Multibeam bathymetry, deep-towed side scan sonar, gravity, and magnetic data over a 50 km square area all suggest the presence of a second-order, nontransform offset on the obliquely spreading Reykjanes Ridge (North Atlantic ocean) near 58¿N latitude. It is the first such offset to be recognized on the Reykjanes Ridge. This region is characterized by a shallow median valley containing en echelon axial volcanic ridges (AVRs) similar to those found on more northerly parts of the Reykjanes Ridge. The side scan sonar shows that all the AVRs are constructional in nature. The one immediately south of the offset basin backscatters strongly, is unmarked by faulting, and so appears extremely young. Other AVRs appear older, having lower backscatter and, off axis, being cut by faults and fissures. At 57¿55'N the progression of AVRs is interrupted by a 600-m-deep, 20 km ¿10 km basin. Residual mantel Bouguer anomalies, corrected for two-dimensional lithospheric cooling, display a high of at least 8 mGal over the basin. Two small, off axis basins occur roughly along the flow line from this basin and are also characterized by gravity highs. The basins are interpreted as regions of crustal thinning and are believed to represent the discontinuous trace of the ridge offset, which has thus been in existence for at least 2 m.y. and has been slowly propagating south. Inversion of the magnetic field shows a very low magnetization (~3--4 A m-1) associated with the offset basin, which is interpreted as indicating a reduced magnetic layer thickness due to poor magma supply to the offset area.

The Brunhes-Matuyama reversal boundary and Anomaly 2 traces are offset about 3 km dextrally across the proposed offset trace. The AVR immediately north of the offset displays high amplitudes of magnetization which steadily increase southward toward its tip adjacent to the offset, suggesting the presence of increasingly highly fractionated basalts toward the AVR tip. Polarity transition widths across the Brunhes-Matuyama boundary are comparable with those on other slow spreading ridges but are wider than individual AVRs, probably because AVRs overlap in the crustal accretion zone. Transition widths are also asymmetrical across the spreading center, being narrower (~4.5 km) to the east than the west (~8 km), suggesting a certain degree of asymmetry in the accretionary and rifting processes at the ridge axis. ¿ American Geophysical Union 1994