LAYERING IN THE TUOLUMNE INTRUSIVE SUITE FORMED BY LIQUID IMMISCIBILITY

New data suggest that layering in ladder structures (LS; a.k.a. ladder dikes) in the Cathedral Peak Granodiorite (Kcp) reflects liquid immiscibility. Because other layered rocks in the Tuolumne Intrusive Suite (TIS) follow the same compositional trend, liquid immiscibility may account for much of the layering in the TIS.

Field observations from Kcp favor the magma tube model of Weinberg et al. (2001) in which LS layers mark successive positions of the trailing wall of a migrating tube. LS compositions trend at a high angle to the TIS trend. The LS trend has been attributed to crystal-liquid separation but that is unlikely. Mafic layers contain extreme concentrations of titanite and apatite which typically appear late in crystallization, and mafic layers have lower Mg numbers than felsic layers. The linear trend nonetheless implies either mixing or unmixing of end members, one near the TIS trend at ~60 wt% SiO₂ and one with very low SiO₂ and very high Fe₂O₃, P₂O₅, and TiO₂. Cm-scale globular fine-grained inclusions found only in or very near LSs were previously interpreted as xenoliths, but they lie on the LS trend at more extreme compositions (e.g., 36 wt% Fe₂O₃) than found in mafic layers. Continuous sheets of the same fine-grained black rock form the margins of at least one LS. We interpret the inclusions to be crystallized blebs of Fe-rich immiscible melt, and the marginal sheets to reflect concentration of immiscible melt along a tube wall. Compositions of felsic rocks formed in tubes at their final locations fit this interpretation. The rocks contain 60-63 wt% SiO2, i.e., less siliceous than typical Kcp, and lie on the LS trend but range off of the TIS trend. The most felsic samples plot on the opposite side of the TIS trend from mafic layers. Felsic rocks in LSs clearly differ from the Kcp and could have formed by removal of Fe-rich immiscible melt from tonalite or low-SiO₂ granodiorite similar to the Kuna Crest or Half Dome units.

Compositional variation in LSs (and perhaps in layered rocks elsewhere) appears to record one or both of: 1) exsolution of immiscible melts at varying temperatures along a solvus, and 2) varying degrees of physical separation of immiscible melts before crystallization. The layering per se may result from shearing of a mixture of higher (felsic) and lower (Fe-rich) viscosity immiscible melts in the magma tube.