The failing spreading system associated with the Galapagos 95.5 ¿W propagator system comprises a complex set of volcanic and structural provinces. Data from detailed surveys using GLORIA SeaMARC II, Sea Beam, Deep-Tow, the submersible Alvin, magnetics, and bottom cameras constrain the plate boundary geometry and evolution of this area. Although the failing spreading system may involve diffuse spreading, the patterns of sedimentation, volcanism, and faulting suggest localization. The failed spreading axis, composed of a set of right-stepping en echelon basins, extends and ages southeastward from offset and overlaping doomed/failing axes. The failing axis is offset ~13 km left-laterally from the doomed axis. Although there may have been a preexisting offset in the doomed axis, lack of structural evicence for this and lack of an offset of this magnitude in the Brunhes/Matuyama magnetic reversal argue against this possibility. A model whereby the offset is transient does not explain the pattern of failed spreading basins which leads directly away from the curently failing axis.

Detailed structural and volcanic patterns suggest that this offset may be caused by a westward progression of northward jumps in the location of the failing axis (like a secondary propagator), probably in association with the migrating strain/stress field of the migrating transform zone of the propagator system. This is modelled as a ''migrating extensional relay zone'' (MERZ) similar to the relay zones observed at some fast-slipping transform faults except that it propagates westward with the overall system. Bathymetric and magnetic evidence for tectonic elements associated with the MERZ supports this model, although the expected extinct trace of the doomed axis is not obvious in the present, incomplete data set. Unlike other models for the plate boundary geometry, the MERZ model does not require serendipity for temporal and spatial juxtaposition of the primary propagator and the offset within the failing spreading system. Observation of a similar geometry in the Lau Basin supports this preferred model. Present data from the Galapagos failing spreading system are inconclusive, however, and the MERZ model remains speculative. If valid, this model implies that in at least some cases, paired propagators act together. The MERZ is too short and deep to drive its own propagation according to standard cracking models. ¿ American Geophysical Union 1989