Our study indicates the following conclusions:
K-feldspar megacrysts in the TB grew from a the liquid portion of a magma at conditions of unusually low nucleation-to-growth ratio, presumably at low degrees of supersaturation. We see no evidence of subsolidus growth, nor that grain coarsening played a significant role beyond the nucleation and microscopic stage of growth.
There is ample evidence of the physical accumulation by several processes (filter pressing, flow sorting, buoyant rise, mixing/mingling) of K-feldspar megacrysts during which the concentration of K-feldspar megacrysts increased from a few volume percent to as high as 80 volume percent and resulted in concentrations in irregular patches, troughs, tubes, plumes, dikes, and enclaves.
Clustering of K-feldspar megacrysts cannot be explained by potassium-rich fluid fluxes, particularly where the megacrysts occur in more mafic layers or in very high in more mafic layers or in very high concentrations. The lack of molding relationships and local evidence of contact melting when megacrysts collide in cluster s support this conclusion.
Evidence of the physical accumulation of the megacrysts has several implications for internal chamber processes. In the TB, numerous, local instabilities preferentially collected certain minerals and formed local compositions and structures. Thus the modes of rocks in the TB do not reflect the original magma compositions [see also Healy et al., 2000, Weibe et al., 2002]. The compositional and structural heterogeneity also argues against widespread convection, at least late in the chamber evolution, and instead implies that local gradients existed in the TB, although the nature of these gradients remain unclear [ [ Zak and Paterson, in review]. Moreover, the megacryst concentrations indicate that the megacrysts and surrounding magmas did not share the same flow history [e.g., Bergantz, 2000] further complicating the physical meaning of magmatic fabrics [Paterson et al. 1998; Healy et al. 2000].