Igneous Rocks

Although names are legion, igneous rocks are formed from a small number of mineral types

This article is intended to present a scheme for remembering the composition of igneous rocks and their relations in connection with their names. Igneous rocks cannot be given names reflecting a specific chemical or mineralogical composition, since these are too variable, although they are composed of minerals of more or less definite composition. They have, instead, been given arbitrary names drawn from geographic names and other sources, with no inherent meanings or mnemonic value. These names have been assigned in great profusion, sometimes reflecting the characteristics of a particular occurrence or suggesting certain associations, and it is very difficult to keep them sorted out in the mind. Moreover, the names reflect texture and mode of occurrence as well as mere composition, adding to the difficulty of mental identification. This account should not be thought of as giving the actual processes of formation of igneous rocks, but simply a logical framework.

An igneous rock is one that has been formed by the reaction of its constituents in an environment that has allowed them to more or less freely associate in forms that minimize the free energy U - TS, or minimum internal energy and maximum entropy, at moderately high temperatures. This need not take place in the liquid state, and indeed it probably seldom does, but also in the presence of active fluids that carry in certain ingredients and carry out others, usually at elevated temperatures. It is too simple to say that igneous rocks crystallized from a liquid magma, though the result is similar. Because different constituents are available, the igneous rocks formed in the crust are different from those formed in the upper mantle. Besides silica, aluminium and the alkalis (K, Na and Ca) are characteristic of the crust, iron and magnesium of the mantle. Accordingly, igneous rocks are classified as sial or acidic, and sima or basic. The terms acidic and basic are not well-chosen, but are traditional. Silicon (the acidic ingredient) is present in all igneous rocks, but free silica, SiO2, only in the more acidic rocks. Rocks typical of the mantle are different from the more common basic igneous rocks, and are called ultrabasic. We shall see what this means in terms of mineral constitution.

The crust of the earth is 59% silica, SiO2, and 15% alumina, Al2O3, so silicon and aluminum must the predominant constituents of any igneous rock. If we mix silica and alumina and heat, the result is pretty much silica and alumina, which have very different crystal structures, and the free energy cannot be lowered by any crafty arrangements. When we add cations, positively-charged atoms like K+ and Na+, the situation changes. Now aluminum tetrahedra, with a composition AlO2, can be associated with the silica tetrahedra, with compositon SiO2, so long as we add one Na or K for each Al to keep things electrically neutral. (Nothing with net charge gives a low free energy). The silicon and aluminium ions are both small (0.039 nm and 0.057 nm, respectively) and fit equally well within four big oxygens. This allows the building of a very stable tectosilicate, which in this case is feldspar. Rocks of this general composition are also called felsic, from feldspar and silica. Felsic rocks are more than 65% silica.

With potassium, we get orthoclase, KAlSi3O8, and with sodium plagioclase, NaAlSi3O8. Note how similar these crystals are. Sodium is a smaller ion than potassium, so slightly different structures result, and the two cations do not substitute for each other. If there is some water around (which can contribute OH ions) it is also possible to make a sheet silicate, with 5 oxygens to every silicon. If we add Al(OH)2·AlO2 (this is electrically neutral) to the orthoclase components, we get the mica muscovite, KAl3Si3O10(OH)2. This formula where everything is packed together is a mess, but we can analyze it into meaningful parts. Now when we cool our magma, we find that all the aluminum is tied up in feldspar and mica, the relative amounts of orthoclase and plagioclase depending on abundances of Na and K. The silica that remains is what was not required in making feldspar and mica, so we get a rock that is mainly crystalline feldspar with silica between the feldspar crystals, interspersed with flakes of mica. This is an example of the rock granite.

Making such a muscovite granite requires a very pure starting mix, with no Ca, Fe or Mg. If there is a little Fe and Mg, it can be soaked up by including (Mg,Fe)3 instead of the Al2 in the muscovite. Mg++ and Fe++ have about the same size (0.078 nm and 0.083 nm, respectively), and so can substitute for one another freely in the crystal. The difference between ions of the same charge and radius is very small, and mixing them up improves the entropy anyway. The result is K(Mg,Fe)3AlSi3O10(OH)2, the black mica biotite, which is very much more common than muscovite. Again, there is some order in the apparently complicated chemical formula. Therefore, starting with silica, alumina, some cations, and a little iron and magnesium, we get a rock composed mainly of feldspar (using up the aluminium), a little biotite (using up the iron and magnesium), with what silica is left over as clear, crystalline free silica. This is a typical pink granite, the pink coming from the feldspar.

