Linear inversion of &tgr;(p) composites for three different profiles suggests that the lower crust beneath eastern Pennsylvania consists of material with high averaged P wave velocities (6.8--7.5 km/s) compared with 6.4--7.1 km/s for most crustal models derived for surrounding areas within the Appalachians and the Grenville Province of the Canadian Shield. Interpretations of these velocities in terms of rock type are limited by the low resolving power of the data and by the nonuniqueness in inferring rock type from P wave velocity alone. The high averaged P wave velocities could be associated with substantial volumes of mafic rock, perhaps emplaced by underplating, or with high-grade metamorphic assemblages of intermediate composition and sedimentary origin, similar to those found in the Ivrea Zone in northern Italy. Because of the sparseness of the S wave data, averaged shear wave velocities are not well enough resolved over most depth ranges to distinguish among the different compositional models. Computations of gravity anomalies for two-dimensional models using various linear velocity-density relations, however, suggest that at least part of the crust beneath the Scranton Gravity High is mafic. Moreover, if the entire basement section beneath this anomaly consists of mafic material, then the gravity studies indicate that the lower crust beneath the Great Valley must be mafic as well. Migrated shot gathers show a concentration of reflectors at depths between 35 and 45 km, a result that is controlled in part by the limited range of offsets used, but that is also consistent with recent observation of a highly reflective lower crust. Reflections that are migrated to depths within the uppermost mantle could be multiples or primary reflections from upper mantle layering. Although the data are rather sparse, the interpretation of these later arrivals as primary reflections is supported by their small ray parameters, which suggests that they are not converted phases, and by their large relative amplitudes, which suggests that they are not multiples. Layering with the upper mantle could be in the form of alternating high- and low-velocity zones, perhaps related to emplacement of eclogites or garnet granulites in the uppermost mantle during the process of underplating. ¿ American Geophysical Union 1992 |