Tuttle & Bowen demonstrated that haplogranites form by magmatic processes and they favored a derivation by fractional crystallization of basalt magma. However it is now widely thought that these felsic granites result more commonly from partial melting of quartzofeldspathic rocks in the crust. The Devonian Yeoval Batholith (1400 km2) of the Lachlan Fold Belt contains haplogranites formed by fractional crystallization, but from a primary magma less mafic than basalt, of potassic intermediate composition. A process of convective fractionation produced mafic cumulate rocks below felsic differentiates, the latter being close to the haplogranite ternary minimum. Separation of cumulates from the fractionating melt was enhanced by an initially low solid fraction in the magma because this initial magma was at a relatively high temperature. Its potassic composition meant that a final haplogranite melt could not be accommodated between the framework of early-formed minerals, and was partly expel. A bimodal separation resulted from what was probably an initially uniform magma. Phenocryst-poor felsic volcanics also formed and represent batches of the differentiates that erupted from a subjacent fractionating pluton. These volcanic rocks include rhyolites that erupted at temperatures ~ 1050 ºC.

Granites that formed from these relatively hot magmas belong to the high-temperature group (GSA Spec. Pap. 389, pp. 1-10). They are relatively uncommon in the Lachlan belt where they have a total exposed area of 2300 km2 and comprise ~ 5% of all granites and ~ 10% of I-type granites. They were derived from source rocks that were deeper and partially melted to higher temperatures than for most Lachlan granites. The products of solidification from the high temperature magmas are striking examples of the Tuttle and Bowen model. However compositional variations seen within most suites of granites in this region, and in low-temperature granite suites more generally, were not produced in this way.