LADDER STRUCTURES IN THE CATHEDRAL PEAK GRANODIORITE, YOSEMITE: TWO-PHASE FLOW IN FINGERED DIKES?
LS layers define linear trends on chemical plots, consistent with differentiation either by mixing or unmixing. The trend leaves the main batholith trend at a high angle at ~60 wt% SiO2, well below the silica content of host Kcp. Xenoliths are nearly absent from Kcp but make up as much as 1% of LSs. These observations indicate that LS magma was not locally derived.
LSs range to extremely low concentrations of felsic components (e.g., <30 wt% SiO2) and extremely high concentrations of everything else (e.g., Fe2O3t>40 wt%, HFSE, REE), characteristics of immiscible liquids. Partitioning of elements between the mafic layers and host granodiorite matches that in natural immiscible pairs quite well. Such a mafic immiscible liquid would be extremely dense and quite fluid. Felsic rock inside terminal tubes also lies on the LS trend, but off the Sierran trend on the opposite side from the LS layers. The rock in terminal tubes thus is the only rock found in the Kcp that is likely to be complementary to LS layers.
Pipe flow of two fluids with different viscosities commonly adopts an annular geometry with the low-viscosity fluid at the margins. If the conduit is gently inclined and the densities of the fluids differ, the high viscosity fluid is located off-center. This predicts that the high-Si magma will be in the center and offset toward the top of the conduit. This is closely consistent with the observed geometry of LSs, favoring an origin by annular flow of immiscible liquids in a tubular conduit.