N. L. Bowen's generally accepted hypothesis that how rocks crystalize from magma--whether they turn out mafic, intermediate, or felsic--depends on the magma's composition and temperature.