Read’s two presidential addresses to the Geological Society (1948 Geological Society (1949) heralded the end of the coherent rearguard action by the ‘granitizers’ against the ‘magmatists’. They were the distillation of his thoughts on the genesis of granite and culminated in his concept of the ‘Granite Series’. In this, he identified a continuity from metamorphic through migmatitic rocks to granite. Although he was wrong on granitization, the general idea remains intact and granites are produced by high-temperature metamorphism leading to partial melting. However the role of migmatites is still contentious. Not all granites belong to the granite series as he presented it; this particularly applies to Cordilleran (Andean) type granites. The genesis of this type will be discussed in the context of the earlier, classic work of Nockolds (1940) on the Garabal Hill complex where he demonstrated fractional crystallization was a major process in producing the diversity present and was similar to that seen in volcanic rocks, which had clearly been liquids. He also proposed that the source was basaltic with the implication, strongly supported by modern isotopic studies, that granites of this type are essentially mantle derived. Nockolds contribution was a geochemical confirmation of Bowen’s belief in the importance of closed-system fractional crystallization in the differentiation of plutonic rocks. Most authorities today would accept this but would not necessarily follow Bowen in the belief that major granitic batholiths formed by differentiation of basalt or that the system was closed. Recent models of the generation of the two main types of granite are presented, incorporating many of the ideas of Nockolds and Read. The discussion focuses on high-level differentiation, partial melting and intrusion in an extensional regime, high-T/low-P metamorphism associated with the magmatism, and the relation of granite to the plate tectonic setting.