Thanks, its starting to make a lot more sense now. How do these orbitals produce an inversion center though? From what I read, the p orbitals become the t1u set in an octahedral field. p orbitals are symmetrically distributed around the atom though so wouldn't they have an inversion center? As for the eg set, it contains a dz2 orbital, I can't really see how an inversion center could apply to that at all.