By using magnetic anomaly data from a detailed geophysical survey west of the Juan de Fuca rise between longitudes 143¿W--134¿W and latitudes 42¿N--48¿N the history of spreading at the Pacific--Farallon spreading center in this region can be reconstructed for the period 35--20 m.y. ago (anomalies 12--5E). During this time period, relative migration of spreading axes separated by transform faults resulted in the elimination of the offset represented by the Surveyor fracture zone. Magnetic anomalies in the southern part of the region require eastward jumps of spreading centers of between 40 and 50 km, and those in the northern part imply westward jumps of up to 70 km. The locations of the spreading center jumps migrate along spreading axes with time, concurrently with northward or southward jumps of transform faults, and leave zones of extensively sheared crust with unidentifiable magnetic anomaly patterns in the crust between old and new spreading centers. Such a process may account for the disturbed zone of magnetic anomalies between the Murray and Molokai fracture zones and could be common to all ridge jumps. If so, it suggests that the new spreading centers do not begin simultaneously over long lengths but instead develop in a manner somewhat similar, but not identical, to a crack propagating through a solid.