The interchangeablility of Fe++ and Mg++ is mirrored by the interchangeability of Na+ and Ca++. The charges are different, but this simply means that we use twice as many sodiums as calciums. This gives a complete series of plagioclase feldspar, from all-sodium albite to all-calcium anorthite, with oligoclase, andesine, labradorite, and bytownite in between. This is a perfect example of the terminological chaos in mineralogy and petrology--none of the arbitrary names gives the slightest clue to its position in the series, and all the names refer to what is essentially the same mineral. Therefore, if our mix contains some calcium, it will probably go into the feldspar. We might also note that OH- and F- ions have the same charge, and about the same size (0.014 nm and 0.0133 nm, respectively), so they readily substitute for one another. This means that micas can contain F as well as OH, and do. All magnesium gives amber mica, phlogopite, while lithium in place of potassium gives pink mica, lepidolite.

The names for these granites reflect the composition of the feldspar. The name granite is reserved for rocks in which potassium feldspar, orthoclase, dominates. If potassium and sodium feldspar are comparable in amount, then it is called granodiorite. With mainly sodium feldspar it is quartz-diorite. These names refer to rocks with visible crystals, or coarse texture. With fine texture, they become rhyolite, rhyodacite and dacite, respectively. The fine-textured rocks were usually erupted as a hot colloidal suspension (sol), the particles of which later coagulated into what appears to be a rock solidified from a liquid, but which was probably never liquid at all. To add to the name confusion, rocks of porphyritic texture, consisting of large crystals in a fine-textured groundmass, are given still other names: quartz porphyry, granodiorite porphyry, and quartz porphyrite, when really the only difference between them is the type of alkali feldspar. Therefore, we have six names for what is really only one rock by composition: quartz-feldspar-mica granite. There are, of course, gelogical reasons for the multiplicity of names.

If the proportion of silica in the mix is smaller, there may not be any free silica left when all the aluminium is used up. This gives us three rocks analogous to the granite sequence: syenite, monzonite and diorite. The fine-grained analogues are trachyte, trachyandesite and andesite, and the porphyritic ones porphyry, monzonite-porphyry and porphyrite. We now have nine names for what is, compositionally, almost the same thing, with varying amounts of silica and a varying K:Na ratio. These names reflect the widespread and multifarious occurrences of this fundamental crustal rock. Porphyry originally referred to a beautiful red Egyptian rock of this texture, but now is used more generally for rocks of any color.

The textures of igneous rocks are often classified by terms that imply the mode of emplacement of the rocks. These are extrusive, hypabyssal or intrusive, and plutonic. They merely reflect the texture of the rock as fine, medium or coarse, not specifically its mode of emplacement, although extruded rocks are usually fine-grained or glassy, plutonic rocks coarse-grained, and hypabyssal rocks graded between these limits. It would be best if these terms were not used in connection with rock types, since they may lead to unwarranted conclusions. As describing modes of emplacement, they are, of course, unexceptionable.

If there is not enough silica to take up all the aluminium, feldspathoids may be formed instead of feldspars. These are likewise tectosilicates. One is leucite, KAlO2·2SiO2, and another is nepheline, KAlO2·SiO2, with only two and one silica units, respectively, instead of the three in a feldspar. Like feldspars, these minerals are hard and of medium density, but weather more readily. They are relatively rare, and rocks containing them do not appear to have any special names. Rocks poor in silica are called undersaturated.

Calcium has not yet been considered, though it is rather common, especially in rocks with a deeper source. When it occurs, it has two main effects. First, it can substitute for Na in plagioclase, as already mentioned, even making the all-calcium feldspar anorthite. Second, it can facilitate the synthesis of chain silicates in cooperation with Mg and OH. With some Al around, it can preferentially form hornblende, an amphibole. In fact, diorite and andesite can be composed almost entirely of plagioclase and hornblende, where it seems that the calcium has soaked up all the silica not used in making feldspar, in the form of hornblende. If there was very little aluminium, we wind up with a rock that is all hornblende, hornblendite or amphibolite. Amphiboles have double chains, and an 11:4 O:Si ratio. Hornblende is hard (5-6), rather heavy (3.2), and can be fibrous or translucent. It is dark green to black, and includes Ca, Na, Mg, ferrous and ferric iron, aluminium, OH, F and, of course, Si.

If there is no aluminium available, calcium and magnesium may form pyroxenes instead. These are single-chain silicates with a 3:1 O-Si ratio, such as CaSiO3 and MgSiO3. The calcium ion is a bit larger than the magnesium ion, 0.106 versus 0.078 nm, but both can be accomodated in the same crystal. A 1:1 mix gives diopside, CaMgSi2O6, or the common augite, Ca(Mg,Fe,Al)(Al,Si)2O6, very similar except for substitutions. Pyroxenes are, like amphiboles, dark and heavy minerals.

When silica is mixed with aluminium, calcium, magnesium and iron oxides, as in material derived from the mantle, the result is calcium plagioclase, typically anorthite, and pyroxene, typically augite. This is a very familiar rock, basalt. It is not a mantle rock, but one that comes from mixing mantle and crustal rocks in varying amounts. Basalt is the fine-grained extrusive form. The less common coarse-grained form is gabbro, and the frequently-seen hypabyssal or intrusive form is dolerite or diabase. These are very important rocks indeed, associated with volcanic eruptions, and arriving near the surface as fluid lavas.

Mantle rocks are typified by olivine, consisting of the substitutional series between MgSiO3, forsterite, and FeSiO3, fayalite, typically dark and heavy, though pure forsterite with a trace of iron gives the attractive gem mineral peridot. If basaltic rocks contain olivine, their names are prefixed with olivine- to show this, and close association with the mantle is indicated. If there is no aluminium, then there is no feldspar, and the rock is called peridotite, and is ultrabasic. Peridotite can contain pyroxene as well, but no free silica, which would react with the other ingredients, and perhaps tremolite, a Mg-Ca amphibole. Amphiboles are weathering products of pyroxenes. Rocks containing olivine, pyroxene and amphibole are often called mafic, from magnesium and ferrous; they contain from 44% to 52% silica. Rocks with from 52% to 65% silica are called mesosilicic or intermediate in composition.

The rock names given above are mainly those in Arthur Holmes's simplified nomenclature (see Reference, p. 116). There are many other rock names of more special application, which are not always used consistently, nor given an obvious definition. For example, a tholeiite has been defined as a basalt that does not contain olivine or similar mantle-like ingredients. However, one prominent book on marine geology refers to an olivine-tholeiite.

To illustrate the important effect that rock texture has on naming, we consider some porphyritic granites. First, however, let's look at graphic granite. A broken surface looks as if it is covered with runic symbols that resemble writing, with parallel alignments. The rock is actually one huge crystal of intergrown silica and feldspar. Both are tectosilicates, made up of silicon and aluminium tetrahedra, so it is not surprising that they can intergrow with the maintenance of crystal order over long distances. This intergrowth can exist on a microscopic scale, when the rock is called microgranite or micropegmatite, since graphic granite is often associated with pegmatites and their huge crystals. The suffix -phyre hints at porphyritic texture, appearing in granophyre, one of the rocks of interest to us. Granophyre has quartz and feldspar crystals, perhaps not very large but still evident to the eye, in a fine-grained groundmass of microgranite. Granophyre is a characteristic result of metamorphic recrystallization in the presence of active fluids, completely changing the texture of a rock while the composition changes very little. Doris Reynolds found all stages in the transformation of very old granodiorite (granite with roughly equal amounts of potassium and soda feldspar) into newer granophyre at Slieve Gullion in Northern Ireland. The old Torridonian sandstone of the Northwest Highlands has also been found transformed into granophyre. The periphery of many granitic batholith-like bodies has also been altered to granophyre, as in the Wichita Mountains of Oklahoma. In composition, granophyre is just a typical granite, but it is the texture that is remarkable.

We should also glance at granite porphyry and quartz porphyry to distinguish them from granophyre. Granite porphyry has the same phenocrysts of quartz and feldspar as granophyre, but the groundmass is just ordinary fine-grained granite. In fact, it often grades into granophyre. Quartz porphyry can be just a medium-grained intermediate between rhyolite and granite, hardly different from granite porphyry, but it is sometimes defined as an extrusive rock, in which large chunks of granite, perhaps in the form of quartz or feldspar crystals, have been erupted with a hot suspension of similar material, which has welded and cooled into a rhyolite-like groundmass. This shows how care must be taken to define a rock carefully before using the name, so that the compositional and textural significance of the name is clear to the reader.


Arthur Holmes, Principles of Physical Geology, 2nd ed. (New York: Nelson, 1965). Chapters IV and V.

C. S. Hurlbut, Jr., Dana's Manual of Mineralogy, 16th ed. (New York: John Wiley & Sons, 1952).

Return to Geology Index

Composed by J. B. Calvert
Created 24 February 2003
Last revised