Magma formation at ocean ridges is the primary process that produces the large scale lithologic and chemical heterogeneity of the Earth. Studies on mid-ocean ridge basalts (MORBs) have documented that their mantle sources are chemically and isotopically heterogeneous at all spatial scales. Pyroxenite components in the mantle likely have a key role in the major processes of MORB formation, including polybaric melting, melt segregation from the source region and melt migration through the lithosphere. Because direct sampling of deep mantle is not accessible, however, the origin of pyroxenite heterogeneities is poorly constrained, thereby leaving their strategic role in the generation of oceanic basalts still controversial. This project wishes to integrate field, petrological and geochemical studies on ophiolitic and oceanic mantle sequences with experiments on solids and melts, to explore the heterogeneity of the MORB mantle. Major goal of the project is to improve our knowledge on: (1) the role of pyroxenites in creating chemical and isotopic heterogeneities in the MORB mantle sources, (2) the consequence of a heterogeneous mantle source on the chemistry of MORBs, and (3) the relationships between mantle heterogeneity and channelized melt migration. To address these issues, the project plans to investigate: (i) the origin and geochemical signature of pyroxenites in fertile and depleted oceanic peridotites, which represent the best available proxies of MORB mantle sources and mantle residues after MORB production, (ii) the chemical/isotopic modifications produced by pyroxenite formation in the host peridotite, (iii) the origin of secondary pyroxenites by melt-peridotite reaction, and (iv) the chemical variability of melts migrating through replacive dunites. The selected sample set encompasses: (i) "aged" pyroxenites from fertile mantle sequences of the Alpine-Apennine ophiolites; (ii) pyroxenites from depleted mantle sequences of the Alpine-Apennine ophiolites and modern oceanic setting (Smoothseafloor region, Southwest Indian Ridge); (iii) abyssal peridotites - MORB suites from Vema Lithosperic Section (Mid Atlantic Ridge) and Smoothseafloor region; (iv) replacive dunites from the Alpine-Apennine ophiolites. The origin of secondary pyroxenites will be simulated by melt-rock reaction experiments. A set of high-pressure experiments (1.5-2.5 GPa) will be devoted to define the distribution of trace elements during interaction between peridotite and pyroxenite-derived melts. In addition, partial melting experiments (1200-1400 °C; 1.0-1.5 GPa) on pyroxenite lithologies will shed light on the contribution of these components in the genesis of MORBs. Key information will be provided on the petrologic processes controlling the generation and evolution of lithological-chemical-isotopic heterogeneities along the mantle column, thus improving our knowledge on magmatic processes ruling the chemical differentiation of the Earth's mantle and oceanic crust.