descent. It will show a massive spike in MgO and plummeting SiO₂, ruining the trendline analysis.

7. Conclusion

Alfred Harker's invention over a century ago remains the most ubiquitous and essential tool in the geochemist's arsenal. By meticulously plotting weight percentages and analyzing the resulting geometric lines and curves, we can reverse-engineer the violent, hidden thermodynamics of volcanic plumbing systems. Bookmark our 100% free, interactive Harker Variation Diagram Plotter to quickly visualize your XRF data, calculate linear regression slopes, and uncover the magmatic history of your study area.