This is in order to be able to consider p parallel meshes that each contain mesh chunks, which are all part of one main computational mesh.