During the past three years the PIs have synthesized major element data from divergent and convergent plate margins into quantitative models of mantle melting that account for the global systematics in a tectonic context. While these models need to be tested, they provide a tool for understanding many related problems of petrogenesis. This project will examine the global data base for ocean island basalts to test how well the major element melting models work in the tectonic setting of within-plate volcanism. In addition, the major element models will be extended to investigate the relationship between the major and trace elements