The genesis of divergent magmatic series that originate from subalkaline, basaltic parent magmas has been a preoccupation for igneous petrologists since Fenner and Bowen, neither of whom appreciated the central role that convergent-margin magmatism plays in the creation and evolution of the continental crust. Petrologic studies of volcanic centers within a portion of the Andean SVZ (Tupungato to Calbuco) have contributed data and ideas pertinent to the consequences of intra-crustal differentiation, mantle heterogeneity, and subduction erosion for continental-arc magmas. Crustal thickness, which nearly doubles from south to north (~30-55 km), and structural arc-segmentation (33.3-34.3°S - 34.8-36°S - 36.2-41.1°S) define the framework for examining contrasting magma evolution paths on regional and local scales. Mantle-source compositions do not vary regularly as a function of latitude. These issues are addressed in terms of factors related to CA versus TH differentiation with a compilation of ~2800 WR analyses (TH:CA ~40:60). Extreme TH differentiation requires nearly closed-system fractional crystallization of H₂O-poor and relatively reduced magmas at low pressures. Late Holocene to historic lavas from Lonquimay (38.4°S) define such a trend. Comparably extreme closed-system CA magmas (e.g.,Holocene Longaví: 36.2°S) are fractionation products of oxidized and H₂O-rich magmas at mid- to lower-crustal depths. The remaining large majority of analyses defines a gradational body of data that occupies the compositional space between these extremes, straddles the Miyashiro Discriminant (MD), and manifests variable open-system contributions. Magma mixing and crustal assimilation are expressed as convergence of evolved compositions on the MD, from both directions. Data subsets for segments and individual volcanoes display tendencies for: (1) higher fractions of CA analyses and (2) greater open-system contributions with increasing crustal thickness. These are consistent with greater average and maximum depths of magma evolution and a higher probability for more extensive crust-magma interactions as ascent-path lengths increase. These first-order observations are derived from combinations of non-traditional plotting parameters that provide new perspectives on this long-standing issue.