INTRODUCTION

Magmas that reach Earth's surface to form lavas are highly varied, ranging in composition from ultramafic komatiites, through basalts and andesites, to rhyolites and feldspathoidal felsic rocks. Although the compositions of these lavas may not represent all magmas formed in the Earth (some may be too dense to rise to the surface, and others may require high pressures to keep volatile fluxes in solution), they do indicate the enormous diversity of magmas. Explaining the origin of this diversity has been the dominant goal of petrology.

Early in the history of the science, most of the different magmas were thought to have independent origins; some were interpreted as the products of magma mixing, and still others the products of magma splitting (immiscible fractions). From early in the twentieth century, however, the trend has been to interpret the wide diversity of igneous rocks as being derived from only a few primary magmas. The process by which these magmas are modified is known as magmatic differentiation. For example, N. L. Bowen (1928), who championed this new interpretation, argued that basalt was the primary magma from which other magmas were derived. So persuasive were his arguments that during the first half of the twentieth century, the interpretation that compositional variations in magmas might reflect primary variations in the source region was almost completely neglected. Although petrologists now recognize the importance of the source region, magmatic differentiation is still considered the major cause for variations in the composition of suites of igneous rocks.