Mid-ocean ridge basalts (MORB) from the well-defined Southern Explorer Ridge segment (SER) in the northeast Pacific Ocean are moderately to strongly enriched in incompatible elements. Enriched MORB were erupted at the highest and widest part of the segment (culmination) where the magma supply may be the greatest and also at the northern and southern ends of the SER. Variations in basalts' incompatible element enrichment occur over short distances and suggest that the underlying mantle is heterogeneous on a small scale. The variations also preclude the existence of a long-lived, well-mixed magma chamber beneath the robust culmination.

Instead, magma chambers are probably temporally and spatially isolated, as evidenced by the presence of highly differentiated lavas. Less enriched (transitional) MORB were erupted along the central part of the SER, 11--27 km south of the culmination, and were supplied to the ridge separately from a less enriched part of the mantle. A more continuous magma chamber cannot be ruled out for this section on geochemical grounds. All of the MORB from the SER have undergone significant amounts of fractional crystallization. Unlike certain segments of the East Pacific Rise, the most evolved MORB were erupted near the culmination of the ridge segment. Aphyric lavas have been recovered only from the vicinity of the culmination, while phyric lavas have erupted along the entire lengtrh of the SER. There is no correlation between composition and total crystal abundances. Despite the requirement of multiple parental magmas, most of the lavas from the culmination and the distal ends of the ridge fall on a common liquid line of descent, indicating that they formed at similar depths and extents of partial melting.

The transitional MORB from the central part of the ridge have probably formed by lesser extents of partial melting and separated from their mantle source at greater depths based on abundances of Na and moderately incompatible elements such as Sm and Zr. Enriched, high-melt fraction lavas at the magma-starved southern end of the SER closely resemble lavas from the culmination. Either the southernmost lavas have flowed laterally within the crust from near the culmination, or there is another melting anomaly that supplies the southern end of the SER. If the latter case is true, correspondence between ridge morphology and mantle thermal states is poor. ¿ American Geophysical Union 1989