Granite plan: granite granitic rocks are classified according to the qapf diagram for coarse

Alphabet classification system
The composition and origin of any magma that differentiates into granite leave certain petrological evidence as to what the granite's parental rock was. The final texture and composition of a granite are generally distinctive as to its parental rock. For instance, a granite that is derived from partial melting of metasedimentary rocks may have more alkali feldspar, whereas a granite derived from partial melting of metaigneous rocks may be richer in plagioclase. It is on this basis that the modern "alphabet" classification schemes are based.
The letter-based Chappell & White classification system was proposed initially to divide granites into I-type (igneous source) granite and S-type (sedimentary sources).^[25] Both types are produced by partial melting of crustal rocks, either metaigneous rocks or metasedimentary rocks.
I-type granites are characterized by a high content of sodium and calcium, and by a strontium isotope ratio, ⁸⁷Sr/⁸⁶Sr, of less than 0.708. ⁸⁷Sr is produced by radioactive decay of ⁸⁷Rb, and since rubidium is concentrated in the crust relative to the mantle, a low ratio suggests origin in the mantle. The elevated sodium and calcium favor crystallization of hornblende rather than biotite. I-type granites are known for their porphyry copper deposits.^[23] I-type granites are orogenic (associated with mountain building) and usually metaluminous.^[9]
S-type granites are sodium-poor and aluminum-rich. As a result, they contain micas such as biotite and muscovite instead of hornblende. Their strontium isotope ratio is typically greater than 0.708, suggesting a crustal origin. They also commonly contain xenoliths of metamorphosed sedimentary rock, and host tin ores. Their magmas are water-rich, and they readily solidify as the water outgasses from the magma at lower pressure, so they less commonly make it to the surface than magmas of I-type granites, which are thus more common as volcanic rock (rhyolite).^[23] They are also orogenic but range from metaluminous to strongly peraluminous.^[9]
Although both I- and S-type granites are orogenic, I-type granites are more common close to the convergent boundary than S-type. This is attributed to thicker crust further from the boundary, which results in more crustal melting.^[23]
A-type granites show a peculiar mineralogy and geochemistry, with particularly high silicon and potassium at the expense of calcium and magnesium^[26] and a high content of high field strength cations (cations with a small radius and high electrical charge, such as zirconium, niobium, tantalum, and rare earth elements.)^[27] They are not orogenic, forming instead over hot spots and continental rifting, and are metaluminous to mildly peralkaline and iron-rich.^[9] These granites are produced by partial melting of refractory lithology such as granulites in the lower continental crust at high thermal gradients. This leads to significant extraction of hydrous felsic melts from granulite-facies resitites.^[28][29] A-type granites occur in the Koettlitz Glacier Alkaline Province in the Royal Society Range, Antarctica.^[30] The rhyolites of the Yellowstone Caldera are examples of volcanic equivalents of A-type granite.^[31]
M-type granite was later proposed to cover those granites that were clearly sourced from crystallized mafic magmas, generally sourced from the mantle.^[32] Although the fractional crystallisation of basaltic melts can yield small amounts of granites, which are sometimes found in island arcs,^[33] such granites must occur together with large amounts of basaltic rocks.^[23]
H-type granites were suggested for hybrid granites, which were hypothesized to form by mixing between mafic and felsic from different sources, such as M-type and S-type.^[34]However, the big difference in rheology between mafic and felsic magmas makes this process problematic in nature.^[35]