Previous studies of the granite of Lankin Dome have led to the conclusion that this granite was a source for the sandstone-type uranium deposits in the basins that surround the Granite Mountains, Wyo. Q-mode factor analysis of 29 samples of this granite shows that five bulk compositions are required to explain the observed variances of 33 constituents in these samples. Models presented in this paper show that the origin of the granite can be accounted for by the mixing of a starting liquid with two ranges of solid compositions such that all five compositions are granitic.
There are several features of the granite of Lankin Dome that suggest derivation by partial melting and, because the proposed source region was inhomogeneous, that more than one of the five end members may have been a liquid. Data for the granite are compatible with derivation from rocks similar to those of the metamorphic complex that the granite intrudes. Evidence for crustal derivation by partial melting includes a strongly peraluminous nature, extremely high differentiation indices, high contents of incompatible elements, generally large negative Eu anomalies, and high initial lead and strontium isotopic ratios. If the granite of Lankin Dome originated by partial melting of a heterogeneous metamorphic complex, the initial magma could reasonably have been composed of a range of granitic liquids.
Five variables were not well accounted for by a five-end-member model. Water, CO 2 , and U0 2 contents and the oxidation state of iron are all subject to variations caused by near-surface processes. The Q-mode factor analysis suggests that these four variables have a distribution determined by postmagmatic processes. The reason for failure of Cs0 2 to vary systematically with the other 33 variables is not known. Other granites that have lost large amounts of uranium possibly can be identified by Q-mode factor analysis.