Résumé
Cores from Holes BA1B, BA3A and BA4A - drilled from the mantle section of the Samail Ophiolite (Oman) - have been systematically sampled and analyzed during Phase 2 of the Oman Drilling Project. 82 samples were selected from Hole BA1B, 62 from Hole BA3A, and 70 from Hole BA4A. Their major, trace and volatile elements and mineralogy were determined aboard D/V Chikyu August-September 2018. For the three holes, harzburgites and dunites are the most common rocktypes recovered with subordinate gabbro and pyroxenite dikes present in all holes, though their proportions vary in each hole. All samples are moderately to highly altered. The geochemical signatures and co-variations seen in each lithological group are very consistent across the three holes suggesting the underlying processes are common among them. Harzburgites and dunites are generally restricted in their composition and fall on differentiation trends defined for Oman peridotites. Harzburgites are relatively depleted (Al (sub 2) O (sub 3) <1%) though some show relative enrichments in Al (sub 2) O (sub 3) and TiO (sub 2) suggesting melt refertilization. Dunites show correlation between Ni and Mg# that suggests coupled Ni loss and Fe-Mg exchange, suggestive of equilibration with a basaltic melt. The peridotites from the three holes are highly to completely serpentinized and have CO (sub 2) contents in excess of estimated mantle concentrations ( approximately 300ppm). Filtering the data to exclude shallow samples gives a mean CO (sub 2) content of 0.21% with little downhole variation, suggesting minor but pervasive carbonation has taken place. Gabbros are primitive (Mg# 80-90) and show compositions intermediate between peridotites and Oman lower crustal gabbros. Pronounced variation in all major element abundances and complete lack of differentiation trends with varying SiO (sub 2) (37-50 wt.%) and Mg# are interpreted as most likely due the result of interaction with wall-rock peridotites during emplacement of the dikes. However, the influence of alteration processes such as rodingitization cannot be ruled out. Key outstanding questions include: the causes of background CO (sub 2) enrichment, the relationship between gabbro dikes and lower crustal gabbros elsewhere and whether the gabbro signatures represent a "missing link" - rarely sampled but naturally ubiquitous - between source peridotites and intruded crustal gabbros.