Reply to the comments on “Xenoliths in ultrapotassic volcanic rocks in the Lhasa block: direct evidence for crust–mantle mixing and metamorphism in the deep crust”