Many previous interpretations of deep seismic reflection profiles across the Great Glen strike-slip fault system have postulated that the fault penetrates the crust and upper mantle as a vertical plane that sharply offsets the Moho discontinuity. Such an interpretation has become a general feature of more recent regional syntheses of the deep geology of the British Isles. Reprocessing of portions of four profiles across the fault improves the resulution of lower crustal and upper mantle structure and calls into question the ''Moho step'' interpretation. Diffraction analysis and seismic migration applied to reflection data north of Shetland across the northward continuation of the Great Glen fault (Walls Boundary fault) indicate a narrow synform or ''keel'' developed on the Moho directly beneath the fault. This keel is itself underlain by a possible mantle reflection that is cut off by the downward projection of the fault. North and west of Ireland, analyses of amplitudes, frequencies, and the geometrical behavior of reflections upon migration show that structures previously interpreted as Moho steps may be better explained as distinct packages of reflectivity that are uppermost mantle or possibly out of the plane in origin.

In one location north of Ireland where some three-dimensional control exists, the vertically downward projection of the Great Glen fault intersects, without disrupting, dipping structure in the upper mantle. This observation leads to a model for displacement on the fault system in which motion is laterally transfered along a dipping ramp (or blind thrust) in the uppermost mantle, somewhat analogous to models develped for the San Andreas fault that indicate displacement along the fault to be laterally offset within the middle crust. One of the principal conclusions of this study, that major vertical steps on the Moho beneath the Great Glen fault are difficult to support from the available seismic data, is consistent with rheologically based studies which predict that Moho ''topography'' such as vertical steps is unlikely to be preserved over long periods of geologic time. ¿ American Geophysical Union 1995