Comments on the use of London's field dependent orbitals