Basalts collected by dredging between 23¿N and 25¿N include samples from the active spreading ridges median valley walls, and older walls of the Kane Fracture Zone transform fault at about 24¿N. Chemical analyses of over 100 basalts and basalt glasses show the depletion in large ion lithophile elements characteristic of 'normal' ocean ridge basalts. Basalt suites recovered from the median valley north and south of the fracture zone are almost identical geochemically. There is no evidence of present or past volcanic activity within the transform zone. Petrogenesis of the basalts, as deduced from both petrographic and geochemical evidence, indicates that simple low-pressure equilibrium crystallization during ascent in shallow vents or flow in seafloor lava tubes can account for a large part of the major element and trace element variation within discrete data subsets. However, more complex models are required to explain relations between samples showing large differences in large ion lithophile element enrichments. A subset of samples of intermediate composition can be explained by simple mixing of highly fractionated basalt with mafic parental basalt. A number of basalts enriched in Al2O3 may reflect accumulation and partial resorption of plagioclase. By comparison to existing phase equilibria experiments, it is shown that a basalt-type common to the area is a logical choice as a mantle-derived primary magma; it is close to the composition of average normal MORB with wt. % TiO2?1.50, Mg♯?66, and normative olivine 10--15 wt. %. This composition is inferred to have separated from the mantle at a pressure of about 8--9 kbar. Rare examples of more mafic 'primary' basalt are present and are inferred to represent melts which escaped from the mantle at higher pressures. We outline a comprehensive hypothesis for melting, fractionation, and mixing which we believe may be adequate to explain compositional variation in these and other suites of 'normal MORB.' An important consequence of this model is that MORB presently at a given position on a low-pressure, multiply saturated cotectic may have reached that point by a variety of paths; some may be nearly direct mantle derivatives, while others may have experienced a complex history of polybaric and low-pressure fractionation and/or mixing.