LADDER DIKES, CRAZY GEOCHEMISTRY, AND LIQUID IMMISCIBILITY(?) IN OTHERWISE SANE GRANODIORITES

Ladder dikes (LDs) continue to confound attempts to explain their field relations and geochemistry. They are commonplace but volumetrically insignificant in many granodiorite plutons worldwide and are particularly well-exposed and abundant in the Late Cretaceous Cathedral Peak Granodiorite of Yosemite National Park, California. Mafic portions of LDs have extremely high abundances of rare-earth and high-field-strength elements, and here we explore the possibility that liquid immiscibility may explain these unusual compositions.

Ladder dikes are roughly tabular bodies, commonly 30-100 cm thick, composed of nested, concave-up, semi-cylindrical, mafic-felsic troughs; stacking of troughs gives them an overall tabular, dike-like form. Trough axes are highly scattered but generally plunge at moderate angles away from the north-south axis of the Tuolumne Intrusive Suite. Cut-off angles between mafic layers suggest upward migration of layers, and LDs commonly terminate laterally/upward in a subcircular mass of leucocratic rock, 50-100 cm across, that fills the highest trough and is rimmed by titanite ± mafic minerals. Although xenoliths are exceedingly rare in the surrounding granodiorite (<0.001 vol%), cm-scale xenoliths composed largely of fine-grained hornblende are common (up to ~0.1 vol%) within LDs. Mafic layers of LDs are enriched over felsic layers 5-25 times in most cations with valences >1, with prominent exceptions being Si, Al, Pb, Sr, and Ba—all of which are concentrated in feldspars and quartz. Such element partitioning can be reasonably explained by crystal-liquid separation, although the phase equilibria of such a process are problematic, and fluid dynamic considerations rule out the usual cross-bedding interpretation of layer cutoffs. Liquid immiscibility can also explain this mafic/felsic partitioning while avoiding phase equilibria problems. Immiscibility would produce an extremely dense and low-viscosity liquid that would percolate downward through the silicate magma. Although attractive from a geochemical standpoint, it is difficult to reconcile immiscibility with field relations. However, layers in LDs may bear as little relation to crystal sedimentation as Fe-oxide layers in picture sandstone bear on sand sedimentation